Citation
Modeling coupled water-heat flow and impacts upon chemical transport in mulched soil beds

Material Information

Title:
Modeling coupled water-heat flow and impacts upon chemical transport in mulched soil beds
Creator:
Shinde, Dilip, 1962-
Publication Date:
Language:
English
Physical Description:
xii, 153 leaves : ill. ; 29 cm.

Subjects

Subjects / Keywords:
Liquids ( jstor )
Moisture content ( jstor )
Mulches ( jstor )
Simulations ( jstor )
Soil water ( jstor )
Soils ( jstor )
Solutes ( jstor )
Vapors ( jstor )
Water flow ( jstor )
Water temperature ( jstor )
Dissertations, Academic -- Soil and Water Science -- UF ( lcsh )
Mulching ( lcsh )
Soil and Water Science thesis, Ph.D ( lcsh )
Soil chemistry ( lcsh )
Soil management ( lcsh )
City of Gainesville ( local )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph.D.)--University of Florida, 1997.
Bibliography:
Includes bibliographical references (leaves 142-152).
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Dilip Shinde.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
028638406 ( ALEPH )
38863565 ( OCLC )

Downloads

This item has the following downloads:


Full Text











MODELING COUPLED WATER-HEAT FLOW AND IMPACTS
UPON CHEMICAL TRANSPORT IN MULCHED SOIL BEDS













By

DILIP SHINDE

















A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA



























I dedicate this work to my parents, father (late) Krishna Rao Shinde and

mother Vatsala Shinde, who instilled in me the importance of education and

knowledge.















ACKNOWLEDGMENTS


I express my sincere gratitude to Professors R.S. Mansell (chair) and A.G.

Hornsby (cochair) for their guidance and support during this research work.

Appreciation is also extended to the members of my advisory committee: Professors

P.S.C. Rao, B.L. McNeal, and F. S. Zazueta; for useful comments, help and support.

I would like to extend my appreciation and thanks to Dr. B.P. Mohanty (Univ.

of California, Riverside) for providing numerous technical literature citations.

Thanks are also due to Professor P.N. Kizza for encouragement and support. I would

like to also thank Dr. A. Akpoji and Ron Jessup for useful discussions on finite

element methods and numerical programming, respectively. I acknowledge Dr. John

L. Neiber (Agric. Eng. Dep., Univ. of Minnesota) for providing the FORTRAN code

used extensively in creating finite element-mesh in this work.

I take this opportunity to thank also 'The Rotary Foundation of The Rotary

International' for the prestigious "FREEDOM FROM HUNGER" scholarship to

enhance my knowledge and join the fight to free the world from hunger.

I would like to acknowledge the sit i- r a r, i r1, r ,:.- l ':. I .rT provided by

Professor A.G. Homsby during the latter part of this research. I would also like to

acknowledge the facilities and partial financial support provided by Professors R S.

Mansell and B.L. McNeal, respectively, during this research.

Lastly, appreciation is extended to my family for their help and support.

iii















TABLE OF CONTENTS




ACKNOWLEDGMENTS ........................ ................ iii

LIST OF TABLES .................... ............ .............. vii

LIST OF FIGURES ......................................... .......... ix

ABSTRACT ............ ................................... xi

INTRODUCTION .................. ........................ ......... 1

Coupled Versus Uncoupled Flow Processes in Soil ..................... 1
Soil Therni F-ri .......................... .... .. .2
Non-isothermal Soil Conditions ............................ 3
Mechanistic Mathematical Modeling .............. .... ...... 4
H ypothesis ..................................... ........ 6. 6
Objectives ....................................... ......... 7

REVIEW OF LITERATURE .................................... .. ..... 8

Isothermal Water Flow in Porous Media ............ ............ 9
Soil Temperature and Heat Flow ............... ............. .11
Soil Thermal and Water Regimes under Mulching ................... .. 15
Coupled Heat and Water Flows ................................... 19
Philip-de Vries Theory (PDT) ................................ 20
Application of Philip-de Vries Theory .............. ... ...... 22
Solute Transport during Coupled Heat and Water Flow ................. 24
Temperature Effects on Solute Transport ........................... 27
Summary ...............................................29

MATHEMATICAL MODEL ............................................. 31

Liquid and Vapor Water Flow in Soil ................... .. ..... 32
Heat Flow in Soil ................ .. .............. .......... 32
Transport of Multiple Species of Solutes in Soil ....................... 33
Energy Balance at the Soil/Atmosphere Interface ................. 34









Initial and Boundary Conditions .................................... 38
Input Parameter Requirements ................................. 39

MODEL VERIFICATION AND VALIDATION ............................... 44

Isothermal Water Flow ............... .................. ...... 45
Uncoupled Heat Flow ................ ......................... 46
Coupled Water and Heat Flow ........................ ........ 47
Isothermal 1-D Solute Transport during Steady Water Flow ............. 48
Isothermal 2-D Solute Transport during Steady Water Flow ............. 49
Isothermal 1-D Solute Transport with Nitrification Chain during Steady
W ater Flow ............................ ...... ......... 51
Isothermal 1-D Solute Transport involving Gas Flow ................... 52
Summary ............... ....................................54

MODEL APPLICATION ................ ............................... 55

Fate and Transport of Methyl Bromide Fumigation in Soils ............. 55
Transport Mechanisms for MBr Fumigant in Soil ................ 59
Fate of MBr Fumigant in Soil .............................. 60
Permeation/Transmission of MBr through Polyethylene Mulch .... 63
Modeling o! IE. TI in ri.:.. n .iuriri. Fumigation ................ 68
Analysis of Fate and Transport of MBr Fumigant .................... 72
Approach ........ ................................... 73
Methodology ................... ............ .. ...... 75
Results and Discussion .......................................... 79
Plastic-Mulching of Soil Beds to Control MBr Loss to the
Atmosphere ............... .......... ... ...... ... 79
Non-Isothermal versus Isothermal Soil-Conditions .............. 85
Variable Water Saturation of Soil .......................... .. 87
Depth of Fumigation Injection ................... ............ 97
Effect of Irrigation (Water Dousing of the Bed Surface) ........... 98
Fumigation Efficiency Analysis ............................. 101
Hydrolysis of MBr (Sensitivity Analysis) ....................... 107
Diurnal Water and Heat Dynamics in a Plastic Covered Soil Bed .. 108
Summary and Conclusions .................................... 119

APPENDIX A SOIL MOISTURE FLOW .................. ............ .. 122

APPENDIX B SOIL HEAT FLOW ................. ........... .. 125

APPENDIX C CHEMICAL TRANSPORT .................. ........... .127

APPENDIX D NUMERICAL APPROACH ................................ 132









LIST OF REFERENCES ................... ..................... .... 142

BIOGRAPHICAL SKETCH ............................................. 153















LIST OF TABLES


Table page

2-1. Average thermal properties of major soil constituents at 20 OC and
1 atm (Ghildyal and Tripathi, 1987) .............................. 12

2-2. Short-wave optical properties of five plastic mulches used in commercial
crop production. Data adapted from Ham et al. (1993) .................. 17

5-1. Earlier estimates for source-contributions of MBr to the atmosphere
(Grojesan, 1991; Khalil et al., 1993; Lobert et al., 1990; Mand and
Andreae, 1994; Yagi et al., 1993; and Zurer, 1993) ................. .... 56

5-2. A recently revised budget for atmospheric MBr (Yvon-Lewis and
Butler, 1997) ......... .. ................. ... ... ...... 57

5-3. Selected chemical and physical properties of MBr (CH3 Br) (Adapted
from Hacherl, 1994) ................... ..................... 58

5-4. Measured degradation rate coefficient and half life of MBr as a function
of soil water content (Adapted from Jin and Jury, 1995) ................ 63

5-5. Rate of escape of MBr through several films at various temperatures
(Adapted from Kolbezen and Abu-El-Haj, 1977) ................... .. 65

5-6. Phi,: i,_J Properties for 5 distinct honzons of a profile of Arredondo fine
sand soil .... ................... ..... ................... 77

5-7. Parameter values for 5 distinct horizons of a profile of Arredondo fine
sand soil ................ ....... .............. ... ........ .. 78

5-8. Material balance (unit: %) of MBr under measured and enhanced levels
of diffusion coefficient for plastic mulch ........................... 84

5-9. Material balance (unit: %) of MBr for fumigation under isothermal and
non-isothermal conditions ..................................... .86









5-10. Material balance of MBr (unit: %) under different water saturations for
one and 3 days after furm gant injection ........................... .. 90

5-11. Matenal balance of MBr (unit: %) during fumigation under different
water saturations for 65-cm injection zone .......... .......... 98

5-12. Material balance (unit: %) of MBr under water dousing and no-dousing 100















LIST OF FIGURES


Ficrure pEae

2-1. Measurements of soil temperature at 10-cm depth beneath five different
plastic mulches and a bare plot near Manhattan, KS on day 195 of the
year 1991. Adapted from Ham et al. (1993) ................... ....... 18

4-1. Observed and computed distributions of water content in a sand column
during constant flux infiltration ................... .. ......... 45

4-2. Observed and computed temperatures for two depths in a field soil ........ 46

4-3. Observed and computed water content after 31 days under the influence
of both matric head and thermal gradients .................. .. 48

4-4. Observed and computed breakthrough curves for boron ................ 49

4-5. Location of solute concentration fronts for three times as estimated
numerically and with the analytical solution .................. ...... 50

4-6. Analytically determined and numerically computed concentration profiles
at 200 hours ................................... ............... 52

4-7. Analytically determined and numerically computed concentration profiles
at 1, 3, and 5 hours ............ ......................... ...... 53

5-1. Temperature dependence of MBr diffusion in water (Data from Maharajh
and Walkley, 1973) ........................................... 60

5-2. Laboratory-measured and estimated Henry's Law constants at different
temperatures for partitioning of MBr into water (Adapted from Mutziger
et al., 1996) ................... .................. ...... ..... 61

5-3. MBr emission fluxes in plastic mulched beds during fumigation at the
University of Florida Green Acres Experimental Station located near
Gainesville, FL ................ ............................ 66

5-4. MBr diffusion through black LDPE of different thicknesses (Data from
Hacherl, 1994; and Kolbezan and Abu-El-Haj, 1977) ....... ........ 68

ix









5-5. Simulated and measured flux density through plastic-mulched surface.
Adapted from Wang et al. (1997) .................. ............ 71

5-6. Schematic of the discretized simulation domain of the soil bed ........... 76

5-7. Measured and elevated levels of diffusion coefficients through 1.25-mil
black LDPE plastic-mulch used in model simulations .................. 80

5-8. Losses of MBr to the atmosphere through soil bed and furrow surfaces
using measured and elevated diffusion coefficients for MBr through
plastic mulch ...................................... ......... 81

5-9. MBr total concentration (pmol cm3 ) contours and gaseous MBr flux
vectors (pmol cm2 hr ') in the bed 6 hrs after injection of MBr fumigant... 82

5-10. Emission fluxes of MBr through bed and furrow under isothermal and non-
isothermal transport of fumigant ............................... 85

5-11. Emission fluxes of MBr through bed and furrow during fumigation
performed under different profile water-saturation scenarios ............ 88

5-12. Contours of water content (cm3 cm-3), temperature (C), and MBr
concentration (umol) with gaseous fluxes (pmol cm"2 hr ') of MBr under
2 different soil-water saturation scenarios after fumigation ............. 91

5-13. Emission fluxes of MBr under different water saturations for a 66-cm
injection zone during the first 3 days after fumigant injection ........... 97

5-14. Effect of water dousing on the MBr emission fluxes from the soil bed
during fumigation ................................................ 99

5-15. Contours of water content, temperature, and MBr concentration with
gaseous fluxes of MBr at 9 hours after fumigation under water dousing
and no-dousing on the bed-surface ....... ........................ 100

5-16. Soil sterilization zones for different scenarios 3 days after MBr fumigation
using different criteria ......................................... 104

5-17. Loss of MBr (unit: %) under different rates of hydrolysis of MBr ......... 107

5-18. Diurnal water and heat dynamics in a mulched bed for a wet soil. ....... 113

5-19. Diurnal water and heat dynamics in a mulched bed for a dry soil. ....... 116

D4-1 Linear triangular element ........... .............. .. .. 136

x















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor o! F r~l..:;.:i ri.

MODELING COUPLED WATER-HEAT FLOW AND IMPACTS
UPON CHEMICAL TRANSPORT IN MULCHED SOIL BEDS

By

Fl l .- i L r- .L,

December 1997

Chair: Robert S. Mansell
Cochair: Arthur G. Homsby
Major Department: Soil and Water Science

Plastic-covered soil bed systems are widely used in the southeastern United

States for vegetable and fruit production. Plastic-mulching results in highly variable

and transient soil thermal regimes both in space and time under atmospheric

influences (e.g., solar radiation, wind, etc.). Much evidence exists that water flow is

affected by non-isothermal conditions and vice-versa. In this work, a two-

dimensional model for simultaneously describing chemical transport and coupled

water-heat flows in plastic-covered soil beds was developed with many

improvements over earlier efforts. Realistic plastic-mulch boundary conditions are

thoroughly described by including optical properties of the plastic mulch into the

energy balance at the soil/atmospheric interface along with other atmospheric

influences of weather components. This model provides opportunity to consider the









effect of coupled water-heat flows upon estimates of chemical transport (both in the

liquid and gaseous phases) in plastic-mulch culture production systems.

Chemical transport in soil beds in the model utilize consideration of

chemical existence in solid, liquid, and gas phases. Temperature dependence of the

rates of chemical reactions and volatilization is described efficiently by

incorporating temperature dependence in fate and transport coefficients. The model

uniquely provides opportunity to simulate coupled water-heat flow and multiple-

solute transport with a complex plastic-mulch atmospheric boundary for realistic

field situations.

The model was used to analyze the fate and transport of methyl bromide

(MBr) during fumigation of mulched soil beds. MBr, a highly volatile pesticide, is

widely used in California, Florida, and Hawaii for its broad-spectrum efficiency in

controlling insects, fungi, and weeds. Unfortunately, MBr is a major environmental

hazard due to its depletion effect on the atmosphere ozone layer. Model analyses

revealed that non-isothermal conditions should be considered in the accurate

description of MBr fate and transport. Modeling results indicate that utilizing a

relatively wet soil profile, deeper injection, and water dousing on the bed-surface

before installation of plastic mulch will decrease atmospheric emissions of MBr.

However, deeper injection of MBr fumigant, although environmentally less

hazardous, may decrease effective sterilization of crop root zones in the soil.















CHAPTER 1
INTRODUCTION



Coupled Versus Uncoupled Flow Processes in Soil



Flow of water, air, and heat in soil is generally assumed to be uncoupled,

that is, the driving forces causing individual flow are generally considered to be

mutually independent. Darcy's equation for liquid water flow, based upon spatial

gradients of soil-water potential, and Fourier's equation for heat flow based upon

spatial gradients of soil temperature, provide two examples of uncoupled flow

processes. In nature, however, water is also known to move in response to

molecular forces, such as thermal and osmotic gradients. Similarly, considerable

heat flow may occur as latent heat by water vapor diffusion in the air-filled soil

pores, and as sensible heat transfer due to liquid and vapor movement. In a partially

saturated soil, flow of water due to thermal gradients occurs through zones

occupied by liquid as well as air, and is assisted by multiple condensation and

evaporation cycles Pri, p aip de Vries, 1957). Thus, transport processes occumng

in the soil under field conditions are intricate and complex.

Diffusion theory indicates that water molecules are in constant thermal

motion, accompanied by jump-like transitions from one temporary equilibrium state

to another (Ghildyal and Tripathi, 1987). The frequency of transition and the number

1











of molecules traversing a certain section of the liquid is highly dependent upon the

temperature gradient and causes self thermo-diffusion of water. As temperature

gradients are accompanied by gradients of surface tension at the air-water

interface, there is a possibility of thermo-capillary flow and thermo-capillary film

flow of water in the liquid phase (Ghildyal and Tripathi, 1987). Thermo-capillary

flow occurs in those capillaries where liquid-water menisci exist. Here, temperature

variation induces the variation of capillary pressure and a corresponding shift of the

liquid column. Thermo-: j 1 ll r, I n i ... is caused due to gradients of surface

tension (change in water affinity with change in temperature) at the water-film

surface. Both of these mechanisms cn.rarrt-uir -i. _nri, arri ,'to the total thermal

water transfer under moderate to high water content conditions. Movement of water

in the liquid state due to changes in the volume of entrapped air with temperature

has also been observed (Ghildyal and Tripathi, 1987). In water-saturated or nearly

saturated soil, the entrapped air may cause direct transport of water through the

pressure it exerts and, in unsaturated soil, through an apparent increase in the

temperature coefficient of the matric potential. Convective flow of water vapor may

also occur due to differences in the densities of cold and warm air.


Soil Thermal Regime


The thermal regime of a soil depends upon radiative solar heat flux into the

soil and upon heat-transfer processes occurring in the soil and between the soil and

air which, in turn, depend on the thermal charactenstics of soil solids, gases, and

water. The soil thermal regime, which greatly modifies the overlying microclimate,









3

is generally characterized in terms of soil temperature. Soil temperature fluctuations

are especially important since they impact physical, chemical, and biological

processes in the soil.

The source of heat for soil is incoming net solar radiation. On a clear day,

Geiger (1957) described the typical distribution of incoming solar radiation as: 1)

global radiation, which is the sum of the direct-beam penetration to the earth (19 %)

and the diffuse-sky radiation (26 %); 2) scattered radiation returned to outer space

(11 %); 3) radiation reflected by clouds (28 %); and 4) radiation absorbed by the

atmosphere (15 %).


Non-isothermal Soil Conditions


Much published literature confirms that soil-water movement is affected by

spatial gradients of both temperature and water potential. Thus, under non-

isothermal conditions, movement of water (both liquid water and vapor tend to

move from regions of high to low temperature) through soil should be considered for

both liquid and vapor phases. Non-isothermal conditions have broader implications

in that solute-transport mechanisms are associated with water movement, and the

fate and transport of volatile organic chemicals are also influenced by the soil

thermal environment.

The influence of thermal gradients upon water movement in soils has largely

been observed in two areas (Jury, 1973): 1) in the hot, dry surface layer of a bare soil

where the interface between the soil and atmosphere causes water movement to be

largely in the vapor phase; and 2) deeper in the soil where thermal gradients persist











over seasons and can slowly transport liquid upward or downward. In the special

case of plastic-mulched soil, imposed thermal gradients tend to be greater at the

soil surface and the mulch also provides a partial barrier to vapor flow between the

soil and atmosphere and causes a return flow of water after condensation under the

mulch.

Both water movement and contaminant transport are influenced by

temperature gradients in soils. Natural temperature gradients are usually highest

during daytime (midday) near the land surface, but the relatively small geothermal

gradient below the level of seasonal temperature variation can be a significant

factor in determining the percolation of water in deep unsaturated zones (Evans and

Nicholson, 1987).


Mechanistic Mathematical Modeling


Mechanistic mathematical models provide powerful tools to describe and

interpret flow and transport processes in soil-water systems. Bear and Verruijt (1987)

defined a model as a 'simplified version of the real system that approximately

simulates the excitation-response relations of the latter'. The first step in modeling

involves the construction of a conceptual model of the problem, consisting of a set

of assumptions that reduce the real problem and the real domain -. -- ;r.- i-,i, -e

versions (ultimately permitting mathematical description of the processes) that are

acceptable in view of the modeling objectives. In the second step, the conceptual

model is expressed in the form of a mathematical model (Bear and Verruijt, 1987).

Properly posed mathematical models for initial-value problems involving mass and











energy transport in porous media require (i) appropriate differential transport

equations, (ii) initial conditions in the flow domain, and (iii) boundary conditions

imposed along the surfaces of the flow domain. Models are essential in performing

complex analyses and in making informed predictions. Friedman et al. (1984) stated

that, in some cases, models have increased the accuracy of estimates of future

events to a level far beyond 'best judgement' decisions. Anderson and Woessner

(1992) emphasized that models provide the best way to make an informed analysis

or prediction concerning consequences for a proposed action. The authors further

stated that models can also be used in an interpretive sense to gain insight into the

controlling parameters in a site-specific setting or as a framework for assembling

and organizing field data and formulating ideas about system dynamics.

Schwarzenbach et al. (1993) defined a model in the following way: 'A model is an

imitation of reality which stresses those aspects that are assumed to be important

and omits all properties considered to be nonessential'. These authors summarized

the role of modeling explicitly in the following ways:

1) Mathematical models allow knowledge acquired for different systems and/ or

different situations to be combined in order to provide the ultimate aim of

constructing general theories. Models permit making predictions, which then must

be verified by measurement.

2) Classic scientific theories begin as models, and it is only long-time positive

experience with some of these models which leads scientists to believe that they are

more true and fundamental than others and should thus bear the name 'law'











3) In environmental sciences, models serve to sort out alternative explanations for

observations made in nature which cannot be controlled in the same way as

conditions for a laboratory experiment can be controlled.

4) Models should always be constructed to transfer the knowledge acquired in one

system to another situation, thus enabling a deeper understanding of the processes

responsible for the measured data. A model should be applied in order to design

new experiments or observation programs that are critical for the testing of

hypotheses.

5) Real observations are always superior to outputs produced by a given model. But

model results can help us examine various scenarios and likely future outcomes

regarding the behavior of environmental systems for which real data would never be

collected as frequently and as ubiquitously as needed.



Hypothesis



The soil thermal regime within plastic-mulched soil beds is well-known to be

influenced by interactions between the overlying atmosphere and the plastic-

covered soil surface. The resulting non-isothermal conditions affect water flow and

redistribution. Hence, coupling of water and heat flows is necessary to model

effectively both components within the vadose zone, especially the crop root zone.

The hypothesis here is that model description of chemical transport within plastic-

mulched soil beds under non-isothermal field conditions requires the utilization of

coupled water and heat flows in order to provide reahstic results for management











practices such as fumigation with methyl bromide to control pathogens. Impacts of

this coupled process upon chemical fate (transformations, microbial degradation,

etc.) and transport are expected to be particularly significant for volatile chemicals

such as methyl bromide that may appear in aqueous, gaseous, and solid phases of

soil water systems. Information obtained here regarding the dynamics of the fate

and transport of methyl bromide during fumigation of agricultural soils should also

be useful in evaluating future fumigation with fumigants other than methyl

bromide, which will no longer be used in this country after the year 2001.



Objectives



This research was performed to model and investigate the interactive flows

of water, heat, and solutes in non-isothermal soil environments (the vadose zone) in

order to minimize empiricism from existing knowledge of such systems. The

following objectives were formulated:

1) Develop a mathematical model describing the coupled processes of water,

heat and chemical transport (2-D) for mulched soil-bed systems, where solar energy

imposes non-isothermal conditions.

2) Verify, validate, evaluate, and refine the model as needed for cases of both

coupled and uncoupled flows of water, heat and chemicals with existing

appropriate models (analytical) and published data.

3) Use model simulations to determine impacts of coupled water and heat

transport due to non-isothermal conditions upon the transport of a volatile pesticide

such as methyl bromide during fumigation of plastic-covered soil beds















CHAPTER 2
REVIEW OF LITERATURE


The rate of liquid water movement in soil is important to the study of solute

transport, nutrient availability, efficacy of pesticides, and also water availability to

crop root systems during agricultural production. Physical properties of the soil

influence water movement in two important ways: the size distribution and

structural arrangement of soil particles determine the pore space configuration

through which water moves; and the interaction between the soil and water gives

rise to water-moving forces. In water-saturated soils such as occur below

groundwater tables where pores are completely filled with water, the fluid is a

single phase and forces originating from gravity, including the pressure gradient,

determine the direction and magnitude of flow. The energy potential for the pore

water under saturated conditions is generally positive, or greater than the

atmospheric pressure. The saturated flow process for the simple isothermal case is

described mathematically by Darcy's equation and Laplace's equation.

Generally, under field conditions, soil pores are incompletely or partially

saturated and the energy potential of water is negative (i.e. a suction), being less

than atmospheric pressure. It depends not only upon position in the gravitational

field, but also upon sorptive forces associated with interfacial boundaries. If the

unfilled pore space is filled with gases such as air, movement in the resulting











composite system of solid, liquid and gaseous phases is referred to as unsaturated

flow. This flow condition in soil is particularly important when the groundwater

table occurs at considerable depth below the soil surface (below the vadose zone).



Isothermal Water Flow in Porous Media



Under saturated conditions where the pores of the soil system are completely

filled with water, the transmission coefficient (i.e., hydraulic conductivity) depends

principally upon the viscosity (u,(T)) and density (p,(T)) (both being functions of

temperature, T) of the water and the nature of the porous media. The volumetric flux

of water flow through soil or other porous material is usually given by Darcy's

equation as

q = -K VP [2.1]

where q, is liquid flux, K is hydraulic conductivity, and P is hydraulic pressure of the

water.

Except for the isothermal saturated condition, Darcy's equation, as originally

postulated, does not apply precisely to water flow through soil. However, this

equation has been used since 1856 for describing liquid flow through porous

material with little reservation.

Under unsaturated conditions when the soil contains both liquid and

gaseous phases, the most frequently used modification of Darcy's equation involves

substituting the matric-potential gradient for the hydraulic pressure gradient as the

driving force. Under these conditions, hydraulic conductivity is no longer a constant











but is a function of capillary potential as well as temperature, which yields


q = -K(h, T)VH ; H = h(L, T)+z K= k ) g [2.2


Here His total potential, h is matric potential, z, is gravitational potential (i.e.,

elevation), 8L is liquid volumetric water content, T is temperature, k, is intrinsic

permeability and a. i, tih acceleration of gravity. Under isothermal conditions the

temperature dependence is eliminated and total potential gradient is the driving

force. Richards (1931) derived the equation of isothermal flow through a porous

medium by using the continuity equation .-1..:n ri-ri ni i-. conservation) and

Darcy's equation. Assuming that the medium is incompressible, the continuity

equation for water flow through a homogenous medium was given as


a(p = -VpqL [2.31
at

where t is time and PL is the liquid density. In very dry soils, the capillary or matnc

potential attains great negative values and, in the unsaturated region above a water

table, the capillary driving force dominates the gravitational force.

Isothermal theory has limited application for soils under actual field

conditions, where heat flow is transient in time and varies in space. It has been

observed that heat is evolved at the wetting front when water infiltrates a very dry

soil of fine texture (Anderson and Linville, 1960). Cooling of soil occurs when water

is evaporated from soil (Wiegand and Taylor, 1962) and warming occurs due to

release of thermal energy during condensation of water vapor. Absorption of solar

radiation at the soil surface and chemical or biological activity within the soil may









11

also result in temperature gradients. Thus, temperature gradients must be included

in the flow equation in order to effectively describe water flow in surface soil

characterized by transient thermal regimes.



Soil Temperature and Heat Flow



Under field conditions, the profiles of soil temperature change rapidly during

a normal day. When incoming solar radiation falls on the soil surface the soil surface

becomes heated, causing a difference of temperature between the surface and the

subsoil. This difference causes heat to flow downwards as a temperature wave. As

the heat wave penetrates slowly downward in response to soil thermal properties,

moisture content, and structure of the soil, a phase shift occurs in the temperature

curve with depth which leads to a time lag in the occurrence of temperature

maxima at different depths in the soil. Also, the amplitude of the temperature wave

decreases with soil depth due to storage of thermal energy. During night time, when

the soil surface cools down, the lower portion of the soil profile remains warmer

than the air temperature. Annual temperature fluctuations extend much deeper into

the soil compared to diumal changes, and are also in general periodic in nature.

Soil texture has a distinct influence on soil heat flux and, hence, on soil

temperature. The amplitude of the daily temperature wave decreases in the order

sand > loam > peat > clay. This order can be attributed to the greater heat

capacity of the finer-textured materials and their normally greater thermal

conductivities.











Table 2-1. Average thermal properties of major soil constituents at 20 C and
1 atm (Ghildyal and Tripathi, 1987).

Volumetric Heat Thermal
Material Specific Heat Thermal Conductivity Thermal
(cal -10o-) Capacity 3 1 1 I Difusivity
(calg oC) (calm- C-1) (10 cal cm- s- C-) Diffusivit

Quartz 0.175 0.46 20.00 43.00
Other Minerals 0.175 0.46 7.00 15.00
Organic Matter 0.46 0.60 0.60 1.00
Water 1.00 1.00 1.42 1.42
Air 0.24 0.00029 0.062 0.21


Sandy soils i-ri-r.ajl ,' hold less water and usually drain faster than finer-

textured soils. At low water contents, air occupies much of the pore space and

hydraulic conductivity of the soil becomes limited. As such, thermal capacity and

conductivity of soil with a low degree of water saturation is governed by the

volumetric content of water, which has the highest heat capacity (Table 2-1) among

the substances commonly found in soils. Because of very small thermal diffusivity,

air provides thermal insulation against rapid changes in temperature.

Three mechanisms by which heat flow occurs in soils include conduction,

convection and radiation (Hillel, 1980). Radiative heat transfer is of importance only

in dry soils at high temperatures and within large pores. It is the least-important

mechanism, and is often neglected. Convection can be separated into two

components, forced and free. Forced convection is mostly negligible, with the

exception of infiltration during irrigation or heavy rain. Free convection occurs as a

result of fluid density gradients. Convective heat transport in the vapor phase can be

Spair, r:.r-r,. ,, r.:, ein i ,l, and latent heat components. The sensible heat capacity of











soil air is very low (Table 2-1) and, as soil air flow is most often small, latent heat

transport is of primary importance. Latent heat transport occurs when a fluid such

as water changes phase and is transported as a new phase (such as vapor).

Vaporization of liquid water requires heat (540 cal g-') that is transported in the

vapor phase through a vapor-phase gradient and may be released during

condensation at a location different from where the liquid originated.

A very simple theory of heat transfer in soils is based upon uncoupled mass

and heat flows and considers only heat conduction, without regard for any other

transport mechanisms which may actually occur. The thermal flux in the soil

(assuming that thermal gradients do not result in water flow) is defined from

Fourier's first law of heat conduction as

q = -XVT [2.4)

where qh is thermal flux and A is thermal conductivity. Invoking the principle of

energy conservation in the form of the thermal continuity equation (Hillel, 1980)

provides
8T x
S= VDH VT; with D, (2.51


where DH is thermal diffusivity and C, is volumetric heat capacity of the soil.

Conduction of heat in soils occurs through the solid, liquid, and gaseous

phases. Since these phases have different heat capacities and thermal

conductivities, the main complication when solving the heat conduction equation in

soils is caused by determining the average soil thermal properties, de Vries (1963)

reported a method to calculate heat capacity of soil material by addition of the heat









14

capacities of the various constituents, weighted according to their volume fractions.

The heat capacity as such can be found by adding the heat capacities of the various

phases present as

C = (xsiC s)+X C +XaC [2.6]

where the subscripts s, w and a refer to solid, liquid water, and air phases

respectively; i distinguishes the various materials in the solid phase; and X denotes

the volumetric fraction of a given phase (de Vries, 1963). Thus, volumetric heat

capacity for soil clearly depends upon water content.

The thermal conductivity of a soil also depends in a rather complicated way

on the composition of the soil, in particular its water content. A physical model,

assuming sohd and gas particles to be suspended in the liquid phase, for estimating

the thermal conductivity in soils was given by de Vries (1963) as


x w+ E (k *X)1t +k.*X/
X = X k X +k-X [2.7]



The summation for this equation extends over the different solid soil constituents,

which are characterized by their thermal conductivities and shapes. The

multiplication factor k' in Eq. [2.7] represents the ratio of the spatial average of the

temperature gradient in the soil grains of kind i and the spatial average of the

temperature gradient in water.

The influence of latent heat transfer in air-filled pores is proportional to the

temperature gradient in these pores and can be taken into account by adding the

thermal conductivity of vapor (Lp) to the thermal conductivity of dry air (X,). The











1,, term represents the increase in gaseous-phase thermal conductivity due to

evaporation and condensation of water vapor. At complete dryness and at very low

water contents, the model assumes air rather than liquid to be the continuous

medium. However, at this stage (i.e. completely dry soil), de Vries (1963) proposed

introduction of a multiplication factor of 1.25 to Eq. [2.7] (i.e. 1.25 1). de Vries

reported that errors in calculating X in this manner would not exceed 10 %, unless

the soil grains have a different shape than those assumed.

Wescott and Wierenga (1972) demonstrated the need for coupling heat and

water movement when predicting heat transfer by conduction and vapor movement

for a dry soil. They found that vapor movement accounted for 40-60 % of the heat

flow in the top 2 cm of the soil pr..,' ii- d u rig certain periods of the day and for

20-25 % of the total heat flow even at the 25-cm depth.



Soil Thermal and Water Recimes under Mulchina



Mulches are used in agriculture not only because they suppress weeds and

reduce water loss from the soil surface, but because they also modify the thermal

microclimate of the soil. Rosenberg et al. (1983) presented a good review of mulch

effects on soil processes and defined mulching as the application or creation of any

soil cover that constitutes a barrier to the transfer of either heat or vapor, or both.

Waggoner et al. (1960) applied black plastic, translucent plastic, aluminum

foil, paper, and hay mulches to soil surfaces and measured the resulting alteration

in surface energy balance and in soil and air temperature profiles. Black plastic











reduced the outgoing radiation as was evidenced by the greater net radiation,

whereas hay and paper (and especially paper) increased it. All mulches decreased

the quantity of energy consumed in evaporation by blocking the transport of vapor

out of the soil. Sensible heat generated at the surface was increased by the black

plastic and the hay, and the temperature of the hay and black plastic was elevated.

Black plastic increased in soil temperature, whereas hay and paper had a cooling

effect and decreased the amount of heat penetrating the soil. Ekem (1967) in Hawaii

found that black plastic mulch reduced water use by pineapples grown in

lysimeters. Note here that the well-known practice of surface mulching (for any

purpose) decreases evaporative water loss by creating a layered condition in the soil

due to differential water contents.

The diurnal supply of solar radiant energy at the soil surface during daylight

hours and loss during night-time hours is the major determinant of the thermal

microclimate of soil. The presence of mulch material at the soil surface creates a

-1,. rr,i ,.nfr. effect (similar to the effect of soil layers within the soil). The use of

nonporous mulch such as polyethylene plastic also creates a barrier to water vapor

flow to the atmosphere.

Optical properties of the plastic mulches play an important role in interaction

of the soil surface with the overlying atmosphere. Ham et al. (1993) reported short-

wave optical properties of five plastic mulches used in commercial crop production

(Table 2-2). These properties exhibited a wide range and thus helped explain earlier

work by Waggoner et al. (1960) and Eker (1967). Black and clear plastic mulches

provide extremes in the absorbance of short-wave radiation. The five mulches











described in Table 2-2 were used under bedded conditions in a field and

temperatures were recorded at the 10-cm depth below the mulched surface in the

center of the bed (Ham et al., 1993).


Table 2-2. Short-wave optical properties of five plastic mulches used in commercial
crop production. Data adapted from Ham et al. (1993).

Mulch Reflectance Transmittance Absorbance
Black 0.03 0.01 0.96
Sunfilm 0.12 0.37 0.51
Al-mulchi 0.39 0.01 0.60
White-on-Black' 0.48 0.01 0.51
Clear 0.11 0.84 0.05

1Al-Mulch and White-on-Black were Black mulch in which the upper surface was
painted with aluminum and white paint, respectively.


Diurnal measurements of soil temperature beneath the 5 mulches and a bare

soil plot as recorded by Ham et al. (1993) for a summer day are given in Fig. 2-1. The

highest temperatures were observed beneath the black mulch, which strongly

absorbed shortwave radiation. Cooler temperatures were observed beneath the

Al-mulch and White-on-Black mulches, which had lower absorbances but higher

shortwave reflectances. Although clear mulches are normally used to maximize soil

heating during cooler seasons, in this experiment the investigators placed the

plastic covers in intimate contact with the soil to ensure a minimal average air gap

between the mulch and the soil. The small air gap greatly minimized thermal

contact resistance at this location. Thus, the thermal insulation due to air between

the plastic sheet and the soil surface was minimized in that investigation.











42
Black
Sunfilm
38 Al-Mulch
U W White
Clear
34 Bare Soil


30


(; 26 -
Day of Year-195
22 0 4 8 12 16 20 24
Local Standard Time (hour)

Figure 2-1. Measurements of soil temperature at the 10-cm depth beneath five
different plastic mulches and a bare plot near Manhattan, KS on
day 195 of the year 1991. Adapted from Ham et al. (1993).


Besides mulching, the geometry of the soil bed, the directional orientation of

the beds, crop geometry (canopy structure) and the directional orientation of crop

rows also influence the radiation-energy exchange with the soil surface. The

amount of solar radiation reaching a horizontal unit of the surface of the earth (at a

particular latitude, longitude, and elevation) depends upon a number of factors

including the intensity of radiation emitted by the sun, astronomical considerations

determining the position of the sun, and the transparency of the atmosphere

F...i -r.-trir.: et al., 1983). The sum of direct solar radiation and diffuse sky radiation

received by a unit horizontal surface is referred to as global radiation. In humid

climates, clouds also play a major part in decreasing the amount of solar radiation

received at a location. The vegetative canopy of growing annual crops tends to

provide increasing shade for the soil surface as the growing season progresses, due









19

to interception of incoming solar radiation. The shadow of the growing crop results

in variable radiation interception at the soil surface, thus affecting the thermal

regime in the upper soil layers.



Coupled Heat and Water Flows



The energy state of water molecules in soil changes when a temperature

gradient is imposed on the system, and water moves in response to this energy

gradient. For the duration of the imposed temperature gradient, an irreversible and

spontaneous flow of heat occurs. Thus, heat and water flow processes interact with

each other and require mutual coupling. Where simultaneous movement of more

than one component (like water and heat) occurs in the system, isothermal flow

equations and simple equilibrium thermodynamics cannot be used to predict or

describe the water flow (Enfield, 1970).

If a soil system initially at equilibrium (uniform water content with a

constant water potential at constant temperature throughout) is subjected to a

temperature gradient, water will flow from a warm region to a cooler region (Gur et

al., 1952; Taylor and Cavazza, 1954; Woodside and Cliffe, 1959). When the degree of

water saturation in the soil is relatively high, water flow occurs primarily in the

liquid phase (Taylor and Cary, 1960). When the degree of water saturation is very

low, water movement occurs primarily in the vapor phase. At intermediate water

contents, however, movement occurs by a combination of liquid and vapor flows

(Cary and Taylor, 1962a, 1962b).











Possible magnitudes of thermally-driven flows have been summarized by

Cary (1966) for various soils and commercial materials. Vapor flow in soils in

response to thermal gradients of 0.01 C were observed to vary from about 0.4 to

2.0-mm day', the same flow rates that would require pressure-head gradients of

0.05 to 2.50 m of water. Since vapor flow out of wet soils (evaporation) may range

from near zero to perhaps 12 mm day' depending on climatic conditions, one can

see that temperature-induced vapor flow should not be ignored.


Philip-de Vries Theory (PDT)


Approximately 40 years ago, in an attempt to describe the phenomenon of

simultaneous flow where water is moved by a thermal gradient, Philip and de Vries

(1957) and de Vnes (1958) proposed theory and equations to describe coupled heat

and water transfer in porous media, in terms of classical mechanisms of vapor

diffusion and liquid movement by capillarity accounting for heat flow and its

resultant effects. Philip and de Vries (1957) were apparently the first to include the

interaction of vapor, liquid, and solid phases and the difference between average

temperature gradients in the air-filled pores and in the soil as a whole. Taking these

factors into account, an approximate analysis was developed which would predict

orders of magnitude and general behavior that was in satisfactory agreement with

the experimental results of Gurr et al. (1952), Jones and Kohnke (1952), Rollins et al.

(1954), Smith (1943), and Staple and Lehane (1954).

Philip and de Vries (1957) separated the isothermal and thermal components

of vapor transfer to show the effect of relative humidity. They separated the vapor









21

flux into two components: 1) one due to temperature gradients; and 2) another due

to moisture gradients. In the saturation range where liquid transfer occurs, the

liquid flux for water was separated into components caused by temperature

gradients, water gradients and gravity. Philip and de Vries (1957) suggested that

interaction between vapor and liquid phases occurs as a series of evaporation and

condensation steps coupled with liquid flow through so-called 'liquid islands'. When

a -ei .,ei.arur aiir ir .- i pi.-1 across a soil, water evaporates at the warm end of

the pores and diffuses as vapor, due to a vapor density gradient, towards the cooler

end where condensation occurs. This process changes the curvature of the liquid

menisci of a liquid island, which continues until capillary liquid flows through the

island, p:..,.:-d b. rhe ,-i.:, in.r difference of both curvatures, equals the rate of

condensation and evaporation. The overall effect is a decrease in the tortuous path

length and an increase in the cross-sectional area available for vapor diffusion.

Philip and de Vries (1957) presented the simultaneous equations of

continuity describing water and heat transfer under the combined influence of

water and temperature gradients as


S=V(DOVT) +v(DV6e) Kk aT = V(XVT) -Lv(Dve) [2.8]
t T -at

where Dr" is thermal water diffusivity, Do" is isothermal water diffusivity, k is the

vertical unit vector, Do' is isothermal vapor diffusivity, and L is the latent heat of

vaporization. Note that the existence of soil-water hysteresis may, however,

invalidate the use of Eq. [2.8] (de Vries, 1958).











Application of Philip-de Vries Theory


Various applications and verifications of PDT have been documented. Barnes

and Allison (1984) applied PDT for non-isothermal conditions to develop a model for

distribution of deuterium and 80 in dry soils. The model was later verified by the

same investigators (Barnes et al., 1989), with the experimental data showing

satisfactory results. However, they emphasized that the assumption of constant

tortuosity for both liquid and vapor movement is questionable. They reported that, m

moderately dry soils the liquid tortuosity factor is probably reduced, and vapor

diffusion may be enhanced by the presence of 'liquid islands'. Scanlon (1992) used a

model developed with PDT to evaluate liquid and vapor water flow in desert soils of

Texas using 3"C1 and 3H tracers. Observed and predicted potentials with the model

showed good agreement. Scanlon (1992) concluded from simulation results that

thermal vapor fluxes were downward-directed during the day and upward at night

during both winter and summer. Net vapor flux was also downward during the day

and upward at night.

Bach (1992) conducted isothermal and non-isothermal laboratory

experiments to investigate the significance of non-isothermal water flow and to

examine theoretical and numerical descriptions of the transport processes. He

concluded from sensitivity analysis that, with adjustment in Ch (temperature

coefficient of the matric potential), PDT provided an adequate description of the

non-isothermal transport processes.











Many other applications of PDT to non-isothermal water movement have

been observed to describe and simulate transport processes (Benjamin, 1989; Chung

and Horton, 1987; de Silans et al., 1989; Mahrer et al., 1984; Milly, 1984; Nassar and

Horton, 1992; Noborio, 1995; Sophocleous, 1979; and Yakirevich et al., 1997). Cassel

et al. (1969) concluded that PDT showed acceptable agreement with observed soil

water movement.

In order to account for soil inhomogeneity and hysteresis of the moisture

retention process, Milly and Eagleson (1980) modified the PDT by using soil matric

pressure-head gradients instead of soil moisture gradients. These authors also

included effects of the heat of wetting on transport processes. A finite-element

one-dimensional formulation of the governing equations showed that the model

performed well in simulating highly coupled, hysteresis-affected, or very nonlinear

problems. Later, Milly (1984) used this modified form to study soil-water evaporation

from silt loam and sandy soils and reported that soil water evaporation was

generally more sensitive to isothermal than to thermal vapor diffusion.

Published evidence thus shows that the mechanistic PDT appears to be

accurate for coupled heat and water transfer in soils. This mechanistic approach to

formulate expressions for the fluxes of water and heat is based upon concepts

derived from fluid mechanics, mass diffusion and heat conduction.












Solute Transport during Coupled Heat and Water Flow



Simultaneous flow of solutes and water occur in most soil-water systems

under field conditions. Salt concentration gradients may also be a major contributor

to water movement in soils located in ard climates. Salt concentration gradients

often occur within the uppermost soil profile, where evaporation selectively removes

water at the soil surface and result in an accumulation of salt. Kemper and Rollins

(1966) reported that, for experiments involving Wyoming Bentonite clay, soil

moisture suction heads needed to be greater than 50 m of water before the osmotic

potential gradient would c ,.- ,1iarfi,.: in r:. ii water movement. Low (1955)

observed that, when the solute was completely restricted by the porous medium,

osmotic potential differences were as effective as pressure potential differences in

causing water to move through clay under isothermal conditions. In studying water

flow through a porous ceramic filter, Jackson (1967) reported that soil water flow

could deviate from Darcy-type flow due to the presence of salt. In laboratory studies

on clay loam and sandy loam soils, Qayyum and Kemper (1962) reported that

diffusive movement of water occurred to the salt-bearing surface at soil moisture

contents less than one-half of field capacity, and that viscous flow occurred at soil

moisture contents higher than one-half of field capacity in closed soil columns.

Kemper and Maasland (1964) found that salt sieving increased with

decreasing pore size of the soil and decreasing valence of the saturating cation.

Letey and Kemper (1969) introduced a theory to describe simultaneous movement of









25

water and solute in soil. The theory included an equation to describe water flow and

another equation to describe solute transfer. The equation for solute transfer implied

that the movement of solute is controlled by Fickian diffusion and salt sieving.

In summary, temperature and osmotic gradients can induce water flow

within soils. As water moves, heat and solute are transported, thus altering the

driving forces (gradients). Interactions among thermal, soil moisture, and osmotic

effects exist, and they differ for the liquid and vapor phases (Nassar and Horton,

1989b).

Nassar and Horton (1989b) extended the PDT, describing water transport

under the combined effects of matric and osmotic pressure head gradients in

unsaturated, non-isothermal, saline soil. Their theory was based upon the concept

of viscous flow of liquid water under the influence of capillary, adsorptive, and

osmotic forces and on a concept of vapor movement by diffusion. They presented

two equations for vapor and liquid transfer as

q,
S-D= VT DevL + DcvVc [2.9]
PL
and
S=-D VT DVeL + DcVC Kk [2.10]


where q, is water vapor flux, D, is thermal vapor diffusivity, DTL is thermal liquid

li rj I -, r.. Dev is isothermal vapor diffusivity, Def is isothermal liquid diffusivity, Dcv

is vapor diffusivity due to solute concentration, DcL is liquid diffusivity due to solute

concentration, and c is solute concentration.









26

Equations [2.9] and [2.10] were summed in order to describe the total water

flux as

q = -DVT DeV + DVC Kk [2.11]
Pr
where q, = q, + q, is total water flux, DT" = D, + D, is thermal moisture

i ffu,; ir r, -,w" = Dev + DOe is isothermal moisture diffusivity, D. = Dcy + DcL is

moisture diffusivity due to solute concentration, and 6 = O, + O6 is total volumetric

water content, where vis water vapor content (expressed as equivalent liquid

volume).

Applying the principle of heat conservation, an equation for transient heat

transfer combining the effects of matric pressure head, osmotic pressure head, and

temperature was formulated. For developing the solute-transport equation (non-

interacting solute transfer), Nassar and Horton (1989b) included the mechanisms of

molecular diffusion, hydrodynamic dispersion, thermal diffusion, mass flow of water

(convection), and salt sieving.

Nassar and Horton (1989a, 1989b) conducted experimental and numerical

studies of coupled heat and mass transport in closed soil columns with three

different soil textures. Experimental observations for unsaturated conditions showed

water accumulation within the low-temperature region of the columns, but solute

accumulation within the high-temperature region. Predicted and observed results

were in good agreement when initial soil-water contents were relatively large and

initial solute concentrations were low. However, for small initial soil-water contents,

the model underestimated soil-water contents close to the low-temperature

boundaries and overestimated than near the high-temperature boundaries. For high









27

initial solute concentration, the theory underestimated liquid transfer toward the hot

boundary temperatures. Unfortunately, the theory proposed by Nassar and Horton

(1989b, 1992) is restricted to homogenous soils and cannot be used in its presented

form for heterogenous soils. The water flow equation is written in terms of water

content and thus does not allow simulation for multilayered soil. Furthermore, the

model is primarily restricted to soil conditions of water unsaturation.



Temperature Effects on Solute Transport



The rates of most chemical reactions increase with temperature, changing

their equilibrium constants (Clark, 1996). Clark further states as a rule of thumb,

that the reaction rate doubles for each increase in temperature of 10 C. Generally,

the understanding of a chemical reaction is not considered complete until the

sensitivity of the reaction to temperature is understood. Many chemical studies thus

include experiments at different temperatures.

Daniels and Alberty (1961) and Clark (1996) discussed the influence of

temperature on reaction rate in the following way. Kinetic data over a range of

temperatures can usually be represented by use of an empirical equation first

proposed by Arrhenius
E
kr= F e RT [2.12]

where k, is reaction rate, F, is a frequency factor, E, is activation energy, and R, is

the universal gas constant. Set in logarithmic form,










E
logk, + log Ff [2.13]
2.303RUT

According to Eq. [2.13], a straight line should be obtained when logarithm of the

rate coefficient (k,) is plotted against the reciprocal of the absolute temperature (T).

The E, can then be estimated as -2.303 R, x the slope of the line. The activation

energy may be envisioned as an energy barrier that must be surmounted for the

reaction to proceed. An increase in E, indicates a slowing down of the reaction (and

vice-versa). The E, thus can be determined by measuring the reaction rate at

different temperatures.

The influence of temperature on chemical processes has also been described

by Stumm and Morgan (1981) through use of the Arrhenius equation. Simfnek and

Suarez (1993) described a function with some modification to account for relative

change in rate coefficients with temperature as



m r re J [2.14]
f(T) = e '[2.14]

where T, is the reference absolute temperature and f (T) = 1 at T = T,.

Cohen and Ryan (1989) and Cohen et al. (1988) used the relationship given by

Bird et al. (1960) to describe the temperature dependence of diffusion coefficients.

The chemical molecular diffusion in air as a function of temperature was estimated

as

Da (T) = DT) 1.823 [2.15]


where D,(T,) is the chemical molecular diffusion coefficient in air at the reference

temperature. Temperature variation of the chemical molecular diffusivity in liquid











water was estimated as


D ( T) = D T T) [2.16]


where p, is the viscosity of liquid water.



Summary



The PDT was proposed !:1, aj- ij:o to mathematically describe coupled

water and heat flows in soil. It has been well-verified and applied by a number of

researchers to date. The theory was extended by Milly and Eagleson (1980) to make

it applicable to heterogenous media, by replacing the water content-based

formulation with a matric potential-based formulation of flow equations. This

advancement was a big step, as it made the theory suitable for the solution of field

problems. Nassar and Horton (1989a, 1992) extended the theory to include salt

effects on water flow and provided insight into additional coupling. But, this theory

cannot be used to solve field problems, because Nassar and Horton used the

original PDT (water content-based) formulation for modification.

Relatively little evidence exists that the potential of PDT and its modification

by Milly and Eagleson (1980) has been fully exploited to use the improved estimates

of water content/flow, and temperature information. Several researchers (Barnes

and Allison, 1984; Barnes et al., 1989; Nassar and Horton, 1989a, 1989b; Noborio,

1995; Scanlon, 1992; and Yakirevich et al., 1997), however, utilized PDT by making

use of improved estimates of water content/flow when exploring solute transport.









30

These investigations, however, did not attempt to use the improved estimates of

temperature by inclusion of its effect on solute transport.

A need, therefore, exists to exploit the full potential of PDT in solving field

problems where temperature could be highly transitional both in space and time.

Furthermore, where chemical (especially volatile-compound) transport is also of

concern, a need exists to use both the improved estimates of water content/flow and

of temperature provided by PDT. The work reported here utilizes PDT in totality to

develop a 2-D numerical model for solving problems involving the fate and transport

of volatile chemicals that may exist in all three phases (viz. liquid, solid, and gas) of

soils. Details for the developed model are presented in Chapter-3.















CHAPTER 3
MATHEMATICAL MODEL



The model developed here considers the soil to be a heterogenous isotropic

medium composed of solid, liquid, and gaseous phases. The latter is assumed to be

a non-interacting binary mixture of water vapor, dry air, and volatile chemicals.

Other general assumptions in the model include

1. The solid matrix is undeformable and immobile.

2. The: i-,jii ... ._er i r ILr: ':. i lP, LIri

3. The density and viscosity of liquid water are not influenced by dissolved

constituents.

4. Hysteresis in the relationship between capillary-pressure and hydraulic

conductivity is neglected.

5. Liquid water and water vapor are in local thermodynamic equilibrium.

6. Water vapor and other gases behave as perfect gases and advective transport

of these components is neglected.

7. Local thermal equilibrium exists among all phases.

8. Radiative heat transfer is negligible.

9. Thermal and matric potential gradients interact to provide heat and water

flow in the porous medium.










More specific assumptions are stated as required. Additional details

concerning the equations governing flow and transport are presented in the

appendices.



Liquid and Vapor Water Flow in Soil



The original water content-based formulation of PDT was modified by Milly

and Eagleson (1980) to provide


[1 P + L eaaPj ah [ +D,,)Vh +D,VT+Kk

[3.1]
pj + e, aPI aT", a Tv
1- +--
p, 9T p, 9T 9t

where p, is vapor water density, 0O is the volumetric air content, and D, is

isothermal vapor diffusivity. Equation [3.1] describes both liquid water and vapor

water flow in response to gradients of matric potential and thermal energy.



Heat Flow in Soil



Coupled heat flow in the soil in response to gradients of temperature and

pressure-head has been described by (de Vries, 1958; Milly and Eagleson, 1980)


C+H +p H aT = VT+p LD h C(T-T q [3.2
S(T 2 T ) t L hO J [3.2










and

C =Cd+CLpL (I+Cvp HI = [Lo+C,(T- To

H2 (CLpL \Cp(T T) p Lo [3.3]

where Cd is volumetric dry heat capacity of the total medium; CL and C, are specific

heat of liquid water and vapor water; respectively; and the subscript implies the

magnitude at a designated reference level.



Transport of Multiple Species of Solutes in Soil



Following the approach of Similnek et al. (1992) and Simfnek and van

Genuchten (1994), in which solutes can exist in all three phases (liquid, solid, and

gaseous) with non-equilibrium transport, -i r.:.- -ir. rg transport equations can be

stated as

a + +-c= + aDELc c 6aDlgg c
a a t -at vLDvc+v Dl g-V(qLc1l) [
[3.4]



+E)LC= + =.) +) D VC DVg -VL i -pL, ri + C -

s, + Ps, ,i p+s 1 ie g ,i-l i1 s, i-lpsi-1 [3.5]

Plg, i.i-1i+YL, iL+s, iP +g, a; i=2,..., n (species)

where s represents adsorbed concentrations; g is the gas concentration; PL, P., and

/, are the first-order rate constants for solutes in the liquid, solid, and gas phases,

respectively; YL, y,, and 7, are zero-order rate constants for the liquid, solid, and gas









34

phases, respectively; p is the soil bulk density; D1 is the dispersion coefficient for the

liquid phase; and D9 is the diffusion coefficient for the gas phase. The subscnpts L,,,

and correspond with the liquid, solid, and gas phases, respectively; the subscript,

represents the i" chain number; n, is the number of solutes involved in a chain

reaction; and p,*, p,*, and p,* are first-order rate constants for solutes in the liquid,

solid, and gas phases, respectively, providing the connection between individual

chain species. The nine zero- and first-order rate constants in Eqs. [3.4] and [3.5] can

be used to represent a variety of reactions or transformations including

biodegradation, volatilization, precipitation, etc., and the effect of temperature on

these coefficients is included through use of Eq. [2.14]. E-u r i.:ri [3.4] alone is

sufficient for a single solute that does not undergo transformation to related

chemical species in a 'chain' reaction. ?..rh E 1- [3.4] and [3.5] are required when the

primary solute undergoes transformations to other species or when multiple

unrelated solutes are present in the soil system.

In order to solve the solute-transport equation, a knowledge of liquid water

content, air content, and volumetric flux is required. Solution of the coupled water

and heat flow equations (Eqs. [3.1] and [3.2]) provides this information.



Enerqv Balance at the Soil/Atmosphere Interface



Interception of solar radiation, interaction with the overlying atmosphere,

and loss of water through evaporation is described via an energy-balance equation

for the soil surface









35

(1-PR +R RLA = H+LE+G [3.61

where p, is soil surface lib-.J i, measured global radiation (taken as positive

downward), R, is longwave sky irradiance (positive downward), Rs is longwave

radiation from the soil surface (positive upward), H, is sensible air heat flux (positive

upward), LE is latent heat of the evaporative flux (positive upward), and G is soil

heat flux (positive downward).

The solution of Eq. [3.6] provides information for soil surface temperature and

atmospheric evaporative demand (excluding transpiration) required to specify the

top boundary condition of the solution domain. This equation is strictly valid only for

a soil surface without a mulch cover.

The presence of a mulch cover alters interaction of the soil surface with the

overlying atmosphere (as discussed earlier). An approach based on the work of

Hares and Novak (1992) and Ham and Kluitenberg (1994) was adapted for

conditions where the soil surface is covered with a plastic mulch such as commonly

used in soil-bed systems for vegetable production in Florida. The energy balance for

a plastic mulch can be defined as

Rnm + Hs Cc = 0 [3.7]

where Rn, is net radiation of the mulch surface and C, represents conduction or free

convection between the mulch and free soil surface. Energy balance for the soil

surface beneath the thin plastic mulch can be defined as

Rn + C+ G = 0 [3.8]

where Rn, is net radiation of the soil surface. These energy-balance equations








36
neglect heat storage within the thin plastic mulch, and the potential impacts of cuts

and planting holes m the mulch is not considered. Thickness for such plastic

mulches typically ranges from 0.02544 to 0.15264 mm.

Net radiation of the plastic mulch is given by (Ham and Kluitenberg, 1994)

Rn. = aRg 1+ P' P P r + 1 ) + Pr msTs

1-e) -2eT, [3.9]
where T,, T,, and T, are the temperatures of the sky (usually taken to be

atmospheric temperature at 1 m above the ground surface), soil surface, and mulch,

respectively; e,,y, e,, and em are the emissivities of the sky, soil surface, and mulch,

respectively; a. is the shortwave absorbance of the mulch; T,, is the transmittance

of the mulch in the longwave spectrum; and a is the Stefan-Boltzmann constant.

The variables p* and p,* represent the internal reflection functions for shortwave

and longwave radiation, respectively. These functions account for multiple

reflections of radiation between the mulch and soil surface, and are defined by Van

de Hulst (1980) as

P' = 1 P ) = 1 + P P p [ 3.10]


S 1- 1- = p, 1- + 2 [3.11]

where p, is the shortwave reflectance of the lower mulch surface and p,, is the

reflectance of the mulch in the infrared spectrum. The infinite series expansion

shows here that the reflection functions account for the initial radiation in the layer

and then amplify the effects via repeated reflections between the soil surface and

the underside of the mulch.











Sensible heat transfer between the plastic mulch and the atmosphere is a

function of the aerodynamic properties of the surface and the temperature gradient

between the plastic and the boundary layer, as defined by


C ( T T )
Hs = C(T- [3.12]
a, h

where Ta is air temperature and r,,h is the aerodynamic resistance to heat transfer.

Heat transfer between the mulch and the underlying soil surface is estimated using

the equation
T T
C [3.13]


where r, represents a thermal contact resistance at the mulch-soil interface and

mainly depends on the air-gap thickness between the soil surface and the mulch.

This determines whether the heat transfer is conductive or convective. Ham and

Kluitenberg (1994) described the estimation procedure and presented representative

values of r. for different air gaps between the soil surface and the mulch for various

plastic mulches.

Net radiation of the soil is dependent upon the global and longwave sky

irradiance transmitted through the mulch, radiation emitted from the mulch and

soil, and multiple scattering of radiation between the soil and mulch as given by

(Ham and Kluitenberg, 1994)

n,= (-I P)TP'Rs pre (Im4 a,,ersoT kPY,,i T ,)-eT [314]


Given the subsurface temperature near the surface at depth z, ground heat flux can

be approximated at the future time step p + 1 (p being the current time) as (Horton










and Chung, 1991)

G=- T' f Ts-- T1 + C s Ts ]z [3.15]
Az s 2At

where T, is soil temperature at the first node below the soil surface and (Tf'i

T,)IAt is the rate of change in temperature with time at the surface between time

steps p and p + 1.






Equations [3.1]-[3.5] require supplementation with appropriate initial (IC) and

boundary (BC) conditions to describe a specific simulation scenario. The IC in

general (at t = 0) are specified as h (x, z, 0) = h, (x, z), T (x, z, 0) = T,(x, z), and

c (x, z, 0) = c, (x, z) respectively for Eqs. [3.1], [3.2], and [3.4], [3.5]. Here h,, T,, and c,

are known values of the state variables at the beginning of the simulation; and x

and z represent the horizontal and vertical locations, respectively.

Boundary conditions may be defined as a specified value of matric

potential/or temperature and/or solute concentration (Dirichlet-type BC), a specified

water/or temperature and/or solute flux (Neumann-type BC), or a combination of the

two (Cauchy-type BC). The top boundary (i.e., the soil surface) in the system under

consideration is a special case of the Neumann-type BC for water (evaporative flux),

dependent upon the overlying atmosphere. In general, the boundary conditions

depend on the specific system being investigated.











Input Parameter Requirements



Soil input data include sand, silt, clay, and organic matter contents; bulk

density; saturated hydraulic conductivity; and soil hydraulic parameters which

describe the water retention curve (Eqs. [3.7]-[3.8]). For the soil/atmospheric

boundary conditions, weather data such as average air temperature, air temperature

daily amplitude, average dew-point temperature, dew-point temperature daily

amplitude, wind speed, surface roughness length, and day-length are required

inputs as well.

Soil hydraulic properties are estimated following van Genuchten (1980) as

e5- r 1
e = e + [ -(: ; FM = 1 -
0+ [ +(.:. I 1

K(L) Kat [3.16]


and,


h 1 -) e -e [317 1

where 6, is residual water content, 0, is saturated volumetric water content, |h is

the absolute value of the matric potential, and K_, is the saturated hydraulic

conductivity. The parameters a, rL, and rr) are empirical constants used to fit the

moisture retention curve. The temperature dependence of matric potential (h) was

described as (Philip and de Vries, 1957)

dh h do,
dT = hdT Chh [3.18]
and hydraulic conductivity (Kh)) a(Constantz 1982)
and hydraulic conductivity (K(h)) as (Constantz, 1982)









40
ref
K(h) Kef(h) [3.19]
w PL, ref

where o, is surface tension of water; Ch ( = -0.0017 K-' by Milly and Eagleson, 1980)

is temperature coefficient of matric potential; p,"' and p T are dynamic viscosities

at the reference temperature and soil temperature, respectively; pL,r and PL, are the

densities of water at reference temperature and soil temperature, respectively; and

K,, (h) is hydraulic conductivity at the reference temperature.

The dependence of 6L upon temperature can be determined as (Milly and

Eagleson, 1980)
ae, 8eL
S h exp[-C,.(T- To [3.20]


Water vapor density is given by Marshall and Holmes (1979) as


Pv = gaPLI u T [3.21]

where p is water vapor pressure and M, is the molecular weight of water. Water

vapor pressure is related to soil water matric potential by the relation (Marshall and

Holmes, 1979)

gahM
S= 0, exp [hM] [3.22]

where cp, is the saturated vapor pressure, given by Noggle (1985) as


S= 237.7 exp '~ -- [3.23]

Differentiating Eq. [3.21] with respect to h yields

ap M mJ [T .324
ah 9 ga- PaL uT exp gahRMT [3.24]
R wT










and with respect to T gives

ap, M
aT D1 *D2- g PL R T [3.25]
where

Dl D2= D3+qsah M D3 ; D3=exp^ gAh
RUT T RT 2 RUT

Variation in saturated vapor pressure with temperature is given by

a0, 237.7ML I ML
aT 2 ex+p [3.26]
TUT RU T Tr

The vapor diffusion coefficients are determined as (Milly and Eagleson, 1980)


Dh ,DTe D T -- [3.27]
with

2 f = for 8L<60 f = 6(+ T- ( for ,L>e, [3.28]


where D, is a molecular diffusion coefficient of water vapor in air, T' is an air-phase

tortuosity factor, fis vapor path correction for liquid islands, 0T is total porosity, 8, is

the water content at which liquid continuity breaks during desaturation, and

S( = 1.7) is a temperature-gradient enhancement factor for vapor.

Specific heat is estimated from the methods of deVries (1963) as

C = X c +X c +Xcc+Xo com +CL + Cve [3.29]

where X,, X,, X, and X.m are the volume fractions of sand, silt, clay, and organic

matter, respectively; and c,, c, c0.,, C, are the volumetric heat capacities of sand,

silt, clay, organic matter, and water vapor, respectively. The specific heat of water

vapor is given by Noggle (1985) as









42
30.54+0.01029T
C 0.0180015 [3.30]

The apparent thermal conductivity can be estimated as a harmonic mean of

the soil constituents (de Vries, 1963)


El Xi. -
1.25 i=st,c,o, k 1+ -I-1
E k*Xi 3 -j= .,4 X gj
1=s,st,c,o,8 J
[3.31]
where X, represents the respective thermal conductivities of the soil constituents, g,

are shape factors for the soil constituents, and X, is the thermal conductivity of air.

The incoming global radiation for the soil/atmospheric boundary condition

must be partitioned into a flux-density distribution for daylight hours. Chung and

Horton (1987) suggested a sine wave distribution

R R sin t-SN+ DL3.32]
g 2 DL 2 DL

where Rd. is daily global radiation, DL is day length, SN is solar noon, and t is the

time of day. The air and dew-point temperature distributions can also be expressed

as sine functions

S ( 2nt __ 2nt
T = T,+A sin 86400 +n = Td+d Asin 4 +0 j [3.331


where the bar indicates average temperature; T. and Td are air and dew point

temperatures, respectively; A, and Ad are air and dew point temperature amphtudes,

respectively; tm is the time from midnight; and the maximum temperature is

assumed to occur at solar noon.









43

The soil surface emissivity, e,, and albedo, p,, are described by the method of

van Bavel and Hillel (1975, 1976) as

es = 0.9+0.186L p, = 0.25 for eL-0.10,
[3.34]
P, = 0.35-eL for O.lO-.e<0.25, pS = 0.10 for eLo0.25















CHAPTER 4
MODEL VERIFICATION AND VALIDATION



Numerical solutions to differential equations require testing to ensure that

numerical instabilities have not been introduced into the solution due to round-off

or truncation errors and that the solution converges on a reasonable estimate of the

exact solution. A numerical scheme is stable if, with increasing time, the difference

between the numerical solution and the exact solution tends uniformly to zero at all

nodes (Benjamin, 1989). Convergence is satisfied if the numerical solution

approaches the exact solution as spatial- and temporal-step sizes approach zero.

Proper testing also ensures the correctness .f ri -ri, i-, ] c3L] r-~pe-e-r1r-1 ,: tr,-.: the

physical system being modeled and the validity of any assumptions made in the

model.

Testing of a model may be achieved in multiple ways. One such way is by

comparing the model results with analytical (exact) solutions for greatly simplified

conditions, and another is by comparing the model results with observed data from

carefully designed experiments. Unfortunately, no analytical solution to the

governing equations for coupled heat and water transport is known to exist. Hence,

portions of the model were tested with observed data and analytical solutions for

simplified conditions, where available.











Isothermal Water Flow


Model results were compared with data for transient one-dimensional,

isothermal water flow through sand as reported by Haverkamp et al. (1977), with

vapor flow ignored. Haverkamp et al. (1977) reported K., 6,, 6,, and soil bulk

density to be 34.0 (cm h '), 0.287, 0.075, and 1.66 g cm-3, respectively. An analytical

model by van Genuchten (1980) was fitted to the soil water content versus matric

potential data in order to provide optimum values for the hydraulic parameters a

and r. A uniform initial water content of 0.10 was specified and water content at

the 0.70 m depth was kept constant at this value. A constant water influx of

13.69 cm h' was specified at the top boundary. Results of the numerical simulation,

and experimental data for different times, are shown in Fig. 4-1. The numerical

solution was stable and converged upon a close approximation of the experimental

values. The model accurately described the movement of the water front in the

column of sand.


,0 5 0.10 0.15 0.20 0.25 0.3
o observed.
-10 Estimated -1o
o 0
-20 0.2 hr -20

-30 -30 Figure 4-1. Observed and computed
S 0.4 r distributions of water content in a
-0 : -40 sand column during constant flux
S-50 -50 infiltration.


Water Content (cm3 cm3)











Uncoupled Heat Flow


Next, the numerical solution for heat flow was verified with field-measured

temperature data (Horton and Chung, 1991) for a Nicollet clay loam soil profile with

a bare surface located near Ames, IA. The effect of water flow was ignored.

Representative values of heat capacity and thermal conductivity for loam soil were

used. Initial conditions as specified by Horton and Chung (1991) were used in the

model and temperature at the 60-cm depth was held constant at 20 "C.

Temperatures observed for the soil surface were used in the model to provide the

surface boundary condition. One-dimensional numerical results (Fig. 4-2) for the

5- and 15-cm soil depths provided satisfactory representation of temporal

temperature profiles for each depth.


40, 20 40 60 80 100 12o
Observed 5-cm
36 Estimated 5-cm 36
o Observed 15-cm
U 32 ------ Estimated 15-cm 32

28 28

24 24
E : /<
20 o -20

16 16
0 20 40 60 80 100 120
Time (hours)

Figure 4-2. Observed and computed temperatures for two depths
in a field soil.











Coupled Water and Heat Flow


In order to test the capacity of the numerical solution to predict coupled heat

and water transfer, simulation using the experimental data of Nassar and Horton

(1989a) was performed. Nassar and Horton (1989a) conducted laboratory

experiments of heat and mass transfer with controlled temperature gradients m

sealed columns (14 cm in length) of silt loam soil with initially uniform water

content (0.117). The column ends were subjected to different temperatures (19.03 C

and 9.36 C) in order to create a temperature gradient. The experiment was run for a

period of 31 days. The observed mean value of saturated hydraulic conductivity for

the soil was 0.954 cm hr-', with a corresponding soil water retention curve having

been determined by Nassar and Horton (1989b).

The van Genuchten (1980) model was fitted to the soil water retention curve

in order to obtain optimum hydraulic parameters for the soil (a = 0.03137 cm',

Ir = 1.33308, 68 = 0.65, 6, = 0.05) for silt loam soil. A numerical simulation was

carried out for 31 days using zero flux boundaries at both ends for water flow and

Dirichlet boundaries for heat flow. Observed and predicted water contents are given

in Fig. 4-3. Good agreement was noted everywhere except in the vicinity of the cold

end of the soil column with elevated water content. The predicted total water

content in the column was slightly higher than the observed value. Since the

computed mass balance error was less than 1.0 % in the numerical simulation, the

possibility exists for a small systematic error in the measurement of either initial or

final water contents during the actual experiment.











0.3U
------- Estimated
0.25 Observed
S-Initial Figure 4-3. Observed
E0.20 and computed water
Content after 31 days
S0.15 under the influence
\ of both matric head
00 and thermal
) 0.10
S01 0o gradients.
S0.05

000 2 4 6 8 10 12 14
Distance (cm)


Isotherm al 1-D 1 -,Iur- Tr .r..... ri. unri _rea.1, ''J r .r l.:....


The model component for solute transport was tested against data reported

by van Genuchten (1981) for the movement of a pulse of boron 'H ,E,-', during steady

water flow through a column of Glendale clay loam. The solute-transport

parameters (imfmnek and van Genuchten, 1994) used were: DL = 2.042 cm2 hr1,

68 = 0.445, p = 1.222 g cm-3, kd = 1.14 cm3 g', P = 1.0, T1 = 0.0, f, = 0.47, and

o) = 1.33x10-2. The initial concentration was zero and a pulse of aqueous solution

containing a boron concentration of 20.0 mmol, L' was applied at a constant flux of

713.0 cm hr-' for a duration of 121.44 hr. Assuming two-site chemical non-

equilibrium sorption, a simulation was run for 20 pore volumes of effluent. A

comparison of numerical and experimental results (Fig. 4-4) verified that the model

described the transport process with satisfactory accuracy.















0.8

00.6

U
S0.4

0.2


Figure 4-4. Observed and computed break through curves for boron.


Isothermal 2-D Solute Transport during Steady Water Flow


The mathematical accuracy of 2-D numerical simulations for isothermal

steady water flow, was verified employing a simple analytical solution. Solute (non-

volatile) transport in a homogenous, isotropic porous medium during steady-state

unidirectional groundwater flow can be written as

ac a2c 2c ac
R = D D R c 4.1]
at TX2 LZ2 az Rc

where D, and DL are transverse and longitudinal dispersion coefficients,

respectively, and v, is average pore-water velocity in the downward vertical

direction. The initially solute-free medium is subjected to a solute source, C,, of unit

concentration. The source covers a length 2a along the inlet boundary at z = 0, and

is located symmetrically about the coordinate x = 0. The transport region of interest

is the half-plane (z 0 0; -- < x s o-). The boundary conditions can be written as










c(x,0,t) = Co -asx a; c(x,0O,t) = 0 other values of x

i, mc ac [4.2]
lim- = 0 ; 1. 0
z-_ az .. .. .

An analytical solution to the above transport problem is given by Leij and Bradford

(1994) as

Cz ( O 2 z2
c(x,z,t) Cz -- exp iR+ z
4(xDL), 2DLJ [ 4D, 4DL0,rW
[4.3]
a -x er a+x jid
2(DT } )l [ 2(Dr.w)a)"



0 _3 ___ ___ _
SOURCE (a =50-m)
-10 Analytical
Numerical
-20 -50 Days

-30 1 ___ _0 ays
2 Days
-40 lConcentration contour level= 0.1
Sv= 0.1 m day'

D = 1.0 m2 day'
R= 1.0
-60- C= 1.0
0 20 40 60 80
Distance from mid-source (m)
Figure 4-5. Location of solute concentration fronts for three times as estimated
numerically and with the analytical solution.


The numerical simulation assumed a source-width of 2a = 100 m. Because of

symmetry, computations were carried out only for the half plane where x 2 0.

Figure 4-5 shows the estimated solute concentration fronts (0.1) at selected times









51

and various values of the parameters as used in the computations. F i- Il p rre.1'- r-.

by the numerical model were similar to those given by the analytical solution.


Isothermal 1-D Solute Transport with Nitrification Chain during Steady Water Flow


To verify the numerical solution capacity in simulating multiple linked

species in a chain, comparison was done with results generated from an analytical

solution published by van Genuchten (1985) for 1-D transport of solutes involved in

sequential first-order decay reactions. The analytical solution holds for a

homogenous, isotropic porous medium with steady-state uni-directional

groundwater flow. The solute-transport Eqs. [3.4] and [3.5] for this case reduces to
ac, = a2c ac
R- D vx PL,1 R cl [4.4]
R1 at ax2I44]
and
ac. 2 82c ac
R = a2 Vx aX PL,-1Ri-i- -1c PL,iRici ; i = 2,3 [4.5]

A simulation was performed here that involves a three-species nitrification chain

NH' (i=1) NO2(i=2) NO, (i=3) [4.6]

and follows van Genuchten (1985) and Cho (1971).

A NH,4 solution was applied to an initially solute-free medium. The input

transport parameters used for the simulation were

v = 1.00 m h' R, = 2.0 [-]

DL= 0.18 m2 h' R = 1.0 [-]

S= 0.01 h- R, = 1.0 [-1

2 = 0.10 h- c, = 0.0 [-]

P3 = 0.0 h-' o, = 1.0 [-]











The simulation was conducted for 200 hours. Figure 4-6 shows the

concentration profiles for all three solutes at time 200 hr, calculated both

numerically and with the analytical solution from the CHAIN code of van

Genuchten (1985). The numerical solution duplicated the analytical results almost

exactly.


1.0 t------------------
-- Numerical NH*
o Analytical NW Figure 4-6.
S0.8 -- Numerical NO;
Analytical NO; Analytically
Numerical NO; determined and
O Analytical NO; numerically
U computed
0.4 concentration
profiles at 200
0.2 hours.

0.0
0 40 80 120 160 200
Distance (m)


Isothermal 1-D Solute Transport involving Gas Flow


The capacity of the numerical solution in simulating gas transport was

tested against an analytical solution given by Crank (1975). Considering uniform

constant water content in a homogenous 1-D soil column, for a chemical species

existing in the gaseous phase undergoing hydrolysis (represented as first-order loss,

pg), the transport equation (Eq. [3.4]) reduces to


ag D -pg 47
Ot ax2









53

For an application of finite source (g,) at one end with zero flux BC and zero

gradient BC at the other end, an analytical solution is (Crank, 1975)


gxt) = e go erf- +erf 2/--- [4.8]
2 n=- 2Dt 2Dt

where a is the initial source length. Prior to deriving the inril,'t~.: ]i ii ir:in the

transformation g = g* exp(-pg t) was used to simplify Eq. [4.7] to

=t D 2 [4.9]

A numerical simulation was conducted for 5 hours with g, = 50 moles,

a = 1 cm, D9 = 47.52 cm2 hr-1, and P = 0.055 hr The numerical solution

duplicated the results of analytical solution at 1, 3, and 5 hours (Fig. 4-7).


Figure 4-7. Analytically determined and numerically computed concentration
profiles at 1, 3, and 5 hours.











Summary



All major components of the numerical model were tested for mathematical

accuracy, the adapted numerical scheme, and its ability to represent the physical

systems. Both analytical solutions for simplified cases and published data were

used to check against simulation results from the numerical model. Excellent

agreement was observed in nearly all the cases undertaken for testing of heat-flow,

water-flow, and solute-transport model components. The coupled water-heat flow

component was also successfully tested. The model could not be tested in its

entirety as a single unit, because no analytical solution exists at present for coupled

water-heat and solute transport. Also, no published data set was found to test all

the features of the model together as a single unit. However, successful testing of

individual model components provides confidence in the application of the model

simulations conducted in this work.















CHAPTER 5
MODEL APPLICATION



a n1-~rij Ti j-,i: L :.r oif T.I,-ih i r.n.,ij- FunuQ ir.,,r ,rl :,,, i, i



Methyl bromide (MBr) is an odorless, colorless toxic gas used in agriculture

throughout the world. The United States uses it on more than 100 crops as a soil

fumigant, a post-harvest treatment, and as a plant quarantine treatment to control a

variety of pests. MBr is also used to fumigate structures such as grain storage

warehouses, flour mills, and ships carrying agricultural commodities. California and

Florida are the primary users of MBr, followed by Hawaii.

MBr, a broad-spectrum pesticide, is urli z-.1 rI-' "rd, as a very effective

agricultural fumigant during vegetable and fruit production in Florida, California,

and Hawaii. MBr is very effective for eradicating insects, nematodes, fungi, bacteria,

and weeds. In the United States, strawberries (16 % of U.S. total) and tomatoes

(24 % of U.S. total) are the crops which use the most MBr, consuming about 6,500

tons annually. Other crops which use this pesticide as a soil fumigant include

tobacco, peppers, grapes, and nut crops (USEPA, 1997).

In the United States, the U.S. Clean Air Act requires that the production and

importation of any substance defined as an ozone depleter by the Montreal Protocol

of 1991 be phased out (USDA ARS, 1997). Therefore, despite its economic value, the

55









56
U.S. Environmental Protection Agency (USEPA) plans to phase out the use of MBr

by January 1, 2001. Since the Clean Air Act allows no exceptions, all uses of MBr,

including plant quarantine, will be banned. In addition to being a toxic gas, MBr is a

-..3r i i.:a'nt .:-z:, r,-depleting substance, with recent scientific evidence estimating

that bromine from this material is at least 50 times more effective at destroying

ozone than chlorine from carboflurocarbons on a per-molecule basis (USEPA, 1997).

MBr has an average residence time of 1.5-2.0 years in the atmosphere (Hacherl,

1994). However, there are also other natural sources, primarily forest fires and

marine organisms besides anthropogenic sources, which contribute significant

quantities of MBr to the atmosphere.


Table 5-1. Earlier estimates for source-contributions of MBr to the atmosphere
(Grojesan, 1991; Khalil et al., 1993; Lobert et al., 1990; Mano and
Andreae, 1994; Yagi et al., 1993; and Zurer, 1993).


MBr Source Quantity in atmosphere
(m ton y )
Oceanic 30,000 (30 %)
Bio-mass burmng 30,000 (30 %)
Industrial 15,000 (15 %)
Agricultural 25,000 (25 %)
Total 100,000


Quantification of MBr release to the atmosphere is extremely difficult, since

limited data exist for the total balance of MBr in the earth's atmosphere. Based upon

information from several sources, atmospheric release can be broadly attributed to

4 major sources i T a~.e I- 1 with agricultural use of MBr being attributed to 25 % of











the atmospheric MBr. Recently, Yagi et al. (1993) estimated that agricultural use

contributes 80 % of anthropogenic MBr in the atmosphere. MBr is used industrially

as a methylating agent and a degreaser.

More recent investigations emphasize new facets (Table 5-2) of the

uncertainty of various source contributions and sinks for atmospheric MBr. New

modeling studies based upon recent monitoring data suggest that the ocean

absorbs more MBr from the atmosphere than previously thought. Rapid removal by

the ocean shortens the average estimated life-time of atmospheric MBr to just

0.7 year, much less than an earlier estimate of 2 years (Friebele, 1997).


Table 5-2. A recently revised budget for atmospheric MBr (Yvon-Lewis and
Butler, 1997).

Emissions Uptake
MBr Sources (Gg yi) MBr Sinks (Gg y-)
(Gg y') (G y')
Oceans 56 Oceans 77
Fumigation (soils, durables, 46 Hydrolysis 86
perishables and structures) and uv
Gasoline, leaded 15 Soils 43
Biomass Burning 20 Plants ?
Total 137 Total 206


These new results with regard to the role of oceans in the level of

atmospheric MBr (Friebele, 1997) raise a question about the contribution that

anthropogenic MBr plays in destroying the ozone layer. A newly revised MBr budget

(Table 5-2) presented by Yvon-Lewis and Butler (1997) estimates that sinks

(206 Gg y ') greatly exceed sources (137 Gg y'1). These estimates imply that MBr











should not accumulate in the atmosphere. Obviously, high uncertainty prevails

regarding sources and sinks for the accumulation of atmospheric MBr. The buffering

role of oceans implies that reducing agricultural sources will not affect atmospheric

MBr to any great extent, and yet could cause unnecessary economic stress on the

agricultural community. Thus, the economic value and efficacy of MBr fumigation

to control pathogens in agricultural soils warrant renewed research efforts.

Modeling MBr fate and transport during agricultural use (soil fumigation) is

particularly needed to provide improved insight of system dynamics with regard to

atmospheric emissions resulting from MBr fumigation and to provide guidelines for

future research needs.


Table 5-3. Selected chemical and physical properties of MBr (CH3 Br) (Adapted
from Hacherl, 1994).

Selected Property Magnitude
Molecular weight (g mole ') 94.94
Boiling point (oC) 4.6
Solubility in water (g kg'') at 20 C 13.4
Gas density (g L-1 at 20 C) 3.974
Diffusion coefficient (cm2 s-1 in free air at 20 OC) 0.0903
Henry's Law constant at 20 C 0.251
Vapor pressure (mm Hg at 20 OC) 1420


MBr is an odorless hydrocarbon (CH3Br) except at extremely high

concentrations, where it has a sweet odor. It has a low boiling point (4.6 C) and

thus remains in gaseous form at temperatures normally encountered under field

conditions. The important chemical properties of MBr are listed in Table 5-3. Lethal











toxicity of MBr to humans has been noted at 1600 ppm in air for 10-20 hours, or at

7900 ppm for 1.5 hours (USEPA, 1986). The lowest inhalation level found to cause

toxicity in humans is 35 ppm in air (EXTONET, 1997).


Transport Mechanisms for MBr Fumiaant in Soil


Although MBr is extremely volatile, it is commonly injected as a liquid from a

pressurized cylinder. The formulation injected is typically a mixture of 98 % MBr

with 2 % chloropicrin (Hacherl, 1994). Chloropicrin is added to act as a lachrymatory

warning agent. Transport of gaseous MBr in the soil environment is mainly via

diffusion through interconnected air-filled soil pores. Mass flow of gases in the soil

environment has been proposed by several mechanisms, primarily controlled by

changes in temperature and barometric pressure, effects of wind, and forced

movement due to infiltrating water (Baver et al., 1972).

The diffusion coefficient for MBr in free air has been reported to be

325 cm2 hr' at 20 OC (Siebering and Leistra, 1979). The temperature-dependence of

the diffusion coefficient in free air can be defined by use of Eq. [2.15]. Estimation of

the diffusion coefficient of gases in a porous soil system is done by taking into

account the air porosity and tortuosity of the soil system. The tortuosity can be

determined by use of Eq. [3.28].

Maharajh and Walkley (1973) measured MBr diffusion in water for

temperatures ranging from 9.9 to 34.9 C (Fig. 5-1). An almost linear increase in the

diffusion coefficient was observed with temperature over the range of investigation.











0.08
o Maharajh and Walkley (1973)
0.07 Linear fit (r = 0.98) o

S0.06 Figure 5-1. Temperature
dependence of MBr
0.05 diffusion in water (Data
0 from Maharajh and
t.0.04 0
0.04 Walldey, 1973).

0.03 0

10 20 40
Temperature (C)3


Fate of I TEr Fi mi q irr .rl til


The fate of MBr is not only dependent upon transport by diffusion in the

gaseous phase, but also upon partitioning between the gaseous and liquid water

phases as well as between the liquid water and solid phases. Reactions in the water

and solid phases greatly influence MBr fate and transport. In general, two ways

exist by which the fate of MBr in soil is affected. The first includes factors that affect

MBr movement in soil in a reversible manner, such as sorption to solids and

dissolution in water. The second includes factors which contribute to the

irreversible chemical degradation or destruction of MBr molecules, such as

hydrolysis.

Gaseous MBr reversibly partitions into both the solid matrix and the liquid

phase. This partitioning may be described by instantaneous linear sorption under

- ri 1, uir, conditions such as (Brown and Rolston, 1980)


S(R- 1)[5.1]
s = kc ;k, e- p p 5.1
e6 + +p











where k, is a partitioning coefficient between the gaseous phase and the

combination of soil + water phase. Goring (1962) reported the Henry's Law constant

as 0.244 [-] at 20 C for the partitioning of gaseous MBr into liquid water. Mutziger et

al. (1996) reported lab-measured Henry's Law constant values over a temperature

range for the partitioning of MBr into water (Fig. 5-2).


1.0
0 Goring (1962)
S0.8 o Observed (1.2 atm)
-- Estimated (1.0 atm) Figure 5-2. Laboratory-
Smeasured and estimated
S0.6 Henry's Law constants at
0 different temperatures for
u 0.4 O the partitioning of MBr
into water (Adapted from
202 Mutziger et al., 1996).
0.2

0.0 10 20 30 40 50 60 70
Temperature (oC)


Reversible partitioning reactions are characterized by the fact that, as

gaseous phase MBr concentrations decrease, MBr molecules will desorb from soil

solids or volatilize from water and thus be returned to the gaseous phase.

Irreversible chemical or biological degradation reactions, on the other hand, are

ones in which MBr is converted into another compound in such a way that it cannot

return to the gas phase and thus is permanently lost from the system.

In an irreversible hydrolysis reaction, MBr is known to react with hydroxide

ions (OHE) in aqueous solutions through both unimolecular and bimolecular

nucleophyllic substitutions to form methanol and the bromide (Br) ion as shown in











the following reactions (Moje, 1960)

SNl: H3C:Br- [H3C] +:Br
[5.21
[H3C] + : OH H C: OH


S2: H3C: Br+ : OH- [HO: CH3: Br] H3C: OH : Br 5.3]

The rate of these hydrolysis reactions is greatly enhanced in the presence of

ultraviolet light, although light from the visible spectrum has no effect (Castro and

Belser, 1981; and Gentile et al., 1989, 1992). Increased pH and increased

temperatures also show an enhancing effect upon the hydrolysis of MBr

(Hacherl, 1994).

Moje (1960) indicated that unimolecular nucleophyllic substitution should be

the major mechanism for the hydrolysis of MBr in water. Thus, the hydrolysis

reaction becomes first-order. Reactions of MBr in soil water with soil organic matter

involve the transfer of methyl groups to the organic matter (Maw and Kempton,

1973). These reactions are expected to be first-order due to the large excess of

organic matter reaction sites (Brown and Rolston, 1980). Brown and Rolston (1980)

measured the appearance of Br in -hl -.i ,, ii ,jiuo rr._- n :i ri nr from the decomposition

of MBr as a function of time for three different soils. The production of Br for a given

introduction of MBr was greatest for organic soil, intermediate for loam soil, and

lowest for sand. The results thus suggest that mechanisms of decomposition

involving clay particles and soil organic matter are significant to the decomposition

of MBr in soils during fumigation. A first-order hydrolysis rate coefficient of

0.014-hr' was reported for Yolo loam soil by Rolston and Glauz (1982). Hornsby et al.











(1996) reported the soil organic carbon sorption coefficient (k_) for MBr to be

22.0 ml g' of organic carbon. In a recent study by Jin and Jury (1995), on a

Pachappa fine sandy loam soil with 0.49 % organic carbon, the degradation rate of

MBr was observed to be first-order and was insensitive to changes in soil water

content (Table 5-4). However the MBr degradation rate showed a slight increase at

water saturation. These results confirmed earlier work showing that organic matter

is the major factor influencing degradation of MBr in soils.


Table 5-4. Measured degradation rate coefficient and half life of MBr as a function
of soil water content (Adapted from Jin and Jury, 1995).

Degree of Water Saturation in % Half-life Rate coefficient
(Water Content, 0,) (days) (hr-')
45 (6, = 11.5) 12.5 0.0023
76 (9L = 19.4) 12.5 0.0023
100 (6, = 25.6) 10.7 0.0027


Although hydrolysis deactivates toxic MBr molecules, toxic levels of bromide

may accumulate in the soil water of treated soils (Macartney and Price, 1988; and

Maw and Kempton, 1982). High levels of bromide accumulation have also been

found in crops such as tomato, carrot, lima bean, lettuce, radish, tobacco, wheat,

orange, and lemon grown on initially fumigated soils (Hoffman and Malkomes, 1979;

and Maw and Kempton, 1973).


Permeation/Transmission of MBr through Polyethylene Mulch


Open-field fumigation without a plastic soil cover (mulch) may provide a

large input of MBr into the atmosphere. In plastic-mulch culture, atmospheric











emission occurs by diffusive transport (or permeation) through the plastic mulch

(within the bed) as well as by diffusion through uncovered soil surfaces in furrows

between parallel soil beds. Several environmental factors contribute to diffusive flux

through the soil in furrows and through the plastic mulch in beds. Environmental

factors with the greatest impact upon diffusive flux include the temperature

gradient across the plastic mulch, the moisture or condensed water layer on the

underside of the plastic mulch, and the sweeping effect of wind over the surface

(overland wind velocity).

The physical dynamics of water and heat transport beneath plastic-mulched

soil surfaces are quite complex (as discussed in Chapter-2.). Temperature increases

at the plastic mulch during the daytime by as much as 15 to 20 oC due to solar

radiation are not uncommon (Ham and Kluitenberg, 1994). Such temperature

increases during daylight hours enhance the water evaporation rate, thereby

increasing relative humidity of the air gap and of soil beneath the plastic mulch.

Since the plastic cools faster than the soil, the condensation of water from the

humid soil air tends to occur on the underside of the plastic mulch during nighttime

hours. As a result, permeability of the plastic mulch to MBr would be reduced due to

a blocking effect by the layer of condensed water. As the mulch temperature

increases during daytime hours, evaporation then removes the condensed water

layer. In cases where permeating MBr is soluble in or sorbs to the plastic material,

the rate of movement by diffusion would further be reduced, thus delaying the

equilibrium. As such, only operational value for the diffusion coefficient can be

envisioned.











Table 5-5. Ratet of escape of MBr through several films at various temperatures
(Adapted from Kolbezen and Abu-El-Haj, 1977).

Temp. LDPE* HDPE-l* HDPE-2" Polybutene Saran Black LDPE
(C) (1-miil) (-mil) (1-mil) (1-mil) (0.5-mil) (6-mil")
23 8.2 2.7 1.4 2.9 Nil 1.6
30 11.1 3.5 1.7 4.0 Nil 2.2
40 15.1 5.0 2.9 6.3 Nil 3.4
50 20.1 6.8 3.9 9.1 0.18 5.9
60 30.6 8.8 5.2 11.8 0.41 9.2

t Units are ml hr' -2 m-2 L',
low density polyethylene,
" high density polyethylene sample 1,
tt high density polyethylene sample 2, and
* 1 mil =2.544E-3 cm.


Few investigations have been conducted to measure the diffusion coefficient

of MBr through polyethylene sheets used for mulching. Kolbezen and Abu-El-Haj

(1977) measured rates of escape of gaseous MBr through six selected films at

different temperatures. The escape rates increased with temperature for all films.

MBr escape rates ranged from 0.41 to 30.6 ml hr-' m-2 ml' L-' for the six materials.

The authors found one of the high-density polyethylenes (Table 5-5) to be an

effective barrier to MBr transport, with minimum escape rate over the temperature

range investigated.

During a field experiment for unbedded mulched soil in California Yagi et al.

(1993) observed that the cumulative emission of MBr decreased as the thickness of

LDPE plastic increased (1.03,1.29,1.37, and 1.6 mil thickness were used).

Unpublished results from field investigations for mulched soil bed system at the











University of Florida Green Acres Experimental Station, near Gainesville, FL, with

black LDPE (1.2 mil) and clear laminated plastic (relatively thicker than black

LDPE) showed emission fluxes differed between the two plastic materials.

Emissions were reported for both the plastic-covered beds and uncovered furrows in

between the parallel beds.


0.25
0.25 Clear Laminated Plastic
0.20 Feb. 1996 Bed: Observed
0.15 Bed: Moving Average
0.10 Furrow: Observed
0.05 "---------- Furrow: Moving Average

,0.05

0.25 Black Plastic
S( Feb. 1996 Bed: Observed
0.15 \ Bed: Moving Average
n0.10- Furrow: Observed
S ----------. Furrow: Moving Average



a 3.0
" 3 Black Plastic
, June 1995 Furrow: Observed
W 2.0 Furrnw: Movino AvrFur


Bed: Observed
--- Bed: Moving Average


- ..-


Figure 5-3. MBr
emission fluxes in
plastic mulched beds
during fumigation at
the University of
Florida Green Acres
Experimental Station
located near
Gainesville, FL.


24 48 72 96 120 144 168
Time after fumigation (hour)


MBr was applied at the rate of 450 kg ha-' with a 3-nozzle applicator at a

depth of 20 and 33 cm during the summer of 1995 and winter of 1996, respectively,

in soil beds of Arredondo Fine Sand soil used for vegetable production. As expected,

higher magnitudes of the emission fluxes were observed in the summer shallow









67

injection over winter deeper injection during fumigation (Fig. 5-3). A diurnal cyclical

pattern was evident in the emission fluxes to the atmosphere due to non-isothermal

conditions. The relatively thicker clear laminated plastic was able to retain slightly

more MBr in the soil bed over black LDPE (Fig. 5-3). However, material balance

estimations from emission fluxes and soil Br residue measurements for MBr applied

during fumigation revealed that 39, 46, and 70 % of applied MBr were

unaccountable in bed systems with black LDPE (June, 1995), black LDPE (Feb.,

1996), and clear laminated plastic (Feb., 1996) mulches, respectively. Hence, definite

conclusions were not feasible from these field investigations of MBr fumigation in

plastic mulched vegetable beds (personal communication with investigators').

Hacherl (1994) conducted laboratory investigations measuring the diffusion

through black LDPE (1.25 mil) film commonly used in agriculture. Figure 5-4 shows

the results of diffusion investigations by Hacherl (1994) and Kolbezen and

Abu-El-Haj (1977) for black LDPE with 1.25-mil and 1.00-mil thicknesses,

respectively. The results revealed a near-linear increase of diffusion coefficient with

temperature, with lower values in the case of Hacherl (1994) corresponding to

thicker LDPE. In another study (Jin and Jury, 1995), from the data presented, the

diffusion coefficient was estimated to be 0.005969 cm2 hr- for black LDPE (1.0 mil).


SA. G. Hornsby, L.T. Ou, D. Dickson, S.J. Locascio, and P.S.C. Rao











S0.o8 Kolbezen and Abu-El-Haj (1977)
Black LDPE: 1.00- mil
0.007 r =0.97
0.006 Figure 5-4. MBr diffusion
0 through black LDPE of
S0.005/ 7 different thicknesses (Data
S0.004 from Hacherl, 1994; and
C. Kolbezan and Abu-El-Haj,
r 0.003
003 o 1977).
0.002 e Hacherl (1994)
So-- 0 Black LDPE: 1.25-mil
0.001 r = 0.99
20 30 40 50 60
Temperature (C)


Modeling of MBr Transport during Fumigation


Only a few prior modeling efforts have been attempted to describe MBr

transport in soil during fumigation. Hemwall (1959) described a two-dimensional

model for gaseous diffusion of MBr in soil with relatively low water content. The

model assumed a small constant water content and isothermal soil conditions.

Typical simulations were presented to demonstrate time-dependent patterns of

fumigant concentration and biological control within an uncovered soil profile after

fumigation.

Siebering and Leistra (1979) presented a one-dimensional model for gaseous

diffusion of MBr under isothermal conditions. Water content was held constant and

small, such that water flow could be ignored. Comparison of model simulations with

experimental results for MBr concentrations during soil fumigation applied to soil

under cover in a greenhouse matched only in the upper part of the soil profile

(to 40 cm) but deviated substantially elsewhere.











Rolston and Glauz (1982) proposed a model describing radial diffusive

transport of gaseous MBr away from injection chisels during fumigation of a field

profile. The model was based on similar work done by Hemwall (1959, 1960) and

Siebering and Leistra (1979). The model included the dissolution-distillation of MBr

gas on soil solids and in soil water, using both kinetic and equilibrium relationships

and hydrolysis to yield Bf ions according to first-order kinetics. The model

considered the case where the soil surface was covered with a barrier totally

impermeable to the diffusion of gas along with the case of a bare soil surface.

Reported simulation results provided a better match with the experimental data of

Abdalla et al. (1974), which involved a mulch barrier, when no barrier was

considered at the soil surface during simulations. Sensitivity analysis of input

parameters showed reaction kinetics to be unimportant, so they could be safely

ignored. Water flow and non-isothermal temperature effects were ignored in all

simulations. Also, in that model, the boundary conditions (BC) at the soil surface

were unrealistically handled by making the plastic mulch impermeable and

applying a zero-concentration fixed BC at the soil surface.

Mignard and Benet (1989) considered a kinetic relationship for MBr

partitioning between the gaseous and solution phases. Synthetic simulations were

conducted and compared with equilibrium models, and profiles were developed to

forecast the treatment zones and behavior of MBr. The authors inferred from their

analysis that a kinetic partitioning relationship is important only for rapidly

changing boundary conditions. This model also ignored water and temperature

variations in the soil.









70

Recently Wang et aL (1997) adopted a two-dimensional (2-D) finite-element

model, CHAIN_2D (Sim~mek and van Genuchten, 1994) to simulate the fate and

transport of MBr fumigant in soil. CHAIN_2D numerically solves the partial

differential equations for 2-D nonlinear, nonequilibrium simultaneous transport of

water and heat, (but not coupled heat-water) and solute in a variably saturated

porous medium. The temperature dependence of various coefficients (such as DI,

K,, etc.) were described in similar fashion to the present work. However, the model

lacked a realistic description of the soil-surface plastic-mulch/atmospheric

boundary. Placement of a plastic mulch significantly changes the energy balance at

the soil/atmospheric interface (as discussed in Chapter-3) and requires proper

redefinition of the energy exchange and balance at the interface to simulate non-

isothermal soil conditions. The transfer of MBr to the atmosphere was described

(Wang et al., 1997) by introducing a transfer coefficient (h,) as

Dg
FlCux =--- (g-g)= hr, (g-g.) [5.41


where db (cm) was considered a composite boundary thickness inclusive of an air

gap below the mulch, the mulch thickness, and an overlying atmospheric boundary

layer; and where h, was described as a transfer coefficient (cm s '). The db

parameter was shown to be temperature-dependent through use of an equation

similar to Eq. [2.14], but the scientific basis for doing so was not explained.

Although a cyclic nature in emission flux in accordance with non-isothermal

conditions was demonstrated, model simulations failed to accurately describe

(Fig. 5-5) the cyclic field-measured emission flux reported by Yates et al. (1996).













120.0 9 -- Observed
SEstimated
100.0 -

80.0


40.0 -


20.0 0 0
0.0 0 6 0 .
.0 10 20 30 40 50 60 70 80 90 100 110 120 130
Time (hour)
Figure 5-5. Simulated and measured flux density through plastic-
mulched surface. Adapted from Wang et al. (1997).


It is thus evident that, in modeling MBr fate and transport, various

investigators have made simplifying assumptions (such as small and constant soil

water content and isothermal conditions) that reflect unrealistic field conditions.

Also, BCs at the soil surface which are more realistically described by a Neumann-

type in the case of MBr fumigation have been improperly designated (such as a zero

concentration fixed BC at nodes exposed to the atmosphere for uncovered soil). The

dissolution (Henry's Law constant) and diffusion (both in air and water) of MBr are

temperature-dependent, and variably saturated soil-water flow is a common

occurrence under field conditions. This is particularly true for humid climates.












Analysis of Fate and Transport of MBr Fumicant



The model presented in this dissertation represents many improvements over

earlier efforts (Hemwall, 1959; Mignard and Benet, 1989; Rolston and Glauz, 1982;

and Siebering and Leistra, 1979), for it includes coupled transient flows of both heat

and water based upon PDT theory. Realistic plastic-mulch boundary conditions are

described thoroughly by inclusion of optical properties of the plastic mulch in the

energy balance at the soil/atmospheric interface. This model provides opportunity to

consider the effect of improved temperature upon estimates of solute transport (both

in the liquid and gaseous phases).

Modeling investigations (synthetic simulations) of the fate and transport of

MBr fumigant were conducted to obtain improved insight into the system dynamics

for MBr fumigation of soils. Modeling simulations of MBr fumigation prior to this

research have utilized overly simplified assumptions, such as gaseous diffusive

transport in air-filled soil pores under isothermal conditions with the effects of

transient liquid-water flow neglected. Simulations reported here utilized an

assumption that was more realistic of actual field conditions; i.e., non-isothermal

soil conditions that fluctuate diurnally due to net solar radiation, thus altering vapor

and liquid water flow as well as MBr diffusion. Objectives were as follows:

1) To investigate important major processes influencing MBr fate and transport

including volatilization and degradation in fumigated soil beds used for

vegetable and fruit production in Florida,











2) To investigate the effect of non-isothermal and variably saturated soil

conditions on material balance of MBr during simulated fumigations,

3) To evaluate the effectiveness of the current management technique-i.e.,

plastic-mulching of soil beds-in minimizing MBr emission from fumigated

soil beds to the atmosphere, and

4) To examine alternative management methods for decreasing MBr emission

during commercial use of soil fumigation.



Aooroach


Simulation analysis was designed considering major factors known to affect

the fate and transport of MBr. The following factors were considered in the design of

simulations:

I. Effect of variable water saturation in the soil bed: Two levels of heterogenous

water saturation were selected to represent a relatively dry soil and a

reasonably wet profile. The effects of hydrolysis (degradation) on MBr

material balance were considered for different levels of soil water saturation.

II. Effect of variable temperature regime in the soil bed:

A. Non-isothermal conditions: Effects of transient thermal conditions

were considered upon the following input parameters for the model

1. MBr diffusion coefficient in the air phase (DI),

2. MBr diffusion coefficient in the water phase (D"),

3. MBr partition coefficient between gaseous and aqueous

phases (Henry's Law constant, K,),









74

4. Diffusion coefficient of MBr through polyethylene film (D"fpc),

B. Isothermal conditions: Uniform thermal conditions were provided for

comparison (average daily temperature was assumed to exist

uniformly over the simulation domain).

III. MBr diffusion through plastic mulch: Due to uncertainty in published data

and based on reported field investigations of MBr emissions from plastic-

mulched soil, different magnitudes for the MBr diffusion coefficient through

plastic film were used to examine the material balance of MBr.

IV. Depth of fumigant injection: Two depths of fumigant injection zones were

investigated:

A. Typical shallow injection zone: 33 cm depth total

B. Deep injection zone: 66 cm depth total (arbitrarily chosen for

comparison)

V. Effect of irrigation (dousing of water on the bed): Effect of irrigation was

examined to evaluate the surface-sealing effect (due to wetness) on

reduction of gaseous emissions for MBr.

VI. Fumigation Efficiency: Soil sterilization during fumigation was investigated

based on the nematoxic level of MBr (20 ppm or greater) and exposure time

(minimum 24 hr) for 2nd-stage juveniles of root knot nematode (Mckenry and

Hesse, 1978).

VII. Hydrolysis of MBr: As no data for the hydrolysis rate of MBr in Florida soils

was available, sensitivity analysis was carried out for 3 rates; 1) 0.014 hr' (for

California, Yolo loam soil, Rolston and Glauz (1980)), 2) 0.007 hr ', and

3) 0.0035 hr '.











VIII. Diurnal water and heat dynamics in soil bed: The influence of atmospheric

interactions with the soil surface (both bare furrow and mulched bed) were

examined by analyzing the water and heat dynamics in the soil bed system

over a diurnal period of 24 hrs.



Methodolovg


Arredondo Fine Sand (Grossarenic Paleudults, loamy, siliceous,

hyperthermic), a coarse-textured soil at the IFAS Green Acres farm near Gainesville

and commonly found in north central Florida, was selected for the modeling

investigation. Representative soil properties were taken from 'Soil Science Research

Report # 88-1, 19881. The soil consisted of five horizons namely; Ap (0-23 cm),

El (23-66 cm), E2 (66-99 cm), Btl (99-165 cm), and Bt2 (165-203 cm). The weather

data for simulation purposes were taken from a weather station at the Green Acres

Farm of IFAS located west of Gainesville. The month of February was selected for

all simulations, since MBr fumigations are normally performed for the Gainesville

area at that time of the year.

A spatial profile discretization of the cross-sectional soil bed resulted in 2782

elements with 1474 nodes (Fig. 5-6). Bed symmetry was assumed, so only a half-

section (AB-BC-CD-DE-FE-AF) of the soil bed needed be considered for simulations.





'Compiled by Univ. of Florida, IFAS, Soil Science Dep. Soil Characterization Lab. in
cooperation with the USDA-SCS.











No flow BCs on the left- and right-hand sides (AF and DE) were defined for heat,

water, and solutes. The DE boundary represents a vertical symmetry line through

the middle of a furrow between two parallel soil beds. The AF boundary represents a

vertical symmetry line through the middle of a soil bed. The lower boundary (FE)

was designated as a zero thermal gradient with gravitational water flow (= K(h, T)).



-x Mulched-Bed




-. -- ---- .. .. ..
Dis c fr:- r Be
Furro%%
i : ... i .:_ ~ :3 .:
.Inection Zre. I .



-75- : 2




-. 7-- ------- ---.
B l .

4I ... ... ....- .


-175 -
t B3 2 .

I[I l I: .:. I.I

Distance from Bed Center (cm)

Figure 5-6. Schematic of the discretized simulation domain of
the soil bed.


For solute transport, both convective (in the liquid phase) and diffusive (both

in the aqueous liquid and gaseous phases) transport were allowed at the bottom

(FE), with the assumption that solute leaving the lower boundary is lost from the










flow domain (i.e., there is no return flow). The emission loss of MBr to the

atmosphere (AB-BC) was described through diffusive flux, assuming a 1-cm

boundary layer and zero concentration on the atmospheric side of the boundary

layer. These emission fluxes may be defined explicitly as

a) For Bed

'd = gasi s gs+ l(air)
S I Zs Zs+ (air) /

b) For Furrow

jFwrow i 8 s(air) +,ni g -l
Jgas free air --, +ZI- O),C z--- z- j

where the subscripts s, +,, and, represent nodes at the soil surface, above the

surface, and below the surface, respectively.


Table 5-6. Physical properties for 5 distinct horizons of a profile of Arredondo fine
sand soil.

Soil Layer Sand Silt Clay OCt BD K at
Soil Layer (%) (%) (%) (%) (g cm 3) (cm hr
Ap 91.9 3.3 4.8 0.83 1.52 6.2
El 92.0 3.3 4.7 0.28 1.54 16.7
E2 91.4 3.0 5.6 0.14 1.52 15.1
Btl 86.0 3.7 10.3 0.11 1.56 6.4
Bt2 75.2 3.0 21.8 0.20 1.57 2.4

SOC- Organic Carbon,
BD- Bulk Density


The data for physical properties of the soil profile used are presented in

Table 5-6, with other related parameter values in Table 5-7. The soil water










characteristic data used for estimating (van Genuchten, 1980) hydraulic-function

parameters were taken from the report mentioned earlier. Average weather

parameters' used in simulations for the month of February were: 1) average daily

temperature 11.6 C, 2) temperature amplitude 7.7 OC, 3) wind velocity 11x104

cm hr1, 4) day-length 11.0 hrs, and 5) global daily radiation 1400 J cm-2 day'.


Table 5-7. Parameter values for 5 distinct horizons of a profile of Anedondo fine
sand soil.

Soil pI I c K [ 1 Kd2 van Genuchten Parameters
Layer (hr-1) (cm) (cmgg-) (cm ( g c) a (cm 0
Ap 0.014 0.1826 0 0.0225 2.124 0.43 0.03
El 0.014 0.0616 0 0.0216 2.762 0.42 0.04
E2 0.014 0.0308 0 0.0271 2.626 0.43 0.04
Btl 0.014 0.0242 0 0.0211 2.091 0.41 0.06
Bt2 0.014 0.0440 0 0.0307 1.557 0.41 0.13

t- hydrolysis rate coefficient for MBr,
soil sorption coefficient for MBr, and
tt- soil sorption coefficient for bromide ion.


Temperature-dependence of the diffusion coefficients for MBr in gaseous

and aqueous phases of the soil were defined by use of Eq. [2.15] and data reported

ir, F-i. 5-1 from Maharajh and Walkley (1973), respectively. As information is

unavailable for the temperature dependence of the hydrolysis rate coefficient lp,), a

fixed value of 0.014 hr-1 was used (Rolston and Glauz, 1982). The diffusion coefficient





'Climatological Data, Agronomy Department and NOAA Cooperating, 1996.











for the plastic mulch was defined using the data of Hacherl (1994). The MBr

fumigant was assumed to be applied at a rate of 450 kg ha-' with a three-shank

applicator, so that 1/ shank (3591 pmoles) injections occur within the flow domain

shown in Fig. 5-6. Unless otherwise noted, the depth of the injection and time of day

for injection for all the simulations were set at 33 cm and 9:00 AM, respectively.

Black LDPE of thickness 1.25 mil was assumed as the plastic mulch m all the

simulations. Mass and energy balance errors were less than 1.0 % for all simulations

reported in this work.



Results and Discussion



Plastic-Mulching of Soil Beds to Control MBr Loss to the Atmosphere


Preliminary simulations conducted using the laboratory-measured data

(Fig. 5-7, Level-1) of Hacherl (1994) for diffusion through plastic (black LDPE,

1.25-mil thick) provided estimates for the atmospheric losses of MBr during

fumigation of the soil bed that were very low (23 % during the 1" day) compared to

reported values of 40 to 60 % for recent investigations (Yagi et al., 1993; Yates et al.

1996). Hence, parallel elevated temperature-dependent levels were designed

(Fig. 5-7) for investigation. The use of an experimentally measured diffusion

coefficient for a 1.25-mil thick black LDPE by Hacherl (1994) in simulations showed

that very little MBr escaped (0.02 % of that applied) through the plastic mulch into

the atmosphere (i.e. from the bed excluding the furrow) during the 1st day. The












diffusive flux (Fig. 5-8: Level-1, and Fig. 5-9: a, Level-1) based upon the measured

diffusion coefficient of plastic mulch from the soil bed was very low and it was also




27 Elevated Level-4 (= Level-I x 104)
22 6430*s^
17
12 _
S 2.7 Elevated Level-3 (= Level-I x 103) A


1' 2.2 1.2 -._

8 0.27 Elevated Level-2 (= Level-1 x 102)
S0.22
017
S0.12 _______________
0.0027 Measured Level-I (Hacherl, 1994)
0.0022
0.0017
0.0012 20
20 25 30 35 40
Temperature (C)


Figure 5-7. Measured and elevated levels of diffusion coefficients
through 1.25-mil black LDPE plastic mulch used in the
model simulations.


demonstrated by the material balance after 1 and 3 days (Table 5-8) that only 23.4

and 45.8 % of the applied MBr, respectively, was lost from the system (mulched soil-

bed plus bare furrow) to the atmosphere. This result is much lower than the 40 to

60 % atmospheric losses earlier reported from field investigations (Yagi et al., 1993;

Yates et al., 1996), which were carried out under unbedded conditions with

complete mulch cover of the soil. Initial average fluxes both from the bed and the

furrow were substantially higher than for later periods (after 24 hours) when the










81

concentrations in soil had decreased due to spreading, hydrolysis, and loss of MBr to

the atmosphere. Also note that emission fluxes from the furrow were nearly 5 times

larger than for the bed, except for elevated plastic-mulch diffusion Level-4 due to

the absence of a mulch barrier over the furrow.



1750-
1400
1050

350






35 00
8 0 7_ ----------------

Bed


\ 5.00
4.00-




I 3A0 2
Furrow

0.00

2, 125 ----- Elevated Level-2
S-- Elevated Level-3
1.50 -- Elevated Level-4
Bed
0.00
Time after fumigation (hour)

Figure 5-8. Losses of MBr to the atmosphere through soil bed and furrow
surfaces using measured and elevated diffusion coefficients for
MBr through plastic mulch'.







'Small steps in some of the figures resulted from the need for plotting more points
within certain time periods.









82

Use of elevated levels of diffusion coefficients for plastic mulch showed more

losses of MBr into the atmosphere from the system (Level-2 = 25.1 %; Level-3 =

34.7 %; and Level-4 = 51.7 %) one day after fumigant injection. In all cases,

emission fluxes from the bare soil furrow were highest near the bed due to close

proximity to the injection zone. Figure 5-9: a shows MBr fluxes in the bed 6 hours

after fumigant injection, when peak emission fluxes occurred from the bed (Fig. 5-8).


(a) Measured, Hacherl (1994) Elevated Level-2
.. .I, m .I .










Elevated Level-3 Elevated Level-4
t '- 1 .--.





-- '*2-- .*0












S 15 30 45 60 75 0 15 30 45 60 75
Distance from bed center (cm)

Figure 5-9. MBr total concentration (pmol cm- ) contours and gaseous
MBr flux vectors (pmol cm2 hr-') in the bed 6 hrs after infection
of MBr fumigant.
a) 4 levels of diffusion coefficients through the plastic mulch,
b) non-mulched bed.











Emission fluxes (Fig. 5-9: a) increased from the mulched bed with elevated

levels of diffusion coefficients for the plastic mulch. Also to be noted here is the fact

that the trend of diffusive flux was reversed in the furrow compared to that for the

bed, and lower diffusion coefficients for plastic mulch showed higher diffusive flux

and more cumulative loss from the furrow (Fig. 5-8 Fig. 5-9: a, and Table 5-8).

Figure 5-9: b shows the emission of MBr in furrow and bed without a plastic mulch,

for comparison purposes. The emission fluxes were higher from soil-surface regions

close to the fumigant injection zone (Fig. 5-9: b).


(b) No Mulch
0 ltt s-l l:1 lmol cm2 hr-'






Fiqure 5-9--continued


-75 '


-0 15 30 45 60 75
Distance from bed center (cm)


Overall, total material losses of MBr from the simulated soil-system to the

atmosphere (Table 5-8) increased when elevated levels of diffusion coefficients for

plastic mulch were used in the simulations. Three-day simulations were

appropriate, as the results showed that most of the MBr losses occur within the

initial 24 to 30 hrs after fumigation (Fig. 5-8 and Table 5-8). Atmospheric losses were










observed to continue in small amounts even after 3 days, as long as undegraded

MBr is present in the soil.

Obviously much uncertainty prevails regarding the diffusion coefficient of

MBr in plastic mulches. The use of a lab-measured diffusion coefficient for

simulating field conditions greatly under-estimated published results from the field

investigations done in the past. However, the use of elevated diffusion coefficients

(Level-2, Level-3, and Level-4) improved simulation results. It was, therefore,

decided to use the Level-3 diffusion coefficient for all other simulations, as results

obtained from this level were close to those from the published field investigations.


Table 5-8. Material balance (unit: %) of MBr under measured and enhanced levels
of diffusion coefficient for plastic mulch.

Atmospheric Emission
Day-1 Total MBr Losses Hydrolyzed Deep Down-
in Profile flow
Bed Furrow Total
Level-1 69.55 0.02 23.39 23.41 7.04 0
I I
Level-2 67.91 1.81 1 23.29 25.10 6.98 0
Level-3 58.54 13.15 21.57 34.72 6.73 0
Level-4 42.24 34.24 17.41 51.65 6.11 0
I I
No-Mulch 37.29 56.92 5.79 0
Day-3
Level-1 36.38 0.20 i 45.63 45.83 17.79 0
Level-2 33.94 4.24 I 44.42 48.66 17.39 0
I I
Level-3 23.46 24.55 1 36.52 1 61.07 15.46 0
II
Level-4 15.05 46.32 1 25.84 1 72.16 12.79 0
N h 13.40 I74.68 111
No-Mulch 13.40 1 74.68 11.91 0











Non-Isothermal versus Isothermal Soil-Conditions


An isothermal simulation was performed in order to compare with

non-isothermal simulations existing under a black LDPE. Thermal effects on solute

transport (involving D9, D",kg, Dgp ,c parameters) were ignored in isothermal (with

respect to solute transport only) simulations. However, the coupled heat and water

flow model was used in its original form, with solute transport being made

independent of temperature fluctuations. The parameters for solute transport were

estimated and kept constant for a daily average temperature of 11.6 C, -.p,,: ji ,

existing during the month of February in the vicinity of Gainesville (diffusion

coefficient in air = 308 cm2 hr-1, Henry's Law constant = 0.15 [-]).



S5.00 Non-Isothermal
S3.75 ----- Isothermal

o Furrow

S0.00
R 0.99

0.66

80.33 Bed
CO~ ~ ------------^
< "000 12 24 36 48 60 72
Time after fumigation (hour)

Figure 5-10. Emission fluxes of MBr through bed and furrow under
isothermal and non-isothermal transport of fumigant.









86

The isothermal and non-isothermal simulations conducted for a 3-day period

showed (Fig. 5-10) that MBr emission fluxes from both bed and furrow were affected

by non-isothermal conditions. Figure 5-10 shows the diurnal fluctuations in MBr

emissions from bed and furrow in the case of non-isothermal simulations for black-

LDPE, whereas a similar cyclical nature of MBr emission was missing m isothermal

simulations. This emission pattern has been reported earlier (Fig. 5-5) for field

investigations (Yagi et al., 1993; and Yates et al., 1996). Higher diffusive flux rates

for MBr were observed initially from the furrow, which subsequently flattened out

due to lower concentration gradients as increasing amounts of MBr were lost to the

atmosphere. The delay in reaching the initial peak emission flux (Fig. 5-10) from the

bed occurred due to a build-up time for MBr concentrations beneath the mulch and

an increase in temperature of the plastic mulch as the day progressed (application

time 9:00 AM), whereas the bare furrow surface offered a quick passageway to the

atmosphere.


Table 5-9. Material balance (unit: %) of MBr for fumigation under isothermal and
non-isothermal conditions.


13.44 21.59 34.73


Total MBr Atmospheric Emission Losses
Day-1
in Profile Bed Furrow Total

Isothermal 61.28 11.42 20.33 31.75
i j


Non-
Isothermal


58.55









87

The results suggest that inclusion of non-isothermal conditions is essential

to correctly estimate the fate and transport of volatile chemicals. Non-inclusion of

temperature effects on solute transport may underestimate (Table 5-9) matenal

losses to the atmosphere in the case of volatile chemicals such as MBr. The

isothermal simulation estimated 260.2 kg ha'1 (57.8 % of application) emission loss

to atmosphere compared to 274.8 kg ha' (61.1 % of application) as a non-isothermal

estimate over a 3-day period.


Variable Water Saturation of Soil


Volumetric water contents present in the soil bed at the time of fumigation

not only affect the hydrolysis of MBr but also its transport in the soil. Two different

wetting regimes were selected for the simulations. An initially saturated soil profile

was allowed to drain and evaporate water for 40 days. Two wetting levels were then

selected to provide the soil water regime:

1) Level-1 (WL-1): 3292 cm3 total water content in the flow domain, and

2) Level-2 (WL-2): 1864 cm3 total water content in the flow domain.

The simulations revealed (Fig. 5-11) that larger amounts of MBr were lost to

the atmosphere when there was less water (WL-2
result was expected, due to less hydrolysis and enhanced diffusion. Higher

temperatures and lower water contents existed (Fig. 5-12) near the surface in the

case of the drier soil system (WL-2). Greater diffusion pathways (larger diffusion

volume due to lower water contents) at higher diffusion rates also existed in the

drier profile. This resulted in a higher MBr emission flux from both the bed and









88

furrow in the case of the drier (WL-2) profile (Fig. 5-12). The emission fluxes from the

bed were also higher (Fig. 5-11) during the later periods (after approximately 34 hrs)

in WL-1. This was due to the higher water contents in WL-1 causing larger amounts

of MBr to partition into the aqueous phase, which were then released to the gaseous

phase as atmospheric losses occurred in order to maintain equilibrium between the

aqueous and gaseous phases (controlled by the Henry's Law constant).



5.00 WL-1
S3.75 ---- WL-2

S2 50 Furrow


o o.oo '1/-----------
1 .25 1- --------------------------
0.99 -






Time after fumigation(hour)
'S 0.66 -
S ;Bed

IF0.33

< .00^ -^-^ --^ -- -- -*f
Time after fumigation(hour)
Figure 5-11. Emission fluxes of MBr through bed and furrow dunng
fumigation performed under different profile
water-saturation scenarios.


Figure 5-12 shows the plots of water content, temperature, MBr

concentration, and MBr fluxes in the bed under different water saturations (WL-1

and WL-2) for 0.5, 3, 6, 12, 18, and 24 hours after fumigant injection. Temperatures

under the bed and at the bare soil surface were higher in WL-2 as compared to

WL-1, whereas water contents were lower in WL-2 as compared to WL-1 at the




Full Text

PAGE 1

MODELING COUPLED WATER HEAT FLOW AND IMPACTS UPON CHEMICAL TRANSPORT IN MULCHED SOIL BEDS By DILIP SHINDE A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1997

PAGE 2

I dedicate this work to my parents, father (late) Krishna Rao Shinde and mother Vatsala Shinde, who instilled in me the importance of education and knowledge.

PAGE 3

ACKNOWLEDGMENTS I express my sincere gratitude to Professors R.S. Mansell (chair ) and A G Hornsby (cochair) for their guidance and support during this research work Appreciation is also extended to the members of my advisory committee: Professors P.S.C. Rao, B.L. McNeal, and F. S. Zazueta; for useful comments, help and support. I would like to extend my appreciation and thanks to Dr. B.P. Mohanty (Univ. of California, Riverside) for providing numerous technical literature citations. Thanks are also due to Professor P N. Kizza for encouragement and support I w o uld like to also thank Dr. A. Akpoji and Ron Jessup for useful discussions on finite element methods and numerical programming, respectively. I acknowledge Dr John L. Neiber (Agric. Eng. Dep., Univ. of Minnesota) for providing the FORTRAN code used extensively in creating finite element-mesh in this work. I take this opportunity to thank also 'The Rotary Foundation of The Rotary International' for the prestigious "FREEDOM FROM HUNGER" scholarship to enhance my knowledge and join the fight to free the world from hunger. I would like to acknowledge the substantial financial support provided by Professor A.G. Hornsby during the latter part of this research. I would also like t o acknowledge the facilities and partial financial support provided by Professors R.S. Mansell and B L. McNeal, respectively, during this research. Lastly, appreciation is extended to my family for their help and support. l.l.l.

PAGE 4

TABLE OF CONTENTS ACKNOWLEDGMENTS . . . . . . . . . . . . . . . . . . . . . . . . iii LIST OF TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . vii LIST OF FIGURES .................................................... ix ABS TM CT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi INTRODUCTION ....................................................... 1 Coupled Versus Uncoupled Flow Processes in Soil ..................... 1 Soil Thermal Regime . . . . . . . . . . . . . . . . . . ... 2 Non-isothermal Soil Conditions ............................... 3 Mechanistic Mathematical Modeling .......................... 4 H -ypothesis . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Objectives .................................................... .. 7 REVIEW OF LITEM TURE ............................................... 8 Isothermal Water Flow in Porous Media . . . . . . . . . . . . . . .. 9 Soil Temperature and Heat Flow ... ... ............................. 11 Soil Thermal and Water Regimes under Mul ching ................... . 15 Coupled Heat and Water Flows ................ .. .................. 19 Philip-de Vries Theory (PDT) ............................. ... 20 Application of Philip-de Vries Theory ...................... . . 22 Solute Transport during Coupled Heat and Water Flow ................. 24 Temperature Effects on Solute Transport ............................ 27 Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 MATHEMATICAL MODEL .. ................... ....... ...... ....... . 31 Liquid and Vapor Water Flow in Soil . . . . . . . . . .............. 32 Heat Flow in Soil . . . . . . . . . . . . . . . . ................ 32 Transport of Multiple Species of Solutes in Soil ........ ............... 33 Energy Balance at the Soil/ Atmosphere Interface ................... . 34 1. V

PAGE 5

Initial and Boundary Condition s .................................... 38 Input Parameter Requirements ............................... ...... 39 MODEL VERIFICATION AND VALIDATION ............................... 44 Isothermal Water Flow ...... .. .................................... 45 Uncoupled Heat Flow . . . . . . . . . . . . . . . . . . . ..... 46 Coupled Water and Heat Flow .................................. ... 47 Isotherrnal 1D Solute Transport during Steady Water Flow . . . ....... 48 Isothermal 2-D Solute Transport during Steady Water Flow ............. 49 Isothermal 1-D Solute Transport with Nitrification Chain during Steady Water Flow . . . . . . . . . . . . . . . . . . . . . . . 51 Isothermal 1-D Solute Transport involving Gas Flow ................... 52 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 MODEL APPLICATION ................................................. 55 Fate and Transport of Methyl Bromide Fumigation in Soils ...... ....... 55 Transport Mechanisms for MBr Fumigant in Soil . . . . . . . . 59 Fate of MBr Fumigant in Soil . . . . . . . . . . . . . ...... 60 Permeation/Transmission of MBr through Polyethylene Mulch .... 63 Modeling of MBr Transport during Fumigation . . . . . . . . 68 Analysis of Fate and Transport of MBr Fumigant .............. ........ 72 Approach . . . . . . . . . . . . . . . . . . . . . . ... 73 Methodology . . . . . . . . . . . . . . . . . . . . . . .. 75 Results and Discussion . . . . . . . . . . . . . . . . . . . . ... 79 Plastic-Mulching of Soil Beds to Control MBr Loss to the Atmosphere . . . . . . . . . . . . . . . . . . . . 79 Non-Isothermal versus Isothermal Soil-Conditions .............. 85 Variable Water Saturation of Soil . . . . . . . . . . . . . .. 87 Depth of Fumigation Injection .. ............................ 97 Effect of Irrigation (Water Dousing of the Bed Surface ) ........... 98 Fumigation Efficiency Analysis . . . . . . . . . . ......... 101 Hydrolysis of MBr (Sensitivity Analysis) ....................... 107 Diurnal Water and Heat Dynamics in a Plastic Covered Soil Bed .. 108 Summary and Canel us ions . . . . . . . . . . . . . . . . . . . 119 APPENDIX A SOIL MOISTURE FLOW .................................. 122 APPEND IX B SOIL HEAT FLOW . . . . . . . . . . . . . . . . . .... 125 APPENDIX C CHEMICAL TRANSPORT ................................. 127 APPENDIX D NUMERICAL APPROACH ................................ 132 V

PAGE 6

LIST O F REFERENCES .......... ....................... ........ .... 14 2 BI O GRAPHICAL SKETCH . . . . . . . . . . . . . . . . . . ......... 1 53 Vl

PAGE 7

LIST OF TABLES Table page 2-1. Average ther1nal properties of major soil constituents at 20 C and 1 atm (Ghildyal and Tripathi, 1987) .................... .. ........... 12 2-2. Short-wave optical properties of five plastic mulches used in commercial crop production. Data adapted from Ham et al. ( 1993) . . . . . ....... 17 5-1. Earlier estimates for source-contributions of MBr to the atmosphere (Grojesan, 1991; Khalil et al. 1993; Lobert et al., 1990; Mano and Andreae, 1994; Yagi et al., 1993 ; and Zurer, 1993) ................. .... 56 5-2. A recently revised budget for atmospheric MBr (Yvon-Lewis and Butler, 1997) .................................................... 57 5-3. Selected chemical and physical properties of MBr (CH 3 Br) ( Adapted from Hacherl, 1 994) . . . . . . . . . . . . . . . . . . . . ...... 58 5-4. Measured degradation rate coefficient and half life of MBr as a function of soil water content (Adapted from Jin and Jury, 1 995) ......... ....... 63 5-5. Rate of escape of MBr through several films at various temperatures (Adapted from Kolbezen and Abu-El-Haj, 1977) ................ ... .. .. 65 5-6 Physical Properties for 5 distinct horizons of a profile of Arredondo fine sand soil . . . . . . . . .. . . . . . . . . . . . . . . ........ 77 5-7. Parameter values for 5 distinct horizons of a profile of Arredondo fine sand soil ................................................ .... .. 78 5-8. Material balance (unit:%) of MBr under measured and enhanced levels of diffusion coefficient for plastic mulch ............................. 84 5-9. Material balance (unit: %) of MBr for fumigation under isothermal and non-isothermal conditions ........................ .. ... ... ....... 86 Vll

PAGE 8

5-10. Material balance of MBr (unit: %) under different water saturations for one and 3 days after fumigant injection .............................. 90 5-11. Material balance of MBr (unit: % ) during fumigation under different water saturations for 65-cm injection zone ........................... 98 5-12. Material balance (unit: %) of MBr under water dousing and no-dousing .. 100 ' Vlll

PAGE 9

LIST OF FIGURES Figure page 2 1. Measurements of soil temperature at 10 cm depth beneath five different plastic mulches and a bare plot near Manhattan, KS on day 195 of the year 1991. Adapted from Ham et al. ( 1993 ) ........................... 18 4-1 Observed and computed distributions of water content in a sand column during constant flux infiltration ................... ................ 45 4-2. Observed and computed temperatures for two depths in a field soil ........ 46 4-3 Observed and computed water content after 31 days under the influence of both matric head and thermal gradients . . . . . . . . . . . ..... 48 4-4. Observed and computed breakthrough curves for boron ....... ......... 49 4-5. Location of solute concentration fronts for three times as estimated numerically and with the analytical solution ........ ................ 50 4-6. Analytically deter1nined and numerically computed concentration profiles at 200 hours . . . . . . . . . . . . . . . . . . . . . . . . . ... 52 47 Analytically determined and numerically computed concentration profiles at 1 3, and 5 hours ....... ................. ..... .. ... .. ..... .. 53 5-1. Temperature dependence of MBr diffusion in water (Data from Maharajh and Walkley, 1973) ............................................... 60 5-2. Laboratory-measured and estimated Henry's Law constants at different temperatures for partitioning of MBr into water (Adapted from Mutziger et al., 1996 ) .................................................... 61 5-3. MBr emission fluxes in plastic mulched beds during fumigation at the University of Florida Green Acres Experimental Station located near Gainesville FL. . . . . . . . . . . . . . . . . . . . . . . . . 66 5-4. MBr diffusion through black LDPE of different thicknesses (Data from Hacherl 1994; and Kolbezan and Abu-El-Haj, 1977 ) ................... 68 lX

PAGE 10

5-5. Simulated and measured flux density through pla stic-mulched surface. Adapted from Wang et al. (1997) .................................... 71 5-6. Schematic of the discretized simulation domain of the soil bed ........... 76 57. Measured and elevated levels of diffusion coefficients through 1. 25mil black LDPE plastic-mulch used in model simulations .................. 80 5-8. Losses of MBr to the atmosphere through soil bed and furrow surfaces using measured and elevated diffusion coefficients for MBr through plastic mulch ................................................... 81 5-9. MBr total concentration (mol cm 3 ) contours and gaseous MBr flux vectors (mol cm 2 hr 1 ) in the bed 6 hrs after injection of MBr fumigant. . 82 5-10. Emission fluxes of MBr through bed and furrow under isothermal and nonisothermal transport of fumigant ..................... .............. 85 5-11. Emission fluxes of MBr through bed and furrow during fumigation performed under different profile water-saturation scenarios ............ 88 5-12. Contours of water content (cm 3 cm 3 ), temperature (C), and MBr concentration (mol) with gaseous fluxes (mol cm 2 ru 1 ) of MBr under 2 different soil-water saturation scenarios after fumigation ............. 91 5-13. Emission fluxes of MBr under different water saturations for a 66-cm injection zone during the first 3 days after fumigant injection ........... 97 5-14. Effect of water dousing on the MBr emission fluxes from the soil bed during fumigation ................................................ 99 5-15. Contours of water content, temperature, and MBr concentration with gaseous fluxes of MBr at 9 hours after fumigation under water dousing and no-dousing on the bed-surface ............................... 100 5-16. Soil sterilization zones for different scenarios 3 days after MBr fumigation using different criteria .................... ....... ................ 104 5-17. Loss of MBr (unit: %) under different rates of hydrolysis of MBr ... .... 107 5-18. Diurnal water and heat dynamics in a mulched bed for a wet soil ........ 113 5-19. Diurnal water and heat dynamics in a mulched bed for a dry soil. ...... 116 D4-1 Linear triangular element . . . . . . . . . . . . . . . . . . . . 136 X

PAGE 11

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy MODELING COUPLED WATER-HEAT FLOW AND IMPACTS UPON CHEMICAL TRANSPORT IN MULCHED SOIL BEDS By Dilip Shinde December 1997 Chair: Robert S. Mansell Cochair: Arthur G. Hornsby Major Department: Soil and Water Science Plastic-covered soil bed systems are widely used in the southeastern United States for vegetable and fruit production. Plastic-mulching results in highly variable and transient soil thermal regimes both in space and time under atmospheric influences (e.g., solar radiation, wind etc.). Much evidence exists that water flow is affected by non-isothermal conditions and vice-versa. In this work, a two dimensional model for simultaneously describing chemical transport and coupled water-heat flows in plastic-covered soil beds was developed with many improvements over earlier efforts. Realistic plastic-mulch boundary conditions are thoroughly described by including optical properties of the plastic mulch into the energy balance at the soil/atmospheric interface along with other atmospheric influences of weather components. This model provides opportunity to consider the Xl

PAGE 12

effect of coupled water-heat flows upon estimates of chemical transport (both in the liquid and gaseous phases) in plastic-mulch culture production systems. Chemical transport in soil beds in the model utilize consideration of chemical existence in solid, liquid, and gas phases. Temperature dependence of the rates of chemical reactions and volatilization is described efficiently by incorporating temperature dependence in fate and transport coefficients. The model uniquely provides opportunity to simulate coupled water-heat flow and multiple solute transport with a complex plastic-mulch atmospheric boundary for realistic field situations. The model was used to analyze the fate and transport of methyl bromide (MBr) during fumigation of mulched soil beds. MBr, a highly volatile pesticide, is widely used in California, Florida, and Hawaii for its broad-spectrum efficiency in controlling insects, fungi, and weeds. Unfortunately, MBr is a major environmental hazard due to its depletion effect on the atmosphere ozone layer. Model analyses revealed that non-isothermal conditions should be considered in the accurate description of MBr fate and transport. Modeling results indicate that utilizing a relatively wet soil profile, deeper injection, and water dousing on the bed-surface before installation of plastic mulch will decrease atmospheric emissions of MBr. However, deeper injection of MBr fumigant, although environmentally less hazardous, may decrease effective sterilization of crop root zones in the soil. Xl.1.

PAGE 13

CHAPTER 1 INTRODUCTION Coupled Versus Uncoupled Flow Processes in Soil Flow of water, air, and heat in soil is generally assumed to be uncoupled that is, the driving forces causing individual flow are generally considered to be mutually independent. Darcy 1 s equation for liquid water flow, based upon spatial gradients of soil-water potential and Fourier 1 s equation for heat flow based upon spatial gradients of soil temperature, provide two examples of uncoupled flow processes. In nature, however, water is also known to move in response to molecular forces such as thermal and osmotic gradients. Similarly, considerabl e heat flow may occur as latent heat by water vapor diffusion in the air-filled soil pores, and as sensible heat transfer due to liquid and vapor movement In a partially saturated soil, flow of water due to thermal gradients occurs through zones occupied by liquid as well as air, and is assisted by multiple condensation and evaporation cycles (Philip and de Vries 1957). Thus transport processes occurring in the soil under field conditions are intricate and complex. Diffusion theory indicates that water molecules are in constant thermal motion, accompanied by jump-like transitions from one temporary equilibrium s ta te to another (Ghildyal and Tripathi 1987). The frequency of transition and the number 1

PAGE 14

2 of molecules traversing a certain section of the liquid is highly dependent upon the temperature gradient and causes self thermo-diffusion of water. As temperature gradients are accompanied by gradients of surface tension at the air-water interface, there is a possibility of thermo-capillary flow and thermo-capillary film flow of water in the liquid phase (Ghildyal and Tripathi, 1987). Thermo-capillary flow occurs in those capillaries where liquid-water menisci exist. Here, temperature variation induces the variation of capillary pressure and a corresponding shift of the liquid column. Thermo-capillary film flow is caused due to gradients of surf ace tension (change in water affinity with change in temperature) at the water-film surface. Both of these mechanisms contribute significantly to the total thermal water transfer under moderate to high water content conditions. Movement of water in the liquid state due to changes in the volume of entrapped air with temperature has also been observed (Ghildyal and Tripathi, 1987). In water-saturated or nearly saturated soil, the entrapped air may cause direct transport of water through the pressure it exerts and, in unsaturated soil, through an apparent increase in the temperature coefficient of the matric potential. Convective flow of water vapor may also occur due to differences in the densities of cold and warm air. Soil Thermal Regime The thermal regime of a soil depends upon radiative solar heat flux into the soil and upon heat-transfer processes occurring in the soil and between the soil and air which in tum, depend on the thermal characteristics of soil solids, gases, and water. The soil therrnal regime which greatly modifies the overlying microclimate,

PAGE 15

3 is generally characterized in terms of soil temperature. Soil temperature fluctua tio ns are especially important since they impact physical, chemical, and biological processes in the soi l The source of heat for soil is incoming net solar radiation. On a clear day Geiger (1957) described the typical distribution of incoming solar radiation as: 1 ) global radiation, which is the sum of the direct-beam penetration to the earth ( 19 %) and the diffuse sky radiation (26 %); 2) scattered radiation returned to outer spa ce ( 1 1 %); 3) radiation reflected by clouds (28 %); and 4) radiation absorbed by the atmosphere (15 %). Non-isothermal Soil Conditions Much published literature confirrns that soil-water movement is aff e cted by spatial gradients of both temperature and water potential. Thus, under non isothermal conditions, movement of water (both liquid water and vapor tend to move from regions of high to low temperature) through soil should be considered for both liquid and vapor phases. Non-isothermal conditions have broader implications in that solute-transport mechanisms are associated with water movement and the fate and transport of volatile organic chemicals are also influenced by the soil therrr1al environment The influence of thermal gradients upon water movement in soils has largely been observed in two areas (Jury 1973) : 1) in the hot, dry surface layer of a bare s oil where the interface between the soil and atmosphere causes wat e r movement t o b e largely in the vapor phase; and 2) deeper in the soil where therrnal gradients pers i st

PAGE 16

4 over seasons and can slowly transport liquid upward or downward. I n the special case of plastic-mulched soil, imposed thermal gradients tend to be greater at the soil surface and the mulch also provides a partial barrier to vapor flow between the soil and atmosphere and causes a return flow of water after condensation under the mulch. Both water movement and contaminant transport are influenced by temperature gradients in soils. Natural temperature grad i ents are usually highest during daytime (midday) near the land surface, but the relatively small geothermal gradient below the leve l of seasonal temperature variation can be a significant factor in deter1-nining t h e percolation of water in deep unsaturated zones (Evans and Nicholson, 1 987). Mechanistic Mathematical Modeling Mechanistic mathematical models provide powerful tools to describe and interpret flow and transport processes in soil-water systems. Bear and Verruijt ( 1987 ) defined a model as a 'simplified version of the real system that approximately simulates the excitation-response relations of the latter'. The first step in modeling involves the construction of a conceptual model of the problem, consisting of a s e t of assumptions that reduce the real problem and the real domain to simplified versions (ultimately permitting mathematical description of the processes) that a re acceptable in view of the modeling objectives. In the second step, the conceptual model is expressed in the form of a mathematical model (Bear and Verruijt, 1987 ). Properly posed mathematical models for initial-value problems involving mass and

PAGE 17

5 energy transport in porous media require (i) appropriate differential transport equations, (ii) initial conditions in the flow domain, and (iii) boundary conditions imposed along the surfaces of the flow domain. Models are essential in perfom1ing complex analyses and in making infor1ned predictions. Friedman et al. (1984) stated that, in some cases, models have increased the accuracy of estimates of future events to a level far beyond 'best judgement' decisions. Anderson and Woessner (1992) emphasized that models provide the best way to make an informed analysis or prediction concerning consequences for a proposed action. The authors further stated that models can also be used in an interpretive sense to gain insight into the controlling parameters in a site-specific setting or as a framework for assembling and organizing field data and formulating ideas about system dynamics. Schwarzenbach et al. (1993) defined a model in the following way : 'A model is an imitation of reality which stresses those aspects that are assumed to be important and omits all properties considered to be nonessential'. These authors summarized the role of modeling explicitly in the following ways: 1) Mathematical models allow knowledge acquired for different systems and/ or different situations to be combined in order to provide the ultimate aim of constructing general theories. Models permit making predictions, which then must be verified by measurement. 2) Classic scientific theories begin as models and it is only long-time positive experience with some of these models which leads scientists to believe that they are more true and fundamental than others and should thus bear the name law

PAGE 18

3) In environmental sciences, models serve to sort out alternative explanations for observations made in nature which cannot be controlled in the same way as conditions for a laboratory experiment can be controlled. 4) Models should always be constructed to transfer the knowledge acquired in one system to another situation, thus enabling a deeper understanding of the processes responsible for the measured data. A model should be applied in order to design new experiments or observation programs that are critical for the testing of hypotheses. 6 5) Real observations are always superior to outputs produced by a given model. But model results can help us examine various scenarios and likely future outcomes regarding the behavior of environmental systems for which real data would never be collected as frequently and as ubiquitously as needed. Hypothesis The soil thermal regime within plastic-mulched soil beds is well-known to be influenced by interactions between the overlying atmosphere and the plastic covered soil surface. The resulting non-isothermal conditions affect water flow and redistribution. Hence, coupling of water and heat flows is necessary to model effectively both components within the vadose zone, especially the crop root zone The hypothesis here is that model description of chemical transport within plastic mulched soil beds under non-isothermal field condition s requires the utilization of coupled water and heat flows in order to provide realistic results for management

PAGE 19

7 practices such as fumigation with methyl bromide to control pathogens. Impacts of this coupled process upon chemical fate (transforrr1ations, microbial degradation etc.) and transport are expected to be particularly significant for volatile chemicals such as methyl bromide that may appear in aqueous, gaseous, and solid phases of soil water systems. Information obtained here regarding the dynamics of the fate and transport of methy l bromide during fumigation of agricultural soils should also be useful in evaluating future fumigation with fumigants other than methyl bromide, which will no longer be used in this country after the year 2001. Objectives This research was performed to model and investigate the interactive flows of water, heat, and solutes in non-isothermal soil environments (the vadose zone) in order to minimize empiricism from existing knowledge of such systems. The following objectives were formulated: 1) Develop a mathematical model describing the coupled processes of water, heat and chemical transport (2-D) for mulched soil-bed systems, where solar energy imposes non-isotherrr1al conditions. 2) Verify, validate, evaluate, and refine the model as needed for cases of both coupled and uncoupled flows of water, heat and chemicals with existing appropriate models (analytical) and published data. 3) Use model simulations to determine impacts of coupled water and heat transport due to non-isothermal conditions upon the transport of a volatile pesticide such as methyl bromide during fumigation of plastic-covered soil beds.

PAGE 20

CHAPTER2 REVIEW OF LITERATURE The rate of liquid water movement in soil is important to the study of solute transport, nutrient availability, efficacy of pesticides, and also water availability to crop root systems during agricultural production. Physical properties of the soil influence water movement in two important ways: the size distribution and structural arrangement of soil particles determine the pore space configuration through which water moves; and the interaction between the soil and water giv e s rise to water-moving forces. In water-saturated soils such as occur below groundwater tables where pores are completely filled with water the fluid is a single phase and forces originating from gravity, including the pressure gradient determine the direction and magnitude of flow The energy potential for the pore water under saturated conditions is generally positive, or greater than the atmospheric pressure. The saturated flow process for the simple isothermal case is described mathematically by Darcy's equation and Laplace's equation. Generally, under field conditions, soil pores are incompletely or partially saturated and the energy potential of water is negative (i.e. a suction), being less than atmospheric pressure. It depends not only upon position in the gravitational field but also upon sorptive forces associated with interfacial boundaries. If the unfilled pore space is filled with gases such as air, movement in the resulting 8

PAGE 21

composite system of solid, liquid and gaseous phases is referred to as unsaturated flow. This flow condition in soil is particularly important when the groundwater table occurs at considerable depth below the soil surface (below the vadose zone). Isothermal Water Flow in Porous Media 9 Under saturated conditions where the pores of the soil system are completely filled with water, the transmission coefficient (i.e., hydraulic conductivity) depends principally upon the viscosity (w(T)) and density (Pw(TJ) (both being functions of temperature, T) of the water and the nature of the porous media. The volumetric flux of water flow through soil or other porous material is usually given by Darcy's equation as [2.1] where q 1 is liquid flux, K is hydraulic conductivity, and Pis hydraulic pressure of the water. Except for the isothermal saturated condition, Darcy's equation, as originally postulated, does not apply precisely to water flow through soil. However, this equation has been used since 1856 for describing liquid flow through porous material with little reservation. Under unsaturated conditions when the soil contains both liquid and gaseous phases, the most frequently used modification of Darcy's equation involves substituting the matric-potential gradient for the hydraulic pressure gradient as the driving force. Under these conditions, hydraulic conductivity is no longer a constant

PAGE 22

10 but is a function of capillary potential as well as temperature, which yields K= [2 2] Here His total potential, his matric potential, zg is gravitational potential ( i e ., elevation), 8 1 is liquid volumetric water content, Tis temperature, k P is intrinsic permeability and g 1 is the acceleration of gravity. Under isothermal conditions the temperature dependence is eliminated and total potential gradient is the driving force. Richards (1931) derived the equation of isothem1a l flow through a porous medium by using the continuity equation (describing mass conservati o n) and Darcy 1 s equation. Assuming that the medium is incompressible, the c o ntinuity equation for water flow through a homogenous medium was given as [2.3] where tis time and p 1 is the liquid density. In very dry soils the capillary or matri c potential attains great negative values and, in the unsaturated region above a water table the capillary driving force dominates the gravitational force Isothermal theory has limited application for soils under actual field conditions where heat flow is transient in time and varies in space. It has been observed that heat is evolved at the wetting front when water infiltrates a very dry soil of fine texture (Anderson and Linvill e, 1960). Cooling of soil o c curs wh e n w a t e r is evaporated from soil (Wiegand and Taylor 1962 ) and warming occurs due t o release of thermal energy during condensation of water vapor. Absorption of solar radiation at the soil surface and chemical or biological activity within the soil may

PAGE 23

11 also result in temperature gradients Thus temperature gradients must b e included in the flow equation in order to effectively describe water flow in surface so i l characterized by transient thermal regimes Soil Temperature and Heat Flow Under field conditions the profiles of soil temperature change rapidly during a normal day. When incoming solar radiation falls on the soil surface the s oi l surfac e becomes heated, causing a difference of temperature between the surface and th e subsoil. This difference causes heat to flow downwards as a temperature wave. As the heat wave penetrates slowly downward in response to soil therrr1al properti e s moisture content and structure of the soil a phase shift occurs in the temp e rature curve with depth which leads to a time lag in the occurrence of temperature maxima at different depths in the soil. Also the amplitude of the temperature wave decreases with soil depth due to storage of thermal energy. During night time, when the soil surface cools down the lower portion of the soil profile remains warmer than the air temperature. Annual temperature fluctuations extend much de e per i nto the soil compared to diurnal changes and are also in general periodic in natur e. Soil texture has a distinct influence on soil heat flux and hence, on s o il temperature. The amplitude of the daily temperature wave decreases in th e order sand > loam> peat> clay. This order can be attributed to the greater heat capacity of the finer-textured materials and their normally greater therrr1al conductivities

PAGE 24

12 Table 2-1 Average thermal properties of major soil constituents at 20 C and 1 atm (Ghildyal and Tripathi, 1987) Volumetric Heat Thermal Material Specific Heat Thermal Conductivity Capacity Diffusivity (cal g 1 0 c 1 ) (10 3 Cal cm l s l C 1 ) (cal cm 3 0 c 1 ) (10 3 cm 2 s 1 ) Quartz 0.175 0 46 20 00 43 00 Other Minerals 0.175 0 46 7.00 15 00 Organic Matte1 0 46 0 60 0.60 1 00 Water 1.00 1 00 1 42 1 42 Air 0.24 0.00029 0.062 0.21 Sandy soils generally hold less water and usually drain faster than finer textured soils. At low water contents, air occupies much of the pore space and hydraulic conductivity of the soil becomes limited As such, thermal capacity and conductivity of soil with a low degree of water saturation is governed by the volumetric content of water, which has the highest heat capacity (Table 2-1) among the substances commonly found in soils. Because of very small thermal diffusivity air provides thermal insulation against rapid changes in temperature Three mechanisms by which heat flow occurs in soils include conduction convection and radiation (Hillel, 1980). Radiative heat transfer is of importance o nly in dry soils at high temperatures and within large pores. It is the least-important mechanism, and is often neglected. Convection can be separated into two components forced and free Forced convection is mostly negligible, with th e exception of infiltration during irrigation or heavy rain. Free convection occurs as a result of fluid density gradients Convective heat transport in the vapor phase can be partitioned into sensible and latent heat components. The sensible heat capacity o f

PAGE 25

13 soil air is very low (Table 2-1) and, as soil air flow is most often small, latent heat transport is of primary importance. Latent heat transport occurs when a fluid such as water changes phase and is transported as a new phase (such as vapor). Vaporization of liquid water requires heat (540 cal g 1 ) that is transported in the vapor phase through a vapor-phase gradient and may be released during condensation at a location different from where the liquid originated. A very simple theory of heat transfer in soils is based upon uncoupled mass and heat flows and considers only heat conduction, without regard for any other transport mechanisms which may actually occur. The thermal flux in the soil (assuming that therrnal gradients do not result in water flow) is defined from Fourier's first law of heat conduction as where qh is thermal flux and A is thermal conductivity. Invoking the principle of energy conservation in the form of the thermal continuity equation (Hillel 1980) provides BT = '7D "vT a t H 1 where DH is thermal diffusivity and Cs is volumetric heat capacity of the soil. [2.4) [2.5] Conduction of heat in soils occurs through the solid, liquid and gaseous phases. Since these phases have different heat capacities and thermal conductivities, the main complication when solving the heat conduction equation in soils is caused by deterrnining the average soil thermal properties. de Vries (1963) reported a method to calculate heat capacity of soil material by addition of the heat

PAGE 26

14 capacities of the various constituents, weighted according to their volume fractions. The heat capacity as such can be found by adding the heat capacities of the various phases present as C = (x C ) + X C + X C s.1 s .1 w w a a l. [2.6] where the subscripts s, wand a refer to solid, liquid water, and air phases respectively; i distinguishes the various materials in the solid phase; and X denotes the volumetric fraction of a given phase (de Vries, 1963). Thus, volumetric heat capacity for soil clearly depends upon water content. The thermal conductivity of a soil also depends in a rather complicated way on the composition of the soil, in particular its water content. A physical model assuming solid and gas particles to be suspended in the liquid phase, for estimating the therrr1al conductivity in soils was given by de Vries (1963) as X +L(k *x )+k*x w 1 1 a a [2 7] 1 The summation for this equation extends over the different solid soil constituents which are characterized by their thermal conductivities and shapes. The multiplication factor k in Eq. [2.7) represents the ratio of the spatial average of the temperature gradient in the soil grains of kind i and the spatial average of the temperature gradient in water. The influence of latent heat transfer in air-filled pores is proportional to the temperature gradient in these pores and can be taken into account by adding the thermal conductivity of vapor (Avap) to the thern1al conductivity of dry air (A 8 ). The

PAGE 27

15 Avap term represents the increase in gaseous-phase thermal conductivity due to evaporation and condensation of water vapor. At complete dryness and at very low water contents, the model assumes air rather than liquid to be the continuous medium. However, at this stage (i.e. completely dry soil), de Vries (1963) proposed introduction of a multiplication factor of 1.25 to Eq. [2.7] (i.e. 1.25 A) de Vries reported that errors in calculating A in this manner would not exceed 10 %, unless the soil grains have a different shape than those assumed. Wescott and Wierenga (1972) demonstrated the need for coupling heat and water movement when predicting heat transfer by conduction and vapor movement for a dry soil. They found that vapor movement accounted for 40-60 % of the heat flow in the top 2 cm of the soil profile during certain periods of the day and for 20-25 % of the total heat flow even at the 25-cm depth. Soil T hermal and Water Regimes under Mulching Mulches are used in agriculture not only because they suppress weeds and reduce water loss from the soil surface but because they also modify the thermal microclimate of the soil. Rosenberg et al. (1983) presented a good review of mulch effects on soil processes and defined mulching as the application or creation of any soil cover that constitutes a barrier to the transfer of either heat or vapor, or both. Waggoner et al. (1960) applied black plastic, translucent plastic aluminum foil, paper, and hay mulches to soil surfaces and measured the resulting alteration in surface energy balance and in soil and air temperature profiles. Black plastic

PAGE 28

16 reduced the outgoing radiation as was evidenced by the greater net radiation, whereas hay and paper (and especially paper) increased it. All mulches decreased the quantity of energy consumed in evaporation by blocking the transport of vapor out of the soil. Sensible heat generated at the surface was increased by the black plastic and the hay, and the temperature of the hay and black plastic was elevated. Black plastic increased in soil temperature, whereas hay and paper had a cooling effect and decreased the amount of heat penetrating the soil. Ekem (1967) in Hawaii found that black plastic mulch reduced water use by pineapples grown in lysimeters. Note here that the well-known practice of surface mulching (for any purpose) decreases evaporative water loss by creating a layered condition in the soil due to differential water contents. The diurnal supply of solar radiant energy at the soil surface during daylight hours and loss during night-time hours is the major deterrrlinant of the thermal microclimate of soil. The presence of mulch material at the soil surface creates a thermal layering effect (similar to the effect of soil layers within the soil). The use of nonporous mulch such as polyethylene plastic also creates a barrier to water vapor flow to the atmosphere. Optical properties of the plastic mulches play an important role in interaction of the soil surface with the overlying atmosphere. Ham et al. (1993) reported short wave optical properties of five plastic mulches used in commercial crop production (Table 2-2). These properties exhibited a wide range and thus helped explain earlier work by Waggoner et al. (1960) and Ekem (1967). Black and clear plastic mulches provide extremes in the absorbance of short-wave radiation. The five mulches

PAGE 29

described in Table 2-2 were used under bedded conditions in a field and temperatures were recorded at the 10-cm depth below the mulched surface in the center of the bed (Ham et al., 1993). 17 Table 2-2. Short-wave optical properties of five plastic mu l ches used in commercial crop production. Data adapted from Ham et a l (1993). I Mulch I Reflectance I Transmittance I Absorbance I Black 0.03 0.01 0.96 Sunfilm 0.12 0.37 0.51 Al-mulcht: 0.39 0.01 0.60 White-on-Blackt: 0.48 0.01 0.51 Clear 0.11 0.84 0.05 t:Al-Mulch and White-on-Black were Black mulch in which the upper surface was painted with aluminum and white paint, respectively. Diurnal measurements of soil temperature beneath the 5 mulches and a bare soil plot as recorded by Ham et al. (1993) for a summer day are given in Fig. 2-1. The highest temperatures were observed beneath the black mulch, which strongly absorbed shortwave radiation. Cooler temperatures were observed beneath the Al-mulch and White-on Black mulches, which had lower absorbances but higher shortwave reflectances. Although clear mulches are normally used to maximize soil heating during cooler seasons, in this experiment the investigators placed the plastic covers in intimate contact with the soil to ensure a minimal average air gap between the mulch and the soil. The small air gap greatly minimized thermal contact resistance at this location. Thus, the therrr1al insulation due to air between the plastic sheet and the soil surface was minimized in that investigation.

PAGE 30

4 2 ---------------------~ ------.... ..... ..... ..... 4 Black Sun.film Al-Mulch White C l e ar Bare S o il / / I / 8 1 2 16 .... Da y o fY e ar-1 95 20 Local Standard Time ( hour) 24 Figure 2-1. Measure m e n ts of soil temperature at the 1 0 cm depth beneath five different p l astic mulches and a bare plot near Manhattan, KS on day 195 of the year 199 1 Adapted from Ham et al. (1993). 18 Besides mulching, the geometry of the soil bed, the directional orientation of the beds, crop geometry (canopy structure) and the directional orientation of crop rows also influence the r adiation-energy e x change with the soil surface. T he amount of solar radiation reaching a horizontal unit of the surface of the earth (at a particular latitude, l ongitude, and elevation) depends upon a number of factors including the intensity of radiation emitted by the sun, astronomical considerations determining the position of the sun, and the transparency of the atmosphere (Rosenberg et al., 1 983). The sum of direct solar radiation an d diffuse sky radiation received by a unit horizontal surface is referred to as global radiation. In humid climates, clouds also play a major part in decreasing the amount of solar radiation r eceived at a location. The vegetative canopy of growing annual crops tends to provide increasing shade for the soil surface as the growing season progresses, due

PAGE 31

19 to interception of incoming solar radiation. The shadow of the growing crop results in variable radiation interception at the soil surface, thus affecting the thermal regime in the upper soil layers Coupled Heat and Water Flows The energy state of water molecules in soil changes when a temperature gradient is imposed on the system, and water moves in response to this energy gradient. For the duration of the imposed temperature gradient, an irreversible and spontaneous flow of heat occurs. Thus, heat and water flow processes interact with each other and require mutual coupling. Where simultaneous movement of more than one component (like water and heat) occurs in the system, isothermal flow equations and simple equilibrium thermodynamics cannot be used to predict or describe the water flow (Enfield, 1970). If a soil system initially at equilibrium (uniform water content with a constant water potential at con stant temperature throughout) is subjected to a temperature gradient, water will flow from a warm region to a cooler region (Gurr et al., 1952; Taylor and Cavazza, 1954; Woodside and Cliffe, 1959). When the degree of water saturation in the soil is relatively high, water flow occurs primarily in the liquid phase (Taylor and Cary, 1960). When the degree of water saturation is very low water movement occurs primarily in the vapor phase. At intermediate water contents, however, movement occurs by a combination of liquid and vapor flows (Cary and Taylor 1962a, 1962b).

PAGE 32

Possible magnitudes of thermally-driven flows have been summarized by Cary (1966) for various soils and commercial materials. Vapor flow in soils in response to them1al gradients of 0.01 C were observed to vary from about 0.4 to 2.0-mm day 1 the same flow rates that would require pressure-head gradients of 0.05 to 2.50 m of water. Since vapor flow out of wet soils (evaporation) may range from near zero to perhaps 12 mm day 1 depending on climatic conditions, one can see that temperature-induced vapor flow should not be ignored. Philip-de Vries Theory (PDT) 20 Approximately 40 years ago, in an attempt to describe the phenomenon of simultaneous flow where water is moved by a thermal gradient, Philip and de Vries (1957) and de Vries (1958) proposed theory and equations to describe coupled heat and water transfer in porous media, in terms of classical mechanisms of vapor diffusion and liquid movement by capillarity accounting for heat flow and its resultant effects. Philip and de Vries (1957) were apparently the first to include the interaction of vapor, liquid, and solid phases and the difference between average temperature gradients in the air-filled pores and in the soil as a whole. Taking these factors into account, an approximate analysis wa s developed which would predict orders of magnitude and general behavior that was in satisfactory agreement with the experimental results of Gurr et al. (1952), Jones and Kohnke (1952), Rollins et al. (1954), Smith (1943), and Staple and Lehane (1954). Philip and de Vries (1957) separated the isothermal and thermal components of vapor transfer to show the effect of relative humidity. They separated the vapor

PAGE 33

21 flux into two components: 1 ) one due to temperature gradients; and 2) another due to moisture gradients. I n the saturation range where liquid transfer occurs, the liquid flux for water was separated into components caused by temperature gradients, water gradients and gravity. Philip and de Vries (1957) suggested that interaction between vapor and liquid phases occurs as a series of evaporation and condensation steps coupled with liquid flow through so-called 'liquid islands'. When a temperature gradient is applied across a soil, water evaporates at the warm end of the pores and diffuses as vapor, due to a vapor density gradient, towards the cooler end where condensation occurs. This process changes the curvature of the liquid menisci of a liquid island, which continues until capillary liquid flows through the island, produced by the growing difference of both curvatures, equals the rate of condensation and evaporation. The overall effect is a decrease in the tortuous path length and an increase in the cross-sectional area available for vapor diffusion. Philip and de Vr i es (1957) presented the simultaneous equations of continuity describing water and heat transfer under the combined influence of water and temperature gradients as ae at [2.8] where Dr w is ther111al water diffusivity, D 8 w is isotherrr1al water diffusivity, k is the vertical unit vector, D 8 v is isothermal vapor diffusivity, and L is the latent heat of vaporization. Note that the existence of soil-water hysteresis may, however, invalidate the use of Eq [2.8] (de Vries, 1 958).

PAGE 34

22 Application of Philip-de Vries Theory Various applications and verifications of PDT have been documented. Barnes and Allison (1984) applied PDT for non-isothermal conditions to develop a model for distribution of deuterium and 18 0 in dry soils. The model was later verified by the same investigators (Barnes et al., 1989), with the experimental data showing satisfactory results. However, they emphasized that the assumption of constant tortuosity for both liquid and vapor movement is questionable. They reported that, in moderately dry soils the liquid tortuosity factor is probably reduced, and vapor diffusion may be enhanced by the presence of 'liquid islands'. Scanlon (1992) used a model developed with PDT to evaluate liquid and vapor water flow in desert soils of Texas using 36 Cl and 3 H tracers. Observed and predicted potentials with the model showed good agreement. Scanlon (1992) concluded from simulation results that them1al vapor fluxes were downward-directed during the day and upward at night during both winter and summer. Net vapor flux was also downward during the day and upward at night. Bach (1992) conducted isothermal and non-isother1nal laboratory experiments to investigate the significance of non-isothermal water flow and to examine theoretical and numerical descriptions of the transport processes. He concluded from sensitivity analysis that, with adjustment in Ch (temperature coefficient of the matric potential), PDT provided an adequate description of the non-isothermal transport processes.

PAGE 35

23 Many other applications of PDT to non-isothermal water movement have been observed to describe and simulate transport processes (Benjamin, 1989; Chung and Horton, 1987; de Silans et al. 1989; Mahrer et al., 1984; Milly, 1984; Nassar and Horton, 1992; Noborio, 1 995; Sophocleous, 1979; and Yakirevich et al., 1997). Cassel et al. (1969) concluded t hat PDT showed acceptable agreement with observed soil water movement. In order to account for soil inhomogeneity and hysteresis of the moisture retention process, Milly and Eagleson (1980) modified the PDT by using soil matric pressure-head gradients instead of soil moisture gradients. These authors also included effects of the heat of wetting on transport processes. A finite-element one-dimensional formu l ation of the governing equations showed that the model performed well in simulating highly coupled, hysteresis-affected, or very nonlinear problems. Later, Milly ( 1 984) used this modified form to study soil-water evaporation from silt loam and sandy soils and reported that soil water evaporation was generally more sensitive to isothermal than to thermal vapor diffusion. Published evidence thus shows that the mechanistic PDT appears to be accurate for coupled heat and water transfer in soils. This mechanistic approach to formulate expressions for the fluxes of water and heat is based upon concepts derived from fluid mechanics, mass diffusion and heat conduction.

PAGE 36

24 Solute Transport during Coupled Heat and Water Flow Simultaneous flow of solutes and water occur in most soil-water systems under field conditions. Salt concentration gradients may also be a major contributor to water movement in soils located in arid climates. Salt concentration gradients often occur within the uppermost soil profile, where evaporation selectively removes water at the soil surface and result in an accumulation of salt. Kemper and Rollins (1966) reported that, for experiments involving Wyoming Bentonite clay, soil moisture suction heads needed to be greater than 50 m of water before the osmotic potential gradient would cause significant soil water movement. Low (1955) observed that, when the solute was completely restricted by the porous medium osmotic potential differences were as effective as pressure potential differences in causing water to move through clay under isothermal conditions. In studying water flow through a porous ceramic filter, Jackson (1967) reported that soil water flow could deviate from Darcy-type flow due to the presence of salt. In laboratory studies on clay loam and sandy loam soils, Oayyum and Kemper (1962) reported that diffusive movement of water occurred to the salt-bearing surface at soil moisture contents less than one-half of field capacity, and that viscous flow occurred at soil moisture contents higher than one-half of field capacity in closed soil columns. Kemper and Maasland (1964) found that salt sieving increased with decreasing pore size of the soil and decrea sing valence of the saturating cation. Letey and Kemper (1969) introduced a theory to describe simultaneous movement of

PAGE 37

25 water and solute in soil. The theory included an equation to describe water flow and another equation to describe solute transfer. The equation for solute transfer implied that the movement of solute is controlled by Fickian diffusion and salt sieving. In summary, temperature and osmotic gradients can induce water fl o w within soils. As water moves, heat and solute are transported, thus altering the driving forces (gradients). Interactions among thern-1al, soil moisture, and osmotic effects exist, and they differ for the liquid and vapor phases (Nassar and Horton 1989b) Nassar and Horton ( 1989b) extended the PDT describing water transport under the combined effects of matric and osmotic pressure head gradients in unsaturated, non-isothermal, saline soil. Their theory was based upon the concept of viscous flow of liquid water under the influence of capillary, adsorptive, and osmotic forces and on a concept of vapor movement by diffusion. They presented two equations for vapor and liquid transfer as [ 2 9) and [2 10] where qv is water vapor flux Drv is thermal vapor diffusivity D r 1 is thermal liquid diffusivity, Dev is isothermal vapor diffusivity D 8 L is isothermal liquid diffusivity D ev is vapor diffusivity due to solute concentration, D c 1 is liquid diffusivity due to solu t e concentration, and c is solute concentration.

PAGE 38

fill){ as 26 Equations [2.9] and [2. 1 0] were summed in order to describe the total water qw = -D;VT D 0 wV8 + D c VC Kk [2.11] PL where qw = q 1 + qv is total water fill){, Drw = Drv + DrL is thermal moisture diffusivity, D 8 w = Dev + D 61 is isothermal moisture diffusivity D e = D ev + D c 1 is moisture diffusivity due to solute concentration, and 0 = 0 v + 0 1 is total volum e tric water content, where 0 v is water vapor content (expressed as equivalent liquid volume). Applying the principle of heat conservation, an equation for transient heat transfer combining the effects of matric pressure head, osmotic pressure head and temperature was formulated. For developing the solute-transport equation ( non interacting solute transfer), Nassar and Horton (1989b) included the mechanisms of molecular diffusion hydrodynamic dispersion, the1111al diffusion mass flow of water (convection), and salt sieving. Nassar and H orton (1989a 1989b) conducted experimental and numerical studies of coupled heat and mass transport in closed soil columns with three different soil textures. Experimental observations for unsaturated conditions showed water accumulation within the low temperature region of the columns but solute accumulation within the high-temperature region. Predicted and observed results were in good agreement when initial soil-water contents were relatively large and initial solute concentrations were low. However, for small initial soil-water cont e n t s the model underestimated soil-water contents close to the low-temperature boundaries and overestimated than near the high-temperature boundaries. For high

PAGE 39

27 initial solute concentration, the theory underestimated liquid transfer toward the hot boundary temperatures. Unfortunately, the theory proposed by Nassar and Horton (1989b, 1992) is restricted to homogenous soils and cannot be used in its presented form for heterogenous soils. The water flow equation is written in terms of water content and thus does not allow simulation for multilayered soil. Furthermore, the model is primarily restricted to soil conditions of water unsaturation. Temperature Effects on Solute Transport The rates of most chemical reactions increase with temperature, changing their equilibrium constants (Clark, 1996). Clark further states as a rule of thumb that the reaction rate doubles for each increase in temperature of 10 C. Generally, the understanding of a chemical reaction is not considered complete until the sensitivity of the reaction to temperature is understood. Many chemical studies thus include experiments at different temperatures. Daniels and Alberty (1961) and Clark (1996) discussed the influence of temperature on reaction rate in the following way. Kinetic data over a range of temperatures can usually be represented by use of an empirical equation first proposed by Arrhenius E a k = F e R u T r f [2 12] where kr is reaction rate, F 1 is a frequency factor, Ea is activation energy, and Ru is the universal gas constant Set in logarithmic forrr1

PAGE 40

28 Ea log kr = 2 303 R T + log Ff [2.13] u According to Eq. [2. 1 3] a straight line should be obtained when logarithm of the rate coefficient (kr) is plotted against the reciprocal of the absolute temperature (T ). T he Ea can then be estimated as -2.303 Rux the slope of the line. The activation energy may be envis i oned as an energy barrier that must be surmounted for the reaction to proceed An increase in EB indicates a slowing down of the reaction (and vice versa). The EB thus can be determined by measuring the reaction rate at different temperatures. The influence o f temperature on chemical processes has also been described V by Stumm and Morgan ( 1 981) through use of the Arrhenius equation. Simunek and Suarez (1993) described a function with some modification to account for relative change in rate coefficients with temperature as f(T) = e [2.14] where Tr is the reference absolute temperature and f (T) = 1 at T = Tr. Cohen and Ryan (1989) and Cohen et al. (1988) used the relationship given by Bird et al. (1960) to describe the temperature dependence of diffusion coefficients. T he chemical molecular diffusion in air as a function of temperature was estimated as D (T) = D (T) a a r 1.823 [2.15] where D 8 (Tr) is the chemical molecular diffusion coefficient in air at the reference temperature. Temperature variation of the chemical molecular diffusivity in liquid

PAGE 41

29 water was estimated as D (T) = D (T) w w r [2.16] where w is the viscosity of liquid water. Summary The PDT was proposed 40 years ago to mathematically describe coupled water and heat flows in soil. It has been well-verified and applied by a number of researchers to date. The theory was extended by Milly and Eagleson (1980) to make it applicable to heterogenous media, by replacing the water content-based formulation with a matric potential-based formulation of flow equations This advancement was a big step, as it made the theory suitable for the solution of field problems. Nassar and Horton (1989a, 1992) extended the theory to include salt effects on water flow and provided insight into additional coupling. But, this theory cannot be used to solve field problems, because Nassar and Horton used the original PDT (water content-based) formulation for modification. Relatively little evidence exists that the potential of PDT and its modification by Milly and Eagleson (1980) has been fully exploited to use the improved estimates of water content/flow, and temperature information. Several researchers (Barnes and Allison, 1984; Barnes et al., 1989; Nassar and Horton, 1989a, 1989b; Noborio, 1995; Scanlon, 1992; and Yakirevich et al., 1997), however, utilized PDT by making use of improved estimates of water content/flow when exploring solute transport

PAGE 42

30 These investigations, however did not attempt to use the improved estimates of temperature by inclusion of its effect on solute transport. A need, therefore, exists to exploit the full potential of PDT in solving field problems where temperature could be highly transitional both in space and time. Furthermore, where chemical (especially volatile-compound) transport is also of concern, a need exists to use both the improved estimates of water content/flow and of temperature provided by PDT. The work reported here utilizes PDT in totality to develop a 2-D numerical model for solving problems involving the fate and transport of volatile chemicals that may exist in all three phases (viz. liquid, solid, and gas) of soils. Details for the developed model are presented in Chapter-3.

PAGE 43

CHAPTER3 MATHEMATICAL MODEL The model developed here considers the soil to be a heterogenous isotropic medium composed of solid, liquid, and gaseous phases. The latter is assumed to be a non-interacting binary mixture of water vapor dry air, and volatile chemicals. Other general assumptions in the model include 1. The solid matrix is undeforrnable and immobile. 2. The liquid water is incompressible. 3. The density and viscosity of liquid water are not influenced by dissolved constituents. 4. Hysteresis in the relationship between capillary-pressure and hydraulic conductivity is neglected. 5. Liquid water and water vapor are in local thermodynamic equilibrium. 6. Water vapor and other gases behave as perfect gases and advective transport of these components is neglected. 7. Local thermal equilibrium exists among all phases. 8. Radiative heat transfer is negligible. 9. Thern1al and matric potential gradients interact to provide heat and water flow in the porous medium 31

PAGE 44

More specific assumptions are stated as required. Additional details concerning the equations governing flow and transport are presented in the appendices. Liquid and Vapor Water Flow in Soil 32 The original water content-based formulation of PDT was modified by Milly and Eagleson (1980) to provide Pv aeL 1---+ p 1 ah Pv 1[3.1] where P v is vapor water density, 0 a is the volumetric air content, and Dhv is isothermal vapor diffusivity. Equation [3.1] describes both liquid water and vapor water flow in response to gradients of matric potential and thermal energy. Heat Flow in Soil Coupled heat flow in the soil in response to gradients of t emperature and pressure-head has been described by (de Vries, 1958; Milly and Eagleson, 1980) ar at (3.2]

PAGE 45

33 and C Cd+ CL p Le L + CV p Ve a Hl = l L O + CV ( T TO) J ea H2 = (cLpL Cvpv)(T To)PvLo [3.3) where Cd is volumetric dry heat capacity of the total medium; CL and Cv are specific heat of liquid water and vapor water; respectively; and the subscript O implies the magnitude at a designated reference level. Transport of Multiple Species of Solutes in Soil V V Foll owing the approach of Simunek et al. (1992) and Simunek and van Genuchten (1994), in which solutes can exist in all three phases (liquid, solid, and gaseous) with non-equilibrium transport, the governing transport equations can be stated as [3.4) aeLc. aps. ae g. ___ .1 + __ .1 + a .1 = at at at * * + ,ps+ eg+ C + PS + s,.1 s,1 i g,i g,.1 a i L,.1 L i l s,.1 i 1 [3.5) wheres represents adsorbed concentrations; g is the gas concentration; L, s, and g are the first-order rate constants for solutes in the liquid, solid, and gas phases, respectively; y L y 5 and y g are zero-order rate constants for the liquid, solid, and gas

PAGE 46

34 phases, respectively; pi s the soil bulk density; dis the dispersion coefficient for the liquid phase; and V is the diffusion coefficient for the gas phase. The subscripts L s and g correspond with the liquid, solid, and gas phases, respectively; the subscript i represents the i th chain number; n 5 is the number of solutes involved in a chain reaction; and L 5 and g are first-order rate constants for solutes in the liquid solid, and gas phases, respectively, providing the connection between individual chain species. The nine zeroand first-order rate constants in Eqs. [3.4) and [3.5) can be used to represent a variety of reactions or transformations including biodegradation, volatilization, precipitation, etc., and the effect of temperature on these coefficients is included through use of Eq. [2.14]. Equation [3.4) alone is sufficient for a single solute that does not undergo transforrr1ation to related chemical species in a 'chain' reaction. Both Eqs. [3.4) and [3.5] are required when the primary solute undergoes transformations to other species or when multiple unrelated solutes are present in the soil system In order to solve the solute-transport equation, a knowledge of liquid water content, air content, and volumetric flux is required. Solution of the coupled water and heat flow equations (Eqs. (3.1] and [3.2)) provides this information Energy Balance at the Soil/Atmosphere Interface Interception of solar radiation interaction with the overlying atmosphere and loss of water through evaporation is described via an energy-balance equation for the soil surface

PAGE 47

35 ( 1 -p )R + R R = H + LE + G s g LA LS s [3.6] where P sis soil surface albedo, Rg is measured global radiation (taken as positive downward), RLA is longwave sky irradiance (positive downward), RLS is longwave radiation from the soil surface (positive upward), H s is sensible air heat flux (positive upward), LE is latent heat of the evaporative flux (positive upward), and G is soil heat flux (positive downward). The solution of Eq. [3.6] provides inforn1ation for soil surface temperature and atmospheric evaporative demand (excluding transpiration) required to specify the top boundary condition of the solution domain. This equation is strictly valid only for a soil surface without a mulch cover. The presence of a mulch cover alters interaction of the soil surface with the overlying atmosphere (as discussed earlier). An approach based on the work of Hares and Novak (1992) and Harn and Kluitenberg (1994) was adapted for conditions where the soil surface is covered with a plastic mulch such as commonly used in soilbed systems for vegetable production in Florida. The energy balance for a plastic mulch can be defined as Rn + H C = 0 m S C [3.7] where Rnm is net radiation of the mulch surface and C 0 represents conduction or free convection between the mulch and free soil surface. Energy balance for the soil surface beneath the thin plastic mulch can be defined as Rn + C + G = 0 S C [3.8] where Rns is net radiation of the soil surface. These energy-balance equations

PAGE 48

36 neglect heat storage within the thin plastic mulch, and the potential impacts of cuts and p l anting holes in the mulch is not considered. Thickness for such plastic mulches typica l ly ranges from 0.02544 to 0.1526 4 mm. Net radiation of t he plastic mulch is given by (Ham and Kluitenberg, 1994) Rn = a R (1 + p i: p )e e k a T :ky(l + p 2 ~ r i: ( 1 e ) ) + p :re e o T 5 4 + m m g m s m s y m, ir s m s (3.9] where Ts1
PAGE 49

37 Sensible heat transfer between the plastic mulch and the atmosphere is a function of the aerodynamic properties of the surface and the temperature gradient between the plastic and the boundary layer, as defined by r a,h [3.12] where Ta is air temperature and ra ,h is the aerodynamic resistance to heat transfer. Heat transfer between the mulch and the underlying soil surface is estimated using the equation T T m s [3.13] where re represents a thermal contact resistance at the mulch-soil interface and mainly depends on the air-gap thickness between the soil surface and the mulch. This determines whether the heat transfer is conductive or convective. Ham and Kluitenberg ( 1994) described the estimation procedure and presented representative values of r e for different air gaps between the soil surface and the mulch for various plastic mulches. Net radiation of the soil is dependent upon the global and longwave sky irradiance transmitted through the mulch, radiation emitted from the mulch and soil, and multiple scattering of radiation between the soil and mulch as given by (Ham and Kluitenberg, 1994 ) ( ) ( 4 4 4 ) n = 1 p -r p R + P.1r t e k a Tsky + e a Tm + p e a Ts e a T s s m s s m, .1r s y m m, .1r s s [3 .1 4] Given the subsurface temperature near the surface at depth z, ground heat flux can be approximated at the future time step p + 1 (p being the current time) as (Horton

PAGE 50

38 and Chung, 1991) G = A* f::;,z +C s 2 !::;, t [3.15] where T 1 is soil temperature at the first node below the soil surface and (TsP+ 1 T :)/ fl t is the rate of change in temperature with time at the surface between time steps p andp+ 1. Initial and Boundary Conditions Equations [3.1)-[3.5] require supplementation with appropriate initial ( IC ) and boundary (BC) conditions to describe a specific simulation scenario. The IC in general (at t = 0) are specified ash (x, z, 0) = h i (x, z), T (x, z, 0) = T i (x, z), and c (x, z, 0) = c i (x, z) respectively for Eqs [3.1), [3.2), and [3.4), (3.5]. Here h j, T i, and ci are known values of the state variables at the beginning of the simulation; and x and z represent the horizontal and vertical locations, respectively. Boundary conditions may be defined as a specified value of matric potential/or temperature and/or solute concentration (Dirichlet-type BC), a specified water/or temperature and/or solute flux (Neumann-type BC), or a combination of the two (Cauchy-type BC). The top boundary (i.e., the soil surface) in the system under consideration is a special case of the Neumann-type BC for water (evaporative flux), dependent upon the overlying atmosphere. In general, the boundary condition s depend on the specific system being investigated.

PAGE 51

39 Input Parameter Requirements Soil input d ata i nc l ude sand, silt, clay and organic matter contents; bulk density ; saturated hydraulic conductivity ; and soil hydraulic paramet e rs which describe the water retention curve ( Eqs. (3.7]-[3.8]). For the soil/atmospheri c boundary conditions, weather data such as average air temperature air temperature daily amplitude, average dew-point temperature, dew-point temperature daily amplitude wind speed, surface roughness length, and day-length are required inputs as well Soil hydraulic properties are estimated following van Genuchten ( 1980 ) as and 8 = L 0 0 8 + s r r [1 +(a l hj) rt ] rTJ 1 ; r1) = 1ri K( 8 ) = K (1-( a h l ) 11 1 +(ah ) rt rTJ) 2 L sat Ll + (a h I )rt rTJ /2 e [3 16] 0 -0 L r 0 -e s r [3. 17 ] where e r is residual water content 0 s is saturated volumetric water content l b I i s the absolute value of the matric potential, and K sa t is the saturated hydraulic conductivity. The parameters a '1. and r1) are empirical constants used to fit th e moisture retention curve. The temperature dependence of matric potential (h ) w a s described as (Philip and de Vries 1957) dh dT h d os 0 5 dT => = C h h and hydraulic conductivity (K ( h)) as (Constantz, 1 982) [ 3. 18 ]

PAGE 52

ref w K(h) -T w p L, T K ( h) ref PL, r ef 40 [3 19] where a s is surface tension of water; Ch ( = -0.0017 K1 by Milly and Eagleson, 1980) is temperature coefficient of matric potential; wret and wr are dynamic viscosities at the reference temperature and soil temperature, respectively; P iret and P L rare the densities of water at reference temperature and soil temperature, respectively; and Kret (h) is hydraulic conductivity at the reference temperature. The dependence of e L upon temperature can be deterrrlined as (Milly and Eagleson, 1980) [3 20] Water vapor density is g i ven by Marshall and H olmes ( 1 979) as [3.21] where


PAGE 53

and with respect to T gives = Dl *D2 where Dl = D2= acp MW _s D3+q> g h--D3 ; DJ = exp BT s a R T 2 u Variation in saturated vapor pressure with temperature is given by ML w exp --Ru 1 1 --41 [3.25] M h w ga R T u [3.26] The vapor diffusion coefficients are determined as (Milly and Eagleson, 1980) D 1 e ap a a a v p 1 ah D f a [3.27] with [3.28] where D 8 is a molecular diffusion coefficient of water vapor in air, 't 8 is an air-phase tortuosity factor f is vapor path correction for liquid islands, 8 r is total porosity e k is the water content at which liquid continuity breaks during desaturation, and ( = 1.7) is a temperature-gradient enhancement factor for vapor. Specific heat is estimated from the methods of deVries (1963) as C = X C +X C +X C +X C +Ce +Ce s s st st C C om om L L V V [3.29] where Xs, Xst, X e and X 0 m are the volume fractions of sand, silt, clay, and organic matter, respectively; and cs, est cc, c 0 m, Cv are the volumetric heat capacities of sand, silt, clay, organic matter, and water vapor, respectively. The specific heat of water vapor is given by Noggle (1985) as

PAGE 54

C = V 30 54 +0 01029T 0 0180015 42 [3.30] The apparent thermal conductivity can be estimated as a harmonic mean of the soil constituents (de Vries, 1963) L kiX.'A.. A. -1 i i 1 A 1 25 i=s,st ,c,o, e k* L 1 + i -1 g i L 3 A J ki X i j=l, .. ,4 a i=s, st, c, o, e [3.31] where A 1 represents the respective ther1nal conductivities of the soil con stituents, g 1 are shape factors for the soil constituents, and A 8 is the thermal conductivity of air. The incoming global radiation for the soil/atmospheric boundary condition must be partitioned into a flux-density distribution for daylight hours. Chung and Horton (1987) suggested a sine wave distribution R = g IT Rdg -sin 2 DL t-SN+ DL 2 IT DL [3.32] where Rdg is daily global radiation DL is day length, SN is solar noon, and t is the time of day. The air and dew-point temperature distributions can also be expressed as sine functions Ta = T +A sin a a 2ITt m --+IT 86400 2ITt m ---+IT 86400 [3.33] where the bar indicates average temperature; T 8 and Td are air and dew point temperatures, respective l y; A 8 and Ad are air and dew point temperature amplitudes, respectively; tm is the time from midnight ; and the maximum temperature is assumed to occur at solar noon.

PAGE 55

43 The soil surface emissivity, e s, and albedo, P s, are described by the method of van Bavel and Hillel ( 1975 1976 ) as es= 0 9+0 188L Ps = 0 25 for eL~O lO Ps = 0 35-eL for 0 10-<8L-<0 25 Ps = 0 10 for eL~0 25 [3.34]

PAGE 56

CHAPTER4 MODEL VERIFICATION AND VALIDATION Numerical solutions to differential equations require testing to ensure that numerical instabilities have not been introduced into the solution due to round-off or truncation errors and that the solution converges on a reasonabl e estimate of the exact solution. A numerical scheme is stable if, with increasing time, the difference between the numerical solution and the exact solution tends uniformly to zero at all nodes (Benjamin, 1989). Convergence is satisfied if the numerical solution approaches the exact solution as spatialand temporal-step sizes approach zero. Proper testing also ensures the correctness of mathematical representation for the physical system be ing modeled and the validity of any assumptions made in the model. Testing of a model may be achieved in multiple ways One such way is by comparing the model results with analytical (exact) solutions for greatly simplified conditions, and another is by comparing the model results with observed data from carefully designed experiments. Unfortunately, no analytical solution to the governing equations for coupled heat and water transport is known to exist. Hence portions of the model were tested with observed data and analytical solutions for simplified conditions, where available. 44

PAGE 57

45 Isothermal Water Flow Model results were compared with data for transient one-dimensional, isothermal water flow through sand as reported by Haverkamp et al. (1977) with vapor flow ignored. Ha v erkamp et al. (1977) reported Ksat, e s, e r, and soil bulk density to be 34.0 (cm h 1 ), 0 287, 0.075, and 1.66 g cm 3 respectively. An analytical model by van Genuchten ( 1 980) was fitted to the soil water content versus matric potential data in order to provide optimum values for the hydraulic parameters a and rt A uniform initial water content of 0. 1 0 was specified and water content at the 0.70 m depth was kept constant at this value. A constant water influx of 13.69 cm h 1 was specified at the top boundary. Results of the numerical simulation and experimenta l data for different times, are shown in Fig. 4-1. The numerical solution was stable and converged upon a close approximation of the experimental values. The model accurately described the movement of the water front in the column of sand. ~.05 0. 1 0 0.15 0.20 0.25 0.38 0 b se rved 10 Es timat e d -1 0 0 0 0 20 0 .2 hr 20 30 30 Figure 4-1. Observed and computed 0.4 h r distributions of water content in a '._/ ro 40 40 ... sand column during constant flux fr c 1,,-,j -50 -50 infiltration. 0 6 hr 0 -60 -60 0 0 0 0 0 0. 8 hr 70 70 8 8.05 0. 1 o 0. 1 5 0.20 0.25 o.3J 0 W a ter C ont e nt ( cm 3 c m 3 )

PAGE 58

46 Uncoupled Heat Flow Next, the numerical solution for heat flow was verified with field-measured temperature data (Horton and Chung, 1991) for a Nicollet clay loam soil profile with a bare surface located near Ames, I A. The effect of water flow was ignored. Representative values of heat capacity and thermal conductivity for loam soil were used. Initial conditions as specified by Horton and Chung (1991) were used in the model and temperature at the 60-cm depth was held constant at 20 C Temperatures observed for the soil surface were used in the model to provide the surface boundary condition. One-dimensional numerical results (Fig. 4-2) for the 5and 15-cm soil depths provided satisfactory representation of temporal temperature profiles for each depth. 40 2 0 4 0 60 8 0 1 0 0 1 2so A Obse r ve d 5 c m 3 6 Es t ima t ed 5-cm 36 0 O b serve d 15 c m ,,,-... [? 32 ---Es t i m a t e d 15 cm 32 A "--" (1.) Z. 2 8 2 8 ,,... 1/j, $.-4 6 0 \ (1.) 9/ a 2 4 I '-, 24 0/ 0 'I 0 (1.) I I cl 2 0 lJ. A t,. / 20 I p 1 6 1 6 0 2 0 40 6 0 8 0 1 00 120 Tim e ( h o ur s) Figure 4-2. Observed and computed temperatures for two depths in a field soil.

PAGE 59

47 Coupled Water and Heat Flow In order to test the capacity of the numerical solution to predict coupled heat and water transfer, simulation using the experimental data of Nassar and Horton (1989a) was performed. Nassar and Horton (1989a) conducted laboratory experiments of heat and mass transfer with controlled temperature gradients in sealed columns (14 cm in length) of silt loam soil with initially uniform water content (0.117) The column ends were subjected to different temperatures (19.03 C and 9. 36 C) in order to create a temperature gradient. The experiment was run for a period of 31 days. The observed mean value of saturated hydraulic conductivity for the soil was 0.954 cm hr 1 with a corresponding soil water retention curve having been determined by Nassar and Horton ( 1989b). The van Genuchten (1980) model was fitted to the soil water retention curve in order to obtain optimum hydraulic parameters for the soil (a = 0.03137 cm1 fl = 1.33308, e s= 0.65, e r= 0 05) for silt loam soil. A numerical simulation was carried out for 31 days using zero flux boundaries at both ends for water flow and Dirichlet boundaries for heat flow. Observed and predicted water contents are given in Fig. 4-3. Good agreement was noted everywhere except in the vicinity of the cold end of the soil column with elevated water content. The predicted total water content in the column was slightly higher than the observed value Since the computed mass balance error was less than 1.0 % in the numerical simulation, the possibility exists for a small systematic error in the measurement of either initial or final water contents during the actual experiment.

PAGE 60

0.30 ....-------------------, r-;i s 0.25 u C"'l S 0.20 u '-" 5 0.15 1= 0 U 0.10 $-c Q) 0.05 0 000 2 ------Estimated \. 0 ....... o Observed Initi a l 0 -0 ---oO -0-0-0-0 4 6 8 1 0 1 2 Distance (cm) 14 Isotherrr1al 1D Solute Transport during Steady Water Flow 48 Figure 4-3. Observed and computed wat er content after 31 days under the influence of both matric head and thermal gradients. The model component for solute transport was tested against data reported by van Genuchten (1981) for the movement of a pulse of boron (H 3 B0 4 ) during steady water flow through a column of Glendale clay l oam. The solute-transport parameters (Sim un ek and van Genuchten 1994) used were: d = 2.042 cm 2 hr 1 e s= 0.445, p = 1.222 g cm 3 k d = 1.14 cm 3 g 1 p = 1 0 11 = 0.0, = 0.47 and w = 1 .3 3x10 2 The initial concentration was zero and a pulse of aqueous solution containing a boron concentration o f 20.0 mmolc 1 was applied at a constant flux of 713.0 cm ru 1 for a duration of 121 44 hr Assuming two-site chemical non equilibrium sorption, a simulation was run for 20 pore volumes of effluent A comparison of numerical and experimental results (Fig. 4-4 ) verified that the model described the transport process with satisfactory accuracy.

PAGE 61

49 1 0 Es tim a t e d Ob serve d c= 1 0 0 0 8 +-J ; .;j 0 8 0. 6 0 u (l) 0 4 Co > 0 o'b 0 .2 > Oooo 0 0 00 4 8 1 2 1 6 20 Pore Volume Figure 4 -4. Observed and computed break through curves for boron. Isothermal 2-D Solute Tr ansport during Steady Water Flow The mathematical accuracy of 2-D numerical simulations for isothermal steady water flow, was verified employing a simple analytical solution. Solute (non volatile) transport in a homogenous, isotropic porous medium during steady-state unidi r ectional groundwater flow can be written as R ac at ac -v--Rc z az L (4 1] where Dr and DL are transverse and longitudinal dispersion coefficients, respectively, and vz is average pore-water velocity in the downward vertical direction. The initially solute-free medium i s subjected to a solute source, C 0 of unit concentration. The source covers a length 2a along the inlet boundary at z = 0, and is located symmetrically about the coordinate x = 0 The transport region of interest is the half-plane (z < O; 00 x 00 ). The boundary conditions can be written as

PAGE 62

50 c(x 0 t) = C 0 -a~x5;,a ; c(x 0 t) = 0 other values ofx lim ac = 0 ; z-oo az 1 ac O [4.2] imx-oo az An analytical solution to the above transport problem is given by Leij and Bradford ( 1 994) as c( x,z t) Cz t / R 2 z2 vz J exp vz 3/2 0 exp z L R + 't '(l ll 4 (1tDL) 1 12 2D L 4D )II 4D L 'tl ll 0 L erf a x +e rf a +x d-r 2(D r 'tw)l /2 2(D X ) 1 12 lll T lll 0 ------------'----...-::---------, -10 SOURCE (a = 50-m) Analytical Numerical -g 10 ...Q 3 0 ~==========;;;:::..-fr 0 -40 '---o ncentration contour level = 0 1 60 0 20 40 60 80 Di stance from mid-source (m) [4.3] Figure 4-5 Lo cat i o n of solute concentration fronts for three times as estimated num erically and with the analytical solu t io n. The numerical simulation assumed a source-width of 2a = 1 00 m. Because of symmetry, computations were carried out only for the half plane where x 0. Figure 4-5 shows the es tim ated solute concentration fronts (0. 1 ) at selected times

PAGE 63

51 and various values of the parameters as used in the computations. Results predicted by the numerical model were similar to those given by the analytical solution. Isothermal 1-D Solute Transport with Nitrification Chain during Steady Water Flow To verify the numerical solution capacity in simulating multiple linked species in a chain, comparison was done with results generated from an analytical solution published by van Genuchten (1985) for 1-D transport of solutes involved in sequential first-order decay reactions. The analytical solution holds for a homogenous, isotropic porous medium with steady-state uni-directional groundwater flow The solute transport Eqs. [3.4] and [3.5] for this case reduces to [4.4] and ac R i .l at a 2 C ac D L .l .l + R R -V -L 1 1 C 1 L . C ; ax2 X ax ,.1 .1 1. ,.1 .l .l i = 2 3 [4 5] A simulation was performed here that involves a three-species nitrification chain NH 4 + ( i = 1 ) N0 2 ( i = 2 ) [4.6] and follows van Genuchten (1985) and Cho (1971). A NH 4 + solution was applied to an initially solute-free medium. The input transport parameters used for the simulation were V = X 1 .00 m h 1 0.18 m 2 h 1 0.01 h1 0.10 h 1 R = 1 c = l 2 0 [-] 1.0[-] 1.0 [-] 0.0[-] 1.0 [-]

PAGE 64

52 The simulation was conducted for 200 hours. Figure 4-6 shows the concentration profiles for all three solutes at time 200 hr, calculated both numerically and with the analytical solution from the CHAIN code of van Genuchten ( 1985) The numerical solution duplicated the analytical results almost exactly. l.O
PAGE 65

53 For an application of finite source (g 0 ) at one end with zer o flux BC and zero gradient BC at the other end, an analytical solution is (Crank, 1975) ( ) 9 t 1 f a + 2 n 1 x f a 2 n 1 + x g x, t = e g er ---+ er 2 0 n=-oo 2Jv gt 2Jv gt where a is the initial source length. Prior to deriving the analytical solution the transforn1ation g = g* exp(-gt) was used to simplify Eq. [4.7) to ag = v g ag at ax2 A numerical simulation was conducted for 5 hours with g 0 = 50 moles, a = 1 cm, IJl = 47 .52 cm 2 hr 1 andg = 0.055 hr1 The numerical solution duplicated the results of analytical solution at 1 3, and 5 hours (Fig. 47). 20 ..------------------------, 0 0 10 0 ------0 -20 30 40 Di stance (cm) Analytical 1 hr Numerical 1 hr Analytical 3 ln Numerical 3 hr Analytical 5 hr Numerical 5 hr 50 60 [4.8) [4.9) Figure 47. Analytically determ1ned and numerically computed concentration profiles at 1, 3, and 5 hours.

PAGE 66

54 Summary All major components of the numerical model were tested for mathematical accuracy, the adapted numerical scheme, and its ability to represent the physical systems. Both analytical solutions for simplified cases and published data were used to check against simulation results from the numerical model. Excellent agreement was observed in nearly all the cases undertaken for testing of heat-flow, water-flow, and solute-transport model components. The coupled water-heat flow component was also successfully tested. The model could not be tested in its entirety as a single unit, because no analytical solution exists at present for coupled water-heat and solute transport. Also, no published data set was found to test all the features of the mode l together as a single unit. However, successful testing of individual model components provides confidence in the application of the model simulations conducted in this work.

PAGE 67

CHAPTERS MODEL APPLICATION Fate and Transport of Methyl Bromide Fumigation in Soils Methyl bromide (MBr) is an odorless, colorless toxic gas used in agriculture throughout the world. The United States uses it on more than 100 crops as a soil fumigant, a post-harvest treatment, and as a plant quarantine treatment to control a variety of pests MBr is also used to fumigate structures such as grain storage warehouses, flour mills, and ships carrying agricultural commodities. California and Florida are the primary users of MBr, followed by Hawaii. MBr, a broad-spectrum pesticide, is utilized heavily as a very effective agricultural fumigant during vegetable and fruit production in Florida, California, and Hawaii. MBr is very effective for eradicating insects, nematodes, fungi, bacteria, and weeds. In the United States, strawberries (16 % of U.S. total) and tomatoes (24 % of U.S. total) are the crops which use the most MBr, consuming about 6,500 tons annually. Other crops which use this pesticide as a soil fumigant include tobacco, peppers, grapes, and nut crops (USEPA, 1997). In the United States, the U.S. Clean Air Act requires that the production and importation of any substance defined as an ozone depleter by the Montreal Protocol of 1991 be phased out (USDA ARS, 1997). Therefore, despite its economic value, the 55

PAGE 68

56 U.S. Environmental Protection Agency (USEPA) plans to phase out the use of MBr by January 1, 2001. Since the Clean Air Act allows no exceptions, all uses of MBr, including plant quarantine, will be banned. In addition to being a toxic gas, MBr is a significant ozone-depleting substance, with recent scientific evidence estimating that bromine from this material is at least 50 times more effective at destroying ozone than chlorine from carboflurocarbons on a per-molecule basis (USEPA, 1997). MBr has an average residence time of 1.5-2.0 years in the atmosphere (Hacherl, 1 994). However, there are also other natural sources, primarily forest fires and marine organisms besides anthropogenic sources, which contribute significant quantities of MBr to the atmosphere. Table 5-1. E artier estimates for source-contributions of MBr to the atmosphere (Grojesan, 199 1 ; Khalil et al., 1993; Lobert et al., 1990; Mano and Andreae, 1994; Yagi et al., 1993; and Zurer, 1993). MBr Source Quantity in atmosphere (m ton y 1 ) Oceanic 30,000 (30 %) Bio-mass burning 30,000 (30 %) Industrial 15,000 (15 %) Agricultural 25,000 (25 %) I Total I 100,000 I Quantification of MBr release to the atmosphere is extremely difficult, since limited data exist for the total balance of MBr in the earth's atmosphere. Based upon information from several sources, atmospheric release can be broadly attributed to 4 major sources (Table 5-1), with agricultural use of MBr being attributed to 25 % of

PAGE 69

57 the atmospheric MBr. Recently, Yagi et al. (1993) estimated that agricultural use contributes 80 % of anthropogenic MBr in the atmosphere. MBr is used industrially as a methylating agent and a degreaser. More recent invest i gations emphasize new facets ( T able 5-2) of the uncertainty of various source contributions and sinks for atmospheric MBr New modeling studies based upon recent monitoring data suggest that the ocean absorbs more MBr from the atmosphere than previously thought. Rapid removal by the ocean shortens the average estimated life-time of atmospheric MBr to just 0.7 year, much less than an earlier estimate of 2 years (Friebele, 1997) Table 5-2. A recently revised budget for atmospheric MBr (Yvon-Lewis and Butler, 1997). MBr Sources Emissions MBr Sin ks Uptake (Gg y 1) (Gg y -1 ) Oceans 56 Oceans 77 Fumigation (soils durables, 46 Hydrolysis 86 perishables and structures) and uv Gasoline, leaded 15 Soils 43 Biomass Burning 20 Plants ? I Total I 137 II Total I 206 I These new results with regard to the role of oceans in the level of atmospheric MBr (Friebele, 1997) raise a question about the contribution that anthropogenic MBr plays in destroying the ozone layer. A newly revised MBr budget (Table 5-2) presented by Yvon-Lewis and Butler (1997) estimates that sinks (206 Gg y 1 ) greatly exceed sources (137 Gg y 1 ). These estimates imply that MBr

PAGE 70

58 should not accumulate in the atmosphere. Obviously, high uncertainty prevails regarding sources and sinks for the accumulation of atmospheric MBr. The buffering role of oceans implies that reducing agricultural sources will not affect atmospheric M Br to any great extent, and yet could cause unnecessary economic stress on the agricultural community. Thus, the economic value and efficacy of MBr fumigation to contro l pathogens i n agricultural soils warrant renewed r esearch efforts. Modeling MBr fate and transport during agricultural use (soil fumigation) is particularly needed to provide improved insight of system dynamics with regard to atmospheric emissions resulting from MBr fumigation and to provide guidelines for future research needs. Table 5-3. Selected chemical and physical properties of M B r (C H 3 Br) (Adapted from H acher!, 1 994). I Selected Property I Magnitude Molecular weight (g mole 1 ) 94.94 Boiling point (C) 4.6 Solubility in water (g kg 1 ) at 20 C 13.4 Gas density (g 11 at 2 0 C) 3.974 Diffusion coefficient (cm 2 s 1 in free air at 20 C) 0.0903 Henry's Law constant at 20 C 0.251 Vapor pressure (mm H g at 20 C) 1420 MBr is an odorless hydrocarbon (CH 3 Br) except at e x tremely high concentrations, where it has a sweet odor. I t has a low boiling point (4.6 C) and thus remains in gaseous form at temperatures normally encountered under field conditions. The important chemical properties of MBr are listed in Table 5-3. Lethal I

PAGE 71

toxicity of MBr to humans has been noted at 1600 ppm in air for 10-20 hours or at 7900 ppm for 1.5 hours (USEPA, 1986). The lowest inhalation level found to cause toxicity in humans is 35 ppm in air (EXTONET, 1997). Transport Mechanisms for MBr Fumigant in Soil 59 Although MBr is extremely volatile, it is commonly injected as a liquid from a pressurized cylinder. The formulation injected is typically a mixture of 98 % MBr with 2 % chloropicrin (Hacherl, 1994). Chloropicrin is added to act as a lachrymatory warning agent. Transport of gaseous MBr in the soil environment is mainly via diffusion through interconnected air-filled soil pores. Mass flow of gases in the soil environment has been proposed by several mechanisms, primarily controlled by changes in temperature and barometric pressure, effects of wind, and forced movement due to infiltrating water (Baver et al 1972). The diffusion coefficient for MBr in free air has been reported to be 325 cm 2 ru 1 at 20 C (Siebering and Leistra, 1979). The temperature-dependence of the diffusion coefficient in free air can be defined by use of Eq. [2.15]. Estimation of the diffusion coefficient of gases in a porous soil system is done by taking into account the air porosity and tortuosity of the soil system. The tortuosity can be deterrnined by use of Eq. [3.28]. Maharajh and Walkley (1973) measured MBr diffusion in water for temperatures ranging from 9.9 to 34.9 C (Fig. 5-1). An almost linear increase in the diffusion coefficient was observed with temperature over the range of investigation.

PAGE 72

60 0.08 Maharajh and Walkley (1973) --.. 0 Linear fit ( r = 0. 98) 0 N 0.07 '-._/ ..... 0.06 Figur e 5-1. Temperature d.) dependence of MBr .... (.) l 0.05 diffusion in water ( Data d.) from Maharajh and 0 u 0.04 Walkley, 1973) 0 ..... 0.03 0 0.02 10 20 30 40 Temperature (C) Fate of MBr F umigant in Soil The fate of MBr i s n ot only dependent upon transport by diffusion in the gaseous phase, but also upon partitioning between the gaseous and liquid water phases as well as between the liquid water and solid phases. Reactions in the water and solid phases greatly influence MBr fate and transport. In general two ways exist by which the fate of MBr in soil is affected. The first includes factors that affect MBr movement in soi l in a reversible manner, such as sorption to solids and dissolution in water The second includes factors which contribute to the irreversible chemical degradation or destruction of MBr molecules, such as hydrolysis. Gaseous MB r reversibly partitions into both the solid matrix and the liquid phase. T his partitioning may be described by instantaneous linear sorpt i on under equilibrium conditions such as ( Brown and Rolston 1980) S = k 5 C k s [5 1]

PAGE 73

61 where ks is a partitioning coefficient between the gaseous phase and the combination of soil+ water phase. Goring (1962) reported the Henry's Law constant as 0.244 [-] at 20 C for the partitioning of gaseous MBr into liquid water Mutziger et al. ( 1 996) reported l ab-measured Henry's L aw constant values over a temperature range for the partitioning of MBr into water (Fig. 5-2). 1 .0 ,--------------------, I i 0.8 D 0 L ---0.6 0 u Cl) 0.4 -~ (1) ::r: 0.2 Goring (1962) Observed ( 1. 2 atm) Estimated ( 1. 0 atm) 0 O O U-.1.--.11 o--'--L......L-.t.....2.L.O...l....l......l.....L.....L3 o....J...J.--'--L.-4L-O .L.....I.....L.....5 ...L.. O-t.-L.....J...J.--.160--'-L......L-J.....J 7 0 Temperature (C) Figure 5-2. Laboratory measured and estimated Henry 's Law constants at different temperatures for the partitioning of MBr into water (Adapted from Mutziger et al., 1996 ). Reversible partitioning reactions are characterized by the fact that, as gaseous phase MBr concentrations decrease, MBr molecules will desorb from soil solids or volatilize from water and thus be returned to the gaseous phase. Irreversible chemical or biological degradation reactions, on the other hand are ones in which MBr is converted into another compound in such a way that it cannot return to the gas phase and thus is permanently lost from the system. In an irreversible hydrolysis reaction, MBr is known to react with hydroxide ions (OH) in aqueous solutions through both unimolecular and bimolecular nucleophyllic substitutions to fonr1 methanol and the bromide (Br ) ion as shown in

PAGE 74

the following reactions (Moje, 1960) H 3 C : Br => [ H 3 C ] + + : Br [H 3 c] + + : OH => H 3 C : OH 62 [5 2] SN2 : H 3 C : Br+ : OH => [Ho : CH3 : Br]~ H 3C : OH+ : Br [5.3] The rate of these hydrolysis reactions is greatly enhanced in the presence of ultraviolet light, although light from the visible spectrum has no effect (Castro and Belser, 1981; and Gentile et al., 1989, 1992). Increased p H and increased temperatures also show an enhancing effect upon the hydrolysis of MBr ( H acherl, 1994). Moje ( 1 960) indicated that unimolecular nucleophy l lic substitution should be the major mechanism for the hydrolysis of MBr in water. Thus, the hydrolysis reaction becomes first-order. Reactions of MBr in soil wate r with soil organic matter involve the transfer of methyl groups to the organic matter (Maw and Kempton, 1 973). These reactions are expected to be first-order due to the large excess of organic matter reaction sites (Brown and Rolston, 1980). Brown and Rolston (1980) measured the appearance of Brin the soil solution resulting from the decomposition of MBr as a function of time for three different soils. The production of Brfor a given introduction of MBr was greatest for organic soil, intermediate for loam soil, and lowest for sand. The results thus suggest that mechanisms of decomposition involving clay particles and soil organic matter are significant to the decomposition of MBr in soils during fumigation. A first-order hydrolysis rate coefficient of 0.0 1 4-hr 1 was reported for Yolo loam soil by Rolston and G l auz ( 1 982). Hornsby et al

PAGE 75

63 (1996) reported the soil organic carbon sorption coefficient (koc) for MBr to be 22.0 ml g 1 of organic carbon. In a recent study by Jin and Jury (1995), on a Pachappa fine sandy l oam soil with 0.49 % organic carbon, the degradation rate of M Br was observed to be first-order and was insensitive to changes in soil water content (Table 5-4). However the MBr degradation rate showed a slight increase at water saturation. These results confirmed earlier work showing that organic matter is the major factor influencing degradation of MBr in soils. Table 5-4. Measured degradation rate coefficient and half life of MBr as a function of soil water content (Adapted from Jin and Jury, 1995). Degree of Water Saturation in % Half life Rate coefficient (Water Content, 0 r) (days) (hr 1) 45 ( 0 L = 11.5) 12.5 0.0023 76 ( 0 L = 19.4) 12.5 0.0023 100 ( 0 r = 25.6) 10.7 0.0027 Although hydrolysis deactivates toxic MBr molecules, toxic levels of bromide may accumulate in the soil water of treated soils (Macartney and Price, 1 988; and Maw and Kempton, 1982). High levels of bromide accumulation have also been found in crops such as tomato, carrot, lima bean, lettuce, radish, tobacco, wheat orange, and lemon grown on initially fumigated soils (Hoffman and Malkomes 1979 ; and Maw and Kempton, 1973). Permeation/Transmission of MBr through Polyethylene Mulch Open-field fumigation without a plastic soil cover (mulch) may provide a large input of MBr into the atmosphere. In plastic-mulch culture atmospheric

PAGE 76

64 emission occurs by diffusive transport (or permeation) through the plastic mulch (within the bed) as well as by diffusion through uncovered soil surfaces in furrows between parallel soil beds. Several environmental factors contribute to diffusive flux through the soil in furrows and through the plastic mulch in beds Environmental factors with the greatest impact upon diffusive flux include the temperature gradient across the plastic mulch, the moisture or condensed water layer on the underside of the plastic mulch, and the sweeping effect of wind over the surface (overland wind velocity). The physical dynamics of water and heat transport beneath plastic-mulched soil surfaces are quite complex (as discussed in Chapter-2.). Temperature increases at the plastic mulch during the daytime by as much as 15 to 20 C due to solar radiation are not uncommon (Ham and Kluitenberg 1994). Such temperature increases during daylight hours enhance the water evaporation rate, thereby increasing relative humidity of the air gap and of soil beneath the plastic mulch. Since the plastic cools faster than the soil, the condensation of water from the humid soil air tends to occur on the underside of the plastic mulch during nighttim e hours. As a result, permeability of the plastic mulch to MBr would be reduced due to a blocking effect by the layer of condensed water. As the mulch temperature increases during daytime hours, evaporation then removes the condensed water layer. In cases where permeating MBr is soluble in or sorbs to the plastic material the rate of movement by diffusion would further be reduced thus delaying the equilibrium. As such, only operational value for the diffusion coefficient can be envisioned.

PAGE 77

Table 5-5. Ratet of escape of MBr through several films at various temperatures (Adapted from Kolbezen and Abu-El-Haj, 1977). 65 Temp. LDPEt HDPE-1tt HDPE-2tt Polybutene Saran Black LDPE (OC) (1 mil) (1-mil) 23 8.2 2.7 30 11.1 3.5 40 15.1 5.0 50 20.1 6.8 60 30.6 8.8 t Units are ml hr 1 m 2 m1 1 L 1 t low density polyethylene, (1-mil) 1.4 1.7 2.9 3.9 5.2 tt high density polyethy l ene sample 1, tt high density polyethylene sample 2, and :i::i: 1 mil =2.544E-3 cm. (1-mil) (0.5-mil) (6-mil t 1 ) 2.9 Nil 1.6 4.0 Nil 2 2 6.3 Nil 3.4 9.1 0.18 5.9 1 1.8 0.41 9.2 Few investigations have been conducted to measure the diffusion coefficient of MBr through polyet hylene sheets used for mulching. Kolbezen and Abu-El-Haj (1977) measured rates of escape of gaseous MBr through six selected films at different temperatures. The escape rates increased with temperature for all films. MBr escape rates ranged from 0.41 to 30.6 ml hr 1 m 2 m1 1 L 1 for the six materials. The authors found one of the high-density polyethylenes (Table 5-5) to be an effective barrier to MBr transport, with minimum escape rate over the temperature range investigated. During a field experiment for unbedded mulched soil in California Yagi et al (1993) observed that the cumulative emission of MBr decreased as the thickness of LDPE plastic increased (1.03, 1.29, 1.37, and 1.6 mil thickness were used). Unpublished results from field investigations for mulched soil bed system at the

PAGE 78

66 University of Florida Green Acres Experimental Station, near Gainesville, FL with black LDPE (1.2 mil) and clear laminated plastic ( relatively thicker than black LDPE) showed emission fluxes differed between the two plastic materials. Emissions were reported for both the plastic-covered beds and uncovered furrows in between the parallel beds. 0.25 ------------C lear Laminated Pla s tic 0.20 Feb. 1996 0 Bed: Observed 0.15 --Bed: Moving Average o Furrow: Observed 0.10 ,, A '~. 0 --------Furrow: Movrng verage ,--.. 0. 05 r /\ ,d0 ~ q~ 1 r ;: ,J ,.. ~~Bel~ -"' o 11 o r t,;;"'i:Ja<1 ,fl o .-1<: ~ ""I l , ,., ll J= 0. 00 L.,.._ _________ u _, __ ':1 1' __ __ o.. _ ~1-<: ~ ---J N I 0 25 .---B la -ck-Plast -ic ----------------, o. 2 o Feb. 1996 S 0.15 50 10 0 ><: 0.05 r0 ,j, .u. ,; er---" O o.oo ,., s== 0 0 u Bed: Observed Bed: Moving Average Furrow: Observed Furrow: Moving Average 0 ., en 3 .o .-----------------------, Cl) a 2.0 1 0 0 Black Plastic Jun e 19 95 0 -------0 Furrow: Observed Furrow: Moving Average Bed: Observed B ed: Moving Average 0. 0 ~ \<, -J~ l:!il:!Jl~~~~~~ !:l ... ~ -._.-,:--,..."l" ~A: ~ .., 1( -( t.-< ~ .. ) of:-l-!.-L....J-:=J24...,..J......I......L.-L.4...,L,,s,-.l---1-...L......L.,,7:,l,,..2..I......J.....~9L..,,.J6 L..J ---!. -L1~2..,.J..o ....1...-L....L.,1+44 -,J-1-..!' ..1.,-16.J,,..,!_8 Time after fumigation (hour) Figure 5-3 MBr emission fluxes in plastic mulched beds during fumigation at the University of Florida Green Acres Experimental Station located near Gainesville, FL MBr was applied at the rate of 450 kg ha 1 with a 3-nozzle applicator at a depth of 20 and 33 cm during the summer of 1995 and winter of 1996, respectively in soil beds of Arredondo Fine Sand soil used for vegetable production. As expected higher magnitudes of the emission fluxes were observed in the summer shallow

PAGE 79

67 injection over winter deeper injection during fumigation (Fig. 5-3). A diurnal cyclical pattern was evident in the emission fluxes to the atmosphere due to non-isothermal conditions. The relatively thicker clear laminated plastic was able to retain slightly more MBr in the soil bed over black LDPE (Fig. 5-3). However, material balance estimations from emission fluxes and soil Br residue measurements for MBr applied during fumigation revealed that 39 46 and 70 % of applied MBr were unaccountable in bed systems with black LDPE (June, 1995), black LDPE (Feb., 1996), and clear laminated plastic (Feb., 1996) mulches, respectively Hence, definite conclusions were not feasible from these field investigations of MBr fumigation in plastic mulched vegetable beds (personal communication with investigators 1 ). Hacherl (1994) conducted laboratory investigations measuring the diffusion through black LDPE (1.25 mil) film commonly used in agriculture. Figure 5-4 shows the results of diffusion investigations by Hacherl (1994) and Kolbezen and Abu-El-Haj (1977) for black LDPE with 1.25-mil and 1.00-mil thicknesses, respectively. The results revealed a near-linear increase of diffusion coefficient with temperature, with lower values in the case of Hacherl (1994) corresponding to thicker LDPE. In another study (Jin and Jury, 1995), from the data presented the diffusion coefficient was estimated to be 0.005969 cm 2 hr1 for black LDPE (1.0 mil). 1 A. G. Hornsby, L.T. Ou, D. Dickson, S.J. Locascio, and P.S.C. Rao

PAGE 80

J;: 0.008 C'-1 8 0.007 u ...._,, +--1 0.006 u 0.005 tE a.> 0.004 0 u Kolbezen and Abu-El-Raj (1977) Black LDPE: 1 00mil r = 0.97 D s:::l 0.003 o _0 D oo 0.002 0Hacher! (1994) 0 -oBlack LDPE: 1 .25 -mil 0. 00 1 c..L.-i....,..,-,. ........... ......1............,...,,......i.....i......--i.....,J.,,--J-r ....1, _o.__ 9_,_9~---'-...r.....1.~..1-.....1...1 20 30 40 0 50 60 Temperature ( C) M odeling of MBr Transport during Fumigation Figure 5-4. MBr diffusion through black LDPE of different thicknesses (Data from Hacherl, 1994; and Kolbezan and Abu-El-Haj 1977). 68 Only a few prior modeling efforts have been attempted to describe MBr transport in soil during fumigation. Hemwall ( 1 959) described a two-dimensional model fo r gaseous diffusion of MBr in soil with relative ly l ow water content. The model assumed a small constant water content and isothermal soil conditions. Typical simulations were presented to demonstrate time-dependent patterns of fumigant concentration and biological control within an uncovered soil profile after fumigation. Siebering and Leistra (1979) presented a one-dimensional model for gaseous diffusion of MBr under isother1r1al conditions Water content was held constant and small, such that water flow could be ignored. Comparison of model simulations with experimental results for MBr concentrations during soil fumigation applied to soil under cover in a greenhouse matched only in the upper part of the soil profile (to 40 cm) but deviated substantially elsewhere.

PAGE 81

69 Rolston and Glauz (1982) proposed a model describing radial diffusive transport of gaseous MBr away from injection chisels during fumigation of a field profile. The model was based on similar work done by Hemwall (1959, 1960) and Siebering and Leistra (1979). The model included the dissolution-distillation of MBr gas on soil solids and in soil water, using both kinetic and equilibrium relationships and hydrolysis to yield Br ions according to first-order kinetics. The model considered the case where the soil surface was covered with a barrier totally impermeable to the diffusion of gas along with the case of a bare soil surface. Reported simulation results provided a better match with the experimental data of Abdalla et al. ( 1974 ) which involved a mulch barrier, when no barrier was considered at the soil surface during simulations. Sensitivity analysis of input parameters showed reaction kinetics to be unimportant, so they could be safely ignored. Water flow and nonisothermal temperature effects were ignored in all simulations. Also, in that model the boundary conditions (BC) at the soil surface were unrealistically handled by making the plastic mulch impermeable and applying a zero-concentration fixed BC at the soil surface. Mignard and Benet (1989) considered a kinetic relationship for MBr partitioning between the gaseous and solution phases. Synthetic simulations were conduc ted and compared with equilibrium models, and profiles were developed to forecast the treatment zones and behavior of MBr. The authors inferred from their analysis that a kinetic partitioning relationship is important only for rapidly changing boundary conditions. This model also ignored water and temperature variations in the soil.

PAGE 82

70 Recently Wang et al (1997) adopted a two-dimensional (2-D) finite-element V model CHAIN 2D (Simunek and van Genuchten, 1994) to simulate the fate and transport of MBr fumigant in soil. CHAIN_ 2D numerically solves the partial differential equations for 2-D nonlinear nonequilibrium simultaneous transport o f water and heat, (but not coupled heat-water) and solute in a variably saturated porous medium. The temperature dependence of various coefficients (such as JYI K g etc.) were described in similar fashion to the present work. However the mod e l lacked a realistic description of the soil-surface plastic-mulch/atmospheric boundary. Placement of a plastic mulch significantly changes the energy balanc e at the soil/atmospheric interface (as discussed in Chapter-3) and requires proper redefinition of the energy exchange and balance at the interface to simulate non isothermal soil conditions. The transfer of MBr to the atmosphere was described (Wang et al., 1997) by introducing a transfer coefficient (hTR) as [5 4] where db (cm) was considered a composite boundary thickness inclusive of an air gap below the mulch, the mulch thickness, and an overlying atmospheric boundary layer; and where hTR was described as a transfer coefficient (cm s 1 ). The d b parameter was shown to be temperature-dependent through use of an equation similar to Eq. [2.14] but the scientific basis for doing so was not explained. Although a cyclic nature in emission flux in accordance with non-isothermal conditions was demonstrated, model simulations failed to accurately describe ( Fig. 5-5) the cyclic field-measured emission flux reported by Yates et al. (1996 )

PAGE 83

,.-.._ I C'I) 120. 0 1 00 0 ~ 8 80.0 O!) S 60.0 '-" ,-J 40.0 20.0 ? II 1 I I f I I I I I r I r I I Go r I l I I \ I I / \ I \ I I Q ,, I I I I I (l) I I 1 Ci) I 1 I I I (:) I I 6 I I I I 1 6 60 0 (!) Q I I / I I I e O b served Estima t ed 1 0 2 0 30 40 50 60 70 8 0 90 1 00 1 1 0 120 130 Time ( hour ) Figure 5-5. Simulated and measured flux density through plastic mulched surface. Adapted from Wang et al. (1997). 71 It is thus evident that, in modeling MBr fate and transport, various investigators have made simplifying assumptions (such as small and constant soil water content and isothern1al conditions) that reflect un r ealistic field conditions. Also, BCs at the soil surface which are more realistically described by a Neumann type in the case of MBr fumigation have been improperly designated (such as a zero concentration fixed BC at nodes exposed to the atmosphere for uncovered soil). The dissolution ( H enry's Law constant) and diffusion (both in air and water) of MBr are temperature-dependent, and variably saturated soil-water flow is a common occurrence under field conditions. This is particularly true for humid climates.

PAGE 84

72 Analysis of Fate and Transport of MBr Fumigant The model presented in this dissertation represents many improvements over earlier efforts (Hemwall, 1959; Mignard and Benet, 1989; Rolston and Glauz, 1982; and Siebering and Leistra, 1979), for it includes coupled transient flows of both heat and water based upon PDT theory. Realistic plastic-mulch boundary condition s are described thoroughly by inclusion of optical properties of the plastic mulch in the energy balance at the soil/atmospheric interface. This model provides opportunity to consider the effect of improved temperature upon estimates of solute transport (both in the liquid and gaseous phases) Modeling investigations (synthetic simulations) of the fate and transport of MBr fumigant were conducted to obtain improved insight into the system dynamics for MBr fumigation of soils. Modeling simulations of MBr fumigation prior to this research have utilized overly simplified assumptions, such as gaseous diffusive transport in air-filled soil pores under isothermal conditions with the effects of transient liquid-water flow neglected. Simulations reported here utilized an assumption that was more realistic of actual field conditions; i.e., non-isothermal soil conditions that fluctuate diurnally due to net solar radiation, thus altering vapor and liquid water flow as well as MBr diffusion. Objectives were as follows: 1) To investigate important major processes influencing MBr fate and transport including volatilization and degradation in fumigated soil beds used for vegetable and fruit production in Florida,

PAGE 85

2) To investigate the effect of non-isothermal and variably saturated soil conditions on material balance of MBr during simulated fumigations 3) To evaluate the effectiveness of the current management technique-i.e., plastic-mulching of soil beds-in minimizing MBr emission from fumigated soil beds to the atmosphere, and 73 4) To examine alternative management methods for decreasing MBr emission during commercial use of soil fumigation. Approach Simulation analysis was designed considering major factors known to affect the fate and transport of MBr. The following factors were considered in the design of simulations: I. Effect of variable water saturation in the soil bed: Two levels of heterogenous water saturation were selected to represent a relatively dry soil and a reasonably wet profile. The effects of hydrolysis (degradation) on MBr material balance were considered for different levels of soil water saturation. II. Effect of variable temperature regime in the soil bed: A. Non-isothermal conditions: Effects of transient thermal conditions were considered upon the following input parameters for the model 1. MBr diffusion coefficient in the air phase (IJ'), 2. MBr diffusion coefficient in the water phase (Dj, 3. MBr partition coefficient between gaseous and aqueous phases (Henry's Law constant, Kg),

PAGE 86

74 4. Diffusion coefficient of MBr through polyethylene film (ll'p 1 asa c ), B. Isother1nal conditions: Uniform ther1nal conditions were provided for comparison (average daily temperature was assumed to exist uniformly over the simulation domain). III. MBr diffusion through plastic mulch: Due to uncertainty in published data and based on reported field investigations of MBr emissions from plastic mulched soil, different magnitudes for the MBr diffusion coefficient through plastic film were used to examine the material balance of MBr. IV. Depth of fumigant injection: Two depths of fumigant injection zones were investigated: A. Typical shallow injection zone: 33 cm depth total B. Deep injection zone: 66 cm depth total (arbitrarily chosen for comparison) V. Effect of irrigation (dousing of water on the bed): Effect of irrigation was examined to evaluate the surface-sealing effect (due to wetness) on reduction of gaseous emissions for MBr. VI. Fumigation Efficiency: Soil sterilization during fumigation was investigated based on the nematoxic level of MBr (20 ppm or greater) and exposure time (minimum 24 hr) for 2 nd -stage juveniles of root knot nematode (Mckenry and Hesse, 1978). VII. Hydrolysis of MBr: As no data for the hydrolysis rate of MBr in Florida soils was available, sensitivity analysis was carried out for 3 rates; 1) 0.014 ru 1 (for California, Yolo loam soil, Rolston and Glauz (1980)), 2) 0.007 ru 1 and 3) 0.0035 hr 1

PAGE 87

75 VIII. Diurnal water and heat dynamics in soil bed: The influence of atmospheric interactions with the soil surface (both bare furrow and mulched bed) were examined by analyzing the water and heat dynamics in the soil bed system over a diurnal period of 24 hrs. Methodology Arredondo Fine Sand (Grossarenic Paleudults, loamy siliceous, hyperthem1ic), a coarse-textured soil at the IFAS Green Acres farm near Gainesville and commonly found in north central Florida, was selected for the modeling investigation. Representative soil properties were taken from 'Soil Science Research Report# 88-1, 1988 1 '. The soil consisted of five horizons namely; Ap (0-23 cm), El (23-66 cm), E2 (66-99 cm), Bt1 (99-165 cm), and Bt2 (165-203 cm). The weather data for simulation purposes were taken from a weather station at the Green Acres Farm of IF AS located west of Gainesville. The month of February was selected for all simulations, since MBr fumigations are norn1ally performed for the Gainesville area at that time of the year. A spatial profile discretization of the cross-sectional soil bed resulted in 2782 elements with 1474 nodes (Fig. 5-6). Bed symmetry was assumed, so only a half section (AB-BC-CD-DE-FE-AF) of the soil bed needed be considered for simulations. 1 Compiled by Univ of Florida, IFAS Soil Science Dep Soil Characterization Lab. in cooperation with the USDA-SCS

PAGE 88

76 No flow BCs on t h e leftand right-hand sides (AF and D E ) were defined for heat, water, and solutes The D E boundary represents a vertical symmetry line through the middle of a furrow between two parallel soil beds. The AF boundary represents a vertical symmetry line through the middle of a soil bed. The lower boundary (FE) was designated as a zero thermal gradient with gravitational water flow(= K(h T)) .---x Mul c h e d-B e d 0 z . ' ;J ~ A . :: Ap B ~ .., ... Furr o w ... -25 C .I , ,,, ., C, -=' .-i ,_ --' ' .,. ' . ,., Inj ec ti o n Zo n es ,.,, h ,, E 1 .,, -, ,,. / .,, I, -.,, .,. -5 0 ,,, --.,, .,, ., .,, ...,, .,. .,, ,. .,. ---, --I _,,, ,. ,., ., ,. .,. I _,,, I ../ --.,. -I / ,../ i / / / 7 :c.7 / / .,, / ,, .,, .,.. -75 i "-" ...d -1 00 fr / -/ -/ / -/ / / / / / / / / / / / / ,,, / / / ,,. / / / ,, / / / ., / / I / .,, ,,. // Q r ,,. / / f-o ,,. / / -125 -, .,. 1 / i-, / / / / / I / I / I / I / , ./ / / I / / / I / / I / 150 ,, / / / / / I / / / / I / I / / / / / / / / / ./ / I / / 1 75 / / / _,. / / I / I/ / I / / / / '/ / / .. F 'I/ / / ._ 20 0 0 I. I I I I / -/ -.,. -----1 .,. ,,. / / J / ,. / / / / / / / / :;,., -.,. .,. -/ .,. --_,,, / ,,. / E2 .,. / -/ / / / ..,. _.,.. / / / / I / .,. / / / / / / / / / , / I / / / / / / / / / / / , / / / / / / / / / ./ / / / / _,,. / / / / / / / I / / / / / / / / , / / / I / / / / / / / / I / / / / / / / / / / / -../ / 1 / / / / / / ./ / / / / / / / _. / / / / / / / / / / / I / .,./ / / / / I / / / / / / / / 7 / / / ..,. ,,. /' / / ,,.. / / / / / / / .,.,. / / / I / / .,., / ,, ... / / / / / / / / / / ,, ,,., / I / ./ / ./ / 1 / / / / / / / / / / / / I / / / 7 ., ., .,. / _/ / I / Bt 1 / I / / / / / / 7 7 i / / / ,,.. / / / / ., / / / / / I / / / I / .,. ,' / / .. / / / / / I / / / / / / / / / ,. ,/ / / / / / / / / / . / I / / / / / / I / / 7 1 7 / I / / / ,_,,, I / / / / / / / / / / / / / ,, ,, / / / / / ,,,. / / / / / / / / / ,,.. / / / / / / / / / I / / / / ,,. / I / ./ / / / / / / I / I /. / / I/ .. 1 ,.. / / / / / I / / ,, / / I /' / Bt 2 ,, / / / / / / / / I / / ,,, / / I/ ,, I / / / / / / / / ,, .,. / / _., / / / / / I / 1 / I ,1 ,. I ./ 1 I / / I I 20 40 60 Di s tance from Bed Center ( cm ) I ..,. / _r / / / / 7 / / / / / ./ ,7 / / / / / / .,, / ./ / D : L I . .,. ,., _,,, ,? I / ..... ~ 1 7 ..,. / .. /' / / / / ,,. : 7 I-' ., 7 / _, / / I _,. 1 7 / '. ,.. I / / / .,. / , I .; _,, 1 ,. ., I ,; I / I / E I I I Figure 5-6 Schematic of the discretized simulation domain o f the soil bed. For solute transport, both convective (in the liquid phase) and diffusive (both in the aqueous liquid and gaseous phases) transport were allowed at the bottom ( F E), with the assumption that solute leaving the lower boundary is lost from the

PAGE 89

flow domain (i.e., there is no return flow). The emission l oss of MBr to the atmosphere (AB-BC) was described through diffusive flux, assuming a 1-cm boundary layer and zero concentration on the atmospheric side of the boundary layer. These emission fluxes may be defined explicitly as a) For Bed J B e d D g gas plasti c b) For Furrow g s g s+ l ( air ) Z s Z s+ l ( air ) J Furrovi1 = .!_ D g gas 2 free air where the subscripts s s+i and s-l represent nodes at the soil surface, above the surface, and below the surface, respectively Table 5 6. Physical properties for 5 distinct horizons of a profile of Arredondo fine sand soil 77 Soil Layer Sand Silt Clay oct BD+ K (g cm 3 ) sat_ 1 ) (%) (%) (%) (%) (cmhr Ap 91.9 E1 92.0 E2 91.4 Bt1 86.0 Bt2 75.2 t OCOrganic Carbon, BDBulk Density 3.3 4.8 3 3 4.7 3.0 5.6 3.7 10 3 3.0 21.8 0.83 1.52 6.2 0.28 1.54 16 .7 0.14 1.52 15.1 0.11 1.56 6 4 0.20 1 .57 2.4 The data for physical properties of the soil profile used are presented in Table 5-6, with other related parameter values in Table 5-7. The soil water

PAGE 90

characteristic data used for estimating (van Genuchten, 1 980) hydraulic-function parameters were taken from the report mentioned earlier. Average weather parameters 1 used in simulations for the month of February were: 1) average daily temperature 11.6 C, 2) temperature amplitude 7.7 C, 3) wind velocity 11x10 4 cm hr 1 4) day-length11.0 hrs, and 5) global daily radiation1400 J cm 2 day 1 Table 57. Parameter values for 5 distinct horizons of a profile of Arredondo fine sand soil Soil ,1 t Kdl:t: Kd 2 tt van Genuchten Parameters Layer (hr 1 ) (cm g-1) (cm :3 g 1) a (cm 1 ) r e er Ap 0.014 0.1826 0 0.0225 2.124 0.43 0.03 El 0.014 0.0616 0 0.0216 2.762 0.42 0 .0 4 E2 0.014 0.0308 0 0.0271 2.626 0.43 0.04 Bt1 0.014 0.0242 0 0.0211 2.091 0.41 0.06 Bt2 0.014 0.0440 0 0.0307 1.557 0.41 0.13 t_ hydrolysis rate coefficient for MBr, t soil sorption coefficient for MBr, and tt soil sorption coefficient for bromide ion. Temperature-dependence of the diffusion coefficients for MBr in gaseous 78 and aqueous phases of the soil were defined by use of Eq. [2.15] and data reported in Fig. 5-1 from Maharajh and Walkley (1973), respectively. As information is unavailable for the temperature dependence of the hydrolysis rate coefficient 1 1 ), a fixed value of 0.014 hr 1 was used (Rolston and Glauz, 1982). The diffusion coefficient 1 Climatological Data, Agronomy Department and NOAA Cooperating, 1996.

PAGE 91

79 for the plastic mulch was defined using the data of Hacherl (1994). The MBr fumigant was assumed to be applied at a rate of 450 kg ha 1 with a three-shank applicator, so that 11/2 shank (3591 moles) injections occur within the flow domain shown in Fig. 5-6. Unless otherwise noted, the depth of the injection and time of day for injection for all the simulations were set at 33 cm and 9:00 AM, respectively. Black LDPE of thickness 1.25 mil was assumed as the plastic mulch in all the simulations. Mass and energy balance errors were less than 1.0 % for all simulations reported in this work. Results and Discussion Plastic-Mulching of Soil Beds to Control MBr Loss to the Atmosphere Preliminary simulations conducted using the laboratory-measured data (Fig. 57, Level-1) of Hacherl ( 1994) for diffusion through plastic (black LD PE, 1.25-mil thick) provided estimates for the atmospheric losses of MBr during fumigation of the soil bed that were very low (23 % during the 1 st day) compared to reported values of 40 to 60 % for recent investigations (Yagi et al., 1993; Yates et al 1996). Hence, parallel elevated temperature-dependent levels were designed (Fig. 57) for investigation. The use of an experimentally measured diffusion coefficient for a 1.25-mil thick black LDPE by Hacherl (1994) in simulations showed that very little MBr escaped (0.02 % of that applied) through the plastic mulch into the atmosphere (i.e. from the bed excluding the furrow) during the 1 s t day. The

PAGE 92

80 diffusive flux (Fig. 5-8: L evel-1, and Fig. 5-9: a, Level-1) based upon the measured diffusion coefficient of plastic mulch from the soil bed was very low and it was also 27 22 17 12 2.7 2.2 l 7 o 1 2 Elevated Level-4 ( = Level-I x I 0 4 ) ?,,-\ Er\ \ 04 D !>I .. '"' ,,. 6 43 V E l evated Leve l3 (= Level-1 x 10 3 ) A I\ o p~ 1 o4-'2. E,-2.' ,,. 6 43 I\ Ll ,-------------. Q.> 0.27 E l evated Level-2 ( = Level-I x 10 2 ) 0 U 0.22 u 0. 12 ,------------------. Q 0.0027 0.0022 0.0017 0.0012 Measured Level ] (Hacher I I 994) ,., 20 \ Q4&5 Cl s'f . ,,,_ 6 4'3 c, 0 p\ 11 UC u ,, I , I ,, I I 25 30 35 40 Temperature ( C) Figure 57. Measured and elevated levels of diffusion coefficients through 1.25-mil black LDPE plastic mulch used in the model simulations. demonstrated by the material balance after 1 and 3 days (Table 5-8) that only 23.4 and 45.8 % of the applied MBr respectively, was lost from the system (mulched soil bed plus bare furrow) to the atmosphere. This result is much lower than the 40 to 60 % atmospheric losses earlier reported from field investigations (Yagi et al., 1993; Yates et al., 1996), which were carried out under unbedded conditions with complete mulch cover of the soil. Initial average fluxes both from the bed and the furrow were substantially higher than for later periods (after 24 hours) when the

PAGE 93

81 conce n trat i ons in soi l had decreased due to sp r eading, hydrolysis, and loss o f MBr to the atmosphere. Also note t hat emission fluxes from the furrow were nearly 5 times larger than for the bed, except for elevated plasticmu l ch diffusion Level-4 due to the absence of a mulch barrier over the furrow. --. 0 E ::t V) V) 0 0 > p CIS ;:I E ::; u ,....._ .E M e 0 0 1750 1 4 00 1 050 700 350 0 1750 1400 1050 700 350 0 5.00 4 00 3.00 2 00 1.00 0.00 G:l C: 3.00 0 V) V) 2.25 a 0 1 .5 0 0 b.() CIS 0 75 '0 > <: 0 000 -7 ------F t 1 rr ow B ed -------------------------------------------------~ Furrow "< <-=--=---~ .. -----=:-.;._ -Meas u red LevelI Hac h e r! ( 1 994) E l evated Leve l2 E l evated Leve l3 Elevated Level -4 Bed -4--~-~--~=~----1 2 24 36 48 60 72 Time afte r fu mi ga t ion ( h o ur ) Figure 5 8. Losses of MBr to the atmosphere through soil bed and furrow surfaces using measured and elevated diffusion coefficients for MBr through plastic mulch 1 1 Small steps in some of the figures resulted from the need for plotting more points within certain time periods.

PAGE 94

82 Use of elevated levels of diffusion coefficients for plastic mulch showed more losses of MBr into the atmosphere from the system (Level 2 = 25.1 %; Level-3 = 34.7 %; and Level-4 = 5 1 .7 %) one day after fumigant injection. In all cases emission fluxes from the bare soil furrow were highest near the bed due to close proximity to the injection zone. Figure 5-9: a shows MB r fluxes in the bed 6 hours after fumigant injection, when peak emission fluxes occurred from the bed (Fig. 5-8). ( a) Mea s ured,Hacher1(1994) Elevated Level-2 lmol cm 2 hr 1 -25 I I I I I I I -75 1111111111, 111111 ~ J.SE.. ..... . I:.:~~--,---;-,~t I I I I I E levated Level3 Elevated Level-4 25 7 '/ "/ I ~;--.;l,..._;IIII// 'I I 11111111111 2 1 . . 3. S'Ec-Q , . .. . . I I I I t I I I l I I f 02 , , . 3 .S'E-. \ . . . I 00 L-._ ._._.._._......._ ................................ .._._........___,_......._.,__.__.. 0 15 30 45 60 75 0 15 30 45 60 75 Distance from bed center (cm) Figure 5-9. MBr total concentration (mol cm 3 ) contours and gaseous MBr flux vectors (mol cm 2 hr 1 ) in the bed 6 hrs after injection of MBr fumigant. a) 4 levels of diffusion coefficients through the plastic mulch, b) non-mulched bed.

PAGE 95

83 Emission fluxes ( F ig. 5-9: a) increased from the mulched bed with elevated levels of diffusion coefficie n ts for the plastic mulch. Also to be noted here is the fact that the trend of diffusive flux wa s reversed in the furrow compared to that for the bed, and lower diffusio n coefficients for plastic mulch sho w e d higher diffusive flux and more cumulative loss from the furrow (Fig. 5-8, Fig. 5-9: a, and Table 5-8) Figure 5-9: b shows the em i ssion of MBr in furrow and bed without a plastic mulch for comparison purposes. The emission fluxes were higher from soil-surface regions close to the fumigant injection zone (Fig. 5-9: b). (b) No Mulch I I I I I I i \ ' 2 ' ' ' , . . 3. SE Q . . . . . . 1 00 ......,_........_.i.....a.....t.....1....i....,..-"--1,,....i.....i,-i......--L..J.....i.,_L-.L...:....... ....... .u 0 15 30 4 5 60 75 Di s tan ce f r o m bed ce nter (c m ) Fig:re 5-9--continued Overall, total material losses of MBr from the simulated soil-system to the atmosphere (Table 5-8) increased when elevated l evels of diffusion coefficients for plastic mulch were used in the simulations Three-day simulations were appropriate, as the resu l ts showed that most of the M Br losses occur within the initial 24 to 30 hrs after fumigation (Fig. 5-8 and Table 5-8). Atmospheric losses were

PAGE 96

84 observed to continue i n small amounts even after 3 days, as long as undeg r aded MBr is present in the soil. Obviously much u ncertainty prevails regarding the diffusion coefficient of MBr in plastic mulches. T he use of a lab-measured diffusion coefficient for simulating field conditions greatly under-estimated published results from the field investigations done in the past. However, the use of elevated diffusion coefficients (Level-2, Level-3, and Level-4) improved simulation results. It was therefore, decided to use the Level-3 diffusion coefficient for all other simulations, as results obtained from this leve l were close to those from the published field investigations. Table 5-8. Material balance (unit:%) of MBr under measured and enhanced levels of diffusion coefficient for plastic mulch. Total MBr Atmospheric Emission Deep DownDay-1 Losses Hydrolyzed in Profile flow Bed I I Total Furrow I I Level-1 69.55 0.02 I 23.39 I 23.41 7.04 0 l I I Level-2 67.91 1.81 I 23.29 I 25.10 6.98 0 I I I I I Level-3 58.54 13.15 I 21.57 I 34.72 6.73 0 I T I Level-4 42.24 34.24 I 17.41 I 51.65 6.11 0 ! I No-Mulch 37.29 I I 56.92 5.79 0 I I Day-3 Level-1 36.38 0.20 I 45.63 I 45.83 17 .7 9 0 I I I Level-2 33.94 4.24 I 44 42 I 48.66 17 .39 0 l l I I Level-3 23.46 24.55 I 36.52 I 61.07 15 46 0 l l I I Level-4 15.05 46.32 I 25.84 I 72.16 12 .79 0 ! I I No-Mulch 13.40 I I 74.68 11 .91 0 !

PAGE 97

85 Non-Isothermal versus Isotherrr1al Soil-Conditions An isothermal simulation was perfor111ed in order to compare with non-isothermal simulations existing under a black LDPE. Thern1al effects on solute transport (involving IY', Dw,kg, IY'pJast1c parameters) were ignored in isothermal (with respect to solute tra nsport only) simulations. However, the coupled heat and water flow model was used in its original form, with solute transport being made independent of temperature fluctuations. The parameters for solute transport were estimated and kept constant for a daily average temperature of 11.6 C, typically existing during the month of February in the vicinity of Gainesville (diffusion coefficient in air= 308 cm 2 hr-1, Henry's Law constant= 0.15 [-]). -~ 5.00 ~ 8 3.1s (.) 0 2.50 1.25 Non Isothermal ------Isothermal Furrow 0.00 .,__ ______________ ___, 0.99 ----------------------, 0 ,-; ti) @ 0.66 Q.) Q.) ~0.33 "-4 I I I I I I I I I I , ' Bed 0. OOo~......__.~:-'=--~~ -'---'-'' ...... -:-'-:,-'----'--'-----:-'=--'' ...... ~~...l.......\.-~ 12 24 36 48 60 72 Time after fumigation (hour) Figure 5-10. Emission fluxes of MBr through bed and furrow under isothermal and non-isother1nal transport of fumigant.

PAGE 98

86 The isothermal and non-isothen11al simulations conducted for a 3-day period showed (Fig. 5-10) that MBr emission fluxes from both bed and furrow were affected by non-isothermal conditions. Figure 5-10 shows the diurnal fluctuations in MBr emissions from bed and furrow in the case of non-isotherrr1al simulations for black LDPE, whereas a similar cyclical nature of MBr emission was missing in isothermal simulations. This emission pattern has been reported earlier (Fig. 5-5) for field investigations (Yagi et al., 1993; and Yates et a l ., 1996). Higher diffusive flux rates for MBr were observed initially from the furrow, which subsequently flattened out due to lower concentration gradients as increasing amounts of MBr were lost to the atmosphere. The delay in reaching the initial peak emission flux (Fig. 5-10) from the bed occurred due to a build-up time for MBr concentrations beneath the mulch and an increase in temperature of the plastic mulch as the day progressed (application time 9:00 AM), whereas the bare furrow surface offered a quick passageway to the atmosphere. Table 5-9. Material balance (unit: %) of MBr for fumigation under isothermal and non-isother111al conditions. Day1 Total MBr Atmospheric Emission Losses H ydrolyzed Deep in Profile i Bed I Total Down-flow I Furrow I I I Isothern1al 6 1 .28 11.42 I 20.33 I 3 1 .75 6.97 0 I I I I NonI I I I 58.55 13.44 I 21.59 I 34.73 6.72 0 Isothermal I I I I I Day-3 Isothern1al 25.96 22.37 I 35.45 I 57.82 16.21 0 I I I NonI I I I 23.46 24.56 I 36.5 1 I 61.07 15.46 0 Isother1nal I I I l I

PAGE 99

87 The results suggest that inclusion of non-isothe11r1al conditions is essential to correctly estimate the fate and transport of volatile chemicals. Non-inclusion of temperature effects on solute transport may underestimate (Table 5-9) material losses to the atmosphere in the case of volatile chemicals such as MBr The isothermal simulation estimated 260.2 kg ha 1 (57 .8 % of application) emission loss to atmosphere compared to 274.8 kg ha 1 (6 1 .1 % of app l ication) as a non-isother1nal estimate over a 3-day period. Variable Water Saturation of Soil Volumetric water contents present in the soil bed at the time of fumigation not only affect the hydrolysis of MBr but also its transport in the soil Two different wetting regimes were selected for the simulations. An initially saturated soil profile was allowed to drain and evaporate water for 40 days. Two wetting levels were then selected to provide the soil water regime: 1 ) Level-1 (WL1 ): 3292 cm 3 total water content in the flow domain, and 2) Level-2 (WL-2): 1864 cm 3 total water content in the flow domain. The simulations revealed (Fig. 5-11) that larger amounts of MBr were lost to the atmosphere when there was less water (WL-2
PAGE 100

88 furrow in the case of the dri e r (WL-2) profile (Fig. 5-12) The emission fluxes from the bed were also higher (Fig. 5-11) during the later periods (after approximately 34 hrs) in WL-1. This was due to the higher water contents in WL-1 causing larger amounts of MBr to partition into the aqueous phase, which were then released to the gaseous phase as atmospheric loss es occurred in order to maintain equilibrium between the aqueous and gaseous phases (controlled by the H enry's Law constant) ",, --;5.00 ~ : -. I \ C.. I I I 3.75 8 ;, ~, .. \' (.) 0 I I, 2 5 '7 0 1.25 --, _____________ J -. -----,_ WL-1 WL 2 Furrow .... ___ --0.00 ..... :) __________________ _, 0 99 I ,. ,. ~ ... Cl) I ", Cl) ,. ... 8 o.66 ,., ___ .---. CL) ... __ \ .. I Bed Q..) I 0.33 J-4 I Q..) I > ,'
PAGE 101

89 same locations. The wetter profile 0/vL-1) maintained a higher water content at th e mulched soil surface, due to vapor movement. This also contributed to lesser emission of MBr through both bed and furrow surfaces due to partia l blockag e o f soil pores. Though not obvious, a cyclical diurnal pattern in the temperature pr o fil e can be identified (Fig. 5-12). This demonstrates that MBr fate and transport would be affected by the non isothermal conditions within the bed. Figure 5-12: a shows the conditions existing 1/2 hour after the MBr fumigant injection To be noted in this plot is the difference in spread of fumigant from the injection point More rapid movement of MBr is clearly noticeable in the ca se of the drier bed ( WL-2 ), wher e as restricted spread of MBr is noticed in the case of the wetter bed (WL-1) due to partitioning with water and restricted passageways due to higher moisture contents. The discontinuity in the water content distributions (Fig. 5-12) was due t o horizons of different porosity and texture in the soil profile. As time progressed after injection (Fig 5-12 : c-f), the injection masses a t th e injection points coalesced showing the center of fumigant mass in the center o f th e bed. Figure 5-12: d-f in isolation would give an uninforrr1ed observer an impression that there was only one fumigant injection point. The coalescence occurs as a resul t of MBr transport and loss to the atmosphere; thus, higher concentrations remained in the center of the bed. Material balances after fumigant injection indicated that more MBr was hydrolyzed in the wetter profile 0/vL-1) as would be expected. During the 1 st day nearly 20 % higher emission loss to the atmosphere (Table 5-10) was observed in th e drier soil system (WL-2) as compared to the wet profile (WL-1). By the end of the

PAGE 102

90 3 rd day, the amount of MBr retained in the wet profile 0/vL-1) was 10 % more than for the drier one (WL 2). However, nearly 12 % more MBr was hydrolyzed in the wet profile (WL-1) than the drier one (WL-2) over a 3-day period. This indicates that the wet profile was able to retain more MBr compared to the dry one, but resulted in more bromide fom1ation as a decomposition product. The higher retention of MBr in WL1 was partly due to smaller quantities of available diffusion volume in the gaseous phase, partly due to more MBr partitioning into the aqueous phase at the higher water contents. Diffusion in the aqueous phase was very slow (by a factor of 10 3 ) compared to the gaseous phase. T able 5-10. Material balance of MBr (unit: %) under different water saturations for one and 3 days after fumigant injection. Total Atmospheric Emission Losses Deep Day-1 MBrin Hydrolyzed Down' I I Profile Bed I Furrow I Total flow ! I I I I WL-1 (Wet) 72.53 5.54 I 9.44 I 14.98 12.48 0 I I I WL-2 (Dry) 58.55 13.44 I 21.59 I 34.73 6.72 0 ! Day 3 I I WL-1 (Wet) 34.45 1 8.16 I 19.78 I 37.94 27.61 0 I I I WL-2 (Dry) 23.46 24.56 I 36.51 I 61.07 15.46 0 !

PAGE 103

91 ( a ) 0 WL-1 WL-2 2 5 a) __,. 5 0 0 r/.J 75 0 4 WL-1 WL-2 1 9 7 s -25 a) 13.4 i 12. J u '-._I f-4 12 5 a) ~5 0 13.9 (1) 0 a) 75 11.6 13.2 100 .___....1....1.. ........... ............ ..................... ,.__,__. ........._ _.__._'--'---'-...I....U 0 1== ========~ i:7 1---0.15 --0.0 4 WL-1 -25 0.12 0 1 1 a) 0 WL-2 0 . ....._ ') 0.11'1 0 06 ... -50 u f-4 0 15 a) 0 1 5 75 ... 0 06 i------0 14----illlllliiiiil ... () 1 I 10 0 0 15 3 0 4 5 60 75 0 1 5 30 45 60 75 Distance from bed center (cm) Figure 51 2. Contours of water content (cm 3 cm 3 ), temperature (C), and MBr concentration (mol) with gaseous fluxes (mol cm 2 hr 1 ) of MBr under 2 different soil water saturation scenarios after fumigation. a) 1/2 hour; b) 3 hours; c) 6 hours; d) 12 hours; e) 18 hours; f) 24 hours

PAGE 104

(b) 0 -50 75 0 ,,-..., S 25 (.) ....._,,. -5 ~50 Q) Q -7 5 100 0 25 50 -75 ' 5 I 2 hr I mo cm 7 Ciu 0 15 . . . I 6 12.s 11.6 0 1 0.1 0.15 0 15 WL-1 WL-1 26 0 WL-1 0 11 i-------0 14 _____ ...,. (I.) B a (I.) 'S (I.) 'S 0 u $..,j (I.) ..,J WL-2 \ \ \ \ \ '............... -. . I I I I I I \ I ~ ~ ()() ''' "' .......... . . 9"( \\\ ,,.._ ...... .. ... \ ,, , .. .. LL U-1-UW---l~~~,,, ,,, ~ \ I \ \ \ \ ) <-tr::, I I I I I I I I \ \ \ \ ~ y 1 1 1 1 1 11 11 11 \ I I .,,,111111,,,,, .. ,.,., .. ,,, .. .. . . . . r: -02 1,--,~ --.-.-.-. 2 0c 15 6 WL-2 29 3 1 3. 2 1 3.9 13.2 ' 0 04 WL-2 p 0 06.: '\. II 11'\ 0 06 p Ir 0 06 '" '" '" fl 11 l 00 ........... ...................... _._ ................................................ ...._._............., ....... 0 1 5 3 0 45 60 75 0 15 30 45 60 75 Distance from bed center ( cm) Figure 51 2-continued 9 2

PAGE 105

( c) 2.5 mol cm 2 hr 1 0 . . i -75 0 s25 u ...._., ~-50 (l) 0 -75 100 0 -25 -50 -75 1. ; 1 111 11 11 tlllll 6 1111111 WL-1 ... ... 1:-.!.. ~~~w....o~O~l~l~l~l~l~I......-...,....-; ,,,,,11 1,111 111,1 . . . . 16 7 I.] ~ ~, WL-1 29.5 v ~ L.. '"CJ.J.o '''L.. 12. 5 -. 1 .6 I I ,, t .,, 1 , 0.14 0 ]4 WL-1 Q, 1 A 0 1 I 0 12 C C C L.. C 0.14 C 0 J4 L.. '''n I ,. I I . I I I ' I ~~~~A..,,". WL-2 5 0'6 1 I I I -?c-Q? ,-,. ....... ...... --__ .,,,,..,,,,. .,, I I\\,,,,,,----------~ 1111,,,,, ,, .................... -----11 I I\ \,c_() \ ,,,,...._,..._ --.---. 6 1?' \ \ \ '' '' :\ .............. . w....-1~1 \\\\\\\\''~~''''' --I \ \ \ \ \ \ \ )Y ' ' .. .. \ \ I \ ,'\ ' ' ' .. . \ \ \ \ \ \ >~ .. . . . 1 1 , , I O.t.1 , , I I I I I I I I I I\, \~ llll~l \\\\\ \ '''' ''''' 1 1, , , ,, t--s:J /J.4 .. ~-0'2. '1 .o . . WL-2 n--29,3 -15 6 i:--------13 .9-----1 13 .2-------1 ' 0 04 WL-2 0 u.us 0 .06 L.. 0.06 _O I I 15 30 45 60 75 0 1 5 30 45 60 75 Distance from bed center ( cm) Figure 5-12-continued 93

PAGE 106

( d) 0 2 l -2 hr -I mo cm WL-1 '''h :\ "---------t : : : t ~1?--~ : j.c...: l:_.:.l~I -',~ i I I \ \ \ -50 I I I I I I I \ \ \ I I I \ \ \ \ \ 11 11,, ~()\ I I l I I I I \ \ \ 'l,.~P~ 1 \ ' I I I I I l G ~ I I l 1 I I l I \ I l \ t_ ~ \_J~ 11111111,, ,, ,;J -7 5t ''......'..'....'.__'_:__ :....: ~...:.... .:.....:-:-_,_-:--: ' I I I 0 15 .3 WL-1 12 .5 s-25 12 .8 (.) '--" 15 7 -5 12 .5 ~-50 Q.) 0 -75 11.6 ---__ _J 0 0 14 0.14 WL-1 -25 0. 1 A 0 1 l .. .. 0 l2 .. -50 -.. 0.14 .. 0 14 -75 .... .. .. .. I i I I 0 16 I , I I Q.) J.,..j Q.) a Q.) , , 1 I I I , , , ; , , I I I , , I I I I I I , ,, ,, ,, I I I I I I I ,, ,, I I I I ,, I WL-2 I I , I I I I I , . . -. -----. -2;~{--........ ':4 .. 1,, .... Ctc> .. ... __ I I t \ \ ':I' "" "' .,. l I \ \ I t \ \ \ I t \ \ \ ------,.,. ... ,.,. ... --------.. .......... 2. -02 I I , I I I , I J f I ; I I I si-91 \ ,,?,,,,, ti\\\\\\ I I \ \ \ \ I I \ t I I I \ \ \ \ \ \ --a ..-0 \ t \ t I t \ t l ...:.I ~l~l~l ~1 --:-1 ~:--:I~ \ .61'"'\ \ \ I \ I \ t ,.,,,,,,. ''' ., ', I l I s.9E-02 1 6.3 ~ 16.6 WL-2 (jjj) 1 3.2 l 6.6:=:::::==1 14 l ........ -------1 3.9---1 13 2 -------l 0 04 l?o~ WL-2 .. ') 0-:-0-5 0.06 .... 0 06 ._ f\ I 1 15 30 45 60 75 0 15 30 45 60 75 Distance from bed center (cm) Figure 5-12 -co ntinued 9 4

PAGE 107

( e ) 2 l 2 ru mo cm 0 25 -50 -75 0 ,.-._ S 25 <.) '-" ~-50 (l) Q 75 ,1,,, ,,,,. .... ,, .. ,,. i--+-: + r ++-1, 2 4 E" i ~, ~ ',, :, ',ll',1 1 '\ Y ,, ,,. ,, ,, ,, 1, ,,1,,1, ,,,,,, ,,,,,,,,,,, 11111,,,, r:--:---:--:-~f/-:....1 f I I I I ' ' ' ' ' o'i:--1.) s . ,,,,,,, ti ' ' ' 1 3 5 1 2 J 0 t========= == :::nu~ 0 14 0. 14 ----1 -25 0 1 2 -50 0 14 0. 1 3 -7 5 0 1 6 ' WL-1 ,. ,. ; I WL-1 5 8 12. I 1 3 .9 WL-1 Q) "S 0 r/). Q) Q) a Q) d Q) d 0 u ... (1) I, . I I I I I I I I I I ' WL-2 . . 2 0E-02 I I I I ~ . i ... . . . ' r-, . . . ' . I I 1,,.,. t I I I I ' ' ' 1 WL-2 14 .5 6 1 4 .5 14 0 1 3 9 13 2 0 0 4 ... WL-2 o~ : ,, ') 0 05 0 06 0 06 n 1 I 15 3 0 45 60 75 0 15 30 45 60 7 5 Distance from bed center (cm) Figur e 5-1 2-co nti nue d 95

PAGE 108

( f) 2 mol cm2 hr 1 0 E-bi ' ' ~ 2 WL-1 I I I I I I ' ' ' ' ' ' ' 25 , ~~ $~ 02 : , , . . : : i,~\ -50 ti.~ . : i : ~~ : . -7 5 . ~-i 6 6~ ' . 100 0 13 2 @ WL-1 12.9 s25 u 12 9 "-.._/ t-50 (1) 12 6 Q -75 11 8 100 0 4 0 14 WL-1 25 0.11 0 12 -50 0 13 0 13 -75 0 15 1 5 30 45 60 75 Q) ::s 0 r/1 Q) i-. Q) a Q) Q) 0 u ' . ~ : ~ .. ...... L .;..,"') . .. .... .... 0 15 . ~\ ',~ '"\, . 13, '/.; -2 c@ 14.0 13.9 13.2 0 04 '' ) 0 06 0 06 .0 11 30 45 Distance from bed center ( cm) Figure 5-12-continued WL-2 . :.01 \ \'1 .. WL-2 14 5 13 2 WL-2 0.0 5 60 75 96

PAGE 109

97 Depth of Fumi g ant In j ection Influence of the depth of the fumigant injection zone upon MBr distribution and loss was i n vestigate d using two injection depths, 33-cm (shallow) and 66-cm (deep). R esults for the s h allow (33-cm) inject i on depth with v ariable water saturation are presented in F ig 5 -11 and Table 5-10. I n case of d eeper i n jection of MBr at 66 cm (Fig. 51 3), the emission fluxes were smaller than those f or the 33 cm injection zone. T his was partly due to higher water contents in the 66-cm injection zone due to the presence of a differential soil horizon (Fig. 51 2), resulting in more MBr in the aqueous phase and partly due to greater distance to the emission surface from the center of M Br mass. r---. 3. 75 -L... ,, ... WL-1 ', I I \ l= b.. ,, ,I ... I I 3. 00 I L -' WL-2 I ,_--------------I N L... s I 2. 2 5 L... I ', I 0 I 'I -,_ Furrow \., II L I, 0 1. 5 0 >II s -, I, -do.1s L.. I .... L... L... ;:j q 0 Cl) Cl) a <1) <1) bl) ;> < L... I 0 00 ""' 0 60 r ' 0 4 5 ; r I -\ \ ; Bed I 0 .3 0 r ,-, I r r ' \ r I ' I -,_ \ 0 15 I I .. ,_r I I 0.00 0 I I I I I 1 2 2 4 3 6 48 6 0 7 2 Time after fumigation (hour ) F igure 51 3. Emission fluxes of MBr under different water saturations for a 66-cm injection zone during the first 3 days after fumigant injection.

PAGE 110

Table 5-11. Material balance of MBr (unit: % ) during fumigation under different water saturations for 65-cm injection zone. Total Atmospheric Emission Losses Deep Day-1 MBrin H ydro l yzed Down' I I Profile Bed I Furrow Total flow I I i I I WL-1 (Wet) 8 1 .30 0.30 I 1.85 I 2.15 16.54 0 I l I I I WL-2 (Dry) 67.57 4 74 I 17.96 I 22.70 9.72 0 I I Day-3 I I WL-1 0/Vet) 46.05 4.77 I 10.41 I 15.18 38.76 0 I I I I I I WL-2 (Dry) 27.21 13.82 I 38.44 I 52.26 20.53 0 I I Day-7 I I WL-1 0/Vet) 13.56 10.05 I 18 36 I 28.41 58.03 0 I l I I I WL-2 (Dry) 6.76 17.21 I 46.91 I 64.12 29.12 0 I I 98 Negligible amounts of MBr were lost to the atmosphere during the 1 st day in the wet profile (WL-1), and only 28.4 % was lost after 7 days following injection. Also in the case of drier soil system 0/VL-2), deeper injection resulted in 8.8 % less emission to the atmosphere compared to the shallower injection depth (33 cm) over the first 3 days after injection. Thus, deeper injection is potentially conducive to controlling atmospheric emissions of M Br irrespective of profile wetness. Effect of I rrigation (Wate r Dousing of the Bed Surface) The large initial losses of MBr to the atmosphere were assumed to possibly be minimized by dousing water on the bed just before unrolling and placement of plastic mulch on the bed surface. A numerical experiment was run by simulating

PAGE 111

99 such a situation. This was achieved by modifying the initial conditions for matric suction (h = -5 cm) on the top nodal layer of the bed surface. An initial water content of 0.42 cm3 cm3 ( 97 % water saturation) on the bed surface was achieved in this way, representing dousing of water compared to the undoused situation (0.03 cm3 cm 3 6 % degree of water saturation). Only the top surface of the bed was doused with water (the sides of the bed, and the furrow, were excluded). _,,_, 6.00 ..------------------------, N 4.80 8 3.60 ........ S 2.40 ::l '-/ 1.20 ------------No -D ousing Dousing Furrow 0.00 ..__ ________________ __, 1 .00 C/.) -~ 0.75 8 0.50 bJ) C'j 0.25 > I I I I I I I Bed o. oo 0 !:'............ '----'-----'---="'=-"'"......._ ............. ~....._.__,_----'--::~---=-...... ......._~__,_....___~:-'--'' ...... ---'-:::! 12 24 36 48 60 72 Time after fumigation (hour) Figure 5 -1 4. Effect of water dousing on the MBr emission fluxes from the soil bed during fumigation. The results for emission flux for both situations revealed that dousing slightly reduced the emission flux from the bed, but only initially (Fig. 5-14). Dousing saturated the pores of the soil under the plastic mulch with water, thus temporarily blocking MBr gas passage to the atmosphere through the plastic mulch (Fig. 5-15). However, this effect lasted only for a short time in the coarse-textured soil, due to rapid redistribution of water. If the soil were fine-textured, one would expect the

PAGE 112

100 reduction in emission flux to last longer. The material balance (Table 5-12) also showed that, during the 1 st day, 5.4 kg ha 1 of MBr emission losses were prevented 1111,,11\ s ....... I I I I I I I I I ; I I I I I I At o~ ~ I 100 i3 O tc::::::n-.o:~ n fr 0.10 -50 -75 Dou s ing 0 7 ~ 0 -~\ L...---~ 0 06 ,-=-=== ~ 1 ---0 0 4 ---1 0 0 5 0 0 6---~ i=---o .06----i 0 .06 B ::s 0 VJ -+-' c:: E 0 u I I 1 I I t b J 1 1 1 \}..:" h--,-.~ 1.t E I ' l.) ~I.) . .. I j I I I o 1 0 I O t I I ,t\. o . . ' 0 04 No Q ~ Dousing Oif ) 0 06 0 06 l l 15 30 45 60 75 0 15 30 45 60 75 Distance from bed ce nter (cm) Figure 5-15 Contours of water content, temperature, and MBr concentration with gaseous fluxes of MBr at 9 hours after fumigation under water dousing and no-dousing on the bed-surface. Table 5-12. Material balance (unit:%) of MBr under water dousing and no-dousing Total MBr Atmospheric Emission Losses Deep Day-1 Hydrolyzed in Profile I I Down flow Bed I Furrow I Total ! i I I NoI I 58.55 13.44 I 21 59 I 34.73 6.72 0 I I Dousing I I l I I Dousing 59. 14 12.25 I 21.67 I 33.92 6.94 0 I I 1: Day-3 I I NoI I 23.46 24.56 I 36.51 I 61.07 15.46 0 I I Dousing I I I l I I I Dousing 23.92 23.48 I 36.67 I 60.15 15.93 0 !

PAGE 113

from the bed when the bed surface was doused with water immediately after fumigation. Fumigation Efficiency Analysis 101 Effective soil fumigation is achieved when the target pathogen and/or pest is successfully eradicated from the desired soil zones. The primary goal of soil fumigation is to provide a pestfree growing environment to the crop plant in order to obta in economically viable yields To achieve this goal, an optimum fumigant dosage must be determined. This is crucial when a fumigant such as MBr is not only environmentally hazardous due to its high toxicity to humans but also is detrimental to the atmosphere's ozone layer. The determination of optimum fumigant dosage is expected to depend upon 1) Fumigant exposure time and concentration required for eradication of a specific target pathogen and/or pest, 2) Level of pathogen and/or pest infestation, 3) Soil texture and structure, 4 ) Fumigant fate and transport in soil, 5) Desired eradication zones (soil volumes), and 6) Extent of sterilized buffer zones (pathogens are mobile and can reenter a sterilized zone again from surrounding uneradicated zones after fumigation). Mckenry and Hesse (1978) reported that 20 ppm or greater concentration of MBr fumigant for a 24-hr period is required to completely eradicate the 2 nd stage of

PAGE 114

102 root knot nematode (RKN) juveniles. A literature search and personal communication with University of Florida nematologists did not reveal additional information on nematoxic levels and exposure periods for MBr. It was, therefore, decided to use the data from Mckenry and Hesse (1978) as a basis for investigating soil sterilization efficiency for different scenarios. Simulations were conducted for 3-day periods to record 24-hour residence times for concentrations ;;:: 10 ppm, > 20 ppm (Mckenry and Hesse, 1978) and~ 30 ppm. The 10and 20-ppm levels were chosen to provide improved insight of eradication patterns if there exists uncertainty in the data reported by Mckenry and Hesse (1978). To identify the eradication zones for RKN, flood contour plots in the soil bed were used to demarcate those zones which satisfied the conditions of exposure time (24 hrs) for the related level of exposure concentration ( 10 ppm, 20 ppm, and 30 ppm). Full-bed plots were generated for clarity, using the mirror image of the simulated domain. Figure 5-16a shows the demarcated zones using criterion based on Mckenry and Hesse (1978). No -mulch ( A) was observed to be very ineffective and full-mulch (bed plus furrow) (B) the most effective. The current practice of mulching the bed (C) provides adequate sterilization if a single crop row were grown in the center of the bed. The same may become inadequately sterilized if the crops were grown in two rows on the same bed, however. It was very clear, from this analysis, that deep injection in either dry or wet soil does not provide adequate soil sterilization in the root zone where it is desired (Fig. 5-16: a). The analysis provided different results (Fig. 5-16: b) when the required MBr exposure concentration for RKN eradication was changed to concentrations 10

PAGE 115

103 ppm. The eradication zone under no-mulch increased substantially, though it still remained inadequate. In all other cases, eradication increased substantially to provide excellent sterilization except for case (E), which remained inadequate A single-row crop seemed feasible in case (D), which received adequate sterilization with exposure concentrations of 10 ppm (Fig. 5-16: b). Lastly, the analysis for exposure concentrations~ 30 ppm showed (Fig. 51 6: c) that adequate sterilization was not achieved in any of the cases, implying that the soil bed was under-fumigated (at the rate of 450 kg ha 1 ). This exercise reveals that reliable information on toxic concentration levels along with exposure-time levels is very necessary to correctly estimate fumigant dosages Infor111ation on pathogen mobility is also required, to determine the size of buffer zones so that a correct estimation of soil zones and/or volumes (in the case of spot treatments for orchard plantings) requiring sterilization can be made.

PAGE 116

(a) 0 -25 s u -50 '-,_/ -B fr-75 Q -100 A } 25 ..... 1 ........... .........,_.___..._ I ._ ........... _._._._ ...... , ..._ '--' ...... 1 l __._ 1 ....... 1 .,_,_ I ...... ........... 0 -25 s u -50 '-,_/ f;j fr-75 Q -100 0 -25 B u -50 '-,_/ ..d fr-75 Q -100 B D A: No Mulch, Shallow (33-cm) Injection Dry Soil B: Full Mulch, Shallow (33-cm) Injection, Dry Soil C: Mulched Bed Shallow (33-cm) Injection Dry Soil D: Mulched Bed Deep (66-cm) Injection E: Mulched Bed Deep (66-cm) Injection Wet Soil C I II I l,,11 E I fllllo I ii 1 0 4 12 ~75 -60 1 -45 -30 -15 1 0 I 15 30 I 45 f 6 1 0 f 75 Distance from Bed Center ( cm) -75 -60 -45 -30 -15 0 15 30 45 60 75 Distance from Bed Center ( cm) Figure 5-16 Soil sterilization zones for different scenarios 3 days after MBr fumigation using different criteria. a) MBr concentration ~ 20 ppm for 24 hrs (Mckenry and Hesse, 1978 ); b) MBr concentration 10 ppm for 24 hrs; c) MBr concentration ~ 10 ppm for 24 hrs.

PAGE 117

105 (b) 0 A -25 A : No Mulch, Shallow (33 -cm) Injection, Dry Soil 8 B : Full Mulch, Shallow (33-cm) u -5 0 Injection, Dry Soil \._,/ ,s C: Mulched Bed, Shallow (33 -cm ) fr-75 Injection, Dry Soil Q D: Mulched Bed, Deep ( 66-cm) -100 Injection E: Mulched Bed, Deep ( 66-cm) -125 1 Ii I I I I I I I I I I I f Injection, Wet Soil 0 B C -25 8 u 50 -._.; ,s fr-75 Q -100 -1 25 llt II 11111 lll ti, 1, I i I I I I I I I I I I 0 D E -25 s u -50 \._,/ ,s fr-75 Q -100 Figure 5-16-continued

PAGE 118

1 0 6 ( C) 0 A -25 A: No Mulch, Shallow (33-cm) s Injection, Dry Soil B: Full Mulch, Shallow (33-cm) u -50 ...._,, Injection, Dry Soil .s C: Mulched Bed, Shallow (33-cm) fr-75 Injection, Dry Soil 0 D: Mulched Bed, Deep (66-cm) -100 Injection -125 I I I I I I I I I E: Mulched Bed, Deep (66-cm) Injection, Wet Soil 0 B C -25 s u -50 ...._,, .s fr-75 0 -100 f1 11 I ,1,1 I I I I I I 0 D E -25 8 u -50 ...._,, .s fr-75 0 -100 -125 1 1 1 , 1 , , -75 -60 -45 -30 -15 0 15 30 45 60 75 -75 -60 -45 -30 -15 0 15 30 45 60 75 Distance from Bed Center (cm) Distance from Bed Center (cm) Figure 5-16-continued

PAGE 119

107 Hydrolysis of MBr (Sens i tivity Analysis) Information on the rate of MBr hydrolysis in Florida soils is lacking. It is expected that sandy soils will have lower hydrolysis rates compared to fine textured soils and soils with higher organic matter. Three-day simu l ations were conducted for three rates of hydro l ysis which were; 1) 0.014 hr 1 (HR-1), 2) 0.007 ru 1 (HR-2), and 3) 0.0035 ru 1 (HR-3) H R-1 represents a value for Yolo loam soil from California (Rolston and Glauz, 1982). The effect of different rates of hydrolysis on material balance of MBr reflected that model simulations are very sensitive to this input parameter (Fig. 5-17). Material balance showed that over a period of 3 days an amount of 69.6 kg ha 1 37.5 kg ha 1 and 19.5 kg ha 1 of M Br was hydrolyzed, respectively for HR-1, H R-2, and HR-3. The amount of material hydrolyzed showed a linear relation with the hydrolysis rate coefficient (Fig. 51 7). Thus the results indicate that both the water saturation of the soil and the hydrolysis rate coefficient play significant roles in determining the fate and transport of MBr fumigant. .. ___ 16 .'o' ~ --~--..__,, Day 1 Day 3 .,. ,, Figure 5-17 Loss of MBr (unit: %) under different rates of hydrolysis of MBr.

PAGE 120

108 Diurnal Water and Heat Dynamics in a Plastic Covered Soil Bed Interactions of the soil surface with overlying atmosphere due to solar radiation, evaporation, and wind cause disturbances in the system by altering the thermal regime and water content. For a soil system at static equilibrium with respect to water content and heat, a change in boundary conditions controlling the thermal or hydraulic state of the medium causes the system to adjust in a transient unsteady state manner both in water and heat if the flows of water and heat are assumed to be coupled. The effects of overlying atmosphere on the bed system were analyzed for a period of 24 hrs starting at 12:00 noon on the second day after fumigant injection and mulch application in cases of both initially 'wet' and 'dry' soils Figure 5-18: a-c and Fig. 5-19: a-c present the cases of a wet soil and a dry soil, respectively. Simulated data exhibiting the contours of water content and temperature along with flux vectors of water, vapor, and heat are presented in these figures for time intervals of 3 hr depicting the status and system dynamics in the bed. Flux vectors for the heat, water, and vapor on the upper boundary nodes are suppressed as these were controlled by imposed boundary conditions. Patterns of thermal energy distributions were similar in both wet and dry soils (Fig. 5-18: a, Fig. 5-19: a) over the diurnal period. Diurnal temperature amplitudes were greater for unmulched furrow surfaces than for the mulched bed surfaces both under wet and dry soils. The bedded soil surface under mulch showed smaller temperature amplitude because of a 0.5 cm air gap assumed to exist

PAGE 121

109 between the soil surface and mulch. This resulted in contact resistance which impeded the flow of thermal energy and thus caused insulation of soil under the bed (Ham and Kluitenberg, 1994). The a ssumption of a 0.5 cm air gap was realistic for field conditions since perfect contact (0 cm air gap) between the mulch and soil surface is never obtained due to surface roughness of the soil. However the actual temperature of the plastic mulch was higher than the underlying soil surface during the day (data not shown). Black plastic has a high absorbance (0.96) and a very low transmittance (0.01) and reflectance (0.03) (Table 2.2). In conditions where the air gap is practically zero, the soil surface and mulch would have similar temperatures (Ham and Kluitenberg, 1994). Whereas, the behavior of the system (i.e., interactions of soil surface with plastic mulch) is entirely opposite for a clear plastic mulch. The high transmittance (0.84) for clear plastic permits direct heating of the soil surface resulting in high temperatures. With an air gap between the surface and mulch, flow of thermal energy back to the atmosphere is impeded. Thus the soil surface beneath a clear plastic mulch (with some air gap in between) would show higher temperatures during daytime hours. However, with no air gap, the soil surface mulched with black plastic would show higher temperatures because of higher absorbance than the clear plastic Thermal energy fluxes were in general inward to the soil system during the day time (due to input of solar radiation energy) and outward to the atmosphere during the night time at the surface. To be noted in Fig. 5-18: a and Fig. 5-19 : a are the presence of 'hot spots' (closed contours regions, Fig. 5-18: a and Fig. 5-19: a; 12:00 PM to 9:00 AM) that develop in the upper comers of the beds. This was due to

PAGE 122

110 larger exposure area towards the edges which had higher energy levels flowing in and out. Location of the hot spots differed slightly between the wet and dry soils. As therrnal conductivity is greatly affected by water content, lower conductivity in a dry soil would result in lesser dissipation of therrnal energy being exhibited by higher temperature. For this reason hot spots in the dry soil showed slightly higher temperatures than the ones in wet soil. Liquid water movement in the wet soil under the bed was mostly downward due to gravitational drainage. However under the furrow surface it was both downward and upward for the depths depending upon time of the day (Fig. 5-18: b). Upward liquid water movement in the furrow occurred as a result of atmospheric evaporative demand. Loss of water due to evaporation generated high matric suction at the surface allowing liquid water to move in response to potential gradients overcoming thermal gradients. The fluxes were higher near the surface in response to higher evaporative demand during peak solar radiation hours (12:00 AM and 3:00 PM) which subsequently reduced during cooler atmospheric periods. A similar behavior was observed for liquid water movement in dry soil in the furrow (Fig. 5-19: b). However, the spatial pattern for flow of liquid water differed both beneath the bed and furrow in that it was mostly upward in the dry soil. This was due to lower matric potentials in the upper profile in the dry soil. The magnitude of the liquid water flux was ten fold (see the reference vectors in Fig. 5-18: band Fig 5-19: b) smaller in the dry soil. The system behavior is also influenced by the vapor movement in response to thermal energy (as vapor movement, water movement and heat flow are highly

PAGE 123

111 coupled). Vapor water losses to the atmosphere occurred only from the unmulched furrow surface. Both downward and upward movement of vapor was observed in wet and dry soils (Fig. 5-18: c and Fig. 5-19: c). The pattern was generally similar except for the magnitude of fluxes. During the day time the vapor fluxes (excluding the evaporative fluxes from the top layer which were imposed as boundary conditions in the solution of flow equations and hence not shown in the figures) were primarily downward due to heating of the uppermost soil layer. The magnitude of these fluxes were highest during the peak solar radiation hours (corresponding to the times 12:00 AM-3:00 PM as shown in Fig. 5-18: c and Fig. 5-19: c). Direction reversal of flux vectors occurred during the night time when the upper soil profile cooled and the deeper soil profile remained warrr1. During these periods, the vapor fluxes were upward. Also to be noted in conjunction with Figs. 5-18: a and 5-19: a (showing thermal status) is that vapor fluxes coincided with hot spots near the bed edges. A transitional phase occurred around 9:00 AM (Fig. 5-18: c and Fig. 5-19: c) when the upper soil surface began warming due to th e onset of solar radiation During this period the fluxes from the upper soil region were downward and fluxes from still warm lower region were upward. The coalescence of this fluxes occurred at a depth of 25-30 cm in the simulation results (Fig 5-18 : c and Fig. 5-19: c). This analysis reveals the complex nature of system behavior and dynamics under a plastic mulched soil bed during diurnal periods. However, the actual behavior would depend upon the specific soil system under consideration (such as soil type and layering within the soil, water saturation, winter versus summer

PAGE 124

112 weather conditions, plastic material, etc). In a closed soil column system in the laboratory the non-isothermal and coupled flow is easy to demonstrate (fo r e.g. Fig. 4-3) where different constant temperatures can be imposed at opposite ends of the column for long periods. I n an open field system such as plastic mulched bed where the boundary conditions change diurnally, the effects are imperceptible and masked due to system driving towards a state of equilibrium with the open surroundings A field system is also more complex due to inherent heterogeneity of the soil system (such as layering present in the soil used in model simulations). This results in variable hydraulic and thermal characteristics of the soil medium exhibiting differences in porosity (affecting vapor movement), differences in water contents for the similar potential (affecting water movement), and differences in soil texture, structure and mineralogy (affecting heat capacity and the11r1al conductivity). Given the complexities of such a system, inference can be drawn that the fate and transport of volatile chemicals would definitely be influenced by interactions of water, vapor, and heat in such a system.

PAGE 125

(a) 0 hr' I 5 -5 ' ' ' I ' '. 13 6 -10 .,. ' ' -15 . . . -20 -25 -30 /3 . / -35 . o----------~~ -5 -10 I I I I I I I I I I I 11'cP I I I I I I I I I I I I I 'V.' J. I I I I I I I I I I I I I I /S.j_ I I 111111111///, /I 111111111,,,,. I I I -15 , , ,,. I I I I I -20 ' -25 ~:.....:_:_....:_:_ . -30 12:00 AM 28 4 18 0 . . . 06:00 PM .. .. . -35L-...-'------'-.......... ~i......~.......:....:lo..!.....1::,..L....Jl..J...JL.J.-L..J.....L..J.U I f I I I I -10 i-;' -';--;' -; -;..' ..:. _;_' 11 1 111 I I I I I I I -15 I I I I I I t I f -20 , -25 , r------.13..3 -30 1 2:00 PM 1 -33--'---< I I I I -3 5 L.....__._ ........ -1..._.._.__....._~_._...__--1.....::::........____,_J....:1: ___. ._J..J O I I I I I I -10 -15 -20 -25 -30 ' I I ~..+-t-+-,r._ 1 l 3.~ I 1 I I I / I I I I I I I , , ,,, ,., . . . . .. 15 2 . .. . . 06 : 00 AM 2 6 f f f I 8 t 0 f t I I I I I I 1 I I I I f ~~...-1 ' I ' ' I ' I I f f lJ.4 13 j 12 9 . I I I i I I f I illllltJ I I t I I Ill Ill lfl Ill 111 1 11 I I I I I 111 1 11, , ... ,. 13 i . . 113 03:00 PM , 2 9 0 ,,. 2 7 ,. 1 09:00 PM .. ' 'Is_~ -.._ .,_1 6. 1 l l ll ttt . I I I I I I I I I I I I I I 1111 ,, 1 ,, 03:00 AM 15 9 . . . . ~ -;""7~ ..;_:_ t, 111 ,,,'""'-<..J' .. ,,,, I I I I I I I I I I I -,,,, , ... ,, "1. ,f.. ... ,.,___..,, ~i 1~ r-t--+-1~9 6 : --+-+-+ -,1,1111 J J 6 I I I I I I I I I I I I I I I I j . l4 J 09:00 AM . I .., 14 3 . . . . 1'1.j 14 3 ..... ... ... I ... ... I ...+--H9 5-+-+-1 . \3 & .,,, , ,, ..., ,JJ I I JO II I I 9,..:_.:......:._:_....:.....i . I I I ,,11 111 -350 15 30 45 6 0 75 0 15 3 0 45 6 0 75 Distance from bed center (cm) Figure 5-18 Diurnal water and heat dynamics in a mulched bed for a wet soil a) Contours of temperature ( C) and heat flux vectors (J hr1 ); b) Contours of liquid water content (cm 3 cm 3 ) and water flux vectors (cm hr 1 ); c) Contours of liquid water content (cm 3 cm 3 ) and vapor flux vectors (mg cm 2 hr 1 ).

PAGE 126

(b) :> ,......., hr -I Q I------------"-= ""-' "-'-"'~ . l 39: .. -5 ' -1 5 Qi--------------, . . . . .. r.-~--,--;--....-Q. I 39 5 -1 0 -1 5 0 ------------, o l36 : : : . -Sr: ~~--:--~"'.~-:-:-:-~:--7 ~....:~ _j I .. ,.,.,,,,,, 1 0 1111,,1,,1\\ \ \\\ ,1,, , l I I I I 0 : Q l 3 6 -5 I I I I I I I I I 0. 139 \ \ \ I I 1 0 I I I I I I \ \ \ \ \ \ \ I i I I I I I I \ \ \ \ \ \ I I ' I I i I \ I \ \ I I I \ I I -1 5 I I I I o, 1 44 \ \ \ \ \ I I l ' I I I \ \ \ \ -20 I I I I I I I \ \ \ \ \ \ \ \ \ \ \ \ 0 1 39 25 12: 0 0 AM 06 : 00 P M -I I I I I 12 :0 0 P M 06 : 00AM "---3 0 ,~ ~:_;.....;_ I O 0. J o ,. 11111 I~\\\\\ , -. ----.. 11111 't~ \\\;1' ---. o t . . .. 3 9 I I I I I I I I I I I I I I I \ 0. 144 \ I I I I I I I I I I I I I I I I l l I I I I I I I I I I I I \ \ I I I I I I I I I I I I I \ \ \ I I 0 1 18\ I \ \ \ \ \ \ \ \ I I I I I I I I 0 .1 44 . . . . . t-, -.--'. ...... ...... ........:.....:..: ...:.: 0 1 3 9 . '" 11111111\1\\\\\\ ll iti\\111\\\\\\ I I I 0 1 1 14 I I \ \ \ I I I I I \ \ l l l I I I I I I I \ \ \ \ I I I I I Id 1' \ I \ \ \ \ \ I I I 48 \ , \ \ \\ \\'\ .14;4 t-'-..._._~..:...: ...:_: Q l3 6 : . : 1111111\\\\\\\\1 I I I I 1 0 14 I \\\\\ I I I I I I I I I I I I I I I I I I I I I \\\\\\\\ I 1 0 I I \ \ \ \ \ \ \ l l l l I I , J 4 5 \ \ \ \ \ I I I I I I I \ \ \ \ \ \ \ \ \ \ \ \ 0 14 1 114 03:00 PM 09:00 PM 03:00 AM 0. 1 36 t\\\\ -~, __ __ ,,,,,, ~-t-1.-t-..,_..,-L:L.o o. l l l l 1 l l l l 0 1 \ "'
PAGE 127

( C) 0.1 'i mo cm 2 I hr 0 i l i 9 ~ 3 ? i 1 : :' 12:00 AM ._ I -5 I I I I I I I I I I I I I I I / ' ' ' , -10 ' 0.144 ' ,_ L.. . ' . . . . . . . ,_ -15 . . -' . 0.148 . -20 -25 -30 -35 . . -L.. . . . ~ u.139 0 144 ,.,,,. ,. / / 04 . . . . . . ~i1 . ._ 9. 1 0 07.._ . . . . i ' ,, " 1 1i/ 'f . . . . . .. . . . 01-------------~ t-:--; -:~;.....,.; ~: -,..' -'-6 t 3 9 .., ,-.... ''-':' -'L..,:(..:'''.,_l 06:00 PM -5 I I I I I I I I I I / I I I I 1111111111////// -1 0 I I I : : : : I o: 144,_/'."-'./~~~~~:: .. ... ,.,,,,,,, -15 ,, .. ,.,,,,~.,,.,,,.,,. ~~:=:=: =: =~ =~:_o ._ 1 ~4-g::~:~~~:~:::: 20 I -----0 144 ............ ~---l--l~ P J 3 9 ) -25 0.104 --~ IIJlllll I I . . -30 0.110 -35 ' . I I 01----------~ 0 i 3 6,-::-7 -,,-,. _ _j -5r ~~--:~~~..:J I f 1 2:00 PM 1 0 111111111,,,,. 0.144 I-'-'--!.__!._!.. ....:... -.: ' , ,0.146-...._.:..._ -15 -2 0 ' -25 -30 '() I I I I I I I I ~. !07 (a, I -35 Qt-----------~ r';..--.;';..--.;'.._;'.._;'......;.' ._:__' 6 I ' ' I I 1 I I 36-~~-l~l~I~~ 5 I:---;-~.,.-,.-.--.-.. I I ' ' I I I 0.139 06 :00 AM -10 . . . . -1 5 0.144 -20 . . 30 1:-;--;---:--:---,--0 1 07 I I 1 I . .. I I I I t Q \3 ?, ;-+J,-.,1,J.,...i.._j \ ///// I I I c-r-,r-,->-+-'-0 144 I I I I I / I If I I I I I.~.~~~--<_\ I ; I f ,' _,,__.. ' --------. r--:-. :---:-~......,_o. J Jo . . , r.._.___..__._-'--L....!.' O J -:i 9 ' t I I I I I I I I ~.ff~-~~J 111 1 11111,,11,,, f-r-ir-, -.---.-....-LO I , I I I l I I I t Jf1.,.t ,-,-.,-.J I I 1 11 1,1111 ,, , ,,,,,,,. i,-.... ....,., ~~o 148 , ---b.i 1 0 . I I I I I I 6 I I . 136 I I I I I . . 141 . O. J 45 0. 14 1 '() O J 07 !( . I 0.136 ' . . . ... ' . . Q. f39 . . . . . ... . . . . . . . . . . q.144 . . . . . . . . . . ~ ... . . . . . 0.139 . ' . . ._ 0 1 04 ' O. f 07 . . I . . ' 15 30 45 60 75 0 1 5 30 45 Distance from bed center ( cm) Figure 5-18-continued 115 03:00 PM " 0 09:00 PM 0,104 I o I I I ( I I I 03:00 AM t 0 1 t 6 I I I I I I I I t I I I ' ' 09:00 AM -0.139 I I I I I 1 I I I I ' ' l' , I I I I I I I I I I I I I I I I I I I I 60 75

PAGE 128

( a) -->0 J hr "' 0 l============::::::::::::::---7 I -5 I 1111111 I f f I f f I 4.3 I -10 r=~::::-11 .J.--: -;; :;= ~ . -151'-\~-1: : . -20 -25 /. . ~ -30 !4.< 1 2:0 0 AM t-==:===2 8 4====1 I I r.--.-+--1-+18.0+-+--+-i . -35'--'-_._.....__--L._._.....;:::,.....__..M,_........__._~~__._..__..=__~__,__._..L.I or,~~:;:::;=:::-;-:~7 I I I I I I I I I , , 5 I I I I I I I I I I I I I I I I I I I I I I I I I I l'cf' I I I I I I I I I ' I I I I ~9'. I -JO I I I I I I / I I I I I I I I I I ;; 6:j_ ,, I I I I I I I I I I I ,; -15 I f I I I I I i I I I # ct'-',,,,,. -20~'-.....::_..'.-.:._:__:__:_ 25 30 r--~ 14.2 06:00 PM 16 .i =Ft=~ .. . -35'--'-_._~._._...._ ........ ~_._...,__..,.,_~...:.....:.......,.~L...IL. ......... .._.._ ........ .u 0 t---,-. ---:::;-:::::,-. -~--:-.,._-, 12:00 PM -5 11 .1 -IO I t I t t /,/IC"'' t ,,,u,-.,,, r-:-~ -;-' ~:--:...;-:...._ I ,S ' I "' -15 ' ' ' -20 -25 4.2 -30 .14.2 -35 15.5,~-< 0 13. 06:00 AM I I I I I I -5 I I I I I I I I I I I I I I 10 . . -1 5 16. 1 Js_< -20 t : ' -25 . -30 l 4 .' 3 I I I I I I I I 1 3. 4 I I -350 15 30 45 60 75 l'-+--+-t-+-+-+-+--+-+]8 I I I I I I I I ,1 .. I, I j 11111111 I I I t I I f 116 03 :00 PM j f I I I I I I I I ~-4 i::-,-~-_.._ 14 4 ... -~ 4 . 14.2 11 11111,, litltft I I I I I I I ..... ,, 1111,, 1111111 '''''''''/'~ I I I ".t:.:; # I 1 I I I I I I I _,!--'-_._'--.. I ' t I I t . .. t t t o I I I t I o 's:.!)_ t t .., 1111111 ,, .. ~/.-\---t---41 I _.__ .... ' 09:00 PM t I I I -16:6 I J I I 03:00 AM (' 1-;-, --:--.--.-.!..._I I I I I ._ -.. 14 I 4. 4 1 1 t t , 1 , t o I I I I t f I ' 09:00 AM \ 4 .3 ' ' ' ' . .,, . '',., . . . 'S.; . 1 4.6 ~~ --4:s.3:;;;;;;;;;~ l U-H--t-t-t-H , , I 1 ,9 ., 5 I I .i> O.y '..!),tfl/11 ' .,,1111 11 111 0 15 30 45 60 75 Distance from bed center ( cm) Figure 5-19 Diurnal water and heat dynamics in a mulched bed for a dry soil. a) Contours of temperature (C) and heat flux vectors ( J hr1 ); b) Contours of liquid water content (cm 3 cm3 ) and water flux vectors (cm hr 1 ); c) Contours of liquid water content (cm 3 cm 3 ) and vapor flux vectors (mg cm2 hr 1 ).

PAGE 129

(b) ,,-.... s u ...._,, fr Q 117 ffi hr -I 0 12 : 00 AM 0.0 I I I 1 0 3: 00 PM I I , I I -5 I I I 1 f t I I I I I 1 11,, ,,, , 'fl.' f I I 1 66 11 ''''-10 I I I ' r r 06. I I I I 111, I I I I I I f I I 1 I r r I I I I I I I d / / I I I I I I I I _,, 15 I I I 07 I I I I / I I I d I I I .,, I I I I I I I I / / I I I ,07 -20 I I I I I I I I I I I II/// I I I f I I , I I f I I I IO. 0 0!4 I 1111111 I I I -25 I I 1 ' I I I I I I I f I I I I I I I I I I I ' I I I I I I I ' t t I t I I I I -30 I I I r r ,0 .. 05 I I I I I I I I I I I I I I I 0.05 I I I I I I t I 1 J I 1 1 I I ' r I I r I r I I I I r I r r 1 ' 1 I 1 I I I -35 0 06:00 PM 09:00 PM I I I I I -5 I 0,05 I I I f I I I I ~0 I I f , I I I I I , I I I I I I I ,:J' ' I I I I I I I I I -10 I ' I I I I I I I I I I I I I I I I I l/1;.,,~ I I I I I I I Q.061 , -15 I 0.o 7 I I I I _, ,I I I I I I I I i I I I I I ,, ,,.,.,, I I I I I I .,, I I I I I I I I 1 I I I I 07' I I 20 I I I I I I I I I I I I I I I I I I I I I // 0,0f I , I I I I I 0 04 I I 11111111 -25 I 0.05 I I I I I I I I ' I I I I I I I I I I I I I I f t t I I t I f I -30 I I r 1 I I I I I I I 0.05 I i010S-t I 1 0:os ' I 1 I I r r 1 I 1 r r I I I I I I I I I 1 I I I I I I I I I I r I I r I I I I I I I r I I I I I I I I I I -35 ' 0 o,.04 f I I 12:00 PM I I 03:00 AM f I I I I I -5 ., .,. {) I I I I I f ' I I I I I I I I -~ -10 I I I I I I I I I
PAGE 130

( C) o: 0 15 mg cm 2 ru 1 I I r I -5 I t f f I f I I t I ; . .. .. ,. . -10 .. . . . . ... ... .. 15 t .. -:. -:. -:. ~ --....:.. . . . 12 :00 AM ~2 0 . : . :-o. o?~ ._ :_ i,.... ..:... ---'----'--0.07 . ) 0 -25 I / 0.04 30 t-_: : : 0 . 05_~ -:. : -.:_ : _:_:~t_ ,~:1,::_, 'D 1 ~' 1 [r!~ 1 ~~"[" !.] 1' 1 1 35 .._,__,_~......._...._...._._..__._..._ _____....__,__...__.__,_ __ _,_ .... Ol=========:::-1 1,,-.:.' ....:.' ...:.' ...:. ...:.' ...:.' ...:.' _, I I I I I I I I 0.05 I I I I -1oi:-.-...__.'-l.. --!-L. lfltltl .... ,, -15 6.66 '. -20 ,,,,,, I I I I I I I , 06:00 PM ., 0, -JO r--,-.,....~--...:-...:_:__:,_,:__.:.~ ' 1 I I I I I I I I I 0. 0 5 -J-+--1-t--t-!--t-t-+-+-+-I 0--------------, 'n_J litl# 5~~~:-:-~0.v<+~~.. 12:00 PM -I 0 ' I f I I ' I I I I I -15 .,,('''''' v. 07 ::::'\,: i-:......:__:c.....:......:._.:.----=-~---o. o7 25 t-............. ~~....--.-...--.-..:;.:.:,.~~ -20 ' ' ' ' 1 I I I I I I I I I -35,..__ ___ ..._ _____ .._..__._ __ __._..__._ __.__.__~_.... . . . . -5 I I I I I I I I I I I I I -10 r.--:--:--:-,-...,_ 0 : 0~ -15 . 20 -25 1--.-~-.-.--.-...,.....,--,-0. 05 06:00 AM '--....,_t-t-+-t-; O ,04-+-1 I f I I I I I I I ,,11 11 I -30 Fo.os--'-'_1_1 _,_1 _ _--1 I I I I 118 0.0~ J i:,-,N~ct=t-~t+t-~tJ4--i_ .,. .. c." ....... I,.,...,, 111111111, / ,,, ., .. ., .. ,,.,, .. .. 03:00 PM .. ....... ,,,,,, .. .. ,, ~~ =-~~--.;:.._ :.__ .... # . Oo .. . .. : : : : . '. 7 : : ,, ....... '-------------, . . r n'7. \ ~\..,...-----0 0,_..--J : ... : : : .. :: :,~~~,~ l~ 1 ~ ~H ~O~ O~:+,-H . o .os-: -: ~ :_ ::0 ~1~~ .!,.' _!_l J_i_!)_! 1 _!-_:~ ' 1 J ' 0 0 I ' '1 1 I f I i i I I I-'-, -LO -'....... -!.: _,_: ....!_O ....!....I Q Q 5: i i 09:00 PM ........ ,,, -.. -,-,-,.-. ..... -.!'-'' I I I t t I I J ~' , , -; V6, I I I I 1111111 .,_,__.._._ ...... () 04..........._L....l .. i .,,, ,, o: 0 ..:_:__:_ ' '.....---.--,...,... ............ -1 .., ~0 05,;1,,1111 'l1~'~~~~ ', I t I t t I I 0 04 I I I 03:00 AM ' . ' '. '. o~ :_ 0.07 -I I r 1 I I 0 1 0 ~ I I _,,/ I f l ,11 I I I I I I I I o . o 5_..._.,-L-.._._-L....,__,_....L...1 , I I f t I ' .. ... ,, 09:00 AM . . i=-:--:--~-,-,~o.os--:... . f I I I I I I I I l I t I I 0 . . -07. . : : .. 1a----=-...:.......:-=-_:_.:.-=-"---'-.....;.....o.07 L....-----v V'i---1 \ i i i i i ~1 f I I I I I I I I 0 5 I I I I f I I I I ,,;11111111 35 01--_._ ....__-:-"1 '="5 ........,_..._,,.30=---------4-:'::5:-'-------.......,6-f:o-.__~15 o 1 5 30 ,,,,111,111 45 60 75 Distance from bed cen ter (cm) Figure 5-19-continued

PAGE 131

119 Summary and Conclusions A numerical investigation including sensitivity analysis was conducted using a coupled heat-water and solute transport model developed in the present work, to analyze the fate and transport of MBr during fumigation of soil beneath plastic-mulched beds. The investigation provided improved insight into the dynamics of MBr fumigation in relation to critical management and environmental impacts. Specific observations made during this investigation include: 1. A need exists for improved determination of diffusion coefficients for MBr diffusion through plastic mulch under field conditions. Use of lab-measured data for field simulations under predicted MBr emission losses implies that polyethylene-plastic mulch is relatively effective in controlling atmospheric losses of MBr. However, this result is in direct contradiction to reported field studies from the scientific literature. 2. Non-isothermal soil conditions were shown to be especially important to solute transport in plastic-mulched soil beds, particularly in cases where highly volatile chemicals like MBr are used for soil fumigation. 3. A wetter soil profile was shown to retain more (approximately 10-20 % depending on injection depth over 3 day period) MBr than a drier one due to partitioning into water, but also to result in more bromide production due to increased hydrolysis. Higher wetness of the soil restricted the MBr passageway from spreading quickly. Substantial reductions (approximately

PAGE 132

120 23-36 % depending on injection depth over 3 day period) in MBr emission losses to the atmosphere were observed due to higher water-saturation of the soil. 4. Deeper injection (66 cm) of MBr fumigant provided less (approximately 10 and 23 % in dry and wet soil, respectively) emission losses to the atmospher e, irrespective of soil wetness, than a shallow injection zone (33 cm). 5. Fumigant efficiency analysis based upon Mckenry and Hesse (1978) data f o r nematoxic concentration and exposure time for root knot nematode showed that current practice of MBr fumigation and dosage used (450 kg ha 1 ) provides adequate soil sterilization in the crop root zone of the bed However it was also revealed that deeper injection (66 cm) is not conducive for sterilization of the crop root zone even though it is environmentally less hazardous. Such analysis is site-specific and does not apply to general recommendations. 6. For a sandy soil, dousing of water on the soil bed immediately after injection and before placing the plastic mulch was shown to slightly decrease MBr emission losses to the atmosphere, by sealing off the soil pores to gaseous diffusion during the initial hours after mulch placement. Dousing of finer textured soil would be expected to provide more effective control of MBr . ermss1ons. 7. Utilizing a slightly wet profile a slightly deeper injection zone and dousing water on the surface of the bed offers potential fumigant management opportunities to decrease MBr atmospheric losses.

PAGE 133

8. Inforrr1ation on hydrolysis of MBr in Florida soils is lacking. Sensitivity analysis showed that a linear relation of hydrolysis rate occurred with the amount of MBr hydrolyzed. This information is critical for improved estimation of MBr fate and transport in wet soils. 121 9. Information on pathogen toxic concentration and exposure time is quite limited. This information is very important to optimize fumigant dosage in order to achieve effective soil sterilization and avoid environmental hazards through overfumigation. 10. Decreased atmospheric emission of MBr due to improved fumigant management in agricultural soils may result in greater MBr retention by soils and ultimate decomposition to bromide ions during hydrolysis. However, this may undesirably increase bromide concentration in crops and groundwater over long periods of time. 11. Carefully-controlled field experiments are needed to validate conclusions based upon the model simulations. 12. Results from model simulations provide improved insight of system dynamics, as needed in the design of future field experiments.

PAGE 134

APPENDIX A SOIL MOISTURE FLOW The total mass of water in the system may be defined as p MW= p 8 +p 8 L L L v a [A-1] where 1vr represents the total water (liquid+ vapor) storage. The liquid is assumed here to be incompressible. The liquid water flux for an isotropic unsaturated media is defined by Darcy's equation as Equation [A-2] applies to media with properties either constant/or continuously varying in space. [A-2] An intuitive extension of Fick's law applied to porous media for vapor water flow (Philip and de Vries, 1957) can be expressed as q = -D vi: 8 v7 p v a a a v [A-3] where v is a mass flow factor and -r a is tortuosi ty of the air-filled pore space. Water vapor distribution in the air phase is found to be (Edle fson and Anderson, 1943) p = p R = V O h hg p ( T) exp a 0 R T [A-4] u where p 0 is saturation water vapor density over a free water surface, Rh is relative 122

PAGE 135

123 humidity, and ga is the acceleration of gravity. Equation [A-4) is based on the assumption of thermodynamic equilibrium between the liquid and vapor phases and is valid for the natural system. Expanding P v( h T) in Eq. [A-3) and with v being generally reported as unity, it follows that ap q = -D 1 e v V a a a ah ap Vh + V VT T aT h [A-5] Introducing a correction factor for 'liquid islands' and the temperature gradient differences between the whole media and the air-filled pores as described by Philip and de Vries (1957) leads to ap ap q = n 1 e v I TVh n le + eL) a v I h~ VT; V a a a ah a\ a T (VT) a = VT where ('vT ) 8 is the average temperature gradient in the air phase. [A-6] For generalizing over a range of e L values, where the liquid phase is continuous, the quantity ( 8 a + 8 L) in Eq. [A-6] is replaced by ( 8 a + f 8 ), where f goes to zero as 8 a goes to zero. The forrn of f is linear with e a in this regime. This leads to where D = hv ap ap -D t 8 vlT Vh -D f~ vlh VT a a a ah a aT D = Tv D ap _a f~ V P aT h L [A-7)

PAGE 136

124 The total mass flux can be defined as [A-8] -(K + D )v h D 'v T K k hv Tv The continuity equation of water mass now can be stated as = V[(K+D )'vh+D 'vT+Kk] hv Tv [A-9) with, a(e +e)= at L a aeL a at at [A-10) i t follows that, P ae e ap lv L+ a v='v[(K+D )'vh+D 'vT+Kk] p at p at hv Tv L L [A-11] Considering that e L = e L(h, T) and P v = P v(h, T), it is seen that [A-12) Substituting Eq. [A-12) into Eq. [A-11) and considering the effect of heat of wetting (Kay and Groeneve l t, 1974) on the pressure-head field and resulting flow leads to aeL ea apv ah P v aeL ea apv aT -a-h IT+-PL_a_h lr_a_t + l--p-L -a-T l h+-PL_a_T l h at [A-13] = v[(K+ Dhv)'vh + (nrv + Dra)VT + Kk] where Dra is a transport coefficient for adsorbed liquid (film) flow due to thermal gradients. However, in sandy soils with the specific surface area being comparatively very low, heat of wetting can be safely neglected as shown by Milly and Eagleson (1980) to obtain the form of Eq. [3.1) which was given in Chapter-3

PAGE 137

APPENDIXB SOIL HEAT FLOW The total flux of sensible and latent heat in a porous medium, considering convection of air and radiative heat transfer negligible, is given by de Vries (1958) as q =-A VT+L q +C (T-T )q +C (T-T )q h Ov v O v L O L in which A accounts for the Fourier heat diffusion. From basic thermodynamic relationships it is observed that L +C(T-T)=L+C(T-T) ~L =L+C(T-T )-c(T-T) 0 V O L O O L O V 0 which gives Substituting Eq. [B-2] into Eq. [B-1) leads to Including advection of the heat of wetting due to pressure-head gradients ( Groenevelt and Kay, 1974), and substituting Eq [A-7] for qv, it follows that qh =-A-LPL (DhvVh+DTvVT) +pL gaT DTa Vh +CL (T-T 0 ) qw ~qh =-(A+ LpLDTv )vT-pL (LDhv + gaT DTa )Vh+CL (T-To) qw qh = -A* VT PL(L Dhv+gaTDTa)Vh+cL(T-To)qw where A* accounts for the combined effect of simple Fourier heat diffusion and latent heat transport by temperature-induced vapor diffusion. 125 [B-1] [B-2) [B-3] [B-4]

PAGE 138

The bulk volumetric heat content of the porous medium, H, is given by H = cd(T-To) + LOpvea + Cvpvea(T-To) + CLpLeL(T-T o ) t d8 -pL f w Ldt dt t o t d8 H = (c + C p 8 + C p 8 ) (TT ) + L p 8 p f W L dt d L L L v v a o o v a L dt t o 126 [B-5) This equation assumes local thermal equilibrium among soil particles, the water, and the air. Equating the change in stored energy, Eq. [B-5], to the sum of the convergence of the heat flux, Eq. [B-4], yields ap aeL c + H1 a; h + H2 a T h aT ap + H V at 1 ah v[ ./\ V T + p L ( L D hv + g a TD Ta ) V h CL ( T T O ) q w ] where C = Cd + CL p Le L + CV p V e a H l = l L O + C V ( T T O ) J 8 a H = (c p C p )(T T )p W p L 2 LL vv O L VO [B-6] W, the differential heat of wetting (de Vries, 1958; Edlefson and Anderson, 1943) is defined as w = 1 h T ah I J ga aT e [B-7] in whichj is a conversion factor relating mechanical work to then11al energy. The quantity Wis the amount of heat released per unit mass of added water when an infinitesimal quantity of free liquid water is added to the medium. In sandy soils, surface area being small, the heat of wetting can be neglected to give the form of Eq. [3.2] presented in Chapter-3.

PAGE 139

APPENDIXC CHEMICAL TRANSPORT V V Simunek et al. (1992) and Simunek and van Genuchten (1994) presented equations for solute fate and transport assuming that solutes can exist in all three phases (liquid, solid, and gaseous) and that the decay and production processes ar e different in each case. Non-equilibrium solute transport was considered to be caused by the kinetic sorption on some (or all) of the soil sorption sites Solutes were considered to be transported by three different mechanisms: by convection and dispersion in the liquid phase, and by diffusion in the gas phase. Following their approach, which is relatively general the governing Eqs [3.4], [3.5] were formulated. Non-equilibrium interaction between the solution (c) and adsorbed (s) concentrations of solutes, and equilibrium interaction between the solution (c) and gas (g) concentrations, can be assumed for solutes in the soil system. Following V Simunek and van Genuchten ( 1 994), the adsorption isotherm relating s i and cj (where j designates the specie number) can be described via a general nonlinear empirical equation as l. at s = l. 131 1 kd ~ Ci l. l. 13 2 l+n i c i .1 ac. l. + at i=l .. n s 13 1 akd C l. I l. 13 at k 213 d C i 1. l+nici i 13 l+n ici .1 127 i=l .. n s an. l. 2 at [C-1]

PAGE 140

128 where kd t, P 1 and 11 1 are empirical constants. The Freund l ich, Lang1r1uir, and linear adsorption equations are special cases of Eq. [C1 ]). When p j = 1 Eq. [C-1]) becomes the Freundlich equation, and when both P 1 = 1 and 11 = 0, Eq. [C-1] there results a linear adsorption isotherm. Solute transport without adsorption is described when kd i = 0. The concept of two reaction sites on soil surfaces (van Genuchten and Wagnet, 1 989) assumes that sorption sites can be divided into two fractions as i = l .. n 5 [C-2] Sorption on the first fraction, s 1 8 is assumed to be instantaneous, while sorption on the second (remaining) fraction, s 1 k, is considered to be time-dependent. At equilibrium, for Type1 (equilibrium) and Type-2 (kinetic) sites, it follows that i = l .. n s [C-3) where~ is the fraction of sorption sites assumed to be at equilibrium with the solution. The time dependence of Type-1 sites represents the sorption on these sites as at as = f l. r at i = l .. n 5 [C-4] The transfer of mass from the solution to Type-2 sorption sites can be described by a first-order kinetic sorption law. The mass balance for Type-2 sites then becomes i = l, .. n [C-5] where w is the fi r st-order rate adsorption constant for the i th solute.

PAGE 141

129 The concentrations gi (gaseous) and ci (solution) can be related by a linear relation as g = k c.; .l g,i .l i=l .. n s [C-6] where kg ,i is an empirica l constant equal to (K~un 1 (Stumm and Morgan, 1981) where KH is obtained from Henry's Law. Substituting Eqs. [C-1] through [C-6] and the continuity equation describing flow of liquid water in a variably saturated medium aeL at = -'\lqL into Eqs. [3.4] and [3 5) yields aci ( ) r SLR.--+ qL'vc -'v 8LD. 'vc. T 1 c. G. 1 at i .l 1 .l 1 = 0 ; i=l, . n 5 The effective dispersion coefficient tensor Di can be described as e v = e 1 v 1 1 + e v 1 .g k ; i a g,i i=l .. n s The coefficients T i' and Gi in Eq. [C-8] are defined as r * [C-7] [C-8] [C-9] Ti C = L + L i' e + s .1 l. i L s, l. + 8 k s g s,g a g,i + k g' l. Bk e g l. a at [C-10] i = l .. n 5 [C-11]

PAGE 142

where gi is defined a s 13 i akd k 213i gi(c i ) Ci d C pbfr .l .l 13 i at 13 2 1 + Il1C1 1 +nici l i=l .. n s The retardation factor R 1 is defined as 13. 1 e R1(ci) Pb f kd r3 Ci .l k 1 + r ,.1 .l + a g, .l 8L 13 2 8L 1 +nici .l k s .l an .l + at i =2 .. n s 13 ar3 kd ln c c i .1 / .l .l .l 13 2 at 1 +n icil i=l .. n s 130 [C 12] f [C-13] [C-14] In order to solve Eq. [C1 0], it is necessary to kno w the liquid water content and the volumet r ic flux Both va r iables are obtained the solutions of the modified Philip and de Vries equation. If i t is assumed that L s * 11 1 and kg are equal to zero, and that I, and P i are equal to one, the entire system of Eqs. (3. 4] [ 3.5], and [C-1] through [C-14] simplifies into a system of mutually-independent solutions as described by '>J Sim un e k et al. (1992). The components of the dispersion tensor in the liquid phase Ii', were give n by Bear ( 1 972) as [C-15] where Dw is the molecular diffusion coefficient in free water, "C w is a tortuosity factor in the liquid phase, I qi is t h e absolute value of the fluid fl ux density, o ii is the Kronecker delta function ( 0 1 i = 1 if i = j, and 0 1 i = 0 if i j), and D L and D r are the

PAGE 143

131 longitudinal and transverse dispersivities, respectively. After adding the diffusion contribution from the gas phase, the individual components of the effective dispersion tensor in the soi l matrix for two-dimensional transport are as follows 8D xx +8DT +8D k-r Lww agga 2 qLX + DT-+ e D 1 + e D k "( Lww agga qL 8 D = (D D ) qLX qLz L xz L T qL I where Dg is the molecular diffusion coefficient in the gas phase and i-a is the tortuosity factor in the gas phase 8 7/3 L 82 s and Ta 8 7/3 a 82 s [C-16) [C-17]

PAGE 144

APPENDIXD NUMERICAL APPROACH The Finite Element Method presented here follows the methodology given by Istok (1989) and Segerlind (1984). A short description of the method as related to the present work is presented here. A more detailed and complete description of the method can be found in Pinder and Gray (1977), Segerlind (1984), and Istok (1989) The first step in the solution of a water/heat flow or solute-transport problem by the finite-element method is to discretize the problem domain (the soil profile). Discretization is done by replacing the problem domain with a collection of nodes and elements referred to as the finite-element mesh. Different element types are available for one-, two-, and three-dimensional problems, and also for problems with axisymmetry. Elements may be of any size, the size and shape of each e lem ent in the mesh can be different, and several different types of elements can be used in a single mesh. Values of the material properties are usually assumed to be constant within each element, but are allowed to vary from one element to the next. When preparing the finite-element mesh it is important to remember that the precision of the solution obtained and the level of computational effort required to obtain a solution are determined to a great extent by the number of nodes in the mesh. A coarse mesh has a smaller number of nodes and will give a lower precision than a fine mesh. However, the larger the number of nodes in the mesh, the greater will be the required computational effort and cost. 132

PAGE 145

133 Each node in the mesh is assigned a unique node number with no skips inbetween. The nodes are then assigned a set of nodal coordinates. These could be (x), (x,y), (x,y,z), or (r,z) coordinates of the nodes. The nodes have to be numbered in a fashion so as to minimize the semi-bandwidth (SBW) of the matrix for the resulting system of linear equations. When the SBW is large and the system of equations is operated on in matrix form, the storage capacity of current desktop computers can be easily exceeded. The SBW for any mesh can be computed from: SBW = R node + 1, where R nocte is the maximum difference between any two node numbers within a single element in the mesh. The Method of Weighted Residuals The second step in solving the governing flow or transport equation involves deriving an integral forrr1ulation. This integral formulation leads to a system of algebraic equations that can be solved for the values of the field variables (h, Tor c for this work). In the method of weighted residuals, an approximate solution to the boundaryor initial-value problem is defined. When this approximate solution is substituted into the governing differential equation, an error or residual occurs at each point in the problem domain. The weighted average of the residuals for each node is then forced to zero. Let us consider a differential equation of the form L(

is the field variable and Fis a known

PAGE 146

A function. We can define an approximate solution cl> of the forrn m (x y z) = N. (x y z) cl> L..J i i i=l where N 1 are interpolation functions,

( X y z)) F (X I y z) J dQ n [D-5] A To evaluate Eq. [D-5], the mathematical form of the approximate solution A and the weighting function W must be specified. ct> is defined in a piece-wise fashion over the problem domain The value of & within any element e i i=l [D-6]

PAGE 147

135 where ~ (e) are interpolation functions and n is the number of nodes within the element. It is to be noted that in Galerkin 1 s Method, the weighting function for a node is identical to the interpolation function used to define the approximate A solution, Galerkin's method is a subset of the method of weighted residuals that is most commonly used to solve water-flow and solute-transport problems for a porous medium. A After specifying the form of and W the integral in Eq. [D-5) is evaluated in order to obtain a system of linear equations of the form l CJ{} = {F} [D-7) that can be solved for the values of at each node in the mesh. Evaluation of the integral containing the time derivative of requires that the time derivative of be defined over the volume of the element. This can be done using interpolation functions and the values of the time derivatives at the nodes, in the same manner that is defined over the volume of the element using the interpolation functions and the values of at the nodes. Depending on the type of interpolation function used, the procedure is called either a 'con sistent or a 'lumped' element formulation. Both formulations are used in practice. However the lumped formulation is less susceptible to problems of numerical oscillation' than is the consistent formulation (Segerlind, 1984) The lumped formulation assumes that the variation of a <1> 1at with respect to x y, and z is constant between the midpoints of adjacent elements and that the weighting functions are the same as used for the spatial derivatives.

PAGE 148

Element Type and Interpolation Function A commonly used 2-D element a linear triangle (Fig. D-1), was used in the present work Linear interpolation functions for this type of element are derived in Segerlind ( 1984 ). The 1 ( X ( e ) z ( e ) ) k k (e) z 136 ] (~ (e) 1 ; (e) ) interpolation functions are also commonly Figure D-1 Linear triangular element referred to as shape functions or as basis functions. For the 2-D triangular element used in this work, the functions were Ni ( e ) (x z) 1 (a +b x+c z) 2A (e) l l i N( e ) (x z) 1 (a +b x+c z ) [D 8] J 2A (el 1 J 1 N )el ( x z ) 1 (ak +b k x + ckz) 2A (el where a ( e ) ( e ) (e ) ( e ) b ( e ) ( e) C ( e) ( e) [D-9] xk zi xi zk z k z X x J J l J l k a k (e) (e) (e) (e) b k ( e) ( e) Ck ( e) ( e) X Z X Z z z X X 1 l J J l l J ] and 1 (e) ( e ) x z I I A r e ; = 1 ( e ) (e) [D-10] x z J J 1 X k (e) z k (e)

PAGE 149

137 Derivati v es of the interpolation functions are aN _( e) b aN _( e) b aN (el bk l l J J k ax 2A (e) ax 2A (e) ax 2A (e) aN _( e) [D-11) C aN _( e) cj aN}e) Ck l l. J az 2A (el az 2A (el az 2A (el The element matr i ces for the linear triangle element can be estimated as (Segerlind, 1984) J N _( e) a N e) b N ( e) c dA l J k A ( el __ a_! _b_ _c_ 2 A ( e) ( a + b + c) [D-12) where a, b, and care exponents of the interpolation functions (el, ~ (el and Nk (e)_ Matrix Equations Application of the finite-element method, as described, results in the linearized matrix form of the water flow governing equation where elements in then x n coefficient matrices and n x 1 vectors are Aw .. = l] ~ e ( b b + c c ) C l J l. J w 4A e ( b b + c c ) D l.J l.J w 4A e (m=l 2) [D-13) [D-14] The values of the coefficients Cw, Dw, and Ew for each element are estimated as the

PAGE 150

138 average of the nodal values. The subscripts i, j, and k on the Aw and Bw coefficients indicate the value of the coefficient at the respective local node of a particular element. The Ow vector is used to satisfy a Neumann boundary condition where Vw is a prescribed or known flux density across the side of the element, Ls is the length of the boundary side, and m represents the local nodes of the element side on the boundary. For an isotropic heterogenous medium the coefficients in Eq. [D-13] are described as (Milly and Eagleson, 1980) A = w B = w e ap + a V h PL aT h C = K+D w hv D = D w Tv E = Kk w [D-13a] [D-13b] [D-13c] [D-13d] [D-13e] A fully implicit finite-difference scheme was used to integrate the time derivatives. The time domain was divided into a sequence of finite intervals, ei.t, and the time derivatives were replaced by finite differences. After rearrangement to combine known terms on the right hand side of the equation, the resultant matrix equation was obtained as [D-15]

PAGE 151

139 where p denotes the current time step for which a solution is sought and p-1 denotes the previous time step. Storage and transport properties of the elements in Aw, Bw, Cw, Dw, and Ew are evaluated at the current time step. An exactly similar procedure was applied to develop the matrix equation for heat flow. The two equations for water flow and heat flow, being highly nonlinear in nature, are solved iteratively and sequentially until a desired level of convergence is reached. The standard Galerkin's method was also applied to solute transport, resulting in [D-16] where [R] is the retardation matrix, [OS] is the convection matrix, [D] is the diffusion matrix, [Tr] is the transformation matrix, [ G] is the temperature effects vector and [ 0] is the force vector. Elements in the n x n coefficient matrices and n x 1 vectors are r T . .l] b.qL +c qL .l X .l Z 6 b.b D +(b.b +c c.)D +c.c.D .l J xx .l J .l J xz .l J zz 4A e A _e (F + F )(1 + o .. ) G. 2 4 .l J .lJ l A _eG 3 l [D-17]

PAGE 152

A fully implicit finite-difference method was used to integrate the time derivative to obtain where the superscript p represents a future time level. 140 [D-18] The matrix equations were programmed in FORTRAN-90 language using Microsoft Fortran PowerStation TM Version 4.0 (Microsoft Corporation, 1995) 1 Matrix solvers provided with the package (Microsoft IMSL mathematical and statistical libraries 2 ) were used for solving the assembled matrices. The simulation results from the model were visualized using TECPLOT (Amtec Eng. Inc., 1996) 3 Mulch Energy Balance Given known meteorological conditions, system optical and thermal properties, and temperatures T 8 and T 1 Eq. [3.7] and [3.8) are unique functions of T s and Tm An algorithm was developed for solving nonlinear equations using the 1 Microsoft is a registered trademark, and PowerStation is-trademark, of Microsoft Corporation in the United States and/or other countries. 2 IMSL is a registered trademark of Visual Numerics, Inc. 3 TECPLOT is a registered trademark of Amtec Engineering, Inc.

PAGE 153

141 Newton-Raphson technique (Press et al., 1986) Partial derivatives required for the Newton-Raphson solution are as follows 1 1 [D-19] where functions f 1 and f 2 represent energy balance Eqs. [3.7] and [3.8), respectively

PAGE 154

LIST OF REFERENCES Abdalla, N., D.J. Raski, B. Lear, and R.V. Schmitt. 1974. Distribution of methyl bromide in soils treated for nematode control in replant vineyards. Pestic. Sci. 5: 259-269. Amtec Eng., Inc. 1996. TECPLOT Version 7: Productive Power for Data Visualization. Bellevue, WA. Anderson, D.M. and A. Linville. 1960 Temperature fluctuations accompanying water movement through porous media. Science 101: 1370-1371. Anderson, M.P. and W.W. Woessner. 1992. Applied Groundwater Modeling: Simulation of Flow and Advecive Transport Academic Press, Inc., San Diego, CA. Bach, L.B. 1992. Soil water movement in response to temperature gradients: experimental measurements and model evaluation. Soil Sci. Soc. Am. J. 56: 37-46. Barnes, C.J. and G.B. Allison. 1984. The distribution of deuterium and 18 0 in dry soils. 3. Theory for non-isothermal water movement. J. Hydrol. 74: 119 -135. Barnes, C.J., G.B. Allison, and M .W. Hughes. 1989. Temperature gradient effects on stable isotope and chloride profiles in dry soils. J. Hydrol. 112: 69-87. Baver, L.D. W.H. Gardner, and W.R. Gardner. 1972. Soil Physics. John Wiley and Sons, Inc., New York Bear, J. 1972 Dynamics of Fluid in Porous Media Elsevier, New York. Bear, J. and A. Verruijt. 1987. Modeling Groundwater Flow and Pollution. D. Reidel Publishing Company, Dordrecht, Holland Benjamin, J.G. 1989. Coupled water and heat transport in nonunifrom soils. Ph.D. Thesis. Iowa State Univ ersity, Ames, Iowa. Bird, R.B., W.E. Stewart, and E.N. Lightfoot. 1960. Transport Phenom enon. J ohn Wiley and Sons Inc ., New York. 142

PAGE 155

143 Brown, B.D. and D.E. Rolston. 1980. Transport and Transforrr1ation of methyl bromide in soils. Soil Sci. 130: 68-75. Cary, J.W. 1966. Soil moisture transport due to thermal gradients: practical aspects. Soil Sci. Soc. Am. Proc 30: 428-433 Cary, J.W. and S.A. Taylor. 1962a. The interaction of the simultaneous diffusion of heat and water vapor. Soil Sci. Soc. Am. Proc 26: 413-416. Cary, J.W. and S.A. Taylor. 1962b Thenr1ally driven liquid and vapor phase transfer of water and energy in soil. Soil Sci. Soc. Am. Proc. 26: 417-420. Cassel, D.K., D.R. Nielsen and J.W. Biggar. 1969. Soil water movement in response to imposed temperature gradients. Soil Sci. Soc. Am. Proc. 33: 493-500 Castro, C.E. and N.O. Belser 1981 Photohydrolysis of methyl bromide and chloropicrin. J. Agric. Food. Chem. 29: 1005-1008. Cho, C.M. 1971. Convective transport of ammonium with nitrification in soil. Can. J. Soil Sci. 51: 339-350. Chung, S.O. and R Horton. 1987 Soil heat and water flow with a partial surface mulch. Water Resour. Res. 23: 2175-2186. Clark, M.M 1996 Transport Modeling for Environmental Engineers and Scientists. John Wiley & Sons, Inc ., New York. Cohen, Y. and P.A. Ryan. 1989. Chemical transport in the top soil zone the role of moisture and temperature gradients. J. Haz. Materials 22: 283-304. Cohen, Y., H. Taghavi and P.A Ryan. 1988 Chemical volatilization in nearly dry soils under non-isotherrr1al conditions. J. Environ. Oual. 17 : 198-204. Constantz, J. 1982 Temperature dependence of unsaturated hydraulic conductivity of two soils. Soil Sci. Soc. Am. J. 46: 466-470. Crank, J. 1975. The Mathematics of Diffusion. Oxford University Press, London, Great Britain Daniels, F. and R.A. Alberty. 1961 Physical Chemistry. John Wiley & Sons, Inc., New York. de Silans, P.A. L. Bruckler, J.L. Thony and M. Vauclin 1989 Numerical modeling of coupled heat and water flows during drying in a stratified bare soil comparisons with field observations. J. Hydro]. 105: 109-138.

PAGE 156

144 de Vries, D A. 1958. Simultaneous transfer of heat and moisture in porous media. Trans. Am. Geophys. Union 39: 909-916. de Vries, D.A. 1963 Thermal properties of soils. In: Physics of Plant Environment (Ed. W.R. van Wijk). North Holland Publishing Co., Amsterdam pp 210-235. Edlefsen, N E. and A.B.C. Anderson. 1943. Therr11odynamics of soil moisture. Hilgardia 15 : 31-298. Ekern, P.C. 1967. Soil moisture and soil temperature changes with the use of black vapor-barrier mulch and their influence on pineapple growth in Hawaii. Soil Sci. Soc. Am. Proc. 31: 270-275. Enfield C.G. 1970. An irreversible ther1r1odynamic approach to simultaneous movement of water, heat, and salt through unsaturated soils. Report UC-51, Geology and Mineralogy Battelle Memorial Institute, Pacific Northwest Laboratories, Richland, WA. Evans, D.D. and T J Nicholson. 1987 Flow and transport through unsaturated fractured rock: An overview. In: Flow and Transport Through Unsaturated Fractured Rock (Eds. D .D. Evans and T.J. Nicholson). Geophysical Monograph 42, pp 1-10 EXTONET. 1997. The EXtension TOXicology NETwork A pesticide information project of cooperative extension offices at Cornell Univ., Michigan State Univ. Oregan State Univ., and Univ. of California at Davis http://ace.ace.orst.edu/info/extoxnet/. Friebele, E. 1997. Partial picture of methyl bromide demands more pieces. EOS, Trans. Am Geophys. Union 78: 245-246. Friedman R., C. Ansell, S. Diamond, and Y.Y. Haimes 1984. The use of models for water resources management, planning and policy. Water Resour. Res. 20: 793-802. Geiger, R 1957. The Climate Near the Ground. Harvard University Press, Cambridge, MA. Gentile. I .A., L Ferraris, S Crespi, and A. Belligno. 1989. The degradation of methyl bromide in some natural fresh water influence of temperature, pH, and light. Pestic. Sci. 25: 261-272. Gentile, I .A., L. Ferraris M. Sanguinetti, M. Tiprigan and G. Fisich ella. 1992 Methyl bromide in natural fresh waters hydrolysis and volatilisation. Pestic. Sci. 34: 297-301.

PAGE 157

145 Ghildyal, B.P. and R P Tripathi. 1987. Soil Physics. Wiley Eastern Limited, New Delhi, India. Goring, C.A.I. 1962. Theory and principles of soil fumigation. Adv. Pest Control Res. 5: 47-84 Groenevelt, P.H. and B.D. Kay. 1974 On the interaction of water and heat transport in frozen and unfrozen soils, 2. The liquid phase. Soil Sci. Soc. Am. Proc. 38: 400-404. Grojesan, D. 1991. Atmospheric chemistry of toxic contaminants. 4. Saturated halogenated aliphatics: methyl bromide, epichlorhydrin, phosgene. J. Air Waste Manage. Assoc. 41 : 56-61. Gurr, C.G., T.J. Marshall, and J.T. Hutton. 1952. Movement of water in soil in response to temperature gradients. Soil Sci. 74: 335-345. Hacher!, E.L. 1994. Diffusion of methyl bromide through polyethylene films and soil: Effects of temperature and moisture content. M.S. Thesis. University of Florida, Gainesville, FL. Ham, J.M. and G.J. Kluitenberg. 1994. Modeling the effect of mulch optical properties and mulch-soil contact resistance on soil heating under plastic mulch culture. Agric. For. Meteorol. 71: 403-424. Ham, J.M., G.J. Kluitenberg, and W.J. Lamont. 1993. Optical properties of plastic mulches affect the field temperature regime. J. Am. Soc. Hort. Sci. 118: 188 193. Hares, M.A. and M .D. Novak 1992. Simulation of surface energy balance and soil temperature under strip tillage: I. Model description. Soil Sci. Soc. Am. J. 56: 22-29. Havercamp R., M Vauclin, J. Tovina, P.J. Wierenga, and G. Vachaud. 1977. A comparison of numerical simulation models for one-dimensional infiltration. Soil Sci. Soc. Am. Proc 41: 285-294. Hemwall, J.B. 1959 A mathematical theory of soil fumigation. Soil Sci. 88: 184-190. Hemwall, J B. 1960 Theoretical considerations of several factors influencing the effectivity of soil fumigants under field conditions. Soil Sci. 90: 157-168. Hillel D. 1980. Fundamentals of Soil Physics. Academic Press Inc., New York.

PAGE 158

146 Hoffman, G.M. and H.P Malkomes. 1979 The fate of fumigants. In: Soil Disinfestation (Ed. D. Mulder). Elsevier Scientific Publishing Co., Amsterdam, The Netherlands. pp. 291-335 Hornsby, A.G., R D Wauchope, and A.E. Herner. 1996. Pesticide Properties in the Environment Springer-Verlag New York, Inc ., New York. Horton R. and S. Chung. 1991. Soil Heat Flow. In: Modeling Plant and Soil Systems (Eds. J. Hanks and J Ritchie) Monograph 31 Am. Soc. of Agronomy Madison, WI. pp. 397-348. Istok, J. 1989. Groundwater Modeling by the Finite Element Method Water Resources Monograph 13 American Geophysical Union, Washington, DC Jackson, R.D. 1967 Osmotic effects on water flow through a ceramic filter. Soil Sci Soc.Am.Proc. 31: 713-715 Jin, Y. and W.A Jury 1995. Methyl bromide diffusion and emission through s o il columns under various management techniques. J. Environ. Oual. 24 : 10021009. Jones H.E. and H Kohnke. 1952. The influence of soil moisture tension on vapor movement of soil water Soil Sci Soc. Am. Proc. 16 : 245-248 Jury, W.A. 1973. Simultaneous transport of heat and moisture through a medium sand PhD Thesis, University of Wisconsin, Madison, WI. Kay, B.D and P H Groenevelt 1974. On the interaction of water and heat in frozen and unfrozen soils, 1. Basic theory; the vapor phase. Soil Sci. Soc. Am. Proc 38: 395-400. Kemper, W.D. and D.E.L. Maasland. 1964. Reduction in salt content of solution o n passing through thin films adjacent to charged surfaces. Soil Sci. Soc Am Proc. 28: 318-823. Kemper W.D. and J B. Rollins. 1966. Osmotic efficiency coefficients across compacted clays. Soil Sci. Soc. Am. Proc 30: 529-534. Khalil, M.A.K., R.A Rasmussen, and R. Gunawardena. 1993. Atmospheric methyl bromide trends and global mass balance J. Geophys. Res. 98 : 2887-2896. Kolbezan, M.J. and Fawzi J Abu-El-Haj. 1977. Permeability of plastic films to fumigants. Proc. International Agn ultural Plastics Congress, San Diego CA, April 11-16

PAGE 159

147 Leij, F.J. and S.A. Bradford. 1994. 3DADE: A computer program for evaluating three dimensional equilibrium solute transport in porous media. Research Report No. 134, U.S. Salinity Lab., USDA, ARS, Riverside, CA. Letey, J. and W.D. Kemper. 1969. Movement of water and salt through a clay-water system experimental verification of Onsager reciprocal relation. Soil Sci. Soc. A:m.Proc. 33: 25-29. Lobert, J.M., D.H. Scharffe, W.M. Hao, and P.J. Crutzen. 1990. Importance of biomass burning in the atmospheric budgets of nitrogen-containing gases. Nature 346: 552-554. Low, P.F. 1955. Effect of osmotic pressure on diffusion rate of water. Soil Sci. 80: 95100. Macartney, L. and T.V. Price. 1988. Bromide residues in glasshouse soils in Victoria following bromomethane fumigation. Soil Bio. Biochem. 20: 393-397. Maharajh, D.M. and J. Walkley. 1973. The temperature dependence of Ar, CO 2 CH 4 CH 3 CL, CH 3 BR, and CHCL 2 F in water. Can. J. Chem. 51: 944-952. Mahrer, Y., 0. Naot, E. Rawitz, and J. Katan. 1984. Temperature and moisture regimes in soils mulched with transparent polyethylene. Soil Sci. Soc. A:m. J. 48: 362-367. Mano, S. and M.O. Andreae. 1994. Emission of methyl bromide rom biomass burning. Science 263: 1255-1257. Marshall, T.J. and J.W. Holmes. 1979. Soil Physics. Cambridge University Press, Cambridge. Maw, G.A. and R.J. Kempton. 1973. Methyl bromide as a soil fumigant. Soils Fert. 36: 41-47. Maw, G.A. and R.J. Kempton. 1982. Bromine in soils and peats. Plant and Soil. 65: 103-109. Mckenry, M.V. and C.O. Hesse. 1978. Preplant fumigations of planting sites. California Agriculture 32: 14-15.

PAGE 160

148 Microsoft Corporation. 1995. Microsoft Fortran PowerStation TM Version 4.0 1 : Development System for Windows 95 TM and Windows NT Workstation. Redmond, WA. Mignard, E. and J.C Benet. 1989. Diffusion of methyl bromide in soil. J. Soil Sci. 40: 151-165. Milly, P .C.D. 1984 A simulation analysis of thermal effect on evaporation from soil. Water Resour. Res. 20: 1087-1098. Milly, P.C.D. and P.S Eagleson. 1980 The coupled transport of water and heat in a vertical soil column under atmospheric excitation. Tech. Rep. no. 258. R.M. Parsons Lab., Dep. Civil Eng., Mass Inst. Technol., Cambridge. Maje, W. 1960. The chemisty and nematocidal activity of organic halides. Adv. Pest Control Res. 3: 181-217. Mutziger, A.J., S.R. Yates, D. Wang, W.F. Spencer, F.F. Ernst, and J. Gan. 1996 Modeling methyl bromide diffusion in a field soil. Soil Sci. Soc. Am. Annual Meeting, Indianapolis, Indiana Nov 3-8, 1996. Nassar, I.N. and R. Horton. 1989a. Water transport in unsaturated non-isothem1al salty soil: I. Experimental results. Soil Sci. Soc. Am. J. 53: 1330-1337 Nassar I.N. and R Horton. 1989b. Water transport in unsaturated non-isothermal salty soil: II. Theoretical development. Soil Sci. Soc Am. J. 53: 1330-1337 Nassar I.N. and R. Horton 1992. Simultaneous transfer of heat, water, and solute in porous media: I. Theoretical development. Soil Sci. Soc. Am. J. 56: 1350-1356. Noborio, K. 1995. A two-dimensional finite-element model for solution, heat, and solute transport in furrow-irrigated soil. Ph.D. Thesis. Texas A & M University, College Station, TX. Noggle, J.H. 1985. Physical Chemistry. Little Brown and Co., Boston. 1 Microsoft and Windows 95 are registered trademarks, and Windows NT and PowerStation are trademarks, of Microsoft Corporation in the United States and/or other countries.

PAGE 161

149 Philip, J.R. and D.A. de Vries. 1957. Moisture movement in porous materials under temperature gradients. Trans. Am. Geophys. Union 38: 222-232 Pinder, G.F. and W.G. Gray. 1977. Finite Element Simulation in Surface and Subsurface Hydrology. Academic Press, New York. Press, W.H., B.P. Flannery, S.A. Teukolsky, and W.T. Vetterling. 1986. Numerical Recipes: The Art of Scientific Computing. Cambridge University Press, New York. Oayyurn, M.A. and W.D. Kemper. 1962. Salt concentration gradients in soils and their effects on moisture movement and evaporation. Soil Sci. 93: 333-342. Richards, L.A. 1931 Capillary conduction of liquids in porous mediums. Physics. 1: 318-333 Rollins, R.L., M.G. Spangler, and D. Kirkham. 1954. Movement of soil moisture under a ther1r1al gradient. Highway Res. Board Proc. 33: 492-508. Rolston, D.E. and R.D. Glauz. 1982. Comparison of simulated with measured transport and transformation of methyl bromide gas in soils. Pestic. Sci. 13 : 653-664. Rosenberg, N.J., B.L. Blaid, and S.B. Verma. 1983. Microclimate: The Biological Environment John Wiley & Sons, Inc., New York. Scanlon, B.R. 1992. Evaluation of liquid and vapor water flow in desert soils based on chlorine 36 and tritium tracers and non-isothe11r1al flow simulations. Water Resour. Res. 28: 285-297. Schwarzenbach, R. P., P. M. Gschwend, and D.M. Imboden. 1993. Environmental Organic Chemistry. John Wiley & Sons, Inc., New York. Segerlind, L.J. 1984. Applied Finite Element Analysis. John Wiley & Sons, Inc ., New York. Siebering, H. and M. Leistra. 1979. Computer simulation of fumigant behaviour in soil. In: Soil Disinfestation (Ed. D. Mulder). Elsevier Scientific Publishing Co., Amsterdam, The Netherlands. pp 135-161. \I Simunek, J. and D.L Suarez. 1993. Modeling of carbon dioxide transport and production in soil: 1. Model development. Water Resour. Res. 29: 487-497 \I Simunek, J. and M Th. van Genuchten. 1994. The CHAIN_ 2D code for simulating two-dimensional variably saturated water flow, heat transport and transport

PAGE 162

v 150 of solutes involved in sequential first-order decay reactions. Research Report No. 136. U.S. Salinity Lab., ARS, USDA, Riverside, CA. Simunek, J., T. Vogel, and M.Th. van Genuchten. 1992. The SWMS 2D Code for simulating water flow and solute transport in 2-D variably saturated media, Ver. 1.1, Research Report No. 126. U.S. Salinity Lab., USDA, ARS, Riverside, CA. Smith, W.O. 1943. Thermal transfer of moisture in soils. Trans. Am. Geophys. Union 24: 511-523. Sophocleous, M. 1979 Analysis of water a nd heat flow in unsaturated-saturated porous media. Water Resour. Res. 15: 1195-1206. Staple, W.J. and Lehane, J.J. 1954. Movement of moisture in unsaturated soils. Can. J. Agric. Sci. 34: 329-342. Stumm, W. and J.J. Morgan. 1981. Aquatic Chemistry: An introduction emphasizing chemical equilibria in natural waters. John Wiley & Sons, Inc., New York. Taylor, S.A. and J.W. Cary. 1960. Analysis of the simultaneous flow of water and heat or electricity with the thermodynamics of irreversible processes. Trans. Intern. Congr. Soil Sci., 7th Cong. Madison 1: 80-90. Taylor, S.A. and L. Cavazza. 1954. The movement of soil moisture in response to temperature gradients. Soil Sci. Soc. Am. Proc. 18: 351-358. USDA ARS. 1997 ARS Methyl Bromide Research newsletter. http://www.ars.usda.gov/is/mb/mebrweb.html. USEPA. 1986. Chemical Fact Sheet for Methyl Bromide. Fact Sheet No. 98. 22 August 1986 USEPA. 1997 Stratospheric Ozone. Methyl Bromide Inforrr1ation. http://www.epa.gov/docs/ozone/MBr/MBrqa.html. Van Bavel, C.H.M. and D.I. Hillel. 1975. A simulation study of soil heat and moisture dynamics as affected by a dry mulch. Proc. Summer Computer Simulation Cont., 1975. pp 815-821. Van Bavel, C.H.M. and D.I. Hillel. 1976. Calculating potential and actual evapotranspiration from a bare soil surface by simulation of concurrent flow of water and heat. Agricultural Meteorology 17: 453-476.

PAGE 163

151 Van de Hulst, H.C. 1980. Multiple Light Scattering: Tables, Formulas, and Applications. Academic Press, New York. 299 pp. van Genuchten, M. Th. 1980. A closed-forn1 equation for predicting the hydraulic conductivity of unsaturated soils. Soil Sci. Soc. Am. J, 44: 892-898. van Genuchten, M Th. 1981. Non-equilibrium transport parameters from miscible displacement experiments. Research Report No. 11 9, U.S. Salinity Laboratory, Riverside, CA. van Genuchten, M Th 1985. Convective-dispersive transport of solutes involved in sequential first-order decay reactions. Computers & Geosciences 11 : 129-147. van Genuchten, M. Th., and R.J. Wagenet. 1989. Two-site/two-region models for pesticide transport and degradation theoretical development and analytical solutions, Soil Sci. Soc. Am. J. 53: 1303-1310 Waggoner, P.E., P.M. Miller and H.C. DeRoo. 1960. Plastic mulching principles and benefits. Bull No 634, Conn. Agric. Exp. Stn., New Hav en. Wang, D., S.R. Yates and J Gan. 1997. Temperature effect on methyl bromide volatilization in soil fumigation. J. Environ. Qual. 26: 1072-1079 Wescott, D.W. and P.J. Wierenga. 1972. Transfer of heat by conduction and vapor movement in a closed soil system. J. Article No. 456, Agri Exp. St., N ew Mexico State Univ, Las Cruses. Wiegand, C.L. and S.A. Taylor 1962. Temperature depression and temperature distribution in drying said columns. Soil Sci. 94: 7579. Woodside, W. and J.B. Cliffe. 1959. Heat and moisture transfer in closed systems of two granular materials. Soil Sci. 87: 75-82. Yagi K ., J. Williams N.Y Wang, and R.J. Cicerone. 1993. Agricultural soil fumigation as a source of atmospheric methyl bromide. Proc. Natl. Acad. Sci. USA. 90: 8420-8423. Yakirevich, A., P. Berliner, and S. Sorek. 1997 A model for numerical simulation of evaporation from bare saline soil. Water Resour. Res. 33: 1021-1033. Yates, S.R., J. Gan, F.F. Ernst F. Gao, and D. Wang. 1996 Methyl bromide emissions from a covered field: III Correcting chamber flux for temperature. J. Environ. Oual. 25: 892-898.

PAGE 164

Yvon-Lewis S.A. and J.H. Butler. 1997. The potential effect of oceanic biological degradation on the lifetime of atmospheric CH 3 Br. Geophysical Research Letters, 24: 1227-1230. Zurer, P.S. 1993. Proposed ban on methyl bromide opposed by producers, users Chem. & Eng. News. February 1: 23-24. 152

PAGE 165

BIOGRAPHICAL SKETCH Dilip Shinde was born on August 29, 1962 in the city of Gwalior, Madhya Pradesh (MP) State, India. He graduated from St. Paul's Higher Secondary School, Indore, MP, at the age of 16 After studying one year of higher mathematics and science at MVM Science College, Bhopal University in the State Capital, Bhopal, he entered the College of Agriculture at Ind ore city, an affiliate of Jawaharlal Nehru Agricultural University at Jabalpur, MP He graduated with honors in 1983 and continued for a master's program in agronomy at the same institution. He obtained his master's degree, also with H ono rs, in 1985. In January 1986, he received an Australian Government Scholarship to pursue a master's program in soil and water engineering at the Asian Institute of Technology (AIT), Bangkok, Thailand After graduating from AIT in August, 1987 he returned to India and accepted a position as scientist with the Jawaharlal Nehru Agricultural University. His work as a scientist was focused on watershed management in dryland-rainfed agriculture until he came to the University of Florida in January 1992 He started a Ph.D. program in the Soil and Water Science Department on an international Freedom Fr om Hunger 1 scholarship awarded by 'The Rotary Foundation of The Rotary Internati onal'. The doctoral program occured within the interdepartmental hydrologic sciences program, with a strong emphasis on soil physics. 153

PAGE 166

I certify that I have read this study and that in my opinion it confom1s to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Robert S. 1 Mansell, Chair Professor of Soil and Water Science I certify that I have read this study and that in my opinion it conforr11s to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Arthur G. Hornsby, Cochalr Professor of Soil and Water Science I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. his S. C. Rao Graduate Research Professor of Soil and Water Science I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Brian L. McNeal Professor of Soil and Water Science

PAGE 167

I certify that I have read this study and that in my opinion it conforms to acceptable standard s of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy . o Zazue a essor of Ag icultu al and Biological En nee ring This dissertation was submitted to the Graduate Faculty of the College of Agriculture and to the Graduate School and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. December, 1997 < ean College f Agriculture Dean Graduate Sch o ol

PAGE 168

LD 1780 199_ 7 ss,_ eo UNIVERSITY OF FLORIDA II l 1111111 111 I I I 3 1262 08554 9516


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EW1EGXD3U_T33EQT INGEST_TIME 2013-01-23T14:21:43Z PACKAGE AA00012978_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES