<%BANNER%>

Evapotranspiration-Based Irrigation Controllers under Dry Conditions in Florida

Permanent Link: http://ufdc.ufl.edu/UFE0021937/00001

Material Information

Title: Evapotranspiration-Based Irrigation Controllers under Dry Conditions in Florida
Physical Description: 1 online resource (230 p.)
Language: english
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: controllers, evapotranspiration, florida, irrigation, turfgrass
Agricultural and Biological Engineering -- Dissertations, Academic -- UF
Genre: Agricultural and Biological Engineering thesis, M.E.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Evapotranspiration-based controllers, or ET controllers, are irrigation controllers that use evapotranspiration (ET) to schedule irrigation. The goal was to determine whether ET controllers could conserve water in Florida. The primary objectives of this research were to evaluate three brands of ET controllers to A) produce savings compared to a time clock schedule intended to mimic homeowner irrigation schedules while maintaining acceptable turfgrass quality, B) estimate reference evapotranspiration (ETo) compared to the ASCE Standardized ETo methodology, and C) schedule irrigation compared to a theoretical soil water balance model. Secondary objectives included a) quantifying the variation between controller replications, b) comparing the performance of ET controllers based on distance to a weather data source, and c) measuring the ET controller performance scores similar to the SWAT testing protocol. Five treatments replicated four times totaled twenty plots measuring 7.62 m x 12.2 m. The plots, located at the UF GCREC, were partitioned into 65% St. Augustinegrass (Stenotaphrum secundatum ?Floratam?) and 35% mixed-ornamentals to represent a typical Florida landscape. The irrigation treatments were as follows: Weathermatic SL1600 controller (T1), Toro Intelli-sense (T2), ET Water Smart Controller 100 (T3), a time-based treatment determined by UF-IFAS recommendations (T4), and a time-based treatment that is 60% of T4 (T5). The study period experienced dry conditions containing 69% dry days. It was found that using a rain sensor with a time-based irrigation schedule conserved 21% of water despite the unusual dry conditions. Average savings compared to the time-based schedule without rain sensor across all seasons ranged from 35% to 42% for the ET controllers. Reducing the time-based schedule by 40% and including a rain sensor resulted in 53% savings showing that updating the time clock settings throughout the year can result in substantial irrigation savings. Turfgrass quality remained above minimally acceptable over the study period for all treatments. The ET controllers under-irrigated compared to the calculated theoretical irrigation requirement, on average, but fell within results seen for the time-based schedules. Nine ET controllers, three replications of each brand being tested, were installed in addition to the main project to determine if there was variability between controllers concerning irrigation scheduling, ETo estimation, and proximity to weather data source. There were no differences between the replications of the controllers for both irrigation scheduling and ETo estimation. The signal-based controllers were affected by the proximity to the weather data source when estimating ETo. These controllers over-estimated ETo by 8% due to the combination of using Hargreaves equation for ETo calculations and over-estimating maximum temperatures.The Toro controller estimated ETo within 1% when the weather station was within 100 m whereas the ET Water controller under-estimated by 12% compared to ETo calculated from the on-site weather station data.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis: Thesis (M.E.)--University of Florida, 2008.
Local: Adviser: Dukes, Michael D.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2010-05-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0021937:00001

Permanent Link: http://ufdc.ufl.edu/UFE0021937/00001

Material Information

Title: Evapotranspiration-Based Irrigation Controllers under Dry Conditions in Florida
Physical Description: 1 online resource (230 p.)
Language: english
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: controllers, evapotranspiration, florida, irrigation, turfgrass
Agricultural and Biological Engineering -- Dissertations, Academic -- UF
Genre: Agricultural and Biological Engineering thesis, M.E.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Evapotranspiration-based controllers, or ET controllers, are irrigation controllers that use evapotranspiration (ET) to schedule irrigation. The goal was to determine whether ET controllers could conserve water in Florida. The primary objectives of this research were to evaluate three brands of ET controllers to A) produce savings compared to a time clock schedule intended to mimic homeowner irrigation schedules while maintaining acceptable turfgrass quality, B) estimate reference evapotranspiration (ETo) compared to the ASCE Standardized ETo methodology, and C) schedule irrigation compared to a theoretical soil water balance model. Secondary objectives included a) quantifying the variation between controller replications, b) comparing the performance of ET controllers based on distance to a weather data source, and c) measuring the ET controller performance scores similar to the SWAT testing protocol. Five treatments replicated four times totaled twenty plots measuring 7.62 m x 12.2 m. The plots, located at the UF GCREC, were partitioned into 65% St. Augustinegrass (Stenotaphrum secundatum ?Floratam?) and 35% mixed-ornamentals to represent a typical Florida landscape. The irrigation treatments were as follows: Weathermatic SL1600 controller (T1), Toro Intelli-sense (T2), ET Water Smart Controller 100 (T3), a time-based treatment determined by UF-IFAS recommendations (T4), and a time-based treatment that is 60% of T4 (T5). The study period experienced dry conditions containing 69% dry days. It was found that using a rain sensor with a time-based irrigation schedule conserved 21% of water despite the unusual dry conditions. Average savings compared to the time-based schedule without rain sensor across all seasons ranged from 35% to 42% for the ET controllers. Reducing the time-based schedule by 40% and including a rain sensor resulted in 53% savings showing that updating the time clock settings throughout the year can result in substantial irrigation savings. Turfgrass quality remained above minimally acceptable over the study period for all treatments. The ET controllers under-irrigated compared to the calculated theoretical irrigation requirement, on average, but fell within results seen for the time-based schedules. Nine ET controllers, three replications of each brand being tested, were installed in addition to the main project to determine if there was variability between controllers concerning irrigation scheduling, ETo estimation, and proximity to weather data source. There were no differences between the replications of the controllers for both irrigation scheduling and ETo estimation. The signal-based controllers were affected by the proximity to the weather data source when estimating ETo. These controllers over-estimated ETo by 8% due to the combination of using Hargreaves equation for ETo calculations and over-estimating maximum temperatures.The Toro controller estimated ETo within 1% when the weather station was within 100 m whereas the ET Water controller under-estimated by 12% compared to ETo calculated from the on-site weather station data.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Thesis: Thesis (M.E.)--University of Florida, 2008.
Local: Adviser: Dukes, Michael D.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2010-05-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0021937:00001


This item has the following downloads:


Full Text

PAGE 1

1 EVAPOTRANSPIRATION-BASED IRRI GATION CONTROLLERS UNDER DRY CONDITIONS IN FLORIDA By STACIA L. DAVIS A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ENGINEERING UNIVERSITY OF FLORIDA 2008

PAGE 2

2 2008 Stacia L. Davis

PAGE 3

3 To my parental units.

PAGE 4

4 ACKNOWLEDGMENTS I am very grateful for the support everyone has shown me over my time here at the University of Florida. I appreciate the encour agement I have received and only hope I can give back a portion of what has been given to me. I would like to thank my family, especially my parents, for their constant support and unconditional love. I would like to acknowledge and thank my committee members for their words of encouragement and constant guidance. Dr. Grady L. Miller imparted to me the mystical ways of introductory SAS and immeasurable respect for turfgrass research. Dr. Dorota Z. Haman continues to be a positive role model for me. Last but certainly not least, Dr. Michael D. Dukes helped me grow into the student and person I didnt know I could be. His persistence at keeping me challenged does not go unnoticed. I also thank the following individuals who were critical to the success of this project: Larry Miller, David Crockett, Daniel Preston, Me lissa Haley, Mary Shedd, Dr. Amy Shober, Dr. Sydney Park Brown, and Gitta Shurberg. Th is research was supported by the Hillsborough County Water Resource Services, the Florida Department of Agricultural and Consumer Services, the Florida Nursery and Landscape Grow ers Association, and th e Florida Agricultural Experiment Station.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ............................................................................................................... 4 LIST OF TABLES ...........................................................................................................................7 LIST OF FIGURES .......................................................................................................................10 LIST OF ABBREVIATIONS ........................................................................................................ 16 ABSTRACT ...................................................................................................................... .............18 CHAP TER 1 INTRODUCTION .................................................................................................................. 20 Water Demand and Use .......................................................................................................... 20 Residential Irrigation System Components ............................................................................21 Irrigation Timers ..............................................................................................................21 Solenoid Valves ...............................................................................................................22 Sprinkler Types ...............................................................................................................22 Rain Shutoff Devices .......................................................................................................23 Irrigation Scheduling ..............................................................................................................23 Irrigation System Performance Analyses ...............................................................................24 Evapotranspiration ..................................................................................................................25 Evapotranspiration-based Irrigation Controllers .................................................................... 31 Historical-based Controllers ............................................................................................ 31 Standalone Controllers .................................................................................................... 31 Signal-based Controllers .................................................................................................32 General Features ..............................................................................................................32 Summary of ET Controller Technologies ...............................................................................33 Previous Research ...................................................................................................................35 2 EVALUATION OF IRRIGATION APPLICATION BY EVAPOTRANSPIRATIONBASED IRRIGATION CONTROLLERS ............................................................................. 40 Introduction .................................................................................................................. ...........40 Materials and Methods ...........................................................................................................43 Results and Discussion ........................................................................................................ ...47 Fall 2006 ..........................................................................................................................47 Winter 2006-2007 ............................................................................................................50 Spring 2007 .....................................................................................................................51 Summer 2007 ................................................................................................................... 53 Fall 2007 ..........................................................................................................................55 Summary and Conclusions .....................................................................................................57

PAGE 6

6 3 REFERENCE EVAPOTRANSPIRATION ESTIMATION BY EVAPOTRANSPIRATION-BASED IRRIGATION CO NTROLLERS ............................... 81 Introduction .................................................................................................................. ...........81 Materials and Methods ...........................................................................................................84 Reference Evapotranspiration Calculations .................................................................... 85 Controller Descriptions ...................................................................................................87 Site Descriptions ..............................................................................................................88 Weather Stations .............................................................................................................. 89 Results and Discussion ........................................................................................................ ...92 Climatic Data Quality Control ......................................................................................... 92 Standalone Controller ...................................................................................................... 94 Signal-based Controllers .................................................................................................97 Overall Comparisons .......................................................................................................98 Summary and Conclusions ...................................................................................................100 4 IRRIGATION SCHEDULING BY EVAPOT RANSPIRATI ON-BASED IRRIGATION CONTROLLERS .................................................................................................................. 121 Introduction .................................................................................................................. .........121 Materials and Methods .........................................................................................................124 Results ...................................................................................................................................131 Discussion .................................................................................................................... .........150 Conclusions ...........................................................................................................................153 5 CONCLUSIONS AND FU TURE WORK ........................................................................... 197 Conclusions ...........................................................................................................................197 Future Work ..........................................................................................................................201 APPENDIX A STATISTICAL ANALYSIS AND RESULTS FOR CHAPTER 2 ..................................... 203 B TURFGRASS QUALITY RATINGS .................................................................................. 209 C STATISTICAL ANALYSIS AND RESULTS FOR CHAPTER 3 ..................................... 221 LIST OF REFERENCES .............................................................................................................225 BIOGRAPHICAL SKETCH .......................................................................................................230

PAGE 7

7 LIST OF TABLES Table page 1-1. Summary of the Weathermatic Toro, and ET W ater controllers ...................................... 39 2-1. Program settings for each brand of ET controller for su mm er 2006, fall 2006, and winter 2006-2007 ...............................................................................................................60 2-2. Runtimes and applicatio n am ounts per irrigation event1 for the time-based treatment (T4) operating on a twice weekly sc hedule for fall 2006 and winter 2006-2007 seasons ...............................................................................................................................61 2-3. Program settings for each brand of ET controller for spring 2007, summe r 2007, and fall 2007 .............................................................................................................................62 2-4. Runtimes and applicatio n am ounts per irrigation event1 for the time-based treatment (T4) operating on a twice weekly schedule for spring, summer, and fall 2007 seasons ...63 2-5. Average water application for the three re p lications of ET controllers located at the Gainesville turfgrass plots .................................................................................................. 63 2-6. Fall 2006 weekly water application a nd savings com pared to the time WORS treatment1 using cumulative season totals ......................................................................... 64 2-7. Fall 2006 two-week water appli cation and turf quality summary .................................... 64 2-8. Winter 2006-2007 weekly water applicat ion and savings com pared to the time WORS treatment1 using cumulative season totals .............................................................65 2-9. Winter 2006-2007 two-week water app lication and turf quality summ ary ....................... 65 2-10. Spring 2007 weekly water application and savings com pared to the time WORS treatment1 using cumulative season totals ......................................................................... 66 2-11. Spring 2007 two-week water application and turf quality summary ................................. 66 2-12. Summer 2007 weekly water applicati on and savings com pared to the time WORS treatment1 using cumulative season totals ......................................................................... 67 2-13. Summer 2007 two-week water ap p lication and turf quality summary .............................. 67 2-14. Fall 2007 weekly water application a nd savings com pared to the time WORS treatment1 using cumulative season totals ......................................................................... 68 2-15. Fall 2007 water application and turf quality summ ary ...................................................... 68 3-1. ET controller codes and experim ental information ......................................................... 102

PAGE 8

8 3-2. Daily ETo between treatments at the Ga inesville turfgrass plots ..................................... 102 3-3. Dependency on temperature using mean daily ETo values for the Weathermatic controller replications at the Gainesville turfgrass plots .................................................. 102 3-4. Dependency on Ra using mean daily ETo values for the Weathermatic controller replications at the Gain esville turfgrass plots .................................................................. 103 3-5. Average daily ETo, maximum temperature, and minimum temperature between treatments at the GCREC location ................................................................................... 103 3-6. Dependency on temperature using mean daily ETo values for the Weathermatic controller replicati ons at the GCREC .............................................................................. 103 3-7. Dependency on Ra using mean daily ETo values for the Weathermatic controller replications at the GCREC ............................................................................................... 104 3-8. Minimum and maximum temperatures betw een the W eathermatic controllers and the on-site weather station at the Gainesville turfgrass plots ................................................. 104 3-9. Totals and percentage diffe rences of average cumulative ETo between treatments at the GCREC location and the measured ETo from the FAWN weather station ............... 104 3-10. Weekly ETo between the ET Water controllers and local weather stations for the Gainesville turfgrass plots location .................................................................................. 105 4-1. Monthly crop coefficients for warm s eason turfgrass used to calculate crop evapotranspiration for the determination of the theoretical irri gation requirement ......... 156 4-2. Program setting differences1 from 2-2 for the summer 2006 season ...............................156 4-3. Weathermatic controller, T1, results fo r average s cheduling efficiency and irrigation adequacy calculated using 30-day moving to tals, percentage difference in irrigation application from the theoretical requireme nt, cumulative rainfall, and number of rainfall events for each season ......................................................................................... 157 4-4. The ET Water controller, T3, results for average scheduling efficiency an d irrigation adequacy calculated using 30-day moving to tals, percentage difference in irrigation application from the theoretical requireme nt, cumulative rainfall, and number of rainfall events for each season ......................................................................................... 157 4-5. Time-based treatment, T4, results for average sch eduling efficiency and irrigation adequacy calculated using 30-day moving to tals, percentage difference in irrigation application from the theoretical requireme nt, cumulative rainfall, and number of rainfall events for each season ......................................................................................... 158 4-6. Reduced time-based treatment, T5, resu lts for ave rage scheduling efficiency and irrigation adequacy calculated using 30-da y moving totals, percentage difference in

PAGE 9

9 irrigation application from the theoretical requirement, cumulative rainfall, and number of rainfall events for each season ........................................................................ 158 4-7. Toro controller, T2, results for average scheduling efficiency and irrigation adequacy calculated using 30-day moving totals, percen tage difference in irrigation application f rom the theoretical requirement, cumulative rainfall, and number of rainfall events for each season .................................................................................................................159

PAGE 10

10 LIST OF FIGURES Figure page 2-1. Three brands of ET controllers were tested on tw enty landscaped plots at the University of Florida Gulf Coast Research and Education Center. ................................... 69 2-2. Nine additional ET controllers, three of each bran d, are installed at the University of Florida Gainesville turfgrass plots. ....................................................................................69 2-3. Research plot layout and contro ller treatments at the University of Florida Gulf Coast Research and Education Center. ..............................................................................70 2-4. The ornamental plants. .......................................................................................................71 2-5. The ET Controllers chosen for the study lo cated at the University of Florida Gulf Coast Research and Education Center. ..............................................................................72 2-6. Irrigation water applicat ion to each plot was m onitored by 11.4 cm V100 w/ pulse output flow meters manufactured by AMCO Water Metering Systems. ........................... 73 2-7. Flow meters are wired to five SDM-SW8A switch closure input m odules that in turn connect to a CR-10X data logger. ...................................................................................... 74 2-8. Individual plot design of the twenty research plots lo cated at the University of Florida Gulf Coast Research and Education Center .......................................................... 75 2-9. Comparison of rainfall for the 2006-2007 st udy period and average hi storical rainfall on a m onthly and cumulative basis for southwest Florida. ................................................ 75 2-10. Fall 2006 cumulative and daily wate r application and daily rainfall. ................................ 76 2-11. Winter 2006-2007 cumulative and dail y water applied and daily rainfall ......................... 77 2-12. Spring 2007 cumulative and daily water applied and daily rainfall .................................. 78 2-13. Summer 2007 cumulative and daily water applied and daily rainfall. .............................. 79 2-14. Fall 2007 cumulative and daily water applied and da ily rainf all. ...................................... 80 3-1. The ET controllers installed on the Univ ersity of Florida Ga inesville Campus ..............105 3-2. Weathermatic SLW10 weathe r m onitors were installed ................................................. 106 3-3. The FAWN measured solar radiation (Rs) and clear-sky solar radiation (Rso). ...............107 3-4. Data from the NOAA weather station was us ed to calculate sola r radiation and clearsky solar radiation ........................................................................................................... .107

PAGE 11

11 3-5. Measured solar radiation and clear-sky solar radiation for the on-site weather station ...108 3-6. The FAWN daily maximum and m inimum relative humidity. ........................................ 108 3-7. The data from the NOAA weather st ation was used for daily m aximum and minimum relative humidity. .............................................................................................109 3-8. Daily maximum and minimum relative hum idity for the on-site weather station ........... 109 3-9. The FAWN daily minimum temperature and calculated dew point tem perature. ........... 110 3-10. The NOAA weather station was used to obtain daily m inimum temperature and calculated dewpoint temperature .....................................................................................110 3-11. Daily minimum temperature and calculat ed dewpoint tem perature for the on-site weather station. ................................................................................................................111 3-12. The FAWN daily mean temperatures cal culated using 24 hours of tem perature data plotted against the average of the maximum and minimum temperatures of that day .... 111 3-13. Daily mean temperatures calculated us ing 24 hours of tem perature data plotted against the average of the maximum and minimum temperatures of that day for the on-site weather station ...................................................................................................112 3-14. The FAWN daily maximum and average wind speed ..................................................... 112 3-15. Daily maximum and average wind sp eed for the on-site weather station ....................... 113 3-16. The FAWN daily gust factor calculated as the m aximum wind speed divided by the average wind speed ..........................................................................................................113 3-17. Daily gust factor calculated as the maxim um wind speed divided by the average wind speed for the on-site weather station ....................................................................... 114 3-18. The NOAA weather station was used to obtain daily average wind speed. .................... 114 3-19. Cumulative ETo for three replications of Weathermatic controllers compared to Hargreaves equation and the ASCE standard ized equation using data collected from an on-site weather statio n in Gainesville, FL...................................................................115 3-20. Cumulative ETo calculated from weather data collected by the Weathermatic controller or FAWN weather station ................................................................................ 115 3-21. Daily maximum temperature compar isons between the th ree Weatherm atic controllers and an on-s ite weather station ........................................................................ 116 3-22. Daily minimum temperature compar isons between the th ree Weatherm atic controllers and an on-s ite weather station ........................................................................ 116

PAGE 12

12 3-23. Daily maximum temperature comparison from the Weathermatic controller and the FAWN on-site weather station. ........................................................................................117 3-24. Daily minimum temperature comparison from the Weathermatic controller and the FAWN on-site weather station. ........................................................................................117 3-25. Cumulative ETo for three replications of Toro controllers compared to the ASCE standard using data collected from an on-si te weather station a nd the ASCE standard using the NOAA weather station data ............................................................................. 118 3-26. Cumulative ETo for the ET Water controller, cal culated using the ASCE method from on-site weather station data, and cal culated using the ASCE method from the NOAA weather station data .............................................................................................118 3-27. Average cumulative daily ETo for the Weathermatic and Toro controllers compared to the ASCE standard using data collect ed from an on-site weather station. .................. 119 3-28. Seven day total ETo shown cumulatively for all three brands of controllers compared to the ASCE method using data fr om the on-site weather station. .................................. 119 3-29. Cumulative ETo for the controllers at GCREC an d calculated using FAWN data and the ASCE method. ........................................................................................................... 120 4-1. Comparison of rainfall for the 2006-2007 st udy period and average hi storical rainfall on a m onthly and cumulative basis. ................................................................................. 159 4-2. The FAWN measured total rainfall and e f fective rainfall determined from the soil water balance model fo r the study period. ....................................................................... 160 4-3. Weathermatic controller (T1) resu lts over the summer 2006 season for cum ulative theoretical irrigation depth applied, daily e ffective rainfall, and 30-day moving totals of irrigation adequacy an d scheduling efficiency. ........................................................... 161 4-4. ET Water controller (T3) results over the summ er 2006 season for cumulative theoretical irrigation depth applied, daily e ffective rainfall, and 30-day moving totals of irrigation adequacy an d scheduling efficiency. ........................................................... 162 4-5. Time-based treatment (T4) results over the summer 2006 season for cum ulative theoretical irrigation depth applied, daily e ffective rainfall, and 30-day moving totals of irrigation adequacy an d scheduling efficiency. ........................................................... 163 4-6. Reduced time-based treatment (T5) results over the summ er 2006 season for cumulative theoretical irrigation depth app lied, daily effective rainfall, and 30-day moving totals of irrigation adequ acy and scheduling efficiency. .................................... 164 4-7. Measured volumetric soil moisture c ontent over the summ er 2006 season for T5, the reduced time-based treatment. ......................................................................................... 165

PAGE 13

13 4-8. Weathermatic controller (T1) results over the fall 2006 seas on for cum ulative theoretical irrigation depth applied, daily e ffective rainfall, and 30-day moving totals of irrigation adequacy an d scheduling efficiency. ........................................................... 166 4-9. Toro controller (T2) results over the f all 2006 season for cumulative theoretical irrigation depth applied, daily effective rainfall, and 30-day moving totals of irrigation adequacy and scheduling efficiency. ...............................................................167 4-10. Time-based treatment (T4) results over the fall 2006 season for cum ulative theoretical irrigation depth applied, daily e ffective rainfall, and 30-day moving totals of irrigation adequacy an d scheduling efficiency ............................................................ 168 4-11. Reduced time-based treatment (T5) re sults over the fall 2006 season for cumulative theore tical irrigation depth applied, daily e ffective rainfall, and 30-day moving totals of irrigation adequacy an d scheduling efficiency. ........................................................... 169 4-12. Volumetric soil moisture content over the fall 2006 season for T5, the reduced tim ebased treatment. ...............................................................................................................170 4-13. Weathermatic controller (T1) re sults over the winter 2006-2007 season for cum ulative theoretical irrigation depth app lied, daily effective rainfall, and 30-day moving totals of irrigation adequ acy and scheduling efficiency. .................................... 171 4-14. Toro controller (T2) results over the winter 2006-2007 season for cum ulative theoretical irrigation depth applied, daily e ffective rainfall, and 30-day moving totals of irrigation adequacy an d scheduling efficiency. ........................................................... 172 4-15. Time-based treatment (T4) results ov er the winter 2006-2007 season for cum ulative theoretical irrigation depth applied, daily e ffective rainfall, and 30-day moving totals of irrigation adequacy an d scheduling efficiency. ........................................................... 173 4-16. Reduced time-based treatment (T5) results over the winter 2006-2007 season for cum ulative theoretical irrigation depth app lied, daily effective rainfall, and 30-day moving totals of irrigation adequ acy and scheduling efficiency. .................................... 174 4-17. Volumetric soil moisture content over the winter 2006-2007 season for T2, the Toro controller. ................................................................................................................... ......175 4-18. Weathermatic controller (T1) resu lts over the spring 2007 season for cumulative theore tical irrigation depth applied, daily e ffective rainfall, and 30-day moving totals of irrigation adequacy an d scheduling efficiency. ........................................................... 176 4-19. Toro controller (T2) results over th e spring 2007 season for cumulative theoretical irrigation depth applied, daily effective rainfall, and 30-day moving totals of irrigation adequacy and scheduling efficiency. ...............................................................177

PAGE 14

14 4-20. ET Water controller (T3) results over the spring 2007 season for cum ulative theoretical irrigation depth applied, daily e ffective rainfall, and 30-day moving totals of irrigation adequacy an d scheduling efficiency. ........................................................... 178 4-21. Time-based treatment (T4) results over the spring 2007 season for cum ulative theoretical irrigation depth applied, daily e ffective rainfall, and 30-day moving totals of irrigation adequacy an d scheduling efficiency. ........................................................... 179 4-22. Reduced time-based treatment (T5) results over the spring 2007 season for cum ulative theoretical irrigation depth app lied, daily effective rainfall, and 30-day moving totals of irrigation adequ acy and scheduling efficiency. .................................... 180 4-23. Volumetric soil moisture content over the spring 2007 season for T5, the reduced tim e-based tr eatment. ....................................................................................................... 181 4-24. Toro controller (T2) results over th e summ er 2007 season for cumulative theoretical irrigation depth applied, daily effective rainfall, and 30-day moving totals of irrigation adequacy and scheduling efficiency. ...............................................................182 4-25. ET Water controller (T3) results over the summ er 2007 season for cumulative theoretical irrigation depth applied, daily e ffective rainfall, and 30-day moving totals of irrigation adequacy an d scheduling efficiency. ........................................................... 183 4-26. Time-based treatment (T4) results over the summer 2007 season for cum ulative theoretical irrigation depth applied, daily e ffective rainfall, and 30-day moving totals of irrigation adequacy an d scheduling efficiency. ........................................................... 184 4-27. Reduced time-based treatment (T5) results over the summ er 2007 season for cumulative theoretical irrigation depth app lied, daily effective rainfall, and 30-day moving totals of irrigation adequ acy and scheduling efficiency. .................................... 185 4-28. Volumetric soil moisture content for the Summer 2007 season for T5, the reduced tim e-based tr eatment. ....................................................................................................... 186 4-29. The position of the Weathermatic weather m onitor when it was damaged. .................... 187 4-30. Weathermatic controller (T1) results over the fall 2007 seas on for cum ulative theoretical irrigation depth applied, daily e ffective rainfall, and 30-day moving totals of irrigation adequacy an d scheduling efficiency. ........................................................... 188 4-31. Toro controller (T2) results over the f all 2007 season for cumulative theoretical irrigation depth applied, daily effective rainfall, and 30-day moving totals of irrigation adequacy and scheduling efficiency. ...............................................................189 4-32. ET Water controller (T3) results over the fall 2007 season for cum ulative theoretical irrigation depth applied, daily effective rainfall, and 30-day moving totals of irrigation adequacy and scheduling efficiency. ...............................................................190

PAGE 15

15 4-33. Time-based treatment (T4) results over the fall 2007 season for cum ulative theoretical irrigation depth applied, daily e ffective rainfall, and 30-day moving totals of irrigation adequacy an d scheduling efficiency. ........................................................... 191 4-34. Reduced time-based treatment (T5) re sults over the fall 2007 season for cumulative theore tical irrigation depth applied, daily e ffective rainfall, and 30-day moving totals of irrigation adequacy an d scheduling efficiency. ........................................................... 192 4-35. Volumetric soil moisture content for the fall 2007 season for T2, th e Toro controller. .. 193 4-36. Weathermatic controller, T1, percen t frequency of irrigation adequacy and scheduling efficiency scores ............................................................................................193 4-37. Toro controller, T2, percent freque ncy of irrigation ade quacy and scheduling efficiency scores ...............................................................................................................194 4-38. ET Water controller, T3, percent freque ncy of irrigation adequacy and scheduling efficiency scores ...............................................................................................................194 4-39. Time-based treatment, T4, percent fre quency of irrigation adequacy and scheduling efficiency scores ...............................................................................................................195 4-40. Reduced time-based treatment, T5, per cent frequency of irrigation adequacy and scheduling efficiency scores ............................................................................................195 4-41. Total daily rainfall and 30-day moving to tals of irrigation adequacy and scheduling efficiency for the theoretical irrigation requirem ent over the entire study period. .......... 196

PAGE 16

16 LIST OF ABBREVIATIONS AEMN Autom ated Environm ental Monitoring Network ANOVA Analysis of variance ASCE American Society of Civil Engineers CIMIS California Irrigation Mana gement Information System CIT Center for Irrigation Technology DU Distribution uniformity ET Evapotranspiration ETc Crop evapotranspiration ETo Reference evapotranspiration FAO Food and Agriculture Organization of the United Nations FAWN Florida Automated Weather Network FDEP Florida Department of Environmental Protection GCREC Gulf Coast Research and Education Center GLM General linear model IA Irrigation Association ICID International Commission for Irrigation and Drainage Kc Crop coefficient MWDSC Metropolitan Water Distri ct of Southern California NOAA National Oceanic and A tmospheric Administration NRCS National Resource Conservation Service NTEP National turfgrass evaluation procedures SWAT Smart water application technology SWFWMD Southwest Florida Water Management District TDR Time domain reflectometry

PAGE 17

17 UF University of Florida UF-IFAS University of Florida Institute of Food and Agricultural Sciences USCB United States Census Bureau USDOI United States Department of Interior USEPA United States Environmental Protection Agency WORS Without rain sensor

PAGE 18

18 Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Engineering EVAPOTRANSPIRATION-BASED IRRI GATION CONTROLLERS UNDER DRY CONDITIONS IN FLORIDA By STACIA L. DAVIS MAY 2008 Chair: Michael D. Dukes Major: Agricultural and Biological Engineering Evapotranspiration-based contro llers, or ET controllers, are ir rigation controllers that use evapotranspiration (ET) to schedule irriga tion. The goal was to determine whether ET controllers could conserve water in Florida. The primary objectiv es of this research were to evaluate three brands of ET controllers to A) produce savings compared to a time clock schedule intended to mimic homeowner irrigation schedul es while maintaining acceptable turfgrass quality, B) estimate refere nce evapotranspiration (ETo) compared to the ASCE Standardized ETo methodology, and C) schedule irrigation compared to a theoretical soil wa ter balance model. Secondary objectives included a) quantifying the variation between controller replications, b) compare the performance of ET controllers based on distance to a weather data source, and c) measure the ET controller performance scores similar to the SWAT testing protocol. Five treatments replicated f our times totaled twenty plot s measuring 7.62 m x 12.2 m. The plots, located at the UF GCREC, were partitioned into 65% St. Augustinegrass ( Stenotaphrum secundatum Floratam) and 35% mixed-ornamentals to represent a typical Florida landscape. The irrigation treatments were as follows: W eathermatic SL1600 controll er (T1), Toro Intellisense (T2), ET Water Smart Controller 100 (T3), a time-based treatment determined by UFIFAS recommendations (T4), and a time-based treatment that is 60% of T4 (T5).

PAGE 19

19 The study period experienced dry conditions co ntaining 69% dry days. It was found that using a rain sensor with a tim e-based irrigation schedule conser ved 21% of water despite the unusual dry conditions. Average savings compared to the time-based schedule without rain sensor across all seasons ranged from 35% to 42% for the ET controllers. Reducing the timebased schedule by 40% and including a rain se nsor resulted in 53% savings showing that updating the time clock settings th roughout the year can result in substantial irrigation savings. Turfgrass quality remained above minimally acceptable over the study period for all treatments. The ET controllers under-irrigated compared to the calcu lated theoretical irrigation requirement, on average, but fell within results se en for the time-based schedules. Nine ET controllers, three replications of each brand being tested, were installed in addition to the main project to determine if th ere was variability between controllers concerning irrigation scheduling, ETo estimation, and proximity to weather data source. There were no differences between the replications of the controllers for both irri gation scheduling and ETo estimation. The signal-based controllers were affected by the proximity to the weather data source when estimating ETo. These controllers over-estimated ETo by 8% due to the combination of using Hargreaves equation for ETo calculations and over-estimating maximu m temperatures.The Toro controller estimated ETo within 1% when the weather station was within 100 m whereas the ET Water controller under-estimated by 12% compared to ETo calculated from the on-site weather station data.

PAGE 20

20 CHAPTER 1 INTRODUCTION Water Demand and Use Water is a lim ited resource and at any given time are areas in the country experiencing water shortages. More specifically, Florida has the second largest withdrawal of groundwater in the United States that is used for public supply (Solley et al. 1998). Also, compared to other states Florida has the largest ne t gain in population with an inflow of approximately 1,108 people per day and fourth in overall population (United States Census Bureau [USCB] 2005). New home construction has increased to accommodate such a large in flux of people. Florida ranked first in the construction of single family residential units totaling 209,162 in 2005 (USCB 2007) and most new homes include in-ground automatic irrigation systems. However, homes with inground systems utilizing automated irrigation time rs increase outdoor water use by 47% (Mayer et al. 1999). The need for landscape irrigation will continually grow w ith increased population and home construction if there is no change in the demand for aesthe tically pleasing urban landscapes. Water supplies for non-consumptive us es, however, are decreasing due to increased demand (Southwest Florida Water Management Di strict [SWFWMD] 2006) and irrigation must become more efficient to maintain landscapes of acceptable quality. Natural climatic cycles ensure periods of critical drought and improvement in water conservation along with efficient water use is ne cessary to protect from water shortages and crises (Florida Department of Environmenta l Protection [FDEP] 2002) Southwest Florida rainfall averages 1,400 mm per year and average y early rainfall typically exceeds average yearly ET (National Oceanic and Atmospheric Associa tion [NOAA] 2005; Carriker 2000). However, irrigation is still necessary in Florida due to sand y soils with little soil wate r holding capacity that can cause plant stress in just a few days with no rain (Natur al Resource Conservation Service

PAGE 21

21 [NRCS] 2006). Proper irrigation management could result in as much as a two-fold reduction in water use (FDEP 2002). Research has shown that most of Florida single family residences over-irrigate in late fall and winter due to the inconvenience of changing th e time clock to reflect actual water needs or a misunderstanding concerning the amount of water necessary during the seasons (Haley et al. 2007). Increased watering during the fall and winter months does not allow dormancy to occur, leading to the need for additional mowing, and increases the likelihood of diseases, insects, weeds, and stresses to the la wn (Harivandi 1984). On a larger scale, increased watering contributes to the depletion of water resources an d the potential to leach soluble chemicals such as fertilizer into the groundwater. Therefore, better irrigation manageme nt could potentially lead to water conservation as reducing proble ms associated with excess watering. The water savings estimated due to the increased watering effici ency is 25-30% and the savings due to peak demand reduction is substantial (United States Environmental Protection Agency [USEPA] 2004). Residential Irrigation System Components Irrigation sy stems, sectioned into zones, are designed by grouping sprinklers in such a way to maximize efficiency and connect to a water supply through appropriately sized piping capable of handling the potential water flow rates (Ham an et al. 1989). Water sources may include municipal supply, private well, or a stationary surface water body such as a lake or pond. The quantity and size of each zone is determined by the flow rate available, the application rate, and the run time (Haman et al. 1989). Irrigation Timers Tim e clocks (i.e., timers) are installed when i rrigation systems are installed to aid the user in operating the system. Irrigation timers range in complexity from a simple on/off switch with

PAGE 22

22 some type of timing mechanism operating only one valve, to a computer that will operate many zones with different programs (Haman et al. 1989). More complex timers can be programmed with an irrigation schedule for each valve (Ham an et al. 1989), but it is believed that many homeowners leave the origin al settings programmed by th e installation contractor. Solenoid Valves The tim er controls water application by opera ting the valves. There are two types for residential irrigation: electrical or hydraulic (Haman et al. 1989) Most residential irrigation systems use electrical valves that must be matc hed to the power output of the controller, usually 24 volts AC. The power is connected through wire s for electrical valves or control tubing for hydraulic valves (Haman et al. 1989). Sprinkler Types Water is dispersed from the conveyance syst em to the landscape through sprinklers or other types of emitters. Residential sprinklers ar e of three basic types: spray heads, rotary sprinklers, and impact sprinklers (Haman et al. 1989). A spray head has a fixed nozzle that can spray any variety of patterns such as 90 degrees to 360 degrees. Generally, they require 100 kPa to 250 kPa of water pressure and have a maxi mum watering radius no more than 5.5 m (Hunter Industries, Inc. 2006; Rain Bird 2007). Rotary sprinklers are sprinklers that rotate, resulting in the commonly used term rotor to describe them, so that a circular area is covered and require 172 kPa to 450 kPa supply pressures. They are used for irrigation of larg er areas, greater than 4.6 m (Hunter Industries, Inc. 2006; Rain Bird 2007) Impact sprinklers can also be used for residential irrigation, though it is less common. Imp act sprinklers are similar to rotors because they require 172 kPa to 414 kPa for a radius of 6.7 m to 13.7 m. A relatively new type of spray head nozzle comb ines the flexibility of a spray head with a rotating distribution pattern and is known as a rotary nozzle. These nozzles have shown average

PAGE 23

23 water conservation poten tial of 31% due to an increased di stribution uniformity compared to fixed spray heads (Solomon et al. 2006). Rain Shutoff Devices As of May 1991, Florida requires that any new i rrigation system m ust maintain and operate a rain sensor (Florida Statutes, Chapter 373.62 n.d.; Florida Statutes 2001). Some counties mandate the use of a sensor regardless of the age of the system (Dukes and Haman 2002b). These sensors act as a switch that turns off irriga tion due to a given threshold of rainfall. They are relatively inexpensive device s that should be mounted in an open area, free from cover and debris. One type of rain sensor is made from expanding disks. Expanding disk are hygroscopic porous disks that expand when wet and open the ci rcuit to cease irrigation and remains in this state until the disks dry out (H unter Industries, Inc. 2006). Irrigation Scheduling There are tw o methods to schedule irrigation: qu antitative or qualitative. The quantitative method measures plant needs from the soil moistu re or evapotranspirati on (ET) loss directly using instruments such as tensiometers or diel ectric probes. The other method commonly used by homeowners, qualitative, invol ves observing the lawn and irri gating when it looks stressed (Wade and Waltz 2004). Typical signs of stress include foot prints remaining for a long time, a bluish-gray cast to the turf grass, or a majority of leaf blades folded length-wise in half. A common rule of thumb irrigation re commendation is to irrigate when 30% of the lawn is at the point of wilting to encourage roots to grow deeper and become healthier and more drought tolerant (Tichenor et al. 2003).

PAGE 24

24 Irrigation System Performance Analyses Im proper irrigation, whether it is under-irrigat ion or over-irrigation, can negatively impact landscapes as well as waste water resources. Ther e are several ways to numerically evaluate the ability of an irrigation system to properly distribute water as well as the quality of the landscape. Irrigation efficiency is the measure of the effectiveness of an irrigation system to supply water to a given landscape area. Methods to ap propriately determine irrigation efficiency are difficult to determine quickly since the data are difficult to obtain (Burt et al. 1997). Therefore, distribution uniformity (DU) testi ng is used to get an idea of irrigation system performance. High efficiency can only be achieved with hi gh DU values; however, DU does not represent efficiency. Thus, to distinguish DU from efficien cy, Burt et al. (1997) recommended expressing as a ratio rather than percentage. DU is esse ntially a measure of variability from the mean application amount (i.e., depth in sprinkler irrigation). Commonly, th e lowest quarter is used to produce the ratio (American Societ y of Civil Engineers [ASCE] 1978) It is represented by the following equation where DUlq (mm/mm) is the low quarter distribution uniformity, dlq (mm) is the average of the lowest 25% of depths, and davg (mm) is the average of all depths (Burt et al. 1997). avg lq lqd d = DU (1-1) Volume can be used to calculate the ratio when identically sized catch ca ns are used to perform the test. The Irrigation Asso ciation (IA; 2005) suggested that DUlq overestimates the effect of nonuniform landscapes. Therefore, the low half distribution uniformity (DUlh) should be used for irrigation scheduling purposes and can be estimated from DUlq as a percentage by using a simple conversion equation.

PAGE 25

25 lq lhDU0.614+38.6=DU (1-2) Turfgrass quality must remain at acceptable le vels when trying to conserve water. The National Turfgrass Evaluation Procedures (NTEP) created standards used to measure quality based on aspects such as color, density, uniformity, texture, and disease or environmental stress (Shearman and Morris 1998). NTEP developed a subjective rating system where turfgrass is rated on a scale from 1 to 9 where 1 represents d ead turfgrass or bare ground, 9 represents an ideal turfgrass without weeds or disease, and 5 is considered to be the minimum acceptable quality for a residential setting. Evapotranspiration Evapotranspiration (ET) is de fined as the evaporation from the soil surface and the transpiration through plant material (Allen et al. 1998). It is part of a balanced energy budget that exchanges energy for outgoing water at the su rface of the plant. The components of ET are solar radiation, temperature, relative humidity, and wind speed (Allen et al. 2005). Reference ET (ETo) is ET found using a hypothetical reference cr op assumed to be similar to an actively growing, well-watered, dense green grass of uniform height (Allen et al. 2005). There are various ways to estimate ETo. Common field measurement methods include lysimeter experiments, soil water studies, and a long-term inflow-outflow water budget for large areas (Fangmeier et al. 2006). Computer simulated models can be used to predict ETo and guide irrigation. Research has shown that influences such as turfgrass species, mowing height, and nitrogen fertility can affect models. Also, they are relative to an area in that a model could be accurate for a certain part of the country but inaccurate for the rest (Wade and Waltz 2004). Typically, climatic data is collected and used as inputs to equations to calculate ETo. There are three types of equations: mass tran sfer, energy balance, and empirical methods

PAGE 26

26 (Fangmeier et al. 2006). Most of the current methods employ a combination of the three. The appropriate ET equation is chos en depending on many factors including geographical location, types of crops, and weather availabil ity (Fangmeier et al. 2006). Two ETo equations are pertinent to this study. The Penman-Monteith equation is a standard ized equation endorsed by many organizations including the American Society of Civil En gineers (ASCE), Intern ational Commission for Irrigation and Drainage (ICID), Food and Agriculture Organizati on of the United Nations (FAO), and the Irrigation Association (Alle n et al. 2005). It is also co nsidered the ASCE standardized reference evapotranspiration equa tion and is given here as: )uC+1( + u)e(e 273+T C +G)(R 0.408 =ET2d 2as n n 0 (1-3) Variables are defined as follows: 23.237T 3.237T T27.17 exp2503 (1-4) 2 TeTe emin o max o s (1-5) 3.237T T27.17 exp6108.0Teo (1-6) 2 100 RH Te 100 RH Te emin max o max min o a (1-7) nlnsnRRR (1-8) s nsR1R (1-9) 2 TT e14.034.0fRmin 4 K max 4 K a cd nl (1-10) 35.0 R R 35.1fso s cd (1-11) a 5 soRz10275.0R (1-12)

PAGE 27

27 s srsc asincoscossinsindG 24 R (1-13) J 365 2 cos033.01dr (1-14) 39.1J 365 2 sin409.0 (1-15) tantanarccoss (1-16) 42.5z8.67ln 87.4 uuw z2 (1-17) ETo = reference evapotranspiration, mm/day = psychrometric constant, 0.067 kPa/C = slope of the saturation vapor pressure-temperature curve, kPa/C T = daily mean air temperature, C es = saturation vapor pressure, kPa eoT = saturation vapor pressure function, kPa ea = actual vapor, kPa RH = relative humidity, % Rn = net radiation, MJ/m2/day Rns = net short-wave radiation, MJ/m2/day Rnl = net outgoing long-wave radiation, MJ/m2/day Rs = incoming solar radiation, MJ/m2/day = albedo or canopy reflec tion coefficient, 0.23 = Stefan-Boltzmann constant, 4.901 x 10-9 MJ/K4/m2/day fcd = cloudiness function, 0.05 fcd 1.0 Rso = calculated clear-sky radiation, MJ/m2/day Ra = extraterrestrial radiation, MJ/m2/day z = station elevation above sea level, m dr = inverse relative distance factor for the earth-sun = solar declination, rad = latitude, rad s = sunset hour angle, rad J = Julian day Gsc = solar constant, 4.92 MJ/m2/hr G = daily soil heat flux density, 0 MJ/m2/day u2 = wind speed at 2 m height, m/s This equation takes into account net radiation, Rn (MJ/m2/day); heat flux, G (MJ/m2/day); vapor pressure, (kPa/C), es (kPa), ea (kPa); temperature, T (C); and wind speed, u2 (m/s). is the psychrometric constant and can be obtained fr om the measured mean atmospheric pressure (Allen et al. 2005). A grass refere nce crop rather than alfalfa is typically used in Florida since

PAGE 28

28 alfalfa is not produced on a large scale (Irmak and Haman 2003). For a grass reference, the constants Cn and Cd are 900 and 0.34, respectiv ely (Allen et al. 2005). The other important ET equation relevant to this study is the Hargreaves equation which is given as: 17.8+TTDR0.0023=ET2/1 a 0 (1-18) RA (MJ/m2/day) represents the extraterrestrial radiat ion, TD (C) is the difference between the mean daily maximum temperature and the mean da ily minimum temperature, and T (C) is the mean ambient air temperature (Jen son et al. 1990). This equati on only requires temperature as the weather input. Extraterrest rial radiation can be calculated from solar radiation using equations or can be found in a table by using the latitude for the location of interest. This equation is generally utilized in the western part of the United States because it was derived from Alta fescue grass (Festuca arundinacea ) lysimeter data over an ei ght-year period in Davis, California (Jenson et al. 1990) Since temperature data ar e required to calculate ETo with this equation, data collection is simplified and mu ch less expensive compared to the PenmanMonteith equation. The disadvantage to the Ha rgreaves equation is that it has been found to overestimate ETo in humid conditions compared to the Penman-Monteith approach (Trajkovic 2007). Once the ETo is calculated from an equation, crop water use can be estimated by using a crop coefficient. The relationship between ETo and crop-evapotranspiration (ETc) is as follows: 0CCETK=ET (1-19) Kc is the crop coefficient that adjusts ETo to account for differences in ground cover, canopy characteristics, and aerodynamic resistance from the reference cr op to calculate crop-specific ET loss (Jenson et al. 1990). Because ETo is calculated using grass as the reference crop, crop

PAGE 29

29 coefficients must be chosen from empirical valu es derived from grass a nd not alfalfa. The crop specific Kc values can be found using curv es or tables (Jenson et al. 1990; Allen et al. 1998). Weather Data Quality Control Weather station data should be subjected to quality control assessments (Allen et al. 2005) before using for other purposes. Comp arisons should be made by comparing solar radiation (Rs), relative humidity (RH), temperature (T), and wind speed (u2) data against relevant physical extremes. Solar radiation and clear-sky solar radiation (Rso) are equal on cloud-free days. Quality control of Rs can be performed by plotting Rs and Rso over time. Rso may be calculated using the equation defined for the ASCE ETo method (Equation 1-12) as well as a more detailed procedure described below (Allen et al. 2005). aDB soRKKR (1-20) 4.0 t Bsin W 075.0 sinK P00146.0 exp98.0K (1-21) 26.5293 z0065.0293 3.101P (1-22) 1.2Pe14.0Wa (1-23) 242.039.1J 365 2 sin3.085.0sinsin (1-24) B DK36.035.0K for KB 0.15 (1-25) B DK82.018.0K for KB < 0.15 (1-26) KB = clearness index for direct beam radiation KD = transmissivity index for diffuse radiation P = atmospheric pressure at the site elevation, kPa Kt = turbidity coefficient, 0 < Kt 1.0 W = precipitable water in the atmosphere, mm = angle of sun above the horizon, rad Measured relative humidity should be in the range of 30% to 100% for humid climates. Relative humidity values less than this range are possible, but it is unreasonable to maintain

PAGE 30

30 values less than 30%. It is not possible to have relative humidity values greater than 100% in the physical environment. Daily maxi mum and minimum values of relati ve humidity were plotted to verify that the data falls in the acceptable range for a majority of the study period. Another way to verify that the relative humidity data was collected correctly is to calculate daily dew point temperature and compare to the daily minimum temperature. Dew point temperature is calculated with the following equation (Allen et al. 2005): a a deweln78.16 eln3.23791.116 T (1-27) Tdew = dew point temperature, C Dew point temperature and minimu m temperature should be approximately the same a majority of the time in humid climates with the exception s of days with changes in air mass, high winds, or cloudiness at night (Allen et al. 2005). Air temperature data is most likely to be consistent and of the best quality data. Temperature data can be checked for quality by plotting the daily average calculated from the 24-hour time period and the average of the maxi mum and minimum temperatures of the same day over time. These averages should be with in 3C unless caused by rainfall events, unusually high wind speeds, or changes in air mass (Allen et al. 2005). The quality of wind speed data is difficult to assess when duplicate instruments are not used. Ways to determine if the correct data is being reported include pl otting daily average wind speed, daily maximum wind speed, and the gust factor. The gust factor is calculated as a ratio of maximum wind speed to average wind speed. If any of these figures e xhibit consistently low values (< 1.0 m/s) or gust factor values of 1.0, there is some t ype of problem with the data (Allen et al. 2005).

PAGE 31

31 Evapotranspiration-based Irrigation Controllers Evapotranspiration-based controllers, also known as ET controlle rs or smart controllers, are irrigation controllers that use estimated ET to schedule irrigation. Each controller works differently depending on manufact urer but typically can be pr ogrammed with site specific conditions such as soil type, plant type, sprinkle r type, sun and shade, et c. The controllers are designed to either replace the typical timer or act as an amendment to the timer. Also, controllers can have accessories that make them more accurate while others come as a complete package and need no additions (Riley 2005). ET controllers receive ETo information in three general ways, consequently dividing ET controllers into three main types: 1) historical-based controllers, 2) standalone controllers, and 3) signal-based controllers. Historical-based Controllers This type of controlle r relies on historical ETo information for the area. Typically, monthly historical ETo is programmed into the controller by the manufacturer or installing contractor. This is not as efficient as other me thods because it does not take into account actual changes in the weather. For exam ple, if there is an unusually rai ny or dry month, the controller will not adjust for that difference from histori cal values. There are a few controllers on the market that use historical ETo only, however, there are attachments such as temperature sensors to adjust monthly ETo to daily ETo. Examples of historical-based ET controllers are AquaConserve, Calsense, Rain Bird, and Rain Master. Standalone Controllers Standalone controllers typically receive climatic data from on-site measurement sensors and calculations to determine ETo are performed by the controller. Even though the controllers might take readings every second or every fifteen minutes, the ETo used for irrigation purposes is

PAGE 32

32 cumulative daily. On-site sensors could include: temperature, solar radiation, an ET gauge, or even a full weather station (Riley 2005). Controllers that use just temperature or solar radiation based sensors do not use the Penman -Monteith equation to calculate ETo, instead another method of ETo calculation would have to be used such as Hargreaves equation. Benefits of using standalone controllers are that th ey use real weather conditions but are not limited to the use of a full weather station and there are no signal fees as sociated with broadcasts from the manufacturer (Riley 2005). Examples of manuf acturers with this type of co ntroller are Weathermatic and Weatherset. Signal-based Controllers The majority of ET controllers on the market are signal-based. These controllers receive ETo information from a company that collects climatic data from weather stations located near the irrigation site usi ng satellite or internet technology. Depending on the manufacturer, the ETo data can be from an average of multiple weathe r stations in the area or from a single weather station. Generally, with th is type of controller the KC value is programmed by the user or contractor so that it can calculate the ETc of the various irrigation z ones. There is typically signal fee (i.e., subscription) for this controller set by the manufact urer that normally ranges from $4 to $15 per month (Riley 2005). AccuWater, Aquasave, ET Water, Hydropoint, Irrisoft, Rainbird, and Water2Save are examples of companies that sell this type of controller. General Features ET controllers can be purchased as an add-on to the existing timer; its function is to obtain ETo data and communicate with th e timer. It can either allo w the timer to bypass scheduled events or schedule run time s depending on the amount of ETc in a day. The timer would still be required to turn the irrigation system on or off. Irrigation could be scheduled as frequently as once a day which might be more than local restrictions allow. In response to locally specific

PAGE 33

33 watering restrictions, ET controllers can be set to irrigate during certain windows such as twice a week during 4pm to 10am. The ET c ontroller will track the amount of ETo lost since the last irrigation event and allow the system to irrigate the summed amount at the appropriate time. ET controllers can also be purchased as a sing le unit to replace the typical timer. This controller does everything an add-on does as well as controls the irrigation system. The benefit to replacing the timer is the guarantee of no problem s with integrating technol ogies. This type of controller would be ideal for new or replacement irrigation systems. Studies have shown that there are perceptio nal barriers for ET controllers entering the market. Typical perceptions are: water savings not justified by the cost, possible decrease in landscape appearance, possibility that technologies will not integrate properly, possible loss of reliability and control, and no aw areness of problems (USEPA 2004). However, commercially available residential ET controllers are generally inexpensive and th e savings are seen within the first few years. Many municipalities and wate r management districts are providing incentives and starting programs that provide rebates fo r using water conserving technology such as ET controllers (Dewey 2003). The other listed perc eptional barriers are related to lack of maintenance and malfunctioning of the controller. Summary of ET Controller Technologies There are many ET controllers on the mark et that vary based on design by the manufacturer. For this reason, three popular controllers will be summarized including specifications, advantages, and disadvantages (Table 1-1). ET Water Systems LLC is a relatively new comp any that specializes in the manufacturing of a signal-based ET controller called the Smart Controller. They began marketing the Smart Controller 100 for residential use in California, Nevada, an d Colorado in July 2004 (United States Department of Interior [USDOI] 2004). ET Water uses public weather stations where

PAGE 34

34 possible, and will adapt to situ ations where it is not possible by using private stations or installing their own. The ET Water controller can receive ETo data through any phone line or cellular signal. All input s specific to the landscap e are entered onto the ET Water internet site using a personal computer. Internet connection at the controller is not necessary (ET Water Systems LLC 2005). There is also a manual scheduling optio n in case of problems with communications to the cont roller (USDOI 2004). Hydropoint Data Systems, Inc. develops software to schedule irrigation called WeatherTRAK and provides a signal-based ETo information service called WeatherTRAK ET Everywhere. This signal-based system is mark eted under the WeatherTRAK trade name as the WeatherTRAK ET Plus and also by Toro as the Intelli-Sense and by Irritrol Systems as the Smart Dial Series. Hydropoint has access to over 14,000 weather stations via National Oceanic and Atmospheric Administration (NOAA), California Irrigation Management Information System (CIMIS), Denvers ET network, and Georgias Automated Environmental Monitoring Network (AEMN). MM5 modeling developed by Penn State University is used to create virtual weather stations by interpolating between data collected between se veral weather station locations (Hydropoint Data Systems, Inc. 2003). ETo calculations are accurate to one square kilometer for 90% of the country with a st andard deviation of 0.254 mm for daily ETo. Because the data collected is from a full scale weather st ation, the Penman-Monteith equation is used to calculate ETo (Hydropoint Data Systems, Inc. 2003). Weathermatic manufactures a standalone ET cont roller that relies on a weather monitor for temperature measurements and ZIP code for sola r radiation to use as inputs to calculate ETo with the Hargreaves equation (Hargreaves and Sa mani 1982). Even though Hargreaves equation tends to overestimate ETo, it is impractical for homeowners to install full scale weather stations

PAGE 35

35 in their backyards (Trajkovic 2007; Weathermatic 2005). There is no historical data preinstalled or manually entered into the controller (USDOI 2004). SWAT Testing Smart Water Application Technologies (SWAT) is a subset of the Irrigation Association that developed a protocol for determining th e effectiveness of irrigation scheduling by ET controllers. The protocol was designed to m easure the ability of ET controllers to schedule irrigation that is adequate and efficient while minimizing run-off. Irrigation adequacy is measured by under-irrigation and scheduling efficiency is measured by over-irrigation determined from a soil water balance model. Testing must meet the requirements of 30 consecutive days of testing with 10.2 mm of total rainfall and 63.5 mm of ETo (IA 2006c). Previous Research ET controllers have been studied frequently in the last five years in the western part of the United States. Savings were usuall y reported in terms of actual or po tential. Potential savings is defined by Hunt et al. (2001) as the difference between actual outdoor water applied and what should have been applied taking weather into account. Actual savings is determined by comparing current use to some reference us e which is usually hi storically-based. A study was conducted in 2002 in west San Fe rnando Valley, California by Los Angeles Department of Water and Power to assess the pe rformance of weather-ba sed technologies and customer acceptance. WeatherTRAK and Wate r2Save LLC controllers were installed professionally and given an initial schedul e based on landscape and irrigation system characteristics, adjusting the schedule when systems showed signs of poor uniformity or tinkering by the landscaper Twenty five sites were chosen 18 of which were WeatherTRAK enabled controllers. Data collection occurred fo r two years before installation as well as one year after installation. The WeatherTRAK enab led controller showed 17.4% of actual savings

PAGE 36

36 relative to a normalized weathe r year found through statistical modeling from the pre-retrofit time period and both controllers, combined, exhibited 78% of potential savings. When homeowners were questioned about the use of the controller, they responded positively but recommended outreach education programs for landscapers as well as homeowners (Bamezai 2004). The Metropolitan Water Dist rict of Southern Californi a (MWDSC) conducted a year-long bench test in 2002 designed to compare the abil ity of ET controllers to determine theoretical water needs for three types of landscapes: cool se ason turf on loam with full sun, shaded annuals on sandy soils, and low water using ground cover on a sunny, 20 degree slope. AquaConserve, WeatherTRAK, and Weatherset controllers were compared by soil moisture depletion analyses. Irrigation run times were recorded by a data logger and actual qua lity of the landscape could not be assessed since this project was a bench test. Co mparisons were made using the methods of maximum allowable water allowance, water balan ce, and percent soil moisture depletion. The WeatherTRAK enabled controller always applied less water than the maximum allowable water allowance resulting in no overwater ing. This controller performed the water balance sufficiently so that water received equals water required ex cept for the summer months where the controller showed a deficit in irrigation. Percent soil moisture depletion for all scenarios except for the sloped one, where over-irrigation occurred, fell within a 30%70% target range as well as minimized runoff (Metropolitan Water Distri ct of Southern California [MWDSC] 2004). A virtual study was conducted in 2003 using Aqua Conserve, Weathe rSet, WeatherTRAK, and Calsense controllers. The st udy was designed to determine the data used by the controllers, ease of setup and operation, and how accurate they were at matching irrigation needs to five types of landscapes consisting of turfgrass, tree s/shrubs, annuals, mixed high water use plants,

PAGE 37

37 and mixed low water use plants. A real-time reference landscape was used to calculate ETc for the turfgrass directly and was adapted for the trees/shrubs and annuals using a combination of real-time information, plant f actors, and previous research. Irrigation equaled the turfgrass reference requirements in April and October only for the WeatherTRAK controller; overirrigation was 21-40% in March, June, and July, over 40% in November, and 11-20% for the rest of the year. Trees/shrubs were over-irrigated by more than 40 % for the entire study period. Annuals were over-irrigated by 2140% in all months ex cept January, March, Ap ril, and June; in those months irrigation was well over-estimated by more than 40%. The mixed high water use scenario was irrigated correctly from April through August, but was over-irrigated 21-40% in September and December, and under-irrigated by mo re than 40% in April. The mixed low water use plants received 15-60% less irrigation than ETo during the study period. It is suspected that these results were due to very general controll er settings. For instance, the uniformity and sprinkler precipitation rates remained as default values based on inputted generic sprinkler type (Pittenger et al. 2004). A residential runoff reduction (R3) study was c onducted using standard Sterling controllers modified for this study to accept a broadcast signal in Irvine California over an 18 month period. The objectives were to: 1) deve lop and expand the use of signalbased technology, 2) determine the effectiveness of an educational program for homeowners, 3) determine the connection between landscape water use and quantities and qualities of dry w eather runoff, and 4) to gauge the acceptance of ET controller water manageme nt. Five similar neighborhoods, each with a single-point storm drain separate from other communities, participated in the study. Three neighborhoods were control groups and unaware of the study, one group received educational materials, and one was an ET cont roller group also with educationa l materials. The educational

PAGE 38

38 materials consisted of postcards sent with change in weather c onditions suggesting days of week and minutes per day of irrigation. The ET controller group was mixed of single unit homes, condos, homeowner associations a nd public properties, totaling 129 sites. This group potentially saved 49% dry weather runoff and saved 71% compar ed to the control groups. The educational materials group totaled 223 homes and increased dry weather runoff directly by 36% and 72% compared to the control groups. Out of the ET controller group, 72% reported that their landscapes maintained or improved since instal lation and enjoyed using them (Diamond 2003). Aquacraft, Inc. performed a three year study in Boulder, Greeley, and Longmont, Colorado from 2000 to 2002 to determine the reliability an d effectiveness of a WeatherTRAK enabled ET controller. There were nine residences and one commercial property committed to the study in 2001, totaling ten sites. A ll savings were determined from comparisons to ETo for the area instead of ETc that is specific to the plant and six sites were already ir rigating below historical ETo. By the end of the ir rigation season, 94% of ETo was replaced by irrigation with % error between sites and 88% of the potential savings wa s captured. Due to various reasons, only seven sites remained through the 2002 season and five of the seven participants were historical under irrigators due to voluntary participation in the study. On average, 71% of ETo was applied over the 2002 irrigation season and 92% of potential savings were captured. The historical under irrigators, though did not see ac tual water savings by using th e system, maintained their historical average. The cont rollers applied the correct ETo reliably and responded well to three different sets of drought restrictions with no replacements necessary during the study period (Aquacraft, Inc. 2002; Aquacraft, Inc. 2003).

PAGE 39

39 Table 1-1. Summary of the Weatherm atic, Toro, and ET Water controllers Manufacturer Weathermatic Toro ET Water Controller SL1600 Intelli-Sen se Smart Controller 100 Initial Cost $359.95 $4801 $589 Service Cost None $48/yr $199/yr Number of Zones 4 6 12 Programmable Inputs Sprinkler Type Plant Type Soil Type Percentage Adjustment Sprinkler Type Plant Type Soil Type Slope Sun and Shade Efficiency Sprinkler Type Plant Type Soil Type Slope Sun and Shade Efficiency Replace Typical Timer? Yes Yes Yes Connectivity for ET Transmissions None Satellite/Paging Technology Cellular Wireless Accessories SLW15 Weather Monitor Bow Tie Antenna None Warranty 2-year 5-year 3-year Advantages No signal service Full weather station not required Less expensive Rain sensor included on monitor Uses standard ETo calculations Most researched Uses standard ETo calculations Internet scheduling could fit busy lifestyle Disadvantages Method overestimates ETo Must have appropriate location for weather monitor Limited research Ongoing service costs Rain sensor extra More expensive Ongoing service costs Limited research Rain sensor extra More expensive 1Initial cost includes two year s of signal service to the controller automatically.

PAGE 40

40 CHAPTER 2 EVALUATION OF IRRIGATION APPLICATI ON B Y EVAPOTRANSPIRATION-BASED IRRIGATION CONTROLLERS Introduction Water is a limited resource and there are ar eas all over the countr y experiencing water shortages. More specifically, Florida has the second largest w ithdrawal of groundwater in the United States that is used for public supply (Solley et al. 1998). Also, compared to other states Florida has the largest net gain in population with an inflow of approximately 1,108 people per day and fourth in overall population (Unite`d St ates Census Bureau [USCB] 2005). New home construction has increased to accommodate such a large influx of people and most new homes include in-ground automated irrigation systems. However, homes with in-ground systems utilizing automated irrigation tim ers alone increase outdoor wate r use by 47% (Mayer et al. 1999). The need for landscape irrigation will co ntinually grow with increased population and home construction if there is no change in the demand for aesthetically pleasing urban landscapes. Florida rainfall averages 1,400 mm per year a nd average yearly rain fall typically exceeds average yearly evapotranspiration (ET) (Carriker 2000). However, irrigation is still necessary in Florida due to sandy soils with little soil water holding capac ities that can cause drought conditions in just a few days with no rain (H aley et al 2007; Haman et al. 1989). Thus, water stress in plants such as turfgrass and orname ntals can occur even during a rainy season (NRCS 2006). Water is a limited resource and irrigation must become more efficient to maintain landscapes of acceptable quality (Southwest Fl orida Water Management District [SWFWMD] 2005). Evapotranspiration (ET) is de fined as the evaporation from the soil surface and the transpiration through plant canopies (Allen et al. 1998). It is part of a balanced energy budget

PAGE 41

41 that exchanges energy for outgoing water at the su rface of the plant. The components of ET are solar radiation, temperature, relative humidity, and wind speed (Allen et al. 2005). Reference ET (ETo) is ET found using a hypothetical reference cr op assumed to be similar to an actively growing, well-watered, dense green grass of uniform height (Allen et al. 2005). Evapotranspiration-based controllers, also known as ET controllers, are irrigation controllers that use an estimation of ET to schedule irrigation. Each controller works differently depending on manufacturer, but typi cally can be programmed with various conditions specific to the landscape making them more efficien t (Riley 2005). ET controllers receive ETo information in three general ways, consequent ly dividing ET controllers into th ree main types: 1) standalone controllers, 2) signal-based controllers, and 3) hist orical-based controllers. Standalone controllers typically receive climatic data from on-site measurement sensors and calculations to determine ETo are performed by the controller. Even though the controllers might take readings every second or every fift een minutes, the ETo used for irrigation purposes is cumulative daily. On-site sensors could include: temperature, solar radiation, an ET gauge, or even a full weather station (Riley 2005). Benefits of standalone controllers are that they are not limited by requiring the use of a full weather station and there are no signal fees associated with broadcasts from the manu facturer (Riley 2005). Signal-based contro llers receive ETo information from a company that collects climatic data from weather stations locate d near the irrigation si te using satellite or internet technology. Depending on the manufacturer, the ETo data can be from an average of multiple weather stations in the area or from a single weather st ation. There is typical ly a signal fee (i.e., subscription) for this controller set by the manufacturer that normally ranges from $4 to $15 per month (Riley 2005).

PAGE 42

42 Historical-based controll ers rely on historical ETo information for the area. Typically, monthly historical ETo is programmed into the controller by the manufacturer or installing contractor. This is not as efficient as other me thods because it does not take into account actual changes in the weather. ET controllers have been studied frequently in the last five years in the western part of the United States. Savings were usuall y reported in terms of actual or po tential. Potential savings is defined by Hunt et al. (2001) as the difference between actual outdoor water applied and what should have been applied taking weather into account. Actual savings is determined by comparing current use to some reference us e which is usually hi storically-based. A study conducted in 2002 in west San Fernando Valley, California by Los Angeles Department of Water and Power showed 17.4% of actual savings by a WeatherTRAK enabled controller relative to a normali zed weather year found through st atistical modeling from the preretrofit time period and 78% of potential savi ngs (Bamezai 2004). A residential runoff reduction study was conducted using a modified Sterling irri gation controller to accept a broadcast signal from the WeatherTRAK ET Everywhere service in Irvine California; the ET controller group potentially saved 49% dry weather runoff and saved 71% compared to the control groups (Diamond 2003). Aquacraft, Inc. performed an ET controller study in Colorado to determine savings compared to ETo for the area and six sites were al ready irrigating be low historical ETo. The first year resulted in 94% of ETo replacement by irrigation with % error between sites and 88% of the captured potential savings wh ile the second year resulted in 71% of ETo replacement and 92% of captured potential saving s (Aquacraft, Inc. 2002; Aquacraft, Inc. 2003). The objective of this study was to evaluate the ability of three brands of ET-based controllers to schedule irrigation by comparing irrigation a pplication to a time clock schedule

PAGE 43

43 intended to mimic homeowner irrigation schedul es, while maintaining acceptable turfgrass quality. Materials and Methods This study was conducted at the University of Florida Gulf Coast Research and Education Center (GCREC) in Wimauma, Florida and at the University of Florida Agricultural and Biological Engineering Department turfgrass plots in Gainesville, Florida (Figure 2-1; Figure 22). There were a total of twenty plots at the GCREC that measured 7.62 m x 12.2 m, bordered by a 15.2 cm tall black metal barrier, with 3.05 m buffer zones between adjacent plots (Figure 23). The buffer zones were covered with a white mate rial that acted as a weed barrier. Each plot consisted of 65% St. Augustinegrass (Stenotaphrum secundatum Floratam) and 35% mixed ornamentals to represent a typica l residential landscape in Florida. The ornamentals were as follows: Crape Myrtle (Lagerstroemia indica Natchez) (Figure 2-4A), Gold Mound Lantana (Lantana camara Gold Mound) (Figure 2-4B), Indian Hawthorne (Raphiolepis indica) (Fig 24C), Cape Plumbago (Plumbago auriculata) (Figure 2-4D), and Big Blue Liriope (Liriope muscari Big Blue) (Figure 2-4E). Landscapes we re maintained through mowing, pruning, edging, mulching, fertilization, and pest and weed control according to current UF-IFAS recommendations (Black and Ruppert 1998; Sa rtain 1991). The treatments set up at the Gainesville turfgrass plots were tested virtually to study the va riability in water application between ET controllers of the same brand. Five treatments were establishe d at the GCREC, T1 through T5, replicated four times for a total of twenty plots in a completely randomized block design The irrigation treatments are as follows: T1, SL1600 controller with SLW15 weat her monitor (Weathermatic Inc., Dallas, TX); T2, Intelli-sense (Toro Company, Inc., Riverside, CA) utilizing the WeatherTRAK ET Everywhere service (Hydropoint Datasystems, Inc., Petaluma, CA); T3, Smart Controller 100

PAGE 44

44 (ET Water Systems LCC, Corte Madera, CA); T4, a time-based treatment determined by UFIFAS recommendations (Dukes and Haman 2002a); and T5, a time-based treatment that is 60% of T4 (Figure 2-5). All treat ments utilized rain sensors se t at a 6 mm threshold. A metal shed housed the controllers on-site a nd a manifold table supported forty solenoid valve and flow meter combinations to supply an d monitor irrigation to each zone of each plot (Fig 2-6). The flow meters (11.4 cm V100 w/ Pulse Output, AMCO Wa ter Metering Systems, Ocala, FL) used to monitor irrigation water ap plication were connected to five SDM-SW8A switch closure input modules (Campbell Scientific, Logan, UT) that in turn connected to a CR10X data logger (Campbell Scie ntific, Logan, UT) (Fig 2-7). The CR-10X data logger monitored switch closures every 18.9 liters from the water meters. The data was also collected manually on a weekly basis at minimum. Each pl ot contained an irrigation zone for turfgrass and mixed ornamentals. Irrigation sprinklers specified for the turfgrass portions of the plots consisted of Rain Bird (Glendora, CA) 1806 15 cm pop up spray bodies a nd Rain Bird R13-18 black rotary nozzles (Figure 2-8). In each plot, there were four sprinklers with a 180 degree arc (R13-18H) and a center sprinkler with a 360 degree arc (R13-18F). Microsprays (Maxijet, Dundee, FL) were installed to irrigate the mixed orna mental plants. A pre ssure regulator was installed at the plot to maintain a constant pressure of 6 kPa on the microsprays during irrigation. Thirty year historical rainfall averages were calculated from monthly rainfall data collected by the National Oceanic and Atmospheric Administration (NOAA 2005) from 1975 through 2005. The closest NOAA weather station from the pr oject site with available rainfall data was located approximately 28 km away, in Parrish, FL.

PAGE 45

45 There were five seasons of data collec tion: 13 August, 2006 through 30 November, 2006 as fall 2006; 1 December, 2006 through 26 February, 2007 as winter 2006-2007; 27 February, 2007 through 31 May, 2007 as spring 2007; 1 June, 2007 through 31 August, 2007 as summer 2007; and 1 September, 2007 through 30 November 2007 as fall 2007. All five treatments observed local watering restrictions during fall 2006 and winter 2006-2007 which included irrigation windows two days per week, Wednesd ay and Saturday, and no watering between 10 am and 4 pm. Also, the ET controller treatmen ts were established based on the site location without accounting for system efficiency (Table 2-1). T1, the Weathermatic controller, was set to apply 100% of the calculated water requireme nt while T2 and T3, the Toro and ET Water controllers, were set to the maximum efficiency of 95%. The monthly irrigation depth for T4, the time-based treatment, was 60% of the net i rrigation requirement derived from historical ET and effective rainfall specific to south Florid a (Dukes and Haman 2002a) and T5 was a reduced treatment, applying 60% of the irrigation depth cal culated from T4 (Table 2-2). Spring, summer, and fall 2007 differed from the previous three seasons in that the ET controller treatments allowed irrigation windows seven days per week and were updated with a system efficiency of 80% determined from irrigation uniformity testing (Table 2-3). The time-based treatment, T4, was increased to apply irrigati on to replace 100% of the net irri gation requirement instead of 60% used during the first three seasons (Table 2-4). Once again, T5 applied 60% of T4 resulting in the reduced treatment applying 60% of the net irrigation requirement. Results were quantified by comparing all trea tments to a time-based treatment without a rain sensor (time WORS). The time-based treatme nt without a rain sensor was derived from T4 by including water application from irrigation events that were bypassed due to rain and was not an actual treatment. Irrigation runtimes for th is treatment were adjusted monthly based on

PAGE 46

46 historical ET and effective rainfall. Data coll ection included: rainfall data at fifteen minute intervals from a Florida Automated Weather Ne twork (FAWN) weather st ation located on-site; irrigation water applied pe r plot from totalizing flow meters; a nd turfgrass quality measurements. Turfgrass quality was measured monthly using the National Turfgra ss Evaluation Program (NTEP) standards (Shearman and Mo rris 2006). The turfgrass was rated on a scale from 1 to 9 where 1 represented dead turfgrass or bare groun d, 9 represented an ideal turfgrass, and 5 was considered minimally acceptable quality for a resi dential setting. Each rating was determined by examining aspects of color, density, uniformity, te xture, and disease or e nvironmental stress. Evaluations were made monthly by the same gr aduate research assistant using the NTEP standards (Appendix B). Volunteer s belonging to the Master Gard eners Association were asked to evaluate the turfgrass quality various times of the year in addition to the ratings completed once per month as a quality control method (Appe ndix B). Approximately six Master Gardeners rated the study plots during the months of December 2006, January 2007, March 2007, May 2007, July 2007, and October 2007. The monthly ra tings taken by the gradua te research assistant were used in the statistical analysis of turfgrass quality for all seasons. Water application was summed into weekly totals for statistical comparisons between treatments using weeks as repeated measures. Water application per plot was summed for 14 days prior to turfgrass rating days for statistical comparisons to turfgrass quality. Two-week water application was used for turfgrass quality comparisons because ratings were taken at irregular intervals thro ughout the season and the water applie d over the two weeks prior to the rating date would have the most effect on the rating. SAS statistical software (SAS Institute, Inc., Cary, NC) was used for all statistical analysis, utilizing the General Linear M odel (GLM) procedure and the mixe d procedure. The confidence

PAGE 47

47 interval was assumed to be 95%. Means se paration was conducted using Duncans multiple range test and least squares means separation was conducted using the Tukey-Kramer test for pairwise comparisons. Results and Discussion All months received less rain than historical average except for the following three months: July 2006, 97% higher than average; April 2007, 53% higher than aver age; October 2007 104% higher than average (Figure 2-9). Overall, both years were drier th an the historical average with a total of 1326 mm of rainfall for the approxima te 16-month study period occurring from August 2006 through November 2007. This was 33% less than the historical total from the local NOAA weather station and 29% less when compared to the Florida average of 1400 mm/year. There were 145 rain events over 472 days; 69% of the study period contained dry days. Irrigation water application data was collected from the three replications of each brand of ET controller at the Gainesvill e turfgrass plots (Table 2-5). It was determined through an ANOVA that there were no differe nces between the Weathermatic replications (P=0.9263), the Toro replications (P=0.9998), or the ET Water replications (P= 0.9989). Therefore, the results found at the Gainesville turfgrass plots for each br and of controller increased the validity of the results from the controllers located at the GCREC. Fall 2006 This season suffered from an infestation of chinch bugs (Blissus insularis Barber) and a fungal disease known as Curvularia. Chinch bugs ar e small pests that inhabit areas of thatch in St. Augustinegrass and live off of plant fluids causing the turfgrass to die (Buss 1993). Curvularia is a pathogen that t ypically attacks stressed plant material (Wong et al. 2005). The chinch bug problem was treated by Ta lstar with an active ingredient of Bifenthrin (7.9%) at a rate of 30 ml per 93 m2 on 14 Sep, 2006. Scotts Lawn Fungus Control was applied to all turf

PAGE 48

48 areas on 19 September, 2006 with an active ingredie nt of Thiophanate-methyl at 2.3% to control the fungal problem. Damaged turfgrass was re placed with new sod during the week following 26 September, 2006; no more than 25% of any pl ot was resodded and most of the damage was located along the edges of the plots. The time-based treatment, T4, irrigated the most by applying 230 mm, whereas the reduced time-based treatment, T5, irrigated the least, applying 144 mm (Figure 2-10). Cumulatively, the Weathermatic (T1) and Toro (T2) applied similar depths over the season totaling 197 mm and 193 mm, respectively. The ET Wate r controller, T3, did not functi on during this season; results could not be reported. All treatments irrigated less than the time WORS treatment, cumulatively totaling 317 mm. The ET controller treatments applied less irri gation than the time WORS treatment except for the month of October as can be seen in the st eeper slopes of the lines (Figure 2-10). October 2006 experienced less time-based irrigation beca use the schedule derived from Dukes and Haman (2002) contained an error for October in south Florida. Irrigat ion application for the time-based treatments should have resembled Sept ember since October had less rainfall and no more than a 4% difference in ET, totaling 119 mm in September and 115 mm in October. Rainfall events occurred within 24 hours of a scheduled irrigation event, causing many of the scheduled events to be bypassed by all tr eatments. The Weathermatic controller, T1, bypassed more events due to the mandatory 48hour bypass period initiated for each rainfall event greater than 6 mm in the early part of the season. Since the controller was only allowed to irrigate two days per week to follow watering restrictions, there were limited opportunities for this controller to allow irrigati on to occur. However, the Weat hermatic controller, T1, usually

PAGE 49

49 calculated larger irrigation depths per event when allowed to irrigate in the latter part of the season resulting in similar cumulative irrigation as the Toro controller, T2. There were differences among treatments (P<0. 0001), but not replicatio ns (P=0.8073) for fall water application (Table 2-6). The time-ba sed treatments, T4 and T5, averaged 15 mm/wk and 9 mm/wk and were different from each other (P=0.0002). The ET controller treatments, T1 and T2, were not different from each other (P=0 .9995), both averaging 12 mm/wk. There were also not differences between the time-based treat ment, T4, and the Weathermatic controller, T1 (P=0.4152), or Toro controller, T2 (P=0.2945). The Weathermatic, T1, and Toro, T2, controllers were not different (T1: P=0.0626; T2: P=0.1063) co mpared to the reduced time-based treatment, T5. All treatments were different compared to the time WORS where this treatment applied an average of 20 mm/wk (P0.0002). Average turfgrass quality ratings were below the minimally acceptable value of 5.0 for all treatments due to pest problems, fungal diseas e, and the reduced October time schedule for the time-based treatments, T4 and T5, as described a bove (Table 2-7). Twoweek water application was different across treatments (P<0.0269) wh ereas turfgrass quality ratings were not (P=0.7279). More specifically, th e Weathermatic controller, T1, applied more weekly irrigation than the reduced time-based treatment, T5, causing di fferences (P=0.0064). Toro controller (T2) and the reduced time-based treatment (T5), howev er, were not different (P=0.2506) despite the Toro controller applying more per week. Wate r application was not correlated with turfgrass quality (P=0.4503). All treatments showed savings compared to the time WORS treatment (Table 2-6). The reduced time-based treatment, T5, showed the mo st savings at 55% due to the extremely low water application in October. The time-base d treatment, T4, showed 28% savings by also

PAGE 50

50 experiencing the low watering schedule in October. Savings from the ET controller treatments, the Weathermatic (T1) and Toro (T2), fell betw een the other treatments by saving 38% and 39%, respectively. Winter 2006-2007 Winter water application was less than any other season due to the reduced climatic demand. Irrigation application ranged from 84 mm by T2, the Toro controller, to 169 mm by T4, the time-based treatment (Figure 2-11). The ET Water controller, T3, did not function during this season and results were not reported. Ra infall totaled 167 mm over the 88 day period. Irrigation events were less frequent for the ET co ntrollers; the Toro (T2) irrigated 12 times and the Weathermatic (T1) irrigated 16 times out of a possible 25 irrigation days compared to 20 events by the time-based treatments, T4 and T5. Water savings were experienced by all treatments compared to the time WORS treatment ranging from 20% to 60%. There were differences among treatments (P<0. 0001), but not plot replications (P=0.9484) for weekly water application (Table 2-8). The W eathermatic controller, T1, and Toro controller, T2, were not different to each other (P=0 .1877) by averaging 7 mm/wk and 6 mm/wk, respectively, but were different to T5, the reduced time-based treatmen t, (T1: P=0.9663; T2: P=0.5401) for this season averaging 7 mm/wk. Ho wever, these three treatments were different from T4, the time-based treatment (P<0.0001), by applying 11 mm/wk. Also, all treatments were different from the time WORS treatment (P 0.0004); water application averaged 14 mm/wk for this season. Significant differences were observed between treatments (P<0.0001), but not plot replications (P=0.9846) for two-week water appl ication (Table 2-9). Water application by all treatments were different with 95% confidence compared to the reduced time-based treatment, T5, except for T1, the Weathermatic controlle r (P=0.8543), and T2, the Toro controller

PAGE 51

51 (P=0.1915). Turfgrass quality ratings ranged fr om 5.7 to 6.0 and were not different across treatments (P=0.4055). Also, there was not a correlation between two-w eek water application and turfgrass quality (P=0.0818). Both ET controller treatments, T1 and T2, a pplied less water than the reduced time-based treatment, T5, unlike any other time of year. Th e ET controller treatments show potential to save over 50% of water applied in subs equent winter seasons. Spring 2007 Irrigation application ranged from 244 mm by T5, the reduce d time-based treatment, to 445 mm by T1, the Weathermatic controller (Fig ure 2-12). The time-based treatments, T4 and T5, bypassed three irrigation events in April that were attributed to the rain sensor, but no rainfall occurred during that time. These events were superimposed in the cumulative irrigation figure (Figure 2-12), but were not included in the weekly irrigation application av erages (Table 2-10). All ET controller treatments irrigated a smaller amount per event, but more frequently than the time-based treatments (Figure 2-12). Ho wever, weekly water applications were not necessarily less by the ET controllers. Averag e weekly water applicat ion by T3, the ET Water controller, was 24 mm/wk (Table 2-10) and was f ound to be different from all other treatments: the Weathermatic controller, T1, averaged 32 mm/wk (P<0.0001); the Toro controller, T2, averaged 30 mm/wk (P=0.0002); T4, the time-bas ed treatment averaged 29 mm/wk (P=0.0053); and T5, the reduced time-based treatment averaged 17 mm/wk (P<0.0001). All treatments except the reduced time-based treatment, T5, appl ied more weekly irrigation than the ET Water controller, T3. The other two ET controller treatments, the Weat hermatic (T1) and Toro (T2) controllers, were not different from each other (P=0.5553); however, they were different from the reduced time-based treatment, T5 (P<0.0001). The Weathermatic controller, T1, and Toro

PAGE 52

52 controller, T2, were not diffe rent from T4, the time-based treatment (T1: P=0.1317; T2: P=0.9638). Irrigation events occurred every day by the Weathermatic, T1, and ET Water, T3, treatments (Figure 2-12). The Weathermatic co ntroller, T1, irrigated every allowable watering day regardless if a sufficient amount of water was calculated to have left the root zone. The ET Water treatment, T3, irrigated everyday becau se it was programmed with a 25% allowable depletion instead of 50% origin ally programmed causing the controller to irrigate when 25% of the water was calculated to have left the root zone. This controller also would not recognize a rain sensor despite repeated attempts with ET Water customer service to repair. The ET Water controller, T3, frequently had poor signal strength and the irrigation schedule was not updated from April 9, 2007 th rough May 23, 2007. When signal problems occur, this controller uses the last schedule unt il communication can be re-established. Thus, the water application rate stayed constant throughout the spring se ason while the other treatments increased the irrigation rate (i.e., frequency) based on increased climatic demand and little rainfall. The 30% irrigation savi ngs attributed to this controll er (Table 2-10) was an overestimate due to the constant irrigation rate in the spring. The average two-week water applicati on ranged from 31 mm by the 60% time-based treatment, T5, to 74 mm by the Weathermatic cont roller, T1 (Table 2-11). Differences were observed between treatments (P<0 .0001), but not replications (P=0.9845) for two-week water application. Differences were not found be tween the Weathermatic, T1, and Toro, T2, controllers (P=0.9293), averaging 74 mm and 70 mm, as well as the ET Water controller, T3, and time-based treatment, T4 (P=1.000), both averaging 51 mm. However, T1 (P 0.0006) and T2 (P 0.0060) were different from T3 and T4. Also, all treatments were different from T5, the

PAGE 53

53 reduced time-based treatment (T1, T2: P<0.0001; T3, T4: P=0.0025). All treatments maintained similar turfgrass quality ratings well above th e minimally acceptable level, averages ranging from 6.1 to 6.4, and were not different from each other (P=0.9636). Turfgrass quality was not correlated with two-week water application (P=0 .7451). Despite the reduced watering by T5, the reduced time-based schedule st ill had an above average tu rfgrass quality rating. Rainfall totaled 109 mm over this season. Th e time-based schedules, T4 and T5, applied irrigation during every scheduled event for the m onths of March and May (Figure 2-12). Each rain event occurring in March was not substantial enough to trig ger the rain sensor to bypass irrigation and as mentioned earlier there was no ra infall in May. Irrigation savings by the ET controller treatments were based purely on their ability to match irrigation application with environmental demand and not affected by the vari ability of the rain se nsor during these two months. Water savings by all treatments compared to the time WORS treatment ranged from 9% by the Weathermatic controller, T1, to 50% by the reduced time-based treatment, T5 (Table 2-10). Average weekly water application for the time WORS treatment was different from the Toro controller, T2 (P=0.0009), but was not differ ent from the Weathermatic controller, T1 (P=0.1646). The time-based treatments, T4 and T5 and ET Water controller, T3, were different (P<0.0001) to the time WORS treatment, also. Summer 2007 Compared to the spring, rainfall was more frequent during the summer of 2007, totaling 446 mm. Irrigation ranged from 228 mm by T5, the re duced time-based treatment, to 425 mm by T4, the time-based treatment (Figure 2-13). Th e ET Water controller, T3, continued to apply irrigation every day without a functional rain sensor.

PAGE 54

54 A power outage occurring on 8 June, 2007 cause d the pump connected to the well, the source for irrigation water, to stop functioning as well as the Weathermatic weather monitor to discontinue taking measurements to calculate ETo. The water source was switched to the pressurized system provided by the farm that could not supply a constant sp ecified pressure until the pump could be replaced. The depth of ir rigation applied per event for the time-based treatments, T4 and T5, were not constant throu ghout the season due to the changing pressure of the water source. All treatments were subject to same pressure issues. Since the Weathermatic controller, T1, did not operate based on an ET schedule, data for this controller was removed for this season. This controller continued to supply irrigation dail y where total weekly irrigation applied equaled T4, the time-based treatment. The ET controllers, T2 and T3, irrigated less depth per event, but applied irrigation more frequently than the time-based treatments (F igure 2-13); however, av erage weekly irrigation applied by the ET controllers, 26 mm/wk by the Toro controller (T2) and 24 mm/wk by the ET Water controller (T3), was greater than T5, the reduced time-based treatment (16 mm/wk; Table 2-12). The ET controller treatments, Toro (T2) and ET Water (T3), were not different from each other (P=0.8021), but were different from T5, th e reduced time-based treatment (P<0.0001). The time-based treatment, T4, was different from th e reduced time schedule, T5 (P<0.0001), the ET Water controller, T3 (P=0.0012), and th e Toro controller, T2 (P=0.0460). Turfgrass quality ratings were not different across treatm ents (P=0.9329) and remained above the minimally acceptable levels (Table 2-13). Differences were observed between treatments (P<0.0001), but not plot replications (P=0.9606) for twoweek water application. The Toro controller, T2, was not different compared to the ET Water controller, T3 (P=0.3126) or the time-based treatment, T4 (P=0.1812). However, the ET Water controller, T3, was found to be

PAGE 55

55 different compared to T4 (P=0.0030). Both ET controller treatments, T2 and T3, were different compared to T5, the reduced time-based treatment (P<0.0001). The time-based treatments, T4 and T5, were also different from each other (P <0.0001). Turfgrass quality was not correlated with two-week water ap plication (P=0.5910). Water savings by all treatments compared to the time WORS treatment (Table 2-12) ranged from 31% by T4, the time-based treatment, to 63% by T5, the reduced time-based treatment. Savings from the ET controller tr eatments, the Toro (T2) and ET Water (T3) controllers, fell between the othe r treatments by saving 41% and 45% respectively. The average weekly water application by the time WORS treatment was 44 mm /wk and was different from all treatments (P<0.0001). Fall 2007 Water application ranged from 209 mm by T2, th e Toro controller, to 427 mm by T4, the time-based treatment in the Fall 2007 (Figure 2-14 ). Rainfall during this period totaled 264 mm. There were differences between treatments (P<0 .0001), but not plot replications (P=0.7412). Average weekly water applicati on (Table 2-14) for the ET Wate r controller, T3, was 18 mm/wk and was not different compared to the other ET controller treatments, Weathermatic, T1 (20 mm/wk; P=0.5516) and Toro, T2 (15 mm/wk; P= 0.1492), as well as the reduced time-based treatment, T5 (18 mm/wk; P=1.000). The Weathe rmatic controller, T1, and Toro, T2, controller were different compared to the time WORS treat ment (P<0.0001) and were different from each other (P=0.0007). All treatments were different compared to the time-based treatment, T4 (P<0.0001). The water source was switched back to the well after the pump was replaced on 31 August, 2007; water application by the time-based treatment s was more constant per event (Figure 2-14). The ET Water controller, T3, was updated to a 50 % allowable depletion before scheduling an

PAGE 56

56 irrigation event on 29 October, 2007 resulting in less frequent events with larger depths applied per event. The weather monitor used to gather weather information for the Weathermatic controller, T1, was knocked off of its mount on 9 October, 2007. It was uncertain the length of time prior to this date that the monitor was not at a proper he ight or in the vertical direction. Irrigation occurred despite rain events due to the misalignment of the rain sensor and runtimes calculated by this controller during this peri od were possibly skewed. There were no differences between any of the ET controller treatments in two-week water application. The ET Water controller, T3, was not different compared to the Weathermatic controller, T1 (P=0.9985) and the Toro contro ller, T2 (P=0.3805) totaling 39 mm. Also, the Weathermatic, T1 (40 mm) was not different (P=0. 2409) compared to the Toro controller, T2 (33 mm). The Weathermatic controller, T1, and ET Water controller, T3, were not different compared to T5, the reduced time-based treatm ent (T1: P=0.4097; T3: P=0.2634) whereas the Toro controller, T2, was different (P=0.0035). All treatments were different compared to the time-based treatment, T4 (P<0.0001). Turfgrass quality was similar across all treatments and higher than the minimally acceptable value of 5, ranging from 6.4 to 7.1; quality was not different between treatments (P=0.1699). Turfgr ass quality was not corr elated with two-week irrigation depth (P=0.1777). The Weathermatic controller, T1, saved 43% compared to time-based irrigation without a rain sensor while the Toro, T2, and ET Water, T3 controllers saved 59% and 50%, respectively. Both time-based treatments, T4 and T5, also sh owed water savings from 15% to 50%. The time WORS treatment was different compared to all treatments (P 0.0006).

PAGE 57

57 Summary and Conclusions The ET controller treatments consistently applied less than the time-based treatment utilizing a rain sensor (T4) for the first two seasons with only the exception of the incorrect irrigation schedule in October. The ET controller treatments al ways applied less than T4 over the last three seasons except for the following three months: 44% for April (Weathermatic, T1), 7% for May (Weathermatic, T1), and 5% for July (Toro, T2). Some excess watering could be attributed to controller installation; the ET Water, T3, applied initial irrigation based on the assumption th at the soil was dry, 0% volumetric moisture content, to account for the wo rst case scenario. Upon activation, the controller applied more water than necessary to ensure that the soil was filled to field capacity despite being well watered during the establishment period. All treatments applied less water compared to cumulative irrigation for the theoreticallyderived time-based treatment without a rain se nsor (Time WORS). Average potential water savings by using a rain sensor at a 6 mm threshold was 21% over the entire study period. Rainfall was much less than the historical aver age resulting in dry c onditions. These savings occurred despite dry conditions due to schedulin g only two irrigation events per week. There was a high probability of rainfall events greater than 6 mm occurring within each season to cause at least one of the irrigation events to bypass, creating water savings. The reduced time-based treatment, T5, averaged 53% savings for the study period. When operating properly, all ET controller tr eatments exhibited considerable savings according to statistical differences compared to time WORS for every season except spring 2007. This occurred because the time-based treatments were developed considering historical effective rainfall. However, the spring 2007 season experienced very little rainfall a nd an increase in the demand for irrigation. Even though more irriga tion occurred compared to the time-based

PAGE 58

58 treatments, the ET controllers were reacting to the plant wa ter needs based on real-time conditions and not histor ical needs. The Weathermatic controller, T1, averaged 35% savings for the entire study period compared to the time WORS treatment. Averag e savings during watering restrictions, fall 2006 and winter 2006-2007, were 44%. This could be attributed to less cumulative irrigation application over the winter months due to more accu rate estimation of water need for the season. Savings for 2007 seasons averaged 26% by this treatment. Savings we re less because spring water requirements were higher than historical needs. Also, irri gation application was higher for the fall due to less rainfall bypassing caused by th e orientation of the weather monitor. The Toro controller, T2, showed considerable savings during both years averaging 50% for the fall 2006 and winter 2006-2007 seasons and 38 % for the 2007 seasons, averaging 43% for the year. Average savings were less for th e 2007 seasons due to increased water demand for spring. The ET Water controller, T3, resulted in 42% savings for the last three seasons. All treatments maintained acceptable turfgrass quality. The Weathermatic and Toro treatments had more water savings for fall 2007 compared to fall 2006. It was likely that water savings were experienced by the ET c ontroller treatments because watering restrictions were removed. Th e ET controllers were able to apply irrigation when calculated as necessary and not accumulating over many days before irrigation can occur. More savings were also possible for fall 2007 du e to increasing the net irrigation requirement replacement for the time-based schedules fr om 60% to 100% after winter 2006-2007. Haley et al. (2007) found that homes in Centra l Florida used an average of 149 mm/month when their time clocks were not adjusted over the year. Compared to this benchmark, fall 2006 and winter 2006-2007 savings for the ET contro ller treatments were 60% and 71% while the

PAGE 59

59 time-based treatments, T4, T5, and time WORS saved 47%, 63%, and 29%, respectively. During the last three seasons, all treatments irri gated less than the average homeowner except for time WORS (29% increase). Sa vings ranged from 6% by the time-based treatment, T4, to 46% by the reduced time-based treatment, T5. ET c ontroller savings were 20%, 26%, and 30% for the Weathermatic, Toro, and ET Water, respectively. The time-based treatment, T4, developed from 100% replacement of the net irrigation requirement, consistently applied more cumula tive irrigation compared to the ET controller treatments. Also, the reduced time-based schedule T5, applied the least amount of water in all seasons except winter 2006-2007 and fall 2007. However, turfgrass quality remained above the minimally acceptable level for both treatments and there were no statistical differences between the ratings. As a result, 60% replacement of net irrigation re quirements is appropriate for effective water application assuming good uniformity. The reduced time-based schedule, T5, resulted in similar savings as ET controllers. Thus, as has been shown in previous research in Fl orida, changing time clock settings throughout the year can result in substantial irrigation savings. Fall 2006 and winter 2006-2007 were scheduled for only 36% replacement (60% reduction of 60% of the net irrigation requirement) of net irrigation requirement for the reduced time-based treatment, but still irrigated more in the winter compared to the ET controller treatments. Time -based treatments were developed from the net irrigation requirement for the area resulting in less water applied than if scheduled without using historical ET and effective rainfall. Howeve r, time-based schedules do not fluctuate with changing weather conditions and typical homeowners will not manually adjust on a regular basis. Thus, the ET controllers are necessa ry for consistent water savings.

PAGE 60

60 Table 2-1. Program settings for each brand of ET controller for summer 2006, fall 2006, and winter 2006-2007 Setting Weathermatic Toro ET Water Sprinkler type1 15.2 mm/hr 15.5 mm/hr 15.5 mm/hr Plant type2 Warm season turfgrass Warm seaso n turfgrass Warm season turfgrass Root depth NA 152 mm 152 mm Soil type3 Sandy Sandy Sandy Slope 0 0 0 Efficiency4 100% 100% 100% Zip code5 33598 NA NA Microclimate NA Full Sun Full Sun Days allowed6 Wed, Sat Wed, Sat Wed, Sat 1Application rate or precipitation rate is term ed sprinkler type for some ET controllers. 2The plant type setting is used to c hoose crop coefficients to calculate plant evapotranspiration. 3The soil type setting is used to determine the de pth of available water for the root zone. 4Scheduling efficiency is used to calculate gross i rrigation once net irrigation is determined. 5Zip code is used to find the latitude to determine the monthly solar radiation for ET calculations. 6Days Allowed refers to the days irrigation was allowed to occur per week.

PAGE 61

61 Table 2-2. Runtimes and application amounts per irrigation event1 for the time-based treatment (T4) operating on a twice weekly sc hedule for fall 2006 and winter 2006-2007 seasons Month Mixed-ornamentals2 Turfgrass Time3 (min) Depth (mm) Time (min) Depth (mm) January 19 6 23 6 February 20 6 24 6 March 28 9 35 9 April 30 10 37 10 May 28 9 34 9 June 25 8 31 8 July 39 12 48 12 August 43 14 53 14 September 26 8 31 8 October 27 8 32 8 November 27 8 33 8 December 24 7 29 7 Total4 672 210 820 210 1Assumed 60% system efficiency and es timated effective rainfall for south Florida with 60% ET replacement. 2Application rate of 0.61 in/hr for turfgrass and 0.75 in/hr for mixed-ornamentals. 3Two irrigation events per week. 4Total was calculated for the year in cluding both irrigation events.

PAGE 62

62 Table 2-3. Program settings for each brand of ET controller for spring 2007, summer 2007, and fall 2007 Setting Weathermatic Toro ET Water Sprinkler type1 15.2 mm/hr 15.5 mm/hr 15.5 mm/hr Plant type2 Warm season turfgrass Warm seaso n turfgrass Warm season turfgrass Root depth NA 152 mm 152 mm Soil type3 Sandy Sandy Sandy Slope 0 0 0 Scheduling efficiency4 80% 80% 80% Zip code5 33598 NA NA Microclimate NA Full sun Full sun Days allowed6 Everyday Everyday Everyday 1Application rate or precipitation rate is term ed sprinkler type for some ET controllers. 2The plant type setting is used to choose crop co efficients to ultimately calculate plant evapotranspiration. 3The soil type setting is used to determine the depth of available water for the root zone. 4Scheduling efficiency is used to calcula te gross irrigation on ce net irrigation is determined. 5Zip code is used to find the latitude to determine the monthly solar radiation for ET calculations. 6Days Allowed refers to the days irri gation was allowed to occur per week.

PAGE 63

63 Table 2-4. Runtimes and application amounts per irrigation event1 for the time-based treatment (T4) operating on a twice weekly schedule for spring, summer, and fall 2007 seasons Month Mixed-ornamentals2 Turfgrass Time3 (min) Depth (mm) Time (min) Depth (mm) January 31 10 39 10 February 33 11 41 11 March 47 15 58 15 April 50 16 62 16 May 46 15 56 15 June 42 13 51 13 July 65 21 80 21 August 72 23 88 23 September 43 14 52 14 October 43 14 53 14 November 44 14 55 14 December 39 12 48 12 Total4 1110 356 1366 356 1Assumed 60% system efficiency and es timated effective rainfall for south Florida with 60% ET replacement. 2Application rate of 0.61 in/hr for turfgrass and 0.75 in/hr for mixed-ornamentals. 3Two irrigation events per week. 4Total was calculated for the year in cluding both irrigation events. Table 2-5. Average daily irrigation water applicat ion for the three replications of ET controllers located at the Gainesville turfgrass plots from May 22, 2007 through November 30, 2007 Replication Weathermatic Toro ET Water A 1.1 a 1.5 a 1.2 a B 1.2 a 1.5 a 1.2 a C 1.1 a 1.5 a 1.2 a Average** 1.1 B 1.5 A 1.2 B Numbers with different letter s indicated differences at the 95% confidence level using Duncans Multiple Range Test. **Statistical analysis was performed on controller brands and results are shown with different letters for average values.

PAGE 64

64 Table 2-6. Fall 2006 weekly water applicati on (August 13 through November 30) and savings compared to the time-WORS treatment1 using cumulative season totals Treatment Controller Average water application (mm/wk) Savings compared to time WORS 1 Weathermatic 12 bc 38% 2 Toro 12 bc 39% 3 ET Water NA2 NA 4 Time 15 b 28% 5 0.6*Time 9 c 55% Time WORS 20 a -* Numbers with different letters in columns indicated differences at the 95% confidence level using Duncans Multiple Range Test. 1The time WORS treatment refers to the time-based treatment without a rain sensor theoretically derived from T4. 2NA is an abbreviation for Not Applicable and was used for treatments that were not working. Table 2-7. Fall 2006 two-week water applicati on and turf quality summary (August 13 through November 30) Treatment Controller Two-week water applied (mm) Turfgrass quality1 1 Weathermatic 30 a 4.8 a 2 Toro 23 ab 4.9 a 3 ET Water NA2 NA 4 Time 20 ab 4.7 a 5 0.6*Time 14 b 4.8 a Numbers with different letters in columns indicated difference s at the 95% confidence level using Duncans Multiple Range Test. 1Turfgrass quality ratings used a 1 to 9 scale where 1 was of lowest quality, 9 was of highest quality, and 5 was minimally acceptable. 2NA is an abbreviation for Not Applicable and was used for treatments that were not working.

PAGE 65

65 Table 2-8. Winter 2006-2007 weekly water appli cation (December 1 through February 26) and savings compared to the time WORS treatment1 using cumulative season totals Treatment Controller Average water application (mm/wk) Savings compared to time WORS 1 Weathermatic 7 c 50% 2 Toro 6 c 60% 3 ET Water NA2 NA 4 Time 11 b 20% 5 0.6*Time 7 c 49% Time WORS 14 a -* Numbers with different letters in columns indicated differences at the 95% confidence level using Duncans Multiple Range Test. 1The time WORS treatment refers to the time-based treatment without a rain sensor theoretically derived from T4. 2NA is an abbreviation for Not Applicable and was used for treatments that were not working. Table 2-9. Winter 2006-2007 two-week water app lication and turf quality summary (December 1 through February 26) Treatment Controller Two-week water applied (mm) Turfgrass quality1 1 Weathermatic 18 b 5.7 a 2 Toro 11 c 5.9 a 3 ET Water NA2 NA 4 Time 26 a 6.0 a 5 0.6*Time 16 bc 5.7 a Numbers with different letters in columns indicate difference s at the 95% confidence level using Duncans Multiple Range Test. 1Turfgrass quality ratings used a 1 to 9 scale where 1 was of lowest quality, 9 was of highest quality, and 5 was minimally acceptable. 2NA is an abbreviation for Not Applicable and was used for treatments that were not working.

PAGE 66

66 Table 2-10. Spring 2007 weekly water applica tion (February 27 through May 31) and savings compared to the time WORS treatment1 using cumulative season totals Treatment Controller Average water application (mm/wk) Savings compared to time WORS 1 Weathermatic 32 ab 9% 2 Toro 30 b 15% 3 ET Water 24 c 30% 4 Time 29 b 18% 5 0.6*Time 17 d 50% Time WORS 35 a -* Numbers with different letters in columns indicate differences at the 95% confidence level using Duncans Multiple Range Test. 1The time WORS treatment refers to the time-based treatment without a rain sensor theoretically derived from T4. Table 2-11. Spring 2007 two-week water app lication and turf quality summary (February 27 through May 31) Treatment Controller Two-week water applied (mm) Turfgrass quality1 1 Weathermatic 74 a 6.2 a 2 Toro 70 a 6.4 a 3 ET Water 51 b 6.3 a 4 Time 51 b 6.2 a 5 0.6*Time 31 c 6.1 a Numbers with different letters in columns indicate difference s at the 95% confidence level using Duncans Multiple Range Test. 1Turfgrass quality ratings used a 1 to 9 scale where 1 was of lowest quality, 9 was of highest qu ality, and 5 was minimally acceptable.

PAGE 67

67 Table 2-12. Summer 2007 weekly water application (June 1 through August 31) and savings compared to the time WORS treatment1 using cumulative season totals Treatment Controller Average water application (mm/wk) Savings compared to time WORS 1 Weathermatic NA2 NA 2 Toro 26 bc 41% 3 ET Water 24 c 45% 4 Time 30 b 31% 5 0.6*Time 16 d 63% Time WORS 44 a -* Numbers with different letters in columns indicate differences at the 95% confidence level using Duncans Multiple Range Test. 1The time WORS treatment refers to the time-based treatment without a rain sensor theoretically derived from T4. 2NA is an abbreviation for Not Applicable and was used for treatments that were not working. Table 2-13. Summer 2007 two-w eek water application and turf quality summary (Jun 1 through Aug 31) Treatment Controller Two-week water applied (mm) Turfgrass quality1 1 Weathermatic NA2 NA 2 Toro 57 ab 6.1 a 3 ET Water 52 b 6.1 a 4 Time 64 a 6.1 a 5 0.6*Time 32 c 5.8 a Numbers with different letters in columns indicate difference s at the 95% confidence level using Duncans Multiple Range Test. 1Turfgrass quality ratings used a 1 to 9 scale where 1 was of lowest quality, 9 was of highest quality, and 5 was minimally acceptable. 2NA is an abbreviation for Not Applicable and was used for treatments that were not working.

PAGE 68

68 Table 2-14. Fall 2007 weekly wa ter application (September 1 through November 30) and savings compared to the time WORS treatment1 using cumulative season totals Treatment Controller Average water application (mm/wk) Savings compared to time WORS 1 Weathermatic 20 c 43% 2 Toro 15 d 59% 3 ET Water 18 cd 50% 4 Time 31 b 15% 5 0.6*Time 18 cd 50% Time WORS 36 a -* Numbers with different letters in columns indicate differences at the 95% confidence level using Duncans Multiple Range Test. 1The time WORS treatment refers to the time-based treatment without a rain sensor theoretically derived from T4. Table 2-15. Fall 2007 water application and tu rf quality summary (September 1 through November 30) Treatment Controller Two-week water applied (mm) Turfgrass quality 1 Weathermatic 40 bc 6.4 a 2 Toro 33 c 7.1 a 3 ET Water 39 bc 7.0 a 4 Time 77 a 6.6 a 5 0.6*Time 45 b 6.5 a Numbers with different letters in columns indicate difference s at the 95% confidence level using Duncans Multiple Range Test. 1Turfgrass quality ratings used a 1 to 9 scale where 1 was of lowest quality, 9 was of highest qu ality, and 5 was minimally acceptable.

PAGE 69

69 Figure 2-1. Three brands of ET controllers were tested on twen ty landscaped plots at the University of Florida Gulf Coast Research and Education Center. Figure 2-2. Nine additional ET cont rollers, three of each brand, are installed at the University of Florida Gainesville turfgrass plot s (Aerial photo from Google Maps).

PAGE 70

70 Figure 2-3. Research plot layout an d controller treatments at the University of Florida Gulf Coast Research and Education Center in Wimauma, FL. Legend: T1 = Weathermatic T2 = Toro T3 = ET Water T4 = Time T5 = 60% of Time

PAGE 71

71 Figure 2-4. The ornamental plants chosen are: (A) Crape Myrtle (Lagerstroemia indica Natchez) (B) Gold Mound Lantana (Lantana camara Gold Mound) (C) Indian Hawthorne (Raphiolepis indica) (D) Cape Plumbago (Plumbago auriculata) (E) Big Blue Liriope (Liriope muscari Big Blue). A. B. C. D. E.

PAGE 72

72 Figure 2-5. The ET Controllers chosen for the st udy located at the Univer sity of Florida Gulf Coast Research and Education Center in Wimauma, FL A) Weathermatic SL1600 B) Toro Intelli-sense C) ET Water Smart Cont roller 100 and irrigation timer D) Rain Bird.

PAGE 73

73 Figure 2-6. Irrigation water app lication to each plot was mon itored by 11.4 cm V100 w/ pulse output flow meters manufactured by AMCO Water Metering Systems (Ocala, FL).

PAGE 74

74 Figure 2-7. Flow meters are wire d to (A) five SDM-SW8A switch closure input modules that in turn connect to a (B) CR-10X data logger. Water applied is recorded by switch closures every 18.9 liters.

PAGE 75

75 Figure 2-8. Individual plot design of the twenty research plots located at the University of Florida Gulf Coast Research and Education Center in Wimauma, FL. Figure 2-9. Comparison of rainfall for the 2006-2007 study period and average historical rainfall on a monthly and cumulative basis for southwest Florida. The sprinklers are Rain Bird 1806 6-inch pop up spray bodies that have Rain Bird R13-18 black rotary nozzles. There are four 180 degree arc (R13-18H) and one 360 degree arc (R13-18F). The emitters are Maxijet WING piece jets. There are two 180 degree arc (MAR180W) and eleven 340 degree arc (MAR340W) emitters spread evenly throughout the ornamental area.

PAGE 76

76 0 50 100 150 200 250 300 350 400Cumulative Water Applied (mm) 0 20 40 60 80 100 0 5 10 15 20 25 30 35 8/138/279/109/2410/810/2211/511/19Daily Rainfall (mm) Water Applied per Event (mm)Date (2006) Rainfall T1, Weathermatic T2, Toro T4, Time T5, 0.6*Time Time w/o RS Figure 2-10. Fall 2006 cumulative and daily wate r application and daily rainfall (August 13 November 30).

PAGE 77

77 0 50 100 150 200 250 300 350 400Cumulative Water Applied (mm) 0 20 40 60 80 100 120 140 0 5 10 15 20 25 30 35 12/112/1512/291/121/262/92/23Daily Rainfall (mm) Water Applied per Event (mm)Date (2006 2007) Rainfall T1, Weathermatic T2, Toro T4, Time T5, 0.6*Time Time w/o RS Figure 2-11. Winter 2006-2007 cumulative and daily water applied and daily rainfall (December 1 February 26).

PAGE 78

78 0 50 100 150 200 250 300 350 400 450 500Cumulative Water Applied (mm) 0 20 40 60 80 100 120 140 0 5 10 15 20 25 30 35 2/273/133/274/104/245/85/22Daily Rainfall (mm) Water Applied per Event (mm)Date (2007) Rainfall T1, Weathermatic T2, Toro T3, ET Water T4, Time T5, 0.6*Time Time w/o RS Figure 2-12. Spring 2007 cumulative and daily wa ter applied and daily rainfall (February 27 May 31).

PAGE 79

79 0 100 200 300 400 500 600 700Cumulative Water Applied (mm) 0 20 40 60 80 100 120 140 0 5 10 15 20 25 30 35 6/16/156/297/137/278/108/24Daily Rainfall (mm) Water Applied per Event (mm)Date (2007) Rainfall T1, Weathermatic T2, Toro T3, ET Water T4, Time T5, 0.6*Time Time w/o RS Figure 2-13. Summer 2007 cumula tive and daily water applied a nd daily rainfall (June 1 August 31).

PAGE 80

80 0 100 200 300 400 500 600Cumulative Water Applied (mm) 0 20 40 60 80 100 120 140 0 5 10 15 20 25 30 35 9/19/159/2910/1310/2711/1011/24Daily Rainfall (mm) Water Applied per Event (mm)Date (2007) Rainfall T1, Weathermatic T2, Toro T3, ET Water T4, Time T5, 0.6*Time Time w/o RS Figure 2-14. Fall 2007 cumulative and daily water applied and daily rainfall (September 1 November 30).

PAGE 81

81 CHAPTER 3 REFERENCE EVAPOTRANSPIRATION ES TI MATION BY EVAPOTRANSPIRATIONBASED IRRIGATION CONTROLLERS Introduction Water is a limited resource and many areas in the world are experiencing water shortages. More specifically, Florida has th e second largest withdr awal of groundwater in the U.S. that is used for public supply (Solley et al. 1998). Also, compared to other states Florida has the largest net gain in population with an inflow of approximately 1,108 people per day and fourth in overall population (United States Census Bur eau [USCB] 2005). New home construction has increased to accommodate such a large influx of people. Florida ranked first in the construction of single family residential units totali ng 209,162 in 2005 (USCB 2007). Most new homes include in-ground automated irrigation systems. However, homes with in-ground systems utilizing automated irrigation tim ers alone increase outdoor wate r use by 47% (Mayer et al. 1999). The need for landscape irrigation will co ntinually grow with increased population and home construction if there is no change in the demand for aesthetically pleasing urban landscapes. Research has shown that Florida single family residences over-irrigate in late fall and winter because time clock schedules are not ad justed to match changing environmental demand (Haley et al. 2007). Increased watering during the fall and winter months can delay dormancy, leading to the need for additional mowing, and increases the likelihood of diseases, insects, weeds, and stresses to the lawn (Harivandi 1984). On a larger scale, the increased watering contributes to the depletion of water resources and can result in leaching of soluble chemicals into shallow groundwater. Therefore, better irrigation management could potentially lead to water conservation, landscape prob lems, and less groundwater pollution.

PAGE 82

82 Evapotranspiration (ET) is defined as the evaporation from a soil surface and the transpiration from plant material (Allen et al. 1998). ET is part of a balanced energy budget that exchanges energy for outgoing water at the surface of the plant. The components of ET are solar radiation, temperature, relati ve humidity, and wind speed (Allen et al. 2005). Reference ET (ETo) is defined as the ET from a hypothetical re ference crop with the characteristics of an actively growing, well-watered, dense green cool season grass of uniform height (Allen et al. 2005). Typically, climatic data is used as inputs to equations to estimate ETo. There are three types of equations: mass transfer energy balance, and empirical methods (Fangmeier et al. 2006). Most of the current methods employ a co mbination of the three. The appropriate ET equation is chosen depending on many factors including geographical location, types of crops, and weather data availability (Fangmeier et al. 2006). Evapotranspiration-based controllers, also known as ET controllers, are irrigation controllers that use an ET value to schedule irrigation. Each controller works differently depending on manufacturer, but typi cally can be programmed with various conditions specific to the landscape making them more efficien t (Riley 2005). ET controllers receive ETo information in three general ways, consequent ly dividing ET controllers into th ree main types: 1) standalone controllers, 2) signal-based controllers, and 3) hist orical-based controllers. Standalone controllers typically receive climatic data from onsite sensors and calculations to determine ETo are performed by the controller. Ev en though the controllers might take readings every second or every fi fteen minutes, cumulative daily ETo is used for irrigation scheduling. On-site sensors coul d include: temperature, solar radi ation, an ET gauge, or even a full weather station (Riley 2005). Benefits of sta ndalone controllers are that they are not limited

PAGE 83

83 by requiring the use of a full w eather station and there are no signal fees associated with broadcasts from the manu facturer (Riley 2005). Signal-based contro llers receive ETo information from a company that collects climatic data from weather stations locate d near the irrigation si te using satellite or internet technology. Depending on the manufacturer, the ETo data can be from an average of multiple weather stations in the area or from a single weather st ation. There is typical ly a signal fee (i.e., subscription) for this controller set by the manufacturer that normally ranges from $4 to $15 per month (Riley 2005). Historical-based controll ers rely on historical ETo information for the area. Typically, monthly historical ETo is programmed into the controller by the manufacturer or installing contractor. Theoretically, this method does not result in as accurate an ETo estimate because site specific weather variability is not considered. Irrigation application by ET controll ers has been studied frequently in the last five years in the western U.S. Studies were conducted by th e Los Angeles Department of Water and Power (Bamezai 2004), Irvine Ranch Water District (Diam ond 2003), Aquacraft, Inc. (Aquacraft, Inc. 2002; Aquacraft, Inc. 2003), Metr opolitan Water District of Southern California MWDSC 2004), and University of California Cooperative Extens ion (Pittenger et al. 2004) and results are detailed in Chapter 1 and summarized in Chapter 2. However, comparisons of ETo estimations have yet to be documented. The objectives of this experiment were the following: A.) compare the ETo estimation by three brands of ET-based ir rigation controllers to the ASCE-EWRI Standardized ET methodology, B.) quantify the variation between cont rollers of the same brand, and C.) compare

PAGE 84

84 the performance of the controllers based on appr oximate distance to a publicly available weather data source. Materials and Methods This study was conducted at the University of Florida Gulf Coast Research and Education Center (GCREC) in Wimauma, Florida and at the University of Florida Agricultural and Biological Engineering Department turfgrass plots in Gainesville, Florida (see Chapter, Figure 21; Figure 2-2). Three ET c ontroller brands were installed at GCREC as follows: SL1600 controller with SLW15 w eather monitor (Weathermatic, Inc., Dallas, TX), Intelli-sense (Toro Company, Inc., Riverside, CA) utilizing the WeatherTRAK ET Everywhere service (Hydropoint Datasystems, Inc., Petaluma, CA), and Smart Controller 100 (ET Water Systems LCC, Corte Madera, CA). These three brands were also inst alled at the Gainesville turfgrass plots in three replications. The replications at the Gainesvi lle site allowed the study of variability between controllers of the same brand. In addition, the Gainesville site was a pproximately 11 km from the closest public weather station located at the Gainesville Regional Airport whereas the site at GCREC had a Florida Automate d Weather Network (FAWN) station within 100 m of the research site with weather data available via the internet. Data collection from the GCREC location includ ed: climate data at fifteen minute intervals such as wind speed, solar radiation, temperature, relative humidity, and rainfall depth from a Florida Automated Weather Network (FAWN) weat her station located on-site. Maximum and minimum temperature data from the Weathermatic SL1600, daily and weekly summed ETo data from the Toro Intelli-sense contro ller, and weekly summed average ETo from the ET Water Smart Controller 100 were recorded manually. Data collection from the Gainesville location was identical to collection at GCREC with the exception of the weather stations being installed and maintained as part of the research.

PAGE 85

85 SAS statistical software (SAS Institute, Inc., Cary, NC) was used for all statistical analysis, utilizing the General Linear Model (GLM) pr ocedure and the mixed procedure with a 95% confidence level. Time was used as a repl ication. Means separation was conducted using Duncans Multiple Range test and least square s means was conducted using the Tukey-Kramer test for pairwise comparisons. Reference Evapotranspiration Calculations The ASCE standardized reference evapotrans piration equation (Allen et al. 2005) was used to calculate ETo as seen below. )uC+1( + u)e(e 273+T C +G)(R 0.408 =ET2d 2as n n 0 3-1 Variables are defined as follows: 23.237T 3.237T T27.17 exp2503 3-2 2 TeTe emin o max o s 3-3 3.237T T27.17 exp6108.0Teo 3-4 2 100 RH Te 100 RH Te emin max o max min o a 3-5 nlnsnRRR 3-6 s nsR1R 3-7 2 TT e14.034.0fRmin 4 K max 4 K a cd nl 3-8 35.0 R R 35.1fso s cd 3-9 a 5 soRz10275.0R 3-10 s srsc asincoscossinsindG 24 R 3-11

PAGE 86

86 J 365 2 cos033.01dr 3-12 39.1J 365 2 sin409.0 3-13 tantanarccoss 3-14 42.5z8.67ln 87.4 uuw z2 3-15 ETo = reference evapotranspiration, mm/day = psychrometric constant, 0.067 kPa/C = slope of the saturation vapor pr essure-temperature curve, kPa/C T = daily mean air temperature, C es = saturation vapor pressure, kPa eoT = saturation vapor pressure function, kPa ea = actual vapor, kPa RH = relative humidity, % Rn = net radiation, MJ/m2/day Rns = net short-wave radiation, MJ/m2/day Rnl = net outgoing long-wave radiation, MJ/m2/day Rs = incoming solar radiation, MJ/m2/day = albedo or canopy reflec tion coefficient, 0.23 = Stefan-Boltzmann constant, 4.901 x 10-9 MJ/K4/m2/day fcd = cloudiness function, 0.05 fcd 1.0 Rso = calculated clear-sky radiation, MJ/m2/day Ra = extraterrestrial radiation, MJ/m2/day z = station elevation above sea level, m dr = inverse relative distance factor for the earth-sun = solar declination, rad = latitude, rad s = sunset hour angle, rad J = Julian day Gsc = solar constant, 4.92 MJ/m2/hr G = daily soil heat flux density, 0 MJ/m2/day u2 = wind speed at 2 m height, m/s The standard reference crop is grass for Florid a (Irmak and Haman 2003) resulting in constants Cn and Cd as 900 and 0.34, respectively (Allen et al. 2005). Controllers that use only a por tion of the climatic parameters listed above do not use the ASCE standardized ETo equation to calculate ET. Another ETo estimation method is the Hargreaves equation (Hargreaves and Sama ni 1982). The equation is as follows:

PAGE 87

87 17.8+TTDR0.0023=ET2/1 a 0 3-16 TD = difference between the daily maxi mum and daily minimum temperature, C This equation relies on solar radiation calculated from extraterrestrial ra diation and temperature measurements. Controller Descriptions The Weathermatic controller, T1, is a standal one controller because it utilizes an on-site weather monitor to collect ambient air temperature used to calculate ETo by the Hargreaves method (Equation 3-16). This controller stores the maximum and minimum daily temperature used to calculate ETo, allowing an independent manual calculation of ETo using the Hargreaves method. The Toro and ET Water controllers are signalbased. According to the manufacturers, climate parameters such as temperature, relati ve humidity, wind speed, and solar radiation are collected from publicly available weather stations and ETo was calculated using the ASCE method (Equation 3-1). The ETo values were sent to the To ro controllers using paging technology based on the microzone, or designated area of similar ETo, determined by Hydropoint Data Systems (Newport Beach, CA). ETo data for the GCREC controller was made available through e-mail by Hydropoint for comparison to calculated ETo. The ET Water controller used a public weather station to calculate ETo and broadcasted the value to the controller daily using cellular technology. This data was also sent by e-mail from ET Water Systems LCC for the controller at the GCREC. ETo data provided by the manufacturers were compared against values read directly on the controller in the case of the Toro and against values gathered from the ET Water website.

PAGE 88

88 ETo data collection at the G CREC occurred 25 May, 2007 th rough 30 November, 2007, for the Weathermatic controller. ETo values sent to the Toro c ontroller were provided by the manufacturer from 13 August, 2006 through 30 November, 2007. ETo values sent to the ET Water controller were provided by the manufact urer from 4 August, 2006 through 30 November, 2007. As a result, ETo comparisons between all controlle rs at the GCREC must begin on 13 August, 2007. ETo comparisons for the controllers at th e Gainesville turfgrass plots began on 22 May, 2007. Periods of unavailable data for any controll er were removed for all controllers when comparing directly. Data for the GCREC cont rollers was unavailable from ET Water from 17 August, 2006 through 28 August, 2006 as well as 18 September, 2006 through 22 September, 2006. The Toro controller also had missing ETo data on 12 August, 2006 and 4 September, 2006. Missing data for the controllers at the Ga inesville turfgrass plot s occurred throughout the period by a few days or less at a time. Site Descriptions Nine ET controllers, three replicates of each treatment brand, were installed on May 22, 2007 (Figure 3-1). Each Weatherma tic controller utilized an individual on-site weather monitor; each monitor was located at the same height abov e the ground, 1.83 m, and they were staggered 0.81 m apart (Figure 3-2). The other six controll ers were connected to Mini-clik rain sensors (Hunter Industries, Inc., San Marcos, CA). These controllers were virtuall y tested as they were not connected to actual irrigation systems. ETo data was collected directly from the controllers without manufacturer assistance. Five treatments were establis hed at GCREC, T1 through T5, replicated four times for a total of twenty plots in a completely randomized block design (see Chapter, Figure 2-3). From those five treatments, T1 through T3 were ET c ontrollers: T1, Weathe rmatic SL1600 controller

PAGE 89

89 with SLW15 weather monitor; T2, Toro In telli-sense utilizing the WeatherTRAK ET Everywhere service; and T3, ET Water Smart Controller 100. These treatments controlled irrigation to plots at this location. The twenty plots measured 7.62 m x 12.2 m, bordered by a 15.2 cm tall black metal barrier, with 3.05 m buffer zones between adjacent plots. Each plot consisted of 65% St. Augustinegrass (Stenotaphrum secundatum Floratam) and 35% mixed ornamentals to represent a typical residential landscape in Florida. The ornamentals were as follows: Crape Myrtle (Lagerstroemia indica Natchez) (see Chapter, Figure 2-3A), Gold Mound Lantana (Lantana camara Gold Mound) (see Chapter, Figur e 2-3B), Indian Hawthorne (Raphiolepis indica) (see Chapter, Figure 2-3C), Cape Plumbago (Plumbago auriculata) (see Chapter, Figure 2-3D), and Big Blue Liriope (Liriope muscari Big Blue) (see Chapter, Figur e 2-3E). Landscapes were maintained through mowing, pruning, edging, mulching, fertilization, and pest and weed control according to current UF-IFAS recommendations (Black and Ruppert 1998; Sartain 1991). Weather Stations On-site weather stations were used to colle ct weather data for comparison purposes using the ASCE method (Equation 3-1) or Hargreaves method (Equation 3-16), where appropriate. The weather station located on-site at the Gaines ville turfgrass plots wa s installed and managed by our team on a regular basis. The weather station located at GCREC was managed by FAWN personnel. Sensor heights for the FAWN station were similar to the Gainesville weather station except for the anemometer that was mounted at a 10 m height; wind speed data was corrected to 2 m (Equation 3-15). Data was collected from a th ird weather station located at the Gainesville Regional Airport. This station, operate d by the National Oceanic and Atmospheric Administration (NOAA), was the closest weather st ation with publically available weather data to the Gainesville turfgrass plots.

PAGE 90

90 Climatic data was collected at 15-minute interv als for the on-site weather stations and daily intervals for the NOAA weather station at the Gain esville Regional Airport. All stations were subjected to quality control a ssessments (Allen et al. 2005) by comparing solar radiation (Rs), relative humidity (RH), temperature (T), and wind speed (u2) data against relevant physical extremes. Solar radiation and clear-sky solar radiation are equal on cloud-free days. Quality control of Rs was performed by plotting Rs and Rso over time. Rso was calculated using the ASCE methodology (Equation 3-10) as well as a more deta iled procedure describe d below (Allen et al. 2005). aDB soRKKR 3-17 4.0 t Bsin W 075.0 sinK P00146.0 exp98.0K 3-18 26.5293 z0065.0293 3.101P 3-19 1.2Pe14.0Wa 3-20 242.039.1J 365 2 sin3.085.0sinsin 3-21 B DK36.035.0K for KB 0.15 3-22 B DK82.018.0K for KB < 0.15 3-23 KB = clearness index for direct beam radiation KD = transmissivity index for diffuse radiation P = atmospheric pressure at the site elevation, kPa Kt = turbidity coefficient, 0 < Kt 1.0 W = precipitable water in the atmosphere, mm = angle of sun above the horizon, rad Measured relative humidity should be in th e range of 30% to 100% for humid climates (Allen et al. 2005). Relative humidity values less than this range ar e possible, but it is unreasonable to maintain values less than 30%. It is not possible to have relative humidity values greater than 100% in the physical enviro nment. Daily maximum and minimum values of

PAGE 91

91 relative humidity were plotted to verify that the data falls in the acceptable range for a majority of the study period. Another way to verify that the relative humidity data was collected correctly is to calculate daily dew point temperature and compare to the daily minimum temperature. Dew point temperature is calculated with the fo llowing equation (Allen et al. 2005): a a deweln78.16 eln3.23791.116 T 3-24 Tdew = dew point temperature, C Dew point temperature and minimum temperat ure should be approximately the same a majority of the time in humid climates with the exceptions of days with changes in air mass, high winds, or cloudiness at night (Allen et al. 2005). Air temperature data is most likely to be consistent and of the best quality data. Temperature data can be checked for quality by plotting the daily average calculated from the 24-hour time period and the average of the maxi mum and minimum temperatures of the same day over time. These averages should be with in 3C unless caused by rainfall events, unusually high wind speeds, or changes in air mass (Allen et al. 2005). The quality of wind speed data is difficult to assess when duplicate instruments are not used. Ways to determine if the correct data is being reported include plotting daily average wind speed, daily maximum wind speed, and the gust factor. The gust factor is cal culated as a ratio of maximum wind speed to average wind speed. If any of these figures e xhibit consistently low values (< 1.0 m/s) or gust factor values of 1.0, there is some t ype of problem with the data (Allen et al. 2005).

PAGE 92

92 Results and Discussion The nine controllers located at the Gainesville turfgrass plots were installed to determine the variability between controllers of the same brand. These contro llers were labeled as replications A, B, and C for each brand of controller. The three contro llers located at GCREC were labeled as T1, T2, and T3 for each brand. Table 3-1 contains the treatment codes for each controller and a summary of pertin ent information concerning the controller such as location, installation date, and method for obtaining ETo. Climatic Data Quality Control Solar radiation and clear-sky solar radiation should be equal on cloud-free days. Quality control of Rs was performed by plotting Rs and Rso against daily timesteps where Rso was calculated using the Equation 3-10 as well as Equa tion 3-17. Both the GCREC and Gainesville locations showed that E quation 3-10 predicted Rso fairly well compared to the more detailed calculations of Equation 3-17 (Figure 3-3; Figure 3-4; Figure 3-5). Both years of FAWN data showed Rs fitting the Rso curve in late fall, winter, and spring months (Figure 3-3). The summer and early fall months rarely fit the curve for both years, due to cloudy conditions common in the summer and early fall. The NOAA data from the Gainesville Regional Airport (Figure 3-4) as well as the Rs data from the Gainesville turfgrass plots (Figure 3-5) showed the same trend as the data coll ected at GCREC in Hillsborough County in that Rs rarely fit the Rso curve until late in the fall season. Daily maximum and minimum values of relative humidity were plotted to verify that the data fell in the acceptable range (3 0% to 100%) for a majority of the study period. None of the weather stations logged relative humidity valu es greater than 100%. The GCREC weather station measured minimum values above 30% most of the time except from mid-January through May for both years (Figure 3-6). These months are considered part of the dry season in Florida;

PAGE 93

93 arid conditions could result in lower values of relative humidity as long as it never falls below the range of 5% to 10%. The minimum relative humidity at this location was 15% and values dropped below 20% only 9 times in 23 months s howing that the quality of the data is appropriate. Minimum relative humidity data collected from the NOAA station only fell below 30% by 12% over seven years of data and fell be low 20% by 2% of the time (Figure 3-7). The Gainesville turfgrass plots weather station also measured minimum relative humidity above 30% for 94% of the time, with only 11 days having mini mum RH values below 30% (Figure 3-8). Dew point temperature was calculated and comp ared to minimum temperature as another way to verify the relative humidity data. Dew point temperature and minimum temperature should be approximately the same on any particular day for a majority of the time in humid climates. Dew point temperature and minimum temperature were ve ry similar to each other at any of the weather statio n locations. Data from the FAWN station at the GCREC (Figure 3-9), Gainesville NOAA station (Figure 3-10), and the Gainesville turf grass plots location (Figure 311) fit to a one to one scale best with higher temperatures. Minimum temperatures were slightly higher than dew point temperat ures in cooler temperatures for all weather data. The temperature data was checked for quality by plotting the daily mean calculated from the 24-hour time period and the average of the maximum and minimum temperatures of the same day against each other (Figure 3-12; Figure 3-13). There was very little variation between the two averages for both the GCREC and Gainesville turfgrass plots locations when looking at the figures. These averages should be within 3C unless caused by rainfall events, unusually high wind speeds, or changes in air mass. Both locat ions had differences less than 3C for 100% of the time period. Only maximum and minimu m temperature data was available for the

PAGE 94

94 Gainesville Regional Airport weat her station. As a result, mean temperature and average temperature could not be co mpared for this location. Daily average wind speed, daily maximum wind speed, and the gust factor were plotted to assess wind speed data quality fo r both locations. The FAWN weather station did not show wind speeds that were consistently low; the averag e wind speed for the GCREC location was 2.4 m/s (Figure 3-14). The wind speed at the Gainesville location averaged 1.1 m/s (Figure 3-15), much lower than the GCREC location, but wind speeds varied daily and data collection was over a much shorter length of time than the GCREC locati on. The gust factor, ca lculated as a ratio of maximum to average wind speed, was not less than 1. 0 at either site (Figure 3-16; Figure 3-17). Daily average wind speed data was plotted fo r the NOAA weather statio n averaging 1.9 m/s (Figure 3-18). Standalone Controller Data collected from the three replicated Weathermatic controllers installed at the Gainesville turfgrass plots was used to calculate ETo using the Hargreaves equation and the onsite weather station data was used to calculate ETo using the ASCE standardized ETo equation and Hargreaves equation. The cumulative ETo calculated for WM-A, WM-B, and WM-C was 1023 mm, 1021 mm, and 995 mm, respect ively (Figure 3-19). Thes e controllers overestimated ETo by 26% to 29% compared to the ASCE met hod (791 mm) and by 18% to 22% compared to the Hargreaves method (842 mm). The daily ETo estimation by the three Weathermatic controllers, averaging 5.4 mm, were not found to be different from each other (P=0.5674). However, these controllers were different (P<0.0001) from the ASCE method and the Hargreaves method, averaging 4.2 mm a nd 4.5 mm, respectively (Table 3-2). The Weathermatic controllers calculated ETo by the Hargreaves method (Equation 3-16) and was based on two parameters: 1.) Ra determined from manufacturer programmed tables (Ra-

PAGE 95

95 WM) and 2.) daily minimum and maximum temperat ures from an on-site weather monitor (TWM). Both parameters had the poten tial to create variability in the ETo estimation. Comparisons made of ETo calculated from the Hargreaves Method (Equation 3-16) included: temperature reported by the Weathermatic controller and Ra estimated from a table (T-WM, RaWM); temperature reported by the Weathermatic controller and Ra determined from on-site weather station data (T-WM, Ra-WS); temperature determined from on-site weather station data and Ra estimated from a table (T-WS, Ra-WM); and temperature and Ra determined from the onsite weather station data (Har greaves Method). These comparisons showed which of the two parameters, temperature (Table 3-3) or Ra (Table 3-4), cause d variability among ETo estimation. It was determined that there was a difference between ETo calculated using temperatures from different sources (P<0.0001) whereas there was no difference between ETo calculated using Ra from a table vs. calculated (P = 0.6007). Temp erature estimation was the critical parameter influencing ETo estimation by the Hargreaves method. Cumulative ETo from the Weathermatic controller (T1) located at the GCREC was compared to ETo calculated by the ASCE method and Hargreaves method using the FAWN weather station data (Figure 3-20) using the co mparisons described above for the Gainesville location. Daily ETo estimations by the Weathermatic cont roller averaged 4.7 mm compared to the ASCE method averaging 4.4 mm and were f ound to be different (P<0.0001; Table 3-5). Cumulative ETo calculated using the ASCE method totaled 2095 mm. ETo calculated using FAWN data for daily maximum and minimum temperatures (T-FAW N) but different Ra methods were similar to each other, totaling 1986 mm and 2002 mm, an average 5% decrease compared to the ASCE method. Temperature collected by the on-site weather monitor (T-WM) estimated the highest cumulative ETo, 2239 mm to 2256 mm, overestimating by 7% on average compared

PAGE 96

96 to the ASCE method. As was in the Gainesville experiment, daily ETo calculated using Hargreaves method at the GCRE C location was influenced by temperature (P<0.0001; Table 36) and not alternative methods of obtaining Ra (P=0.5908; Table 3-7). Maximum and minimum temperatures collected by the three controllers at the Gainesville turfgrass plots were compared to the on-site weat her station (Figs. 3-21; 3-22). According to the figures, maximum temperatures varied from the on-site weather stati on temperatures while minimum temperatures did not. It was found that the daily maxi mum temperatures were higher than the weather station data (P<0.0001) wher eas the daily minimums were not different (P=0.7798). Also, the three co ntrollers resulted in daily Tmax and Tmin values that were not different from each other (Table 3-8). As was seen at the Gainesville location, temperature impacted the variability of ETo calculations using Hargreaves method. The da ily maximum temperature was typically higher when measured from the Weathermatic on-site weather monitor compared to the FAWN station (Figure 3-23). The maximum temperature meas ured by the Weathermatic controller was found to be different (P<0.0001) than the FAWN w eather data (Table 3-5). The daily minimum temperature data did not show the same variability as the da ily maximum temperature data (Figure 3-24). Minimum temperatures were not statistically different (P=0.1425) when comparing the FAWN weather data and the Weathermatic measurements (Table 3-5). It was shown by Trajkovic (2007) that the Hargreaves equation overpredicts ETo compared to the ASCE method under humid conditions. Hargreaves equation did cumulatively overestimate ETo compared to the ASCE method at the Gainesville turfgrass plots (Figure 3-19). However, this was not the case for the GCREC location. As was described in Chapter 4, there were only two short periods of rainfall in the la te summer and early fa ll months resulting in

PAGE 97

97 relatively dry years and unusual arid conditions in Florida. The ETo was possibly underestimated from the on-site weather station at th e Gainesville turfgrass plots due to lower wind speeds possibly caused from its urban location. Signal-based Controllers Cumulative ETo totaled 713 mm for the replicated Toro controlle rs at the Gainesville turfgrass plots (Figure 3-25). These controllers overestimated ETo by 15% compared to 618 mm from the ASCE method using the we ather data from the on-site weather station. However, these controllers estimated ETo at 662 mm or within 4% of the cumulative ETo calculated from the NOAA weather station, totaling 639 mm, which was a similar valu e (P=0.8890; Table 3-2). The average daily ETo estimation by the Toro controllers, 5. 2 mm, was different compared to the ASCE method averaging 4.2 mm (P<0.0001), but was not diffe rent compared to the NOAA weather station (P=0.9999). According to Hydropoint Data Systems, Inc., the chosen location for these controllers set by their company was different from the actual location, putting the controllers in a different microzone. The microzone error was corrected on November 28, 2007; hence the entire treatment period was affected by th is error. As a result, the ETo estimation by these controllers could have affected the data. However, Hydropoi nt Data Systems representatives indicated that the change in microzones would resu lt in a minor influence on the ETo data. The FAWN weather station located at the G CREC is on-site and the data are publically available. Cumulative ETo for the Toro controller at this location, totaling 2017 mm, was within 1% of the FAWN data using the ASCE method (Table 3-9). Daily ETo estimations averaged 4.3 mm for the Toro controller and 4.4 mm for the ASCE method and were not different (P=0.1555; Table 3-2).

PAGE 98

98 There were no variations between the three replications of the ET Water controllers at the Gainesville turfgrass plots, so the mean ETo of the three controllers was equal to the ETo used by all three controllers. ET Wate r provided a seven day summed ETo value via their website. ETo calculated from the on-site weathe r station data as well as the NOAA weather station was also summed into seven day totals for comparison purposes. The ET Water controllers at the Gainesvill e turfgrass plots resu lted in 733 mm of cumulative weekly ETo (Figure 3-26), over-estimati ng by 6% compared to the ETo calculated from the on-site weather station (693 mm). The average weekly ETo (Table 3-10) for the ET Water controllers was not different (P<0. 1470) compared to the average weekly ETo calculated from the on-site weather station data However, the average weekly ETo for the ET Water controllers was different compared to the NOAA weather station data (P<0.0001). The ET Water controller located at the GCREC underestimated ETo by 12%, calculating 1664 mm compared to 1900 mm of the ASCE method (Table 3-9). The ET Water controller at the G CREC under-estimated average daily ETo compared to the FAWN weather station (Table 3-2). When compared directly, th ere were differences between the estimations (P<0.0001). Overall Comparisons Since the Weathermatic and Toro replications located at the Gainesville turfgrass plots were not statistically different, the replicates of both brands were averaged and expressed as one value for comparison between type s of controllers. Both th e Weathermatic and the Toro overestimated ETo compared to the ASCE method by 24% and 16%, respectively. Also, the Weathermatic overestimated ETo by 7% compared to the Toro c ontroller (Figure 3-27). Also, the daily ETo (Table 3-2) calculated for the Weathermatic (5.4 mm) and Toro (5.2 mm)

PAGE 99

99 controllers were considered di fferent (P<0.0001) compared to the ASCE method (4.2 mm) as well as each other (P<0.0001). Daily ETo values for the Weathermatic and Toro controllers were summed into rolling seven day totals for comparison with the ET Wa ter controller as well as the ASCE method (Figure 3-28). When one day of data was not recorded, the rolling seven day totals for the Weathermatic and Toro controlle rs could not be calculated for seven days. This caused many gaps in the data; only approximately eight cumulative ETo data points could be calculated for all treatments. The Weathermatic controller consistently calculated higher estimated ETo. The Toro controller calculated similar cumulative weekly ETo compared to the NOAA weather station. Also, the ET Water controller cal culated cumulative weekly ETo similarly to the on-site data using the ASCE method. Direct comparisons of cumulative ETo were made between the three brands of controllers at the GCREC location when compared during ti me periods of available data from every controller (Figure 3-29). The Weathermatic controller, T1, overestimated cumulative ETo by 3% (1601 mm) compared to the estimation of 1548 mm using the AS CE method and FAWN weather data for this time period. ETo calculated by T2, the Toro controller, totaled 1545 mm; there was no cumulative difference in ETo when compared to the ASCE method over the study period. The ET Water controller T3, calculated 1334 mm of ETo, 14% less than the ASCE method. Daily ETo calculated by each treatment was found to be different from each other (Table 35). ET Water was different from all other trea tments (P<0.0001) and the Weathermatic was statistically different from the Toro (P =0.0001) and the ASCE method (P=0.0001).

PAGE 100

100 Summary and Conclusions Weather data collected from all three locati ons (the GCREC FAWN weather station, the NOAA weather station, and the Gainesville turf grass plots) were quality checked according to standardized recommendations. It was determined that solar radiation, temperature, relative humidity, and wind speed data met acceptable quality standards. The data was used to calculate ETo using either the ASCE method (Equation 31) or Hargreaves method (Equation 3-16). When analyzing daily ETo estimations using different temperature and Ra sources, it was determined that temperature was a driving fact or in the Hargreaves method whereas the method was not affected by the source of Ra. Statistical analysis of the temperatures collected from the Weathermatic compared to the local weathe r station (FAWN for GCREC and on-site for Gainesville turfgrass plots) showed that there are differe nces in maximum temperature measurements where the Weathermatic values were higher and not minimum temperature measurements. However, maximu m and minimum temperatures we re not significantly different among the replicated Weathermatic controllers. Thus the temperat ure sensor performance of the Weathermatic controller ha d a substantial impact on ETo estimation. Within the replications, controllers of all brands did not result in different daily ETo estimations. Also, the trends observed by the co ntrollers at both locations were similar. Therefore, the replications installed at the Gaines ville turfgrass plots for each brand of controller increased the validity of the results from the controllers located at the GCREC. The Toro controller at the GCREC estimated cumulative ETo approximately equal to the cumulative ETo calculated using the on-site FAWN weather sta tion data over a 16-month period. The ET Water controller, on the other hand, underestimated ETo by 12% compared to the same weather station and time period. This trend was also seen at th e Gainesville turfgrass plots. The NOAA weather station located at the Gainesvill e Regional Airport was the closes t public weather station to the

PAGE 101

101 Gainesville project site. The Toro contro ller was within 4% of the cumulative ETo calculated from the NOAA weather station data; whereas the ET Water controller under-estimated cumulative ETo by 8%. Data collection period for thes e controllers occurred over 7 months. Despite being able to accurately estimate ETo at the closest weather station, the Toro controller over-estimated ETo compared to the ETo calculated from the Gainesville turfgrass plots weather station by 15%. The proximity of th e weather station to the controller location is an important factor in determin ing the representativeness of ETo at the controller location. The ET Water controller resulted in similar ETo estimates as the Gainesville Airport, the closest publically available weather data to the research site. The Toro controllers at the Gain esville location were receiving ETo calculated for a location that was 1 km North and 1 km East away from the actual location of the controllers. It is highly probable that the publically avai lable weather data used to calculate ETo was still from the Gainesville Regional Airport weather stati on because the Airport is also northeast from the Gainesville turfgrass plots. When comparing ETo estimated by all brands, it was determined that they were all different from each other. The Weathermatic and ET Water controllers over-estimated and under-estimated ETo, respectively, compared to the ASCE method whereas the Toro controller was not statistically different from the ETo estimated from public weather stations at all locations. The trend in ETo when looking at daily and cumulative estimations was the same for the GCREC location (Table 3-7; 3-9). The Weathe rmatic controller consistently over-estimated ETo while the ET Water controller consistently under-estimated ETo. The Toro controller was not different in daily ETo estimations and was within 1% for cumulative comparisons.

PAGE 102

102 Table 3-1. ET controller code s and experimental information Code Controller Rep1 Location Installation Date ETo Equation Method to Obtain ETo WM-A Weathermatic 1 Gainesville May 2007 3-162 Standalone WM-B Weathermatic 2 Gaines ville May 2007 3-16 Standalone WM-C Weathermatic 3 Gaines ville May 2007 3-16 Standalone T-A Toro 1 Gainesville May 2007 3-13 Signal-based T-B Toro 2 Gainesville May 2007 3-1 Signal-based T-C Toro 3 Gainesville May 2007 3-1 Signal-based ETW-A ET Water 1 Gainesville May 2007 3-1 Signal-based ETW-B ET Water 2 Gainesville May 2007 3-1 Signal-based ETW-C ET Water 3 Gainesville May 2007 3-1 Signal-based T1 Weathermatic NA4 GCREC May 2006 3-16 Standalone T2 Toro NA GCREC Aug 2006 3-1 Signal-based T3 ET Water NA GCREC Aug 2006 3-1 Signal-based 1Rep refers to replication defined as multip le controllers at the same location. 2Equation 3-16 was used to calculate ETo using Hargreaves Equation (H argreaves and Samani 1982). 3Equation 3-1was used to calculate ETo using the ASCE standardiz ed equation (Allen et al. 2005). 4NA indicates that there are no replicat ions of controllers at that location. Table 3-2. Daily ETo between treatments at the Gainesville turfgrass plots Treatment n ETo (mm) Weathermatic 558 5.4 a Toro 414 5.2 b NOAA ASCE Method 162 5.2 b Hargreaves Method 192 4.5 c On-site ASCE Method 192 4.2 d Numbers with different letters in columns i ndicate differences at th e 95% confidence level using Duncans Multiple Range Test. **Time wa s used as a replication in the statistical analysis. Table 3-3. Dependency on temperature using mean daily ETo values for the Weathermatic controller replications at the Gainesville turfgrass plots Treatment Data Origination ETo (mm) WM-A Weather Monitor 5.44 a WM-B Weather Monitor 5.43 a WM-C Weather Monitor 5.30 a Hargreaves Method Weather Station 4.53 b *Numbers with different letters in columns i ndicate differences at th e 95% confidence level using Duncans Multiple Range Test. **Time wa s used as a replication in the statistical analysis.

PAGE 103

103 Table 3-4. Dependency on Ra using mean daily ETo values for the Weathermatic controller replications at the Gain esville turfgrass plots Treatment Data Origination ETo (mm) WM-A Weather Monitor 5.01 a WM-B Weather Monitor 5.01 a WM-C Weather Monitor 4.94 a Hargreaves Method Weather Station 4.96 a *Numbers with different letters in columns i ndicate differences at th e 95% confidence level using Duncans Multiple Range Test. **Time wa s used as a replication in the statistical analysis. Table 3-5. Average daily ETo, maximum temperature, and minimum temperature between treatments at the GCREC location Treatment N ETo Tmax Tmin (mm) (C) (C) T1 487 4.7 a 29.7 a 16.2 a T2 467 4.3 b NA NA T3 439 3.7 c NA NA ASCE Method 552 4.4 b 28.1 b 17.0 a NA indicates that data collection of this para meter was not applicable for the controller. ** Numbers with different letters in columns indi cate differences at the 95% confidence level using Duncans Multiple Range Test. ***Time wa s used as a replication in the statistical analysis. Table 3-6. Dependency on temperature using mean daily ETo values for the Weathermatic controller replications at the GCREC Treatment Data Origination ETo (mm) T1 Weather Monitor 4.64 a Hargreaves Method Weather Station 4.12 b *Numbers with different letters in columns i ndicate differences at th e 95% confidence level using Duncans Multiple Range Test. **Time wa s used as a replication in the statistical analysis.

PAGE 104

104 Table 3-7. Dependency on Ra using mean daily ETo values for the Weathermatic controller replications at the GCREC Treatment Data Origination ETo (mm) T1 Weather Monitor 4.40 a Hargreaves Method Weather Station 4.36 a *Numbers with different letters in columns i ndicate differences at th e 95% confidence level using Duncans Multiple Range Test. **Time wa s used as a replication in the statistical analysis. Table 3-8. Minimum and maximum temperatures between the Weathermatic controllers and the on-site weather station at th e Gainesville turfgrass plots Treatment n Tmax Tmin (C) (C) WM-A 186 33.7 a 19.5 a WM-B 187 33.8 a 19.9 a WM-C 187 33.2 a 19.7 a Measured Temperature 192 30.2 b 19.6 a Numbers with different letters in columns i ndicate differences at th e 95% confidence level using Duncans Multiple Range Test. **Time wa s used as a replication in the statistical analysis. Table 3-9. Totals and percentage di fferences of average cumulative ETo between treatments at the GCREC location and the measured ETo from the FAWN weather station Treatment TMT ETo ASCE Method Difference (mm) (mm) (%) T1 2256 2095 8 T2 2017 1990 1 T3 1664 1900 -12 ASCE Method totals vary due to start and end da tes of data collection as well as random days of unavailable data within each dataset.

PAGE 105

105 Table 3-10. Weekly ETo between the ET Water controllers and local weather stations for the Gainesville turfgrass plots location Treatment N ETo (mm) ET Water 129 30.5 a On-site Measured 192 29.8 a NOAA Measured 162 36.5 b *Numbers with different letters in columns i ndicate differences at th e 95% confidence level using Duncans Multiple Range Test. **Time wa s used as a replication in the statistical analysis. Figure 3-1. The ET controllers installed on the Univ ersity of Florida Gainesville Campus. They are A) Toro A, B) Toro B, C) Toro C, D) Weathermatic A, E) Weathermatic B, F) Weathermatic C, G) ET Water A, H) ET Water B, and I) ET Water C.

PAGE 106

106 Figure 3-2. Weathermatic SLW10 weather monito rs were installed 1.83 m above ground level and 0.81 m apart for A.) Weathermatic A, B.) Weathermatic B, and C.) Weathermatic C.

PAGE 107

107 Figure 3-3. The FAWN meas ured solar radiation (Rs) and clear-sky solar radiation (Rso) for 2006 and 2007 using the weather station in Balm, FL. Figure 3-4. Data from the NOAA weat her station in Gainesville, FL was used to calculate solar radiation ( min maxarssTTRKR ) and clear-sky solar radiation (Rso) from May through October 2007.

PAGE 108

108 Figure 3-5. Measured solar radia tion and clear-sky solar radiati on from May 22 to November 30, 2007 for the on-site weather st ation in Gainesville, FL. Figure 3-6. The FAWN daily maximum and minimum relative humidity for 2006 and 2007 using the weather station located in Balm, FL.

PAGE 109

109 Figure 3-7. The data from the NOAA weather station located in Gainesville, FL was used for daily maximum and minimum relative humidity from May through October 2007. Figure 3-8. Daily maximum and minimum relative humidity for the on-site weather station located in Gainesville, FL from May 22 through November 30, 2007.

PAGE 110

110 Figure 3-9. The FAWN daily mi nimum temperature and calculat ed dew point temperature for 2006 and 2007 using the weather station located in Balm, FL. Figure 3-10. The NOAA weather station located in Gainesville, FL was used to obtain daily minimum temperature and calculated de wpoint temperature from May through October 2007.

PAGE 111

111 Figure 3-11. Daily minimum temperature and calcu lated dewpoint temperature for the on-site weather station located in Gainesville, FL from May 22 through November 30, 2007. Figure 3-12. The FAWN daily mean temperatures calculated using 24 hours of temperature data plotted against the average of the maximum and minimum temperatures of that day for 2006 and 2007 using the weather station located at Balm, FL.

PAGE 112

112 Figure 3-13. Daily mean temperatures calculate d using 24 hours of temp erature data plotted against the average of the maximum and minimum temperatures of that day for the on-site weather station located in Gain esville, FL from May 22 through November 30, 2007. Figure 3-14. The FAWN daily maximum and average wind speed (at 2 m) for 2006 and 2007 using the weather station located at Balm, FL.

PAGE 113

113 Figure 3-15. Daily maximum and average wind speed (at 2 m) for the on-site weather station located in Gainesville, FL from May 22 through November 30, 2007. 0 1 2 3 4 5 6 7 8 1/12/264/236/188/1310/812/31/283/255/207/159/911/4Gust FactorDate (2006 2007) Maximum Wind Speed/Average Wind Speed Figure 3-16. The FAWN daily gus t factor calculated as the ma ximum wind speed divided by the average wind speed for 2006 using the w eather station located at Balm, FL.

PAGE 114

114 0 1 2 3 4 5 6 7 8 5/226/127/37/248/149/49/2510/1611/611/27Gust FactorDate (2007) Maximum Wind Speed/Average Wind Speed Figure 3-17. Daily gust factor calculated as the maximum wind speed divided by the average wind speed for the on-site weather station located in Gainesville, FL from May 22 through November 30, 2007. Figure 3-18. The NOAA weather station located in Gainesville, FL was used to obtain daily average wind speed (at 2 m) from May through October 2007.

PAGE 115

115 Figure 3-19. Cumulative ETo for three replications of Weathermatic controllers compared to Hargreaves equation and the ASCE standard ized equation using data collected from an on-site weather station in Gainesville, FL. Figure 3-20. Cumulative ETo calculated from weather data collected by the Weathermatic controller or FAWN weather station. T-WM and R-WM represent values used by the Weathermatic controller, and T-FAWN and Ra-FAWN represent values measured by the FAWN weather station. All ETo calculations were performed using Hargreaves equation except for ASCE Method using FAWN.

PAGE 116

116 Figure 3-21. Daily maximum temperature comp arisons between the three Weathermatic controllers and an on-site weathe r station in Gainesville, FL. Figure 3-22. Daily minimum temperature comp arisons between the three Weathermatic controllers and an on-site weathe r station in Gainesville, FL.

PAGE 117

117 Figure 3-23. Daily maximum temperature comparis on from the Weathermatic controller and the FAWN on-site weather station. Figure 3-24. Daily minimum temperature comparis on from the Weathermatic controller and the FAWN on-site weather station.

PAGE 118

118 Figure 3-25. Cumulative ETo for three replications of Toro controllers compared to the ASCE standard using data collected from an on-site weather station a nd the ASCE standard using the NOAA weather station data in Gainesville, FL. Figure 3-26. Cumulative ETo for the ET Water controller, cal culated using the ASCE method from on-site weather stati on data, and calculated using the ASCE method from the NOAA weather station data in Gainesville, FL.

PAGE 119

119 Figure 3-27. Average cumulative daily ETo for the Weathermatic and Toro controllers compared to the ASCE standard using data collect ed from an on-site weather station in Gainesville, FL. Figure 3-28. Seven day total ETo shown cumulatively for all three brands of controllers (Weathermatic, Toro, and ET Water) compared to the ASCE method using data from the on-site weather station in Gainesville.

PAGE 120

120 Figure 3-29. Cumulative ETo for the Weathermatic, Toro, and ET Water controllers at GCREC and calculated using FAWN data and the ASCE method.

PAGE 121

121 CHAPTER 4 IRRIGATION SCHEDULING BY E VAPOTRANSPIRATI ON-BASED IRRIGATION CONTROLLERS Introduction Natural climatic cycles ensure periods of critical drought and improvement in water conservation is necessary since drought is not the criteria for effici ent water use (Florida Department of Environmental Protection [FDEP] 2002). Florida continues to grow rapidly and traditional sources of water are limited. Florida has the largest net gain in population with an inflow of approximately 1,108 people per day an d fourth in overall popul ation (United States Census Bureau [USCB] 2005). New home construction has increased to accommodate such a large influx of people and most new homes include in-ground automated irrigation systems. However, homes with in-ground systems utilizing automated irrigation timers increase outdoor water use by 47% (Mayer et al. 1999). Proper irri gation management could result in as much as a two-fold reduction in water us age (FDEP 2002). Also, improper irrigation, whether it is underirrigation or over-irrigation, can negatively impact landscapes as well as waste water resources (Burt et al. 1997). Irrigation scheduling can be done using quant itative or qualitative methods. The method commonly used by homeowners involves observi ng the lawn and irrigating when it looks stressed (Wade and Waltz 2004). However, research has shown that single families in Florida over-irrigate their landscapes due to the misunderstanding of seasonal water needs or the inconvenience of updating the irriga tion time clock to reflect th e actual water needs of the landscape (Haley et al. 2007). A lternatively, the quant itative method measures plant needs from soil moisture levels using instruments such as tensiometers or dielectric probes or evapotranspiration loss (Wade and Waltz 2004).

PAGE 122

122 Evapotranspiration (ET) is defined as the evaporation from a soil surface and the transpiration from plant material (Allen et al. 1998). ET is part of a balanced energy budget that exchanges energy for outgoing water at the surface of the plant. The components of ET are solar radiation, temperature, relati ve humidity, and wind speed (Allen et al. 2005). Reference ET (ETo) is defined as the ET from a hypothetical re ference crop with the characteristics of an actively growing, well-watered, dense green cool season grass of uniform height (Allen et al. 2005). Typically, climatic data is used as inputs to equations to estimate ETo. There are three types of equations: mass transfer energy balance, and empirical methods (Fangmeier et al. 2006). Most of the current methods employ a co mbination of the three. The appropriate ET equation is chosen depending on many factors including geographical location, types of crops, and weather data availability (Fangmeier et al. 2006). Evapotranspiration-based c ontrollers, also known as ET controllers, are irrigation controllers that use ETo to schedule irrigation. Dependi ng on manufacturer, ET controllers can be programmed with various conditions specific to the landscape making them more efficient (Riley 2005). ET controllers receive ETo information in three general ways, consequently dividing ET controllers into three main types: 1) standalone controll ers, 2) signal-based controllers, and 3) histori cal-based controllers. Standalone controllers typically receive climatic data from onsite sensors and calculations to determine ETo are performed by the controller. Ev en though the controllers might take readings every second or every fi fteen minutes, cumulative daily ETo is used for irrigation scheduling. On-site sensors coul d include: temperature, solar radi ation, an ET gauge, or even a full weather station (Riley 2005). Benefits of sta ndalone controllers are th at they are not limited

PAGE 123

123 by requiring the use of a full w eather station and there are no signal fees associated with broadcasts from the manu facturer (Riley 2005). Signal-based contro llers receive ETo information from a company that collects climatic data from weather stations locate d near the irrigation si te using satellite or internet technology. Depending on the manufacturer, the ETo data can be from an average of multiple weather stations in the area or from a single weather st ation. There is typical ly a signal fee (i.e., subscription) for this controller set by the ma nufacturer that normally ranges from $4 to $15 per month (Riley 2005). Historical-based controll ers rely on historical ETo information for the area. Typically, monthly historical ETo is programmed into the controller by the manufacturer or installing contractor. Theoretically, this method does not result in as accurate an ETo estimate because site specific weather variability is not considered. Bench testing or virtual studies have been conducted where results were determined from whether the controllers would have accura tely irrigated based on scheduling and ETo estimation. The Metropolitan Water District of Southern California conducte d a year-long bench test in 2002 designed to compare the ability of ET controllers to determine theoretical water needs for three types of landscapes: cool season turf on loam with full sun, shaded annuals on sandy soils, and low water using ground cover on a sunny, 20 degree slope. The WeatherTRAK enabled controller always applied less water than the maximum allowable water allowance resulting in no overwatering. This controller performed the water balance sufficiently so that water received equals water required except for the summer mont hs where the controller showed a deficit in irrigation. Percent soil mois ture depletion for all scenario s except for the sloped one, where over-irrigation occurred, fell within a 30%-70% ta rget range and minimized runoff (Metropolitan

PAGE 124

124 Water District of Southern California [MWD SC] 2004). A virtual study was conducted in 2003, also using a WeatherTRAK enabled controller, designed to determine the data used by the controllers, ease of setup and operation, and how accurate they were at matching irrigation needs to five types of landscapes cons isting of turfgrass, trees/shrubs, annuals, mixed high water use plants, and mixed low water use plants. Irriga tion equaled the turfgrass requirements in April and October only; over-irrigation wa s 21-40% in March, June, and July, over 40% in November, and 11-20% for the rest of the year. It was conc luded that poor results we re due to very general controller settings includ ing using default uniformity and pr ecipitation rates (Pittenger et al. 2004). Smart Water Application Technologies (SWAT) is a subset of the Irrigation Association that developed a protocol for determining th e effectiveness of irrigation scheduling by ET controllers. The protocol was designed to m easure the ability of ET controllers to schedule irrigation that is adequate and efficient while minimizing run-off. Adequacy is a measure of under-irrigation and scheduling efficiency is a m easure of over-irrigation determined from a soil water balance model. Testing must meet the requ irements of 30 consecutive days of testing with 10.2 mm of total rain fall and 63.5 mm of ETo (Irrigation Association [IA] 2006c). The objective is to determine the capability of three brands of ET-based irrigation controllers to schedule irri gation compared to a theoretically derived soil water balance model. Materials and Methods This study was conducted at the University of Florida Gulf Coast Research and Education Center (GCREC) in Wimauma, Florida (see Chapter, Figure 2-1). There were a total of twenty plots that measured 7.62 m x 12.2 m, bordered by a 15.2 cm tall black metal barrier, with 3.05 m buffer zones between adjacent plots (see Chapter, Figure 2-3). The buffer zones were covered with a white material that act ed as a weed barrier. Each plot consisted of 65% St.

PAGE 125

125 Augustinegrass (Stenotaphrum secundatum Floratam) and 35% mixed ornamentals to represent a typical residential landscape in Florida. The ornamentals were as follows: Crape Myrtle (Lagerstroemia indica Natchez) (see Chapter, Figure 2-4A), Gold Mound Lantana (Lantana camara Gold Mound) (see Chapter, Figur e 2-4B), Indian Hawthorne (Raphiolepis indica) (see Chapter, Fig 2-4C), Cape Plumbago (Plumbago auriculata) (see Chapter, Figure 2-4D), and Big Blue Liriope (Liriope muscari Big Blue) (see Chapter, Figure 2-4E). Landscapes were maintained through mowing, pruning, edging, mulching, fertilization, and pest and weed control according to current UF-IFAS recommendations (Black and Ruppert 1998; Sartain 1991). Five treatments were established, T1 through T5, replicated four times for a total of twenty plots in a completely randomized block design. The irrigation tr eatments are as follows: T1, SL1600 controller with SLW15 weather monitor (W eathermatic, Inc., Dallas, TX); T2, Intellisense (Toro Company, Inc., Riverside, CA) utilizing the Weathe rTRAK ET Everywhere service (Hydropoint Datasystems, Inc., Petaluma, CA); T3, Smart Controller 100 (ET Water Systems LCC, Corte Madera, CA); T4, a time-based treatment determined by UF-IFAS recommendations (Dukes and Haman 2002a); and T5, a time-based tr eatment that is 60% of T4 (see Chapter, Figure 2-5). All treatments utili zed Mini-clik rain sensors (Hunter Industries, Inc., San Marcos, CA) set at a 6 mm threshold. A metal shed housed the controllers on-site a nd a manifold table supported forty solenoid valve and flow meter combinations to supply an d monitor irrigation to each zone of each plot (see Chapter, Fig 2-6). The flow meters (11.4 cm V100 w/ Pulse Output, AMCO Water Metering Systems, Ocala, FL) us ed to monitor irrigation water application were connected to five SDM-SW8A switch closure input modules (C ampbell Scientific, Logan, UT) that in turn connected to a CR-10X data logger (Campbell Scientific, Logan, UT) (see Chapter, Fig 2-7).

PAGE 126

126 The CR-10X data logger monitored switch closures every 18.9 liters from the water meters. The data was also collected manually on a weekly ba sis at minimum. Each plot contained an irrigation zone for turfgrass and mixed ornamentals. Irrigation sprinklers specified for the turfgrass portions of the plots consisted of Rain Bird (Glendora, CA) 1806 15 cm pop up spray bodies and Ra in Bird R13-18 black rotary nozzles (see Chapter, Figure 2-8). In each plot, there were four sprinklers with a 180 degree arc (R13-18H) and a center sprinkler with a 360 degree arc (R 13-18F). Microsprays (Maxijet, Dundee, FL) were installed to irrigate the mi xed ornamental plants. A pressure regulator was installed at the plot to maintain a constant pressure of 6 kPa on the microsprays during irrigation. It was determined that the application rate specified by the manufacturer for the turfgrass plots of 15.5 mm/hr was not the actual rate durin g the entire treatment period. The application rate was determined to be 20.3 mm/hr on average. As a result, all treatments over-applied irrigation due to calculating a larg er runtime for the same theoretical depth to be applied. The measured water application for each treatment was corrected fo r this error by calculating the runtime specified for each treatment determined from the measured water application totals and then calculating a corrected depth us ing the new average application rate. The type of soil located at the project site was mapped as Zolfo fine sand (NRCS 1989). According to the soil survey, the Zolfo series is a somewhat poorly drained soil composed of sandy, siliceous, hyperthermic Grossarenic En tic Haplohumods. The field capacity and permanent wilting point for Zolfo fine sand was determined from laboratory samples to be 13% and 3% (all soil moisture values are presented on a volumetric basi s), respectively (Carlisle et al. 1985). Time domain reflectometry (TDR) probes (Campbell Scientific, Inc., Logan, UT) were buried in turfgrass and mixed-ornamental areas of each plot to monitor soil moisture in the root

PAGE 127

127 zone represented as 10 cm to 18 cm for St. A ugustinegrass and the upper r oot zone of the mixed ornamentals. According to the manufacturers (Hydropoint Data Systems, Inc. 2003; ET Water 2005), ETo was calculated by the Toro and ET Water c ontrollers using the ASCE standardized ETo equation (Allen et al. 2005). The Weathermatic controller utili zed the Hargreaves equation to estimate ETo (Hargreaves and Samani 1982). The ASCE standardized ETo equation was used in the soil water balance mode l for comparison purposes. Plant-specific ET can be calculated for a plan t material by applying a crop coefficient (Kc), using the following equation: occET*KET 4-1 Kc values chosen for the theoretical soil water balance model were developed from warm season turfgrass in central Florida (Jia et al. 2007) and were adjusted duri ng the winter months to reflect common weather conditions for sout hwest Florida (Table 4-1). The Weathermatic controller used a fixed Kc value of 0.60 for each month. Kc values for the Toro and ET Water controllers were considered proprietary informa tion and were not made available. A daily soil water balance model was used to calculate the theoretical irrigation requirements for comparison with actual irrigation water applied. The balance is defined as: 0RODETIPeSC 4-2 where S (mm) is the change in soil water storage within the root zone, Pe (mm) is the effective rainfall, I (mm) is the irrigation depth, ETc (mm) is the crop evapotranspiration, D (mm) is drainage, and RO (mm) is surface runoff (Fangmeier et al. 2006). Due to the flat topography and relatively high permeability on site (NRCS 1989), it was assumed that there is negligible surface

PAGE 128

128 runoff and irrigation is scheduled so that, ideally, there is negligible drainage. These assumptions reduce equation 4-2 to the equation us ed to calculate the ir rigation depth required: PeETIc 4-3 Effective rainfall is the amount of rainfall that is stored in the root z one. The remainder of rainfall is considered runoff or drainage below the root zone Rainfall causing the soil water content to exceed field capacity in the soil wate r balance model was assumed to be lost due to runoff or drainage. The amount of water able to be held by the root zone and is available to the plant is called available water, AW (mm). Available water is calculated from soil parameters using the equation: 100 RZ*)PWPFC( AW 4-4 where FC (cm3 of water/cm3 of soil) is the field capacity, PWP (cm3 of water/cm3 of soil) is the permanent wilting point, and RZ (mm) is the ro ot zone depth (Irrigati on Association 2005). To prevent plant stress, available water should not be allowed to reach the PWP before irrigation is scheduled; irrigation should be applied when the water level drops by a percentage known as the maximum allowable depletion (MAD), chosen as 50% for warm season turfgrass (Allen et al. 1998). The amount of water allowed to be used be fore irrigation is requ ired is called readily available water, RAW (mm), and is calcula ted using the following equation (IA 2005): MAD*AW RAW 4-5 The net irrigation depth to be applied is determined from the change in soil water level occurring due to ETC loss and effective rainfall. However, the theoretical gross irrigation depth is necessary to compare to the amount of water applied by the treatments. The gross irrigation depth is calculated from an efficiency factor ultimately determined from the low quarter

PAGE 129

129 distribution uniformity (DUlq) of the system (IA 2005). The DUlq was determined to be 0.70 from on-site catch-can testing. The low half distribution uniformity (DUlh) was calculated using DUlq in percentage form as follows: lq lhDU*614.06.38DU 4-6 which, in turn, is used to calculate the efficiency factor (E) using the equation: lhDU 100 E 4-7 The gross irrigation is calculated by multiplying th e net irrigation depth by the efficiency factor, determined from Equation 4-6 and Equation 4-7 to be 1.25 (IA 2005). Scheduling efficiency was defined as the ab ility of a controller to schedule irrigation without applying excess irrigation th at results in drainage or runo ff (IA 2006c). It was calculated in 30-day running totals with the following equation: 100 I Surplus I Enet net 4-8 where Inet (mm) refers to the sum of net irrigation applied over the 30 days and Surplus (mm) refers to the summed depth of wa ter above the field capacity. Irrigation adequacy, on the other hand, quantifie d the ability of the controller to supply sufficient irrigation so that it met the demand for water (IA 2006c). It was also calculated in 30day running totals using the following equation: 100 ET Deficit ET AC C 4-9 where Deficit represents the sum of the depth of water below the maximum allowable depletion over the 30-day period in mm.

PAGE 130

130 Ranges of irrigation adequacy and scheduling e fficiency were large when considered over a long period of time. As a result, perfor mance ranges were designated to determine the frequency of certain scores. Designations were as follows: 90% to 100%, 70% to 90%, 50% to 70%, and below 50%. There were six seasons of data collec tion: May 25, 2006 through August 12, 2006 as summer 2006; August 13, 2006 through November 30, 2006 as fall 2006; December 1, 2006 through February 26, 2007 as winter 2006-2007; February 27, 2007 through May 31, 2007 as spring 2007; June 1, 2007 through August 31, 2007 as summer 2007; and September 1, 2007 through November 30, 2007 as fall 2007. Applied irrigation depths were compared to depths calculated using a daily soil water balance with inputs similar to user-defined inputs programmed into the controllers for the first three seasons (see Chapter, Table 2-2) a nd the last three seasons (see Chapter, Table 2-3). All five treatments observed 2 d/wk watering restrictions during summer 2006, fall 2006, and winter 2006-2007, Wednesday and Saturday, and no watering between 10 am and 4 pm. Also, the ET controller treatments were establis hed based on the site location without accounting for system efficiency. T1, the Weathermatic cont roller, was set to apply 100% of the calculated water requirement while T2 and T3, the Toro and ET Water controllers, were set to the maximum efficiency of 95%. The monthly irriga tion depth for T4, the time-based treatment, was 60% of the net irrigation requirement deri ved from historical ET and effective rainfall specific to south Florida (Duke s and Haman 2002a) and T5 was a reduced treatment, applying 60% of the irrigation depth calculated from T4 (see Chapter 2, Table 2-1). Spring, summer, and fall 2007 differed from the previous three seasons in that the ET controller treatments allowed irrigation windows se ven days per week and were updated with a

PAGE 131

131 system efficiency of 80% determined from unifo rmity testing discussed in Chapter 5. The timebased treatment, T4, was increased to apply i rrigation to replace 100% of the net irrigation requirement instead of 60% used during the first three seasons. Once again, T5 applied 60% of T4 resulting in the reduced treatment applying 60% of the net irrigation requirement. Data collection included: climate data at fift een minute intervals such as wind speed, solar radiation, temperature, relati ve humidity, and rainfall depth from a Florida Automated Weather Network (FAWN) weather station located onsite and soil moisture content from TDR probes. Results Thirty year historical rainfall averages were calculated from monthly rainfall data collected by the National Oceanic and Atmospheric Administration (NOAA 2005) from 1975 through 2005 approximately 28 km away, in Parrish, FL. All months received less rain than historical average except for the following three months: July 2006, 97% higher than average; April 2007, 53% higher than average; October 2007 104% higher than average (Figure 4-1). Overall, both years were drier than the hist orical average with a total of 1,971 mm of rainfall for the approximate 19-month study period, May 2006 th rough November 2007, compared to 2,458 mm for the same historical period. High intensity and large rainfall events can lead to runoff or drainage below the root zone. The portion of rainfall stored in the root zone is considered effective in that this precipitation can contribute to plant water needs. The cumulative depth of effective rainfall from the daily soil water balance, 514 mm, was 73% less than the total cumulative rainfall, 1,934 mm, over the treatment period (Figure 4-2). Only a fraction of each event was able to be stored in the root zone on a regular basis due to the limited turfgr ass root zone of 15 cm and the low soil water holding capacity of 10% by volume. There were two distinct wet periods occurring approximately from June through September for 2006 and June through October for 2007. The

PAGE 132

132 wetter period in 2007 lasted a month longer than in 2006, but had at leas t three distinct dry periods within the wet period resulting in similar cumulative rainfall depths for both years, 204 mm for 2006 and 228 mm for 2007. In the soil water balance model, measured rainfa ll depth was used as an input and effective rainfall was calculated based on the depth required to fill the root zone to field capacity. However, ET controllers use a variety of methods to handle rainfall depending on the manufacturer. The Weathermatic controller incorporated rainfa ll by using an expanding disk rain sensor. This controller bypassed irrigation for 48 hours when the rain sensor sensed rainfall based on a 6 mm set threshold; whereas, the rain sensors on the remaining controllers have been shown to dry out between 68% to 85% of the time 24-30 hours af ter rainfall (Cardena s-Lailhacar and Dukes 2008). This controller maintained an accumulating deficit total based on ETC losses and irrigates to refill the deficit total regardless of soil wate r holding capacity. The cont roller was designed to operate in areas with mandatory watering restrict ions. When enough rain fell to trigger the rain sensor (6 mm setting), the deficit was reduced by 25.4 mm/hr until reset to zero (Weathermatic 2005). The Toro controller was connected to a rain se nsor, similar to all of the treatments, but the controller treats the rain sensor bypass mode as a non-wateri ng day. When the rain sensor bypasses irrigation, the controller keeps track of the number of days and then applies irrigation as if rain never occurred. This results in more irrigation applied than required since irrigation supplements rainfall. This controller, however, is sent a rainfall sign al by the WeatherTRAK ET Everywhere system to incorporate rainfall into th e determination of irrigation applied. This is done by setting certain rainfall depths to the number of days the controller should wait until

PAGE 133

133 irrigation should be resumed. Ra infall depths are collected from the public weather station that the manufacturer uses to calculate ETo and may or may not reflect the rainfall amount at the controller location. During the entire treatment period, the ET Wa ter controller did not recognize the rain sensor and did not bypass irriga tion according to localized rainfall events. Rainfall was taken into account with the soil water balance when sc heduling irrigation, but th e weather station used was over 10 miles away from the project site (ET Water Website 2006). Rainfall can vary substantially over short distances in Florida. The Weathermatic and ET Water controllers began treatments on May 25, 2006; however, hardware issues with the ET Water controller arose late in the summer season causing the controller to be nonfunctional. As a result, the ET Water contro ller did not control irrigation during the fall 2006 and winter 2006-2007 seasons. Once the controlle r was repaired, the programmed settings were updated to reflect se ttings described for spring 2007. However, maximum allowable depletion was set to 25% in stead of 50% for unknown reasons and remained that way for the spring, summer, and fall 2007 seas ons. The Toro controller was not installed until August 13, 2006. The time-based treatments applied irrigation twice per week for every season unless bypassing occurred due to rainfa ll or time-clock malfunction. The irrigation schedule was developed from the net irrigation requirement de termined from historical ET and effective rainfall and was adjusted monthly. T4 was set for 60% replacement for summer 2006, fall 2006, and winter 2006-2007 and 100% replacement for spring, summer, and fall 2007. The irrigation schedule for T5, the reduced time-based treatm ent, was set for 36% replacement for summer 2006, fall 2006, and winter 2006-2007 and 60% re placement for spring, summer, and fall 2007.

PAGE 134

134 As a result, T4 functioned simila rly to a historical ET controller with a crop coefficent the same as the Weathermatic controller for the first th ree seasons while T5 did so for the last three seasons. The first three seasons for T5 could po ssibly be considered a de ficit treatment due to 36% being approximately half of any of the average crop coefficients. Summer 2006 The summer 2006 season varies from other seas ons because it was during the setup of the treatments resulting in different settings thr oughout the season (Table 4-2). The soil water balance model used to determine the theoretical irrigation requirement incorporates these differences. Summer was relatively wet with rainfall occuri ng every few days. Irrigation application for the Weathermatic controller was 159 mm resulting in 24% under-irrigation compared 208 mm calculated as the theoretical irrigation re quirement (Figure 4-3). The Weathermatic scheduled larger irrigation dept h per application, averaging 23 % greater than the theoretical requirement, but had fewer irriga tion events in the season. Ra infall was frequent enough during this season to limit the number of events to 13 out of 23 possible. The Weathermatic controller cal culated irrigation adequacy averaging 59% (Table 4-3). Adequacy scores were higher when rainfall filled the root zone to field capacity and irrigation did not occur. This was because rainfall occurred often during this season so that deficits could not accumulate from lack of water. However when irrigation was necessary, the controller scheduled irrigation on the next av ailable watering day causing a deficit to develop before the watering could occur. This controller averaged -57% in scheduling efficiency (Table 4-3). Scheduling efficiency was lower when the depth applied per irrigation event exceeded field capacity according to the soil water balance mode l or irrigation occurred on a morning where

PAGE 135

135 rainfall occurred later the same day. The 13 irrigation events applied more water than the root zone could theoretically hold due to the combin ation of no soil water ho lding capacity concepts of the controller and watering restrictions causing very low scheduling efficiency scores. The ET Water controller applied 203 mm of irrigation compared to 186 mm of the theoretical irrigation requirement, only over-irri gating by 9% on a cumulative basis (Figure 4-4). Depth applied per event averaged 3% greater than the theoretical requirement and had one additional irrigation event. Irrigation adequacy scores av eraged 50%, while scheduling efficiency averaged 82% (Table 4-4). Adequ acy suffered from following watering restrictions; deficits accumulated between watering days when ra infall did not occur. Scheduling efficiency scores were less than 100% due to slightly larg er irrigation depths per event resulting in overirrigation. The T4 irrigation schedule resulted in 13% over-irrigation, totaling 236 mm compared to 208 mm by the theoretical irrigation requirement (Figure 4-5) whereas T5 under-irrigated by 27%, totaling 152 mm (Figure 4-6). The rain sens or failed during the last week of June causing irrigation to occur when not necessary throughout th e rest of the rainy se ason. Under-irrigation would have occurred if the rain sensor would have functioned appropriately. Compared to the theoretical amount, irrigation ap plied per event was 19% greater by T4 and 23% less by T5. Irrigation was applied for the same number of ev ents by T4, T5, and the theoretical requirement due to a combination of the faulty rain sensor for T4 and T5 and no watering restrictions for the theoretical requirement. The time-based treatment, T4, averaged 64% in scheduling efficiency and 65% in irrigation adequacy (Table 4-5). Initially, this treatment bypassed irrigation events due to rainfall causing some deficit conditions and lower irrigati on adequacy due to watering restrictions. Once

PAGE 136

136 the rain sensor ceased to function, irrigation and rainfall kept the water level above the maximum allowable depletion level. However, this co mbination caused the scheduling efficiency to decrease over time. The sc heduling efficiency percentage never reached 100% like the ET controllers sometimes did. This effect was due to consistently applying more irrigation than required on watering days. The reduced time-based treatment, T5, perfor med similarly to T4; scheduling efficiency and irrigation adequacy averaged 78% and 54% respectively (Table 4-6). The irrigation adequacy increased when the rain sensor malf unctioned and the schedu ling efficiency trend slightly decreased. Scheduling efficiency was higher than T4 for this treatment because less water was applied per event; however, scheduling efficiency suffered because irrigation depth applied was sometimes greater than required to fill the root zone to field capacity. Cumulative season irrigation application was the lowest by T5, the reduced time-based treatment. The soil moisture content for this tr eatment, however, rarely fell below field capacity and never fell below the maximum allowable deplet ion (Figure 4-7). Thus, this treatment was well-watered for the summer 2006 season. Since th e other treatments applied more irrigation than T5, it can be assumed that all treatme nts remained well-watered for the summer 2006 season. Fall 2006 September of the fall 2006 season experienced mo st of the rainfall wher eas the rest of the season only had 8 rainfall events. The Weatherm atic controller under-i rrigated by 13%, applying 150 mm compared to the theoretical requireme nt of 172 mm (Figure 48). This treatment averaged 13% less than the theoretical irrigation depth for each event.

PAGE 137

137 Scheduling efficiency and irrigation adequacy averaged -62% and 85%, respectively, for the Weathermatic controller (Table 4-3). Sche duling efficiency was low during the rainy portion of the season because the controller only bypassed some irrigation events even though rainfall filled the root zone to field capacity. It was shown by Cardenas-Lailhacar and Dukes (2008) that the sensivity in rain sensor bypassing was variab le, especially for sensors set for less than 13 mm. Scheduling efficiency increased once ir rigation became the primary water supply and rainfall was infrequent. Irrigat ion adequacy was generally good with scores decreasing in the middle of the season. This was due to one irri gation event skipped compared to the soil water balance requirements causing the controller to accumulate a small deficit until rainfall filled the root zone to capacity. Because the irrigation adequacy and scheduling efficiency results were determined from 30 day totals, one day of deficit or surplus a ffected 30 scores. Irrigation applied by the Toro controller for fall 2006 wa s 147 mm, under-irrigating by 15% compared to the theoretical irrigation re quirement, calculated as 172 mm (Figure 4-9). Over-irrigation occurred at the beginning of the season when it was generally wet causing the rain sensor to frequently bypass ev ents due to rainfall. However, inefficiency in the rain sensor or inconsistency in the rainfall data collected by the controller caused the controller to schedule more events than the theoretical requirement. The Toro applied irrigation in 7 events compared to the 2 events calculated as the theoretical re quirement. Also, this controller will schedule irrigation as if there is no wate r in the root zone when first installed as a precaution to protect poorly-managed landscapes despite this treatment being well-watered prior to treatment commencement. The controller applied 13.5 mm of irrigation during its first irrigation event when the overall average application per even t was 6.7 mm. Once frequent rainfall events stopped on September 22, the slope of the cumu lative irrigation applic ation for the Toro

PAGE 138

138 controller was much less than the slope of th e theoretical requirement resulting in overall savings. The average application per event by th e Toro controller was 26% less than the depth per event by the theoretical irrigation requirement. Irrigation adequacy averaged -69% for the Toro controller (Table 4-7). Deficits were not an issue in the early part of the fall season due to frequent rainfall as well as continued irrigation. Once irrigation became the primary source for wate r in the root zone, water levels decreased regularly below the maximum allowable deple tion causing a severe d ecline in irrigation adequacy attributable to wate ring restrictions. Scheduling e fficiency showed opposite trends, averaging 84% (Table 4-7). The depth of irriga tion applied during the rainy period of the season was greater than the root zone could hold causi ng a surplus and decreasing scheduling efficiency according to the daily soil water balance. Ir rigation occurring in ear ly morning could not account for rainfall occurring later in the day, an inherent error in daily timesteps of the model. When adequacy declined, scheduling efficiency wa s nearly 100% most of the time because overirrigation did not occur often. The time-based irrigation schedules for Oct ober in fall 2006 were in correct based on an error in the document used to de velop the schedule. The irriga tion schedule should have been similar to both September and November. Thus, the theoretical application for October was calculated by averaging application per event fo r September and November, and substituting that average depth for events that actually occurred in October. As a result, T4 and T5 would have applied 22% (210 mm) more than the theoretical requirement (Figure 4-10) and 25% (130 mm) less than the theoretical requirement of 172 mm (Figure 4-11), respectively. Irrigation application per event was only 1% greater than the theoretical requirement for T4 and 38% less

PAGE 139

139 than the theoretical requirement for T5. Both ti me-based treatments applied irrigation for just 4 more events than the theoretical requirement, totaling 23 events. The time-based treatment, T4, irrigated adequa tely during the first ha lf of the season but did not perform as well during the latter half, averaging -57% (Table 4-5). Irrigation supplemented rainfall so that deficits never occurred. The irrigation schedule for this treatment averaged 75% in scheduling efficiency (Table 45), suffering during the first half of the season, but not during the latter ha lf. Reduced irrigation in October allowed the irrigation events to not exceed field capacity causing this treatment to consistently score 100% for all of the 30-day periods including most of this month. However, the rest of the irrigation events applied more water than necessary to fill the root zone to field capacity resulting in lower scheduling efficiency scores. The reduced time-based treatment (T5) generally resulted in a higher scheduling efficiency than T4, averaging 86% (Table 4-6). This was due to the smaller irrigation depth per event resulting in less cumulative surplus and a longer time peri od of 100% scores. Irrigation adequacy for T5 averaged -101% (Table 4-6). I rrigation was fairly adequate for the majority of the first half of the season due to frequent rainfall. However, adequacy suffered because the reduced time schedule did not regularly apply enough irrigation to keep the water level above the maximum allowable depletion due to the incorrect irrigation schedule in October and little to no rainfall. The reduced time-based treatment cumulatively under-irrigated the most for the fall 2006 season. Water levels were maintained near field capacity for most of the season except for October where the incorrect irriga tion schedule caused soil moistu re to decline (Figure 4-12). Despite October, this treatment was well-watered for the fall season.

PAGE 140

140 Winter 2006-2007 Rainfall during the winter 2006-2007 seas on occurred occasiona lly throughout the season, totaling 17 events over 88 days (Figur e 4-13). The Weathermatic controller overirrigated by 8% compared to the theoretical irrigation requirement, a pplying 82 mm and 76 mm, respectively. The controller averaged 39% less than the theoretical requirement for irrigation depth per event, but scheduled irrigation 16 even ts compared to 9 events for the theoretical requirement. Scheduling efficiency averaged -98% and irrigation adequacy averaged 100% for the Weathermatic controller (Table 4-3). This treatme nt had perfect irrigation adequacy during all of the winter season. This controller never under-ir rigated. However, the controller was generally less efficient at applying irrigation. This exces s was a result of applying irrigation before the water level dropped below the maximum allowable depletion since the irri gation depth per event averaged less than the theoretical requirement. The Toro controller applied 15% less than the theoretical requirement during the winter 2006-2007 season, totaling 64 mm and 76 mm, respectively (Figure 414). The Toro controller applied 36% less per event on average compared to the theoretical requirement resulting in less cumulative application. The smaller irrigation depth per event allowe d the Toro treatment to apply irrrigation without over-irrigating during the first half of the season, scori ng 100% in scheduling efficiency (Table 4-7). The remainder of the season varied in scheduling efficiency, from 42% back to 100% by the end of the season; average schedulin g efficiency was 71%. Irrigation adequacy scores averaged from 22%. Irrigation did not adequately suppleme nt rainfall during the first half of the season resulting in constant cumulative deficit and lower adequacy However, irrigation

PAGE 141

141 was adequate when supplementing rainfall over the last portion of the season resulting in very little cumulative deficit. The rain sensor for the time-based treatments was less sensitive than some because it did not bypass many events during the winter 20062007 season. The irrigation applied by the theoretical irrigation requirement was 76 mm; T4 over-irrigated by 70%, applying 129 mm (Figure 4-15), and T5 over-irrigated by 8%, a pplying 82 mm (Figure 4-16). The reduced timebased treatment applied irrigation similarly to the theoretical requirement due to the crop coefficient of the theoretical requirement bei ng similar to the reduction percentage of the treatment. Cumulative over-irrig ation occurred because additional events took place that were not bypassed by the rain sensor. There was twice th e total number of even ts for the time-based treatments while applying 23% and 51% less irri gation per event compared to the theoretical requirement. The time-based treatment, T4, resulted in irri gation adequacy of 98% on average (Table 45). This treatment applied ade quate irrigation over the winter season due to applying irrigation consistently and not allowing a deficit to accrue However, scheduling efficiency averaged 55%; this treatment cumulatively over-irrigated causing th e scheduling efficiency to remain lower. The reduced time-based treatment scored 95% for irrigation adequacy and 83% for scheduling efficiency (Table 4-6). Adequacy and scheduling efficiency were so high for the latter half of the season due to small and freque nt irrigation events. Adequacy increased to 100% once enough rain fell to bring the water leve l above maximum allowable depletion so that irrigation became supplemental again. The Toro controller treatment, T2, cumulativ ely irrigated the least for this season. However, soil moisture levels were maintained near field capacity and above the maximum

PAGE 142

142 allowable depletion (Figure 4-17) This treatment, despite applying less than other treatments, was well-watered for the winter 2006-2007 season. Spring 2007 Rainfall was infrequent during the spring 2007 se ason, totaling 22 mm of effective rainfall from 9 events over 94 days; there was no rainfall in May. The Weathe rmatic controller underirrigated by 25%, applying 339 mm of irrigation compared to 453 mm of irrigation calculated by the theoretical requirement (Figure 4-18). Irri gation occurred everyday by the Weathermatic controller unless in bypass mode due to the rain sensor. Th e controller averaged 67% less irrigation depth per event, but applied irrigation for 48 additional events compared to the soil water balance totaling 84 events. Irrigation adequacy and schedul ing efficiency averaged 69% and 70%, respectively, for the Weathermatic controller treatment (Table 4-3). Adequacy decreased when the turfgrass water requirements changed, creating deficit accumulation. Irrigation applicati on was slightly greater than required initially, causing lower schedul ing efficiencies until the deficit accumulated enough so that irrigation depths per event were not greater than field capacity. The Toro controller applied 317 mm of i rrigation, under-irrigating by 30%, compared to 453 mm calculated for the theoretical irrigation requi rement (Figure 4-19). Irrigation application per event was half for the Toro controller, aver aging 7 mm/event, compared to the theoretical requirement, but there were approximately twice as many events applied by the Toro than the theoretical requirement, totaling 62 events. Overall, the Toro controller applied 59% less per event than was scheduled by the theoretical irrigation requirement. The Toro controller adequately applied irrigation at the beginning of the season, achieving 100% until rainfall and change in seasonal water needs cau sed under-irrigation to

PAGE 143

143 occur, ending at an irrigation adequacy of 98%, averaging 6% (Table 4-7). Irrigation was scheduled efficiently with scor es ranging from 77% to 100%, av eraging 90%. Similarly to the Weathermatic, scheduling efficiency increased to 100% in the latter part of the season due to the accumulated deficit reflected in th e irrigation adequacy score. Irrigation occuring over the season for the ET Water controller totaled 259 mm, underirrigating by 43% compared to the theoretica l irrigation requirement which totaled 451 mm (Figure 4-20). The theoretical irrigation requirement increased application per event to match perceived seasonal need whereas the ET Water controller did not update its schedule for an extended period of time. The controller failed to update its irrigation schedule due to signal issues from April 9, 2007 until May 23, 2007. It is likely that the controller would have updated crop coefficients and fluctuated with weather conditions if updated daily. The ET Water controller scored 28% in irriga tion adequacy where the deficit accumulated from the lack of change with seasonal water re quirements (Table 4-4). Scheduling efficiency, however, averaged 97%. This controller was generally able to schedule irrigation without applying more than the root zone could hold on a regular basis. The spring 2007 season experienced very little rainfall; however, the rain sensor caused irrigation to stop for an extended period of time re sulting in less irrigation than representative of the treatments. The additional irrigation events we re added into the treatments as theoretically would have been applied if the malfunction had not occurred. The application per event for the theoretical treatment was derived from the average of similar irrigation applications in the same month. The time-based treatments, T4 and T5, applied 21% (356 mm) and 52% (215 mm) less than the theoretical irrigation requirement, cal culated as 453 mm (Figure 4-21; Fig 4-22). The irrigation applied per event by T4 was 9% greater than the theoretical requirement and T5 was

PAGE 144

144 34% less than the requirement; both treatments ap plied irrigation for a less number of events, totaling 26 events compared to 36 by the theoreti cal irrigation requirement Under-irrigation by both treatments can be attributed to less number of events as we ll as higher crop coefficients compared to the reductions from historical of the time-based treatments. Irrigation adequacy for the time-based treatment, T4, and the reduced time-based treatment, T5, averaged 30% and -13%; adequacy dropped in the latter part of the season due to the high theoretical requirement (Table 4-5; Tabl e 4-6). Scheduling efficiency averaged 77% for T4 and 97% for T5. Scheduling efficiency was lower in the beginning of the season because the depth applied per event filled th e root zone past field capacity. However, scheduling efficiency increased as the deficit increased and adequacy decreased. Spring 2007 resulted in the most under-irrigat ion compared to the theoretical requirement by T5. According to the soil moisture data, th e volumetric water content only fell below the maximum allowable depletion during the period wher e the rain sensor did not allow irrigation to occur even though there was no rainfall (Figure 4-23) Otherwise, the treatment remained wellwatered indicating that all treatments we re well-watered for this season. Summer 2007 The summer 2007 season was rainy, bringing light ning into the area. On June 8, 2007, a lightning storm affected the irrigation equipment including the pump for the irrigation well. The GCREC maintenance crew immediat ely transferred the water source of the project to the farm system. However, pressure problems were appare nt in August. The pump was replaced at the end of August and the water source transferred b ack to the irrigation well, fixing the pressure problems.

PAGE 145

145 The Weathermatic controller defaulted to a ti me-based schedule during this season. The weather monitor used to gather data to calculate ETo was inoperable after being affected by the power outage; the weather monitor was replaced for the fall 2007 season. Rainfall was much more frequent, beginning the wet period in the summer 2007 season. However, the rain sensor connected to the Toro c ontroller became much less sensitive to rainfall causing the rain sensor to bypass irrigation for inte nse rainfall events. Irrigation application per event was 58% less for the Toro controller than the theoretical irrigation requirement, but applied irrigation for 55 events compared to 24 events for the theoretical requirement (Figure 4-24). This summer season resulted in 3% under-irr igation by applying 278 mm compared to the theoretical requirement that applied 287 mm. However, under-irrigation was only possible due to the lack of pressure suppl y during the month of August. Irrigation adequacy for the Toro controller scored 18% to 98%, averaging 71% (Table 47). This controller applied irri gation consistently to climb out of the deficit created from underirrigation in the previous season. However, once Augusts reduced irri gation application from the lack of pressure occurred, adequacy droppe d again due to accumulating deficit. Scheduling efficiency for the Toro contro ller ranged from 79% to 100%, averaging 85%, as a result of smaller, more frequent irrigation events. This controller only ove r-irrigated minimally throughout the entire season to cause the scheduling efficiency score to be less than 100%. Cumulative irrigation over summer 2007 was 260 mm, 18% less than the theoretical requirement that totaled 319 mm (Figure 4-25). Application dept h per event for the ET Water controller was 54% less than irriga tion depth per event for the theoretical requirement. However, the ET Water controller scheduled twice as many ev ents as the theoretical requirement. Similar

PAGE 146

146 to the Toro controller, under-i rrigation was only possible due to the lack of pressure supply resulting in reduced irrr igation in August. Also similar to the Toro controller were its irrigation adequacy and scheduling efficiency scores. Irrigation adequacy for the ET Wate r controller scored 66%, applying irrigation consistently to reduce the deficit created last season (Table 4-4). However, once irrigation decreased from the lack of pressure, adequacy sc ores fell again due to accumulating deficit. Scheduling efficiency for this controller averag ed 93% as a result of smaller, more frequent irrigation events. This controll er also over-irrigated minimall y throughout the entire season to cause the scheduling efficiency sc ore to be less than 100%. Irrigation application for the rainy su mmer 2007 season by T4 and T5 was 324 mm and 174 mm, respectively (Figure 4-26; Figure 4-27). These treatments applied 13% greater than the theoretical irrigation requirement and 40% less than the theoretical irrigation requirement, calculated to apply 287 mm. Application per even t by T4 was 51% greater than the theoretical requirement while T4 applied 15% less than th e theoretical requirement on average. Overirrigation and under-irrigation occurred because th e time-based treatments applied irrigation for less number of events in combination with the di fference in average irrigation depth per event. Also, the time-based treatments were affected by th e lack of pressure in August just as the other treatments. Irrigation adequacy averaged 72% over the summer season for T4, the time-based treatment (Table 4-5). Initia lly, the irrigation adequacy score suffered from the deficit accumulated over the latter part of the spring seas on. However, rainfall was frequent over the season to increase the water level so that less accumulation below the maximum allowable depletion occurred. Scheduling efficiency, how ever, suffered from the increased water level

PAGE 147

147 causing irrigation depths to be gr eater than field capacity and decreasing scheduling efficiency scores, averaging 48%. The reduced-time-based treatment, T5, applied less irrigation per event than T4. This affected the irrigation adequacy and scheduling efficiency scores appropriately by accumulating more deficit totals before irrigation became supplemental to rainfall as well as increasing scheduling efficiency ranges by not applying more than field capacity as much as T4 did. Irrigation adequacy and scheduling efficiency aver aged 34% and 78%, respect ively (Table 4-6). The reduced time-based treatment, T5, was well watered during the summer 2007 season. The volumetric soil moisture content remained a bove the maximum allowable depletion level for the entire season and water levels frequently exceeded field capacity (Figure 4-28). This treatment cumulatively app lied the least amount of i rrigation for the season. Fall 2007 Results from the Weathermatic controller for this season were not necessarily representative of the treatment because the weather monitor was damaged and left in an undesirable position (Figure 4-29). The weather monitor was not able to detect rainfall in this position resulting in excess irrigatio n. It is unknown the length of time this controller was affected by the situation and whether the heig ht and orientation affected the temperature measurements; the weather monitor was replaced to the proper position on October 9, 2007. Fall 2007 experienced much more rainfall during most of the season since it was included in the wet period. Rainfall events mostly occu rred prior to November; only 1 event out of 32 occurred in November. The Weathermatic cont roller applied 220 mm of irrigation compared to 167 mm calculated as the theoretical irrigati on requirement, over-irrigating by 32% (Figure 430). The controller did not bypass irrigation from September 27 through October 20, applying

PAGE 148

148 irrigation during 5 rainfall events. Irrigation adequacy was high for this season, averaging 99% (Table 4-3), due to frequent rainfall and excess irrigation during the rainy period; the root zone of this treatment experienced very little under-irrigation. However, irrigation occurring when not necessary severely affected the schedu ling efficiency, averaging 23%. Cumulatively, the Toro controll er applied the same amount of irrigation as the theoretical irrigation requirement, totaling 166 mm (Figure 4-31). Once again, the Toro controller applied less irrigation per event by 32%, but applied irrigation more often totaling 23 events compared to 16 events calculated for the theoretical requirem ent. Periods of fre quent irrigation events resulted in high irrigation adequacy scores wher e scores averaging 55% (T able 4-7). Irrigation still occurred when there were frequent rain fall events at the beginning of the season. Scheduling efficiency suffered due to the excess irrigation, but recovered later in the season when rainfall bypassing resumed. Scheduling e fficiency averaged 79% over the fall season. Irrigation application by the ET Water controller in the fall 2007 season was cumulatively equal to the calculated theoreti cal irrigation requirement, totaling 199 mm (Figure 4-32). Overirrigation occurred during the fi rst half of the season due to the rainy conditions and not recognizing a rain sensor. The la tter half of the season applied less irrigation per event compared to the theoretical irrigation requirement, appare nt in the steeper slope of the theoretical requirement. Irrigation adequacy for the ET Water controlle r averaged 79% (Table 4-4). Adequacy scores were affected by the redu ced irrigation over the rainy peri od in late October. Rainfall depths used by the manufacturer to incorporate into the controllers soil water balance were from a different weather station than the one used fo r the soil water balance model. The irrigation depths were calculated using the alternate rainfall data causing the irrigation depths to vary in

PAGE 149

149 adequacy over that time peri od. Irrigation scheduling effici ency averaged 92% over the fall season. Irrigation depths per event were sm all enough to over-irrigate cumulatively by a small amount keeping the scores generally high. Irrigation occurred despite significant rainfall events during the first portion of the fall 2007 season by the time-based treatments. T4 and T5 over-irrigated by 95% (326 mm) and 17% (194 mm), respectively, compared to 166 mm by th e theoretical irrigation requirement (Figure 433; Figure 4-34). Compared to the theoretical re quirement, more irrigation per event was applied for T4 by 40% and less irrigation per event wa s applied for T5 by 17%. There were more irrigation events scheduled by the time-based treatments than the theoretical irrigation requirement due to the less sensitive rain sensor bypassing less events as well as the theoretical irrigation requirement taking rain fall into account when calculating the depth to be applied. The time-based treatment, T4, and the reduced time-based treatment, T5, only had small amounts of under-irrigation due to watering day re strictions resulting in irrigation adequacy averaging 94% and 89%, respectively (Table 4-5, 4-9). Irrigation adequacy was sometimes not as high for T5 compared to T4 due to the re duced irrigation depth per event. Scheduling efficiency suffered, however, because the irri gation applied over the 30 day periods was in excess of water requirements causing consistent ov er-irrigation; scheduling efficiency averaged 36% for T4 and 61% for T5. Scheduling efficiency was so low in the first part of the season due to irrigation application desp ite frequent rainfall for both treatments. The Toro controller, T2, applied the least amount of irrigation for the fall 2007 season compared to the other treatments. The treatment maintained volumetric soil water content above field capacity most of the time resulting in well-watered conditions (Figure 4-35).

PAGE 150

150 Discussion It was determined in Chapter 3 that the W eathermatic controller over-estimated ETo used for scheduling irrigation. However, cumulative irrigation applied was less than the theoretical requirement during summer 2006, fall 2006, and sp ring 2007. This was due to the combination of using a lower crop coefficient (0.6 fixed over time) than was representa tive of the actual site and the 48-hour rain sensor bypass period which accumulated irrigation deficits. The 48 hour bypass is longer than recent resear ch has indicated where expanding disk rain sensors dry out with 24 hrs most of the time. The crop coe fficient used by the controller during winter 20062007, when over-irrigation occurred, was greater than the one used by the soil water balance model. Fall 2007 most likely over-irrigated due to the position of the w eather monitor during the rainy period. The Weathermatic controller regularly overirrigated as proven by the generally low scheduling efficiency scores, 65% of the scores were less than 50%, an d irrigated frequently enough to keep adequacy scores above 50% for 88% of the time (Figure 4-36). Watering restrictions cut irrigation days to two per week for the first three seasons. Scheduling efficiency scores were much lower for these seasons, averag ing -72%, because irrigation occurred to refill the entire deficit each time despite what the root zone could actually hold. This was because the Weathermatic controller did not use a traditional soil water balance with a root zone. Adequacy scores were high because irrigation occurred ever yday since watering restrict ions were lifted and the controller replaced the wate r loss from the previous day wi thout depleting to a certain maximum allowable depletion. The most unde r-irrigation compared to the theoretical requirement for this controller occurred duri ng the spring 2007 season when average scheduling efficiency and adequacy were both high values a nd the controller was allowed to irrigate seven days per week.

PAGE 151

151 Chapter 3 showed that the cumulative ETo estimation by the Toro controller compared to the ASCE method using on-site weather data we re nearly identical, estimating within 1% of each other. Settings used to calculate the theore tical irrigation requirement were identical to the Toro controller settings except for crop coeffi cient and method for rainfall estimation. The calculated theoretical requirement likely used crop coefficients that were much larger than the coefficients used by the Toro controller since th ey were developed for California and not Florida, resulting in the under-estimation of the actual requirement of the study site. Also, the Toro controller translated ra infall depths into number of days to pause irrigation without including rainfall into irrigation scheduling by the controller. The rain pause feature for the Toro controller sometimes caused the controller to bypass irriga tion for up to 5 days due to a considerable amount of rainfall. However, sandy soils in Flor ida have small soil water holding capacities and most rainfall was lost to drainage. The soil water balance model only considers effective rainfall and irrigation bypassing never occurr ed for more than 3 days. This difference also caused the Toro to under-irrigate compared to the theoretical requirement. Scheduling efficiency values were relative ly high for every season where 79% of the scores were above 70% for the study period, showing that the Toro controller was able to accurately judge the depth of irrigation required to fill the root zone to field capacity (Figure 437). Only 4% of the scores were below 50%. Average irrigation adequacy scores were negatively impacted over the firs t half of the study period due to day of week watering restrictions. Irrigation adequacy increased to pos itive values, but was still low for the last half of the study period; 51% of the scores were below 50%. Under-irrigation was also seen in the cumulative totals of irrigation application for the seasons. However, it was determined in Chapter 2 that turfgrass quality remained above acceptable quality standards despite low average

PAGE 152

152 adequacy scores for every season. Under-irriga tion for short periods of time will not negatively affect the quality of the landscape. Chapter 3 showed that that ETo estimation by the ET Water controller was 8% less than the ETo calculated using weather data from the onsite weather station. As a result, this controller under-irrigated compared to the theoretical requirement during the dry periods. The spring 2007 season resulted in the most savings, includi ng using an early April schedule to irrigate throughout late April and May. Irrigation wa s over-applied during wet periods because the controller failed to utilize the rain sensor. Scheduling efficiency scores were above 90% for 85% of the time for all seasons showing that the ET Water controller was also able to acc urately judge the depth of irrigation required to fill the root zone to field capacity (Figure 4-38) Irrigation adequacy scores were frequently lower than the scheduling efficiency, but were also relatively high where performance was greater than 50% for 67% of the study period. Under-irriga tion was apparent in the underirrigation compared to the theoretical require ment for spring and summer 2007, calculated as 43% and -18%, respectively, where adequacy averaged the lower values of 28% and 66%. Similarly to the Toro controller, turfgrass quali ty maintained acceptable levels for the entire study period as was determined in Chapter 2. The time-based treatments, T4 and T5, followed the theoretical requirement trends when water needs reflected the historical net irrigati on requirement. However, these treatments were not able to adjust for real weather conditions especially during periods of irregular weather changes from historical averages. These treat ments maintained wateri ng restrictions throughout the entire study period ensuring average irriga tion adequacy measurements less than 100%; scores were greater than 50% fo r 68% of the time for T4 and 96% of the time for T5 (Figure 4-

PAGE 153

153 39, 4-40). Scheduling efficiency scores were greater than 50% for 67% of the time and 56% of the time for T4 and T5, respectively. As would be expected, irrigation adequacy scores were greater for T4 than T5 while scheduling effici ency scores were greater for T5 than T4. The theoretical irrigation requirement did not always result in ir rigation adequacy and scheduling efficiency scores of 100% (Figure 441). The irrigation adeq uacy performance never reached 100% because the readily available water was allowed to deplete below the maximum allowable depletion before irri gation occurred. Scheduling effi ciency suffered from rainfall where irrigation events scheduled from the wate r needs of the previous day occur on the same day as rainfall. Despite the performance results of the theoretical requirement, the published SWAT scores for these controlle rs were within 5% of 100% fo r both irrigation adequacy and scheduling efficiency. Conclusions Rainfall in Florida is localized and important in determining how well these controllers schedule irrigation. The Weathermatic and Toro controllers both utiliz e a rain pause feature where the controller pauses irrigation for a certain number of days determined by the manufacturer. The Weathermatic controller test ed here (newer controller models have an adjustable rain pause from 1 to 7 days) alwa ys pauses for 48 hours despite whether there was enough rainfall to maintain adequate soil mois ture levels. The Toro controller uses a predetermined scale to choose the number of days based on depth of rainfall, whether or not that depth was effective. Though it is unclear exactly how it is done, rainfall is factored into the scheduling of the ET Water. However, they use rainfall at a weather station that may not be representative of the depth of rain at the site location. Inputs to the ET controllers, both manufactur er and user programmed, are extremely important to proper irrigation scheduling. Crop coefficients used by the ET controllers were

PAGE 154

154 either average for the entire year as in the W eathermatic or unknown, but were not necessarily representative of the values measured in Florid a. The crop coefficients developed for Florida by Jia et al. (2007) were used in the soil water ba lance model and could explain some of the lower scheduling efficiency and adequacy scores. Crop coefficients for the signal-based controllers were developed for California a nd not specific to Florida. For example, all ET controllers irrigated much less than the theoretical requiremen t for May where the crop coefficient was 0.90. The known inputs to the controllers were us ed to calculate the theoretical irrigation requirement; however, most of the time, the irrigation treatments maintained well-watered conditions according to the soil moisture conten t data even when the theoretical comparison showed under-irrigation and cumulative defi cits through low adequacy scores. The depth of the readily available water fo r the soil water balance model was 7.6 mm. This depth was most likely much smaller than what was used by the ET controllers to schedule irrigation. The smallest value for this depth is 14 mm for the SWAT testing protocol. As a result, it would be considerably harder for thes e controllers to maintain high irrigation adequacy and scheduling efficiency scores at the same time by not irrigating too much or too little. The SWAT testing protocol is the only accep ted way to test whether these controllers adequately schedule irrigation. The landscape scenario was published and the manufacturers could program the controllers before the test be gan so that their maximum allowable depletion depth was within the published values to ensure nearly perfect adequacy and scheduling efficiency scores. There are also no differences between the controller weather data and the weather data used for the soil water balance m odel due to the close pr oximity of the weather station to the test site; however, measurement of the ability of the controller to properly schedule irrigation depends on how well site appropriate whet her data used to estimate ETo are collected.

PAGE 155

155 SWAT currently allows manufacturers to con tinually test their controllers until they score what the manufacturer finds acceptable, tweaking be tween 30-day periods; they also allow the manufacturer to decide whether to publish the resu lts. It is likely that some controllers will be tested for six months to a year before achieving scores that are acceptable and the scores usually end as being within 5% of 100%. However, as was shown above in 30-day moving totals, controllers will schedule differently in a shor t period of time depending on weather conditions, rainfall, and time of year. As was seen in the scheduling efficiency and irrigation adequacy results for the theoretical requirement, rainfall was an important factor for determining the irrigation adequacy and scheduling efficiency results. Currently, onl y 10.2 mm of rainfall wa s required to complete the test with publishable results. However, su ch a small amount of rainfall does not allow the controller to show how it will perform in wet periods distinctive to Floridas historical climate. A properly managed time-based schedule with ra in sensor could provide the same water savings as an ET controller while maintaining adequacy and scheduling efficiency; however, it must be regularly adjusted to match climatic demand. A properly pr ogrammed ET controller could work better than manual irrigation scheduli ng so that irrigation would consistently be supplemental to rainfall and to minimize extraneous watering. However, the results show that ET controllers perform irrigation scheduling as well as the program settings and how they measure and incorporate rainfall.

PAGE 156

156 Table 4-1. Monthly crop coeffi cients for warm season turfgrass used to calculate crop evapotranspiration for the determination of the theoretical irrigation requirement Month Theoretical1 SWAT2 January 0.46 0.52 February 0.46 0.64 March 0.57 0.70 April 0.83 0.73 May 0.90 0.73 June 0.77 0.71 July 0.73 0.69 August 0.72 0.67 September 0.69 0.64 October 0.65 0.60 November 0.60 0.57 December 0.46 0.53 Average 0.65 0.64 1Theoretical refers to the crop coefficients used in the soil water balance model to determine the theoretical irrigation requir ement (Jia et al. 2007). 2SWAT refers to the crop coefficients used for ET controller testing by the Smart Water A pplication Technology grou p under the Irrigation Association (IA 2006c) developed in California. Table 4-2. Program setting differences1 from Table 2-2 for the summer 2006 season Start Date End Date Setting Treatment Difference from Table 2-2 May 25 Jul 10 Sprinkler Type2 Weathermatic 25.4 mm/hr May 25 Jul 10 Sprinkler Type ET Water 18.0 mm/hr May 25 Jun 8 Scheduling efficiency ET Water 45% May 25 Jun 8 Root Depth ET Water 305 mm May 25 Aug 10 Sprinkler Type Time-based4 19.1 mm/hr Jun 20 Jun 29 MAD3 ET Water 35% 1Settings outside of these dates are the same as listed in Table 2-2. 2Application rate or precipitation rate is termed sprinkler type for some ET controllers. 3MAD refers to the maximum allowable depletion. 4Time-based treatment refers to both the time-based treatment, T4, and the reduced time-based treatment, T5.

PAGE 157

157 Table 4-3. Weathermatic controll er, T1, results for average sche duling efficiency and irrigation adequacy calculated using 30-day moving to tals, percentage difference in irrigation application from the theoretical requireme nt, cumulative rainfall, and number of rainfall events for each season Season Scheduling Efficiency (%) Irrigation Adequacy (%) Difference from Theoretical1 (%) Rainfall Avg2 CV3 Avg CV Cum4 (mm) Events5 Sum 06 -57 -234 59 42 -24 641 32 Fall 06 -62 -342 85 16 -13 308 33 Win 06-07 -98 -114 100 0 8 167 16 Spr 07 70 32 69 58 -25 109 9 Sum 07 NA6 NA NA NA NA 446 37 Fall 07 23 24 99 2 32 264 32 Average -25 -278 82 22 -4 1Difference from theoretical is the difference between cumulative water application for the season compared to the theoretic al irrigation requirement. 2Avg is the average value calculated from all 30-day moving totals for the season. 3CV is the coefficient of variation calculated from all 30-day moving totals for the season. 4Cum is the cumulative total rainfall for the season. 5Events is the number of rainfall events that occurred over the season. 6NA is an abbreviation for Not Applicable and indicates seasons where the treatment was not working. Table 4-4. The ET Water controller, T3, results for average scheduling efficiency and irrigation adequacy calculated using 30-day moving to tals, percentage difference in irrigation application from the theoretical requireme nt, cumulative rainfall, and number of rainfall events for each season Season Scheduling Efficiency (%) Irrigation Adequacy (%) Difference from Theoretical1 (%) Rainfall Avg2 CV3 Avg CV Cum4 (mm) Events5 Sum 06 82 21 50 56 9 641 32 Fall 06 NA6 NA NA NA NA 308 33 Win 06-07 NA NA NA NA NA 167 16 Spr 07 97 2 28 290 -43 109 9 Sum 07 93 1 66 43 -18 446 37 Fall 07 92 3 79 25 0 264 32 Average 91 7 56 39 -13 1Difference from theoretical is the difference between cumulative water application for the season compared to the theoretic al irrigation requirement. 2Avg is the average value calculated from all 30-day moving totals for the season. 3CV is the coefficient of variation calculated from all 30-day moving totals for the season. 4Cum is the cumulative total rainfall for the season. 5Events is the number of rainfall events that occurred over the season. 6NA is an abbreviation for Not Applicable and indicates seasons where the treatment was not working.

PAGE 158

158 Table 4-5. Time-based treatment, T4, results fo r average scheduling efficiency and irrigation adequacy calculated using 30-day moving to tals, percentage difference in irrigation application from the theoretical requireme nt, cumulative rainfall, and number of rainfall events for each season Season Scheduling Efficiency (%) Irrigation Adequacy (%) Difference from Theoretical1 (%) Rainfall Avg2 CV3 Avg CV Cum4 (mm) Events5 Sum 06 64 12 65 36 13 641 32 Fall 06 75 34 -57 -211 2 (22)7 308 33 Win 06-07 55 20 98 8 70 167 16 Spr 07 77 21 30 176 -32 (-21) 109 9 Sum 07 48 20 72 10 13 446 37 Fall 07 36 24 94 3 95 264 32 Average 59 27 50 115 27 (32) 1Difference from theoretical is the difference between cumulative water application for the season compared to the theoretic al irrigation requirement. 2Avg is the average value calculated from all 30-day moving totals for the season. 3CV is the coefficient of variation calculated from all 30-day moving totals for the season. 4Cum is the cumulative total rainfall for the season. 5Events is the number of rainfall events that occurred over the season. 6NA is an abbreviation for Not Applicable and indicates seasons where the treatment was not working. 7Parentheses represent what the treatment would have applie d if the rain sensor had not malfunctioned. Table 4-6. Reduced time-based treatment, T5, results for average scheduling efficiency and irrigation adequacy calculat ed using 30-day moving totals percentage difference in irrigation application from the theoretical requirement, cumulative rainfall, and number of rainfall events for each season Season Scheduling Efficiency (%) Irrigation Adequacy (%) Difference from Theoretical1 (%) Rainfall Avg2 CV3 Avg CV Cum4 (mm) Events5 Sum 06 78 10 54 41 -27 641 32 Fall 06 86 23 -101 -135 -36 (-25) 308 33 Win 06-07 83 14 95 13 8 167 16 Spr 07 97 5 -13 -497 -59 (-52) 109 9 Sum 07 78 12 34 136 -40 446 37 Fall 07 61 16 89 8 17 264 32 Average 80 15 26 282 -23 1Difference from theoretical is the difference between cumulative water application for the season compared to the theoretic al irrigation requirement. 2Avg is the average value calculated from all 30-day moving totals for the season. 3CV is the coefficient of variation calculated from all 30-day moving totals for the season. 4Cum is the cumulative total rainfall for the season. 5Events is the number of rainfall events that occurred over the season. 6NA is an abbreviation for Not Applicable and indicates seasons where the treatment was not working.

PAGE 159

159 Table 4-7. Toro controller, T2, results for averag e scheduling efficiency and irrigation adequacy calculated using 30-day moving totals, percen tage difference in irrigation application from the theoretical requirement, cumulative rainfall, and number of rainfall events for each season Season Scheduling Efficiency (%) Irrigation Adequacy (%) Difference from Theoretical1 (%) Rainfall Avg2 CV3 Avg CV Cum4 (mm) Events5 Sum 06 NA6 NA NA NA NA 641 32 Fall 06 84 25 -69 -156 -15 308 33 Win 06-07 71 25 22 509 -15 167 16 Spr 07 90 9 6 1386 -30 109 9 Sum 07 85 4 71 41 -3 446 37 Fall 07 79 11 55 62 0 264 32 Average 82 9 17 324 -13 1Difference from theoretical is the difference between cumulative water application for the season compared to the theoretic al irrigation requirement. 2Avg is the average value calculated from all 30-day moving totals for the season. 3CV is the coefficient of variation calculated from all 30-day moving totals for the season. 4Cum is the cumulative total rainfall for the season. 5Events is the number of rainfall events that occurred over the season. 6NA is an abbreviation for Not Applicable and indicates seasons where the treatment was not working. Figure 4-1. Comparison of rainfall for the 2006-2007 study period and average historical rainfall on a monthly and cumulative basis.

PAGE 160

160 Figure 4-2. The FAWN measured to tal rainfall and effective rainfall determined from the soil water balance model for the study period us ing the weather sta tion in Balm, FL.

PAGE 161

161 0 2 4 6 8 10 12 14 16 18 20 0 50 100 150 200 250 5/256/86/227/67/208/3Daily Effective Rainfall (mm) Irrigation Applied (mm)Date (2006) Irrigation Required Irrigation Applied 0 20 40 60 80 100Percentage (%) Irrigation Adequacy Scheduling Efficiency Figure 4-3. Weathermatic cont roller (T1) results over the summer 2006 season for cumulative theoretical irrigation depth applied, daily effective rainfall, and 30-day moving totals of irrigation adequacy an d scheduling efficiency.

PAGE 162

162 0 2 4 6 8 10 12 14 16 18 20 0 50 100 150 200 250 5/256/86/227/67/208/3Daily Effective Rainfall (mm) Irrigation Applied (mm)Date (2006) Irrigation Required Irrigation Applied 0 20 40 60 80 100Percentage (%) Irrigation Adequacy Scheduling Efficiency Figure 4-4. ET Water controller (T3) results over the summer 2006 season for cumulative theoretical irrigation depth applied, daily effective rainfall, and 30-day moving totals of irrigation adequacy an d scheduling efficiency.

PAGE 163

163 0 2 4 6 8 10 12 14 16 18 20 0 50 100 150 200 250 5/256/86/227/67/208/3Daily Effective Rainfall (mm) Irrigation Applied (mm)Date (2006) Irrigation Required Irrigation Applied 0 20 40 60 80 100Percentage (%) Irrigation Adequacy Scheduling Efficiency Figure 4-5. Time-based treatment (T4) resu lts over the summer 2006 season for cumulative theoretical irrigation depth applied, daily effective rainfall, and 30-day moving totals of irrigation adequacy an d scheduling efficiency.

PAGE 164

164 0 2 4 6 8 10 12 14 16 18 20 0 50 100 150 200 250 5/256/86/227/67/208/3Daily Effective Rainfall (mm) Irrigation Applied (mm)Date (2006) Irrigation Required Irrigation Applied 0 20 40 60 80 100Percentage (%) Irrigation Adequacy Scheduling Efficiency Figure 4-6. Reduced time-based treatment (T 5) results over the summer 2006 season for cumulative theoretical irrigation depth applied, daily effective rainfall, and 30-day moving totals of irrigation adequacy and scheduling efficiency.

PAGE 165

165 Figure 4-7. Measured volumetric soil moisture content over the summer 2006 season for T5, the reduced time-based treatment.

PAGE 166

166 0 2 4 6 8 10 12 14 16 18 20 0 50 100 150 200 8/138/279/109/2410/810/2211/511/19Daily Effective Rainfall (mm) Irrigation Applied (mm)Date (2006) Irrigation Required Irrigation Applied 0 20 40 60 80 100Percentage (%) Irrigation Adequacy Scheduling Efficiency Figure 4-8. Weathermatic contro ller (T1) results over the fa ll 2006 season for cumulative theoretical irrigation depth applied, daily effective rainfall, and 30-day moving totals of irrigation adequacy an d scheduling efficiency.

PAGE 167

167 0 2 4 6 8 10 12 14 16 18 20 0 50 100 150 200 8/138/279/109/2410/810/2211/511/19Daily Effective Rainfall (mm) Irrigation Applied (mm)Date (2006) Irrigation Required Irrigation Applied 0 20 40 60 80 100Percentage (%) Irrigation Adequacy Scheduling Efficiency Figure 4-9. Toro controller (T2) results over the fall 2006 seas on for cumulative theoretical irrigation depth applied, da ily effective rainfall, and 30-day moving totals of irrigation adequacy and scheduling efficiency.

PAGE 168

168 0 2 4 6 8 10 12 14 16 18 20 0 50 100 150 200 8/138/279/109/2410/810/2211/511/19Daily Effective Rainfall (mm) Irrigation Applied (mm)Date (2006) Irrigation Required Irrigation Applied Theoretical T4 0 20 40 60 80 100Percentage (%) Irrigation Adequacy Scheduling Efficiency Figure 4-10. Time-based treatment (T4) resu lts over the fall 2006 season for cumulative theoretical irrigation depth applied, daily effective rainfall, and 30-day moving totals of irrigation adequacy and scheduling effici ency. Actual water application was used in the calculations of i rrigation adequacy and irrigation scheduling efficiency.

PAGE 169

169 0 2 4 6 8 10 12 14 16 18 20 0 50 100 150 200 8/138/279/109/2410/810/2211/511/19Daily Effective Rainfall (mm) Irrigation Applied (mm)Date (2006) Irrigation Required Irrigation Applied Theoretical T5 0 20 40 60 80 100Percentage (%) Irrigation Adequacy Scheduling Efficiency Figure 4-11. Reduced time-based treatment (T5) results over the fall 2006 season for cumulative theoretical irrigation depth applied, daily effective rainfall, and 30-day moving totals of irrigation adequacy and scheduling effici ency. Actual water application was used in the calculations of i rrigation adequacy and irrigation scheduling efficiency.

PAGE 170

170 Figure 4-12. Volumetric soil moisture content over the fall 2006 season for T5, the reduced timebased treatment.

PAGE 171

171 0 2 4 6 8 10 12 14 16 18 20 0 25 50 75 100 125 12/112/1512/291/121/262/92/23Daily Effective Rainfall (mm) Irrigation Applied (mm)Date (2006 2007) Irrigation Required Irrigation Applied 0 20 40 60 80 100Percentage (%) Irrigation Adequacy Scheduling Efficiency Figure 4-13. Weathermatic co ntroller (T1) results over the winter 2006-2007 season for cumulative theoretical irrigation depth applied, daily effective rainfall, and 30-day moving totals of irrigation adequacy and scheduling efficiency.

PAGE 172

172 0 2 4 6 8 10 12 14 16 18 20 0 25 50 75 100 125 12/112/1512/291/121/262/92/23Daily Effective Rainfall (mm) Irrigation Applied (mm)Date (2006 2007) Irrigation Required Irrigation Applied 0 20 40 60 80 100Percentage (%) Irrigation Adequacy Scheduling Efficiency Figure 4-14. Toro controller (T 2) results over the winter 2006-2007 season for cumulative theoretical irrigation depth applied, daily effective rainfall, and 30-day moving totals of irrigation adequacy an d scheduling efficiency.

PAGE 173

173 0 2 4 6 8 10 12 14 16 18 20 0 25 50 75 100 125 12/112/1512/291/121/262/92/23Daily Effective Rainfall (mm) Irrigation Applied (mm)Date (2006 2007) Irrigation Required Irrigation Applied 0 20 40 60 80 100Percentage (%) Irrigation Adequacy Scheduling Efficiency Figure 4-15. Time-based treatment (T4) result s over the winter 2006-2007 season for cumulative theoretical irrigation depth applied, daily effective rainfall, and 30-day moving totals of irrigation adequacy an d scheduling efficiency.

PAGE 174

174 0 2 4 6 8 10 12 14 16 18 20 0 25 50 75 100 125 12/112/1512/291/121/262/92/23Daily Effective Rainfall (mm) Irrigation Applied (mm)Date (2006 2007) Irrigation Required Irrigation Applied 0 20 40 60 80 100Percentage (%) Irrigation Adequacy Scheduling Efficiency Figure 4-16. Reduced time-based treatment (T 5) results over the winter 2006-2007 season for cumulative theoretical irrigation depth applied, daily effective rainfall, and 30-day moving totals of irrigation adequacy and scheduling efficiency.

PAGE 175

175 Figure 4-17. Volumetric soil moisture content over the winter 2006-2007 season for T2, the Toro controller.

PAGE 176

176 0 2 4 6 8 10 12 14 16 18 20 0 50 100 150 200 250 300 350 400 450 500 2/273/133/274/104/245/85/22Daily Effective Rainfall (mm) Irrigation Applied (mm)Date (2007) Irrigation Required Irrigation Applied 0 20 40 60 80 100Percentage (%) Irrigation Adequacy Scheduling Efficiency Figure 4-18. Weathermatic cont roller (T1) results over the spring 2007 season for cumulative theoretical irrigation depth applied, daily effective rainfall, and 30-day moving totals of irrigation adequacy an d scheduling efficiency.

PAGE 177

177 0 2 4 6 8 10 12 14 16 18 20 0 50 100 150 200 250 300 350 400 450 500 2/273/133/274/104/245/85/22Daily Effective Rainfall (mm) Irrigation Applied (mm)Date (2007) Irrigation Required Irrigation Applied 0 20 40 60 80 100Percentage (%) Irrigation Adequacy Scheduling Efficiency Figure 4-19. Toro controller (T2) results over the spring 2007 s eason for cumulative theoretical irrigation depth applied, da ily effective rainfall, and 30-day moving totals of irrigation adequacy and scheduling efficiency.

PAGE 178

178 0 2 4 6 8 10 12 14 16 18 20 0 50 100 150 200 250 300 350 400 450 500 2/273/133/274/104/245/85/22Daily Effective Rainfall (mm) Irrigation Applied (mm)Date (2007) Irrigation Required Irrigation Applied 0 20 40 60 80 100Percentage (%) Irrigation Adequacy Scheduling Efficiency Figure 4-20. ET Water controller (T3) results over the spring 2007 season for cumulative theoretical irrigation depth applied, daily effective rainfall, and 30-day moving totals of irrigation adequacy an d scheduling efficiency.

PAGE 179

179 0 2 4 6 8 10 12 14 16 18 20 0 50 100 150 200 250 300 350 400 450 500 2/273/133/274/104/245/85/22Daily Effective Rainfall (mm) Irrigation Applied (mm)Date (2007) Irrigation Required Irrigation Applied Theoretical T4 0 20 40 60 80 100Percentage (%) Irrigation Adequacy Scheduling Efficiency Figure 4-21. Time-based treatment (T4) result s over the spring 2007 season for cumulative theoretical irrigation depth applied, daily effective rainfall, and 30-day moving totals of irrigation adequacy an d scheduling efficiency.

PAGE 180

180 0 2 4 6 8 10 12 14 16 18 20 0 50 100 150 200 250 300 350 400 450 500 2/273/133/274/104/245/85/22Daily Effective Rainfall (mm) Irrigation Applied (mm)Date (2007) Irrigation Required Irrigation Applied Theoretical T5 0 20 40 60 80 100Percentage (%) Irrigation Adequacy Scheduling Efficiency Figure 4-22. Reduced time-based treatment (T5) results over the spring 2007 season for cumulative theoretical irrigation depth applied, daily effective rainfall, and 30-day moving totals of irrigation adequacy and scheduling efficiency.

PAGE 181

181 Figure 4-23. Volumetric soil moisture content ov er the spring 2007 season for T5, the reduced time-based treatment.

PAGE 182

182 0 2 4 6 8 10 12 14 16 18 20 0 50 100 150 200 250 300 350 400 450 6/16/156/297/137/278/108/24Daily Effective Rainfall (mm) Irrigation Applied (mm)Date (2007) Irrigation Required Irrigation Applied 0 20 40 60 80 100Percentage (%) Irrigation Adequacy Scheduling Efficiency Figure 4-24. Toro controller (T 2) results over the summer 2007 season for cumulative theoretical irrigation depth applied, da ily effective rainfall, and 30-day moving totals of irrigation adequacy and scheduling efficiency.

PAGE 183

183 0 2 4 6 8 10 12 14 16 18 20 0 50 100 150 200 250 300 350 400 450 6/16/156/297/137/278/108/24Daily Effective Rainfall (mm) Irrigation Applied (mm)Date (2007) Irrigation Required Irrigation Applied 0 20 40 60 80 100Percentage (%) Irrigation Adequacy Scheduling Efficiency Figure 4-25. ET Water controller (T3) results over the summer 2007 season for cumulative theoretical irrigation depth applied, daily effective rainfall, and 30-day moving totals of irrigation adequacy an d scheduling efficiency.

PAGE 184

184 0 2 4 6 8 10 12 14 16 18 20 0 50 100 150 200 250 300 350 400 450 6/16/156/297/137/278/108/24Daily Effective Rainfall (mm) Irrigation Applied (mm)Date (2007) Irrigation Required Irrigation Applied 0 20 40 60 80 100Percentage (%) Irrigation Adequacy Scheduling Efficiency Figure 4-26. Time-based treatment (T4) results over the summer 2007 season for cumulative theoretical irrigation depth applied, daily effective rainfall, and 30-day moving totals of irrigation adequacy an d scheduling efficiency.

PAGE 185

185 0 2 4 6 8 10 12 14 16 18 20 0 50 100 150 200 250 300 350 400 450 6/16/156/297/137/278/108/24Daily Effective Rainfall (mm) Irrigation Applied (mm)Date (2007) Irrigation Required Irrigation Applied 0 20 40 60 80 100Percentage (%) Irrigation Adequacy Scheduling Efficiency Figure 4-27. Reduced time-based treatment (T 5) results over the summer 2007 season for cumulative theoretical irrigation depth applied, daily effective rainfall, and 30-day moving totals of irrigation adequacy and scheduling efficiency.

PAGE 186

186 Figure 4-28. Volumetric soil moisture content for the Summer 2007 season for T5, the reduced time-based treatment.

PAGE 187

187 Figure 4-29. The position of th e Weathermatic weather mon itor when it was damaged.

PAGE 188

188 0 2 4 6 8 10 12 14 16 18 20 0 50 100 150 200 250 300 350 400 9/19/159/2910/1310/2711/1011/24Daily Effective Rainfall (mm) Irrigation Applied (mm)Date (2007) Irrigation Required Irrigation Applied 0 20 40 60 80 100Percentage (%) Irrigation Adequacy Scheduling Efficiency Figure 4-30. Weathermatic cont roller (T1) results over the fall 2007 season for cumulative theoretical irrigation depth applied, daily effective rainfall, and 30-day moving totals of irrigation adequacy an d scheduling efficiency.

PAGE 189

189 0 2 4 6 8 10 12 14 16 18 20 0 50 100 150 200 250 300 350 400 9/19/159/2910/1310/2711/1011/24Daily Effective Rainfall (mm) Irrigation Applied (mm)Date (2007) Irrigation Required Irrigation Applied 0 20 40 60 80 100Percentage (%) Irrigation Adequacy Scheduling Efficiency Figure 4-31. Toro controller (T 2) results over the fall 2007 season for cumulative theoretical irrigation depth applied, da ily effective rainfall, and 30-day moving totals of irrigation adequacy and scheduling efficiency.

PAGE 190

190 0 2 4 6 8 10 12 14 16 18 20 0 50 100 150 200 250 300 350 400 9/19/159/2910/1310/2711/1011/24Daily Effective Rainfall (mm) Irrigation Applied (mm)Date (2007) Irrigation Required Irrigation Applied 0 20 40 60 80 100Percentage (%) Irrigation Adequacy Scheduling Efficiency Figure 4-32. ET Water controller (T 3) results over the fall 2007 season for cumulative theoretical irrigation depth applied, da ily effective rainfall, and 30-day moving totals of irrigation adequacy and scheduling efficiency.

PAGE 191

191 0 2 4 6 8 10 12 14 16 18 20 0 50 100 150 200 250 300 350 400 9/19/159/2910/1310/2711/1011/24Daily Effective Rainfall (mm) Irrigation Applied (mm)Date (2007) Irrigation Required Irrigation Applied 0 20 40 60 80 100Percentage (%) Irrigation Adequacy Scheduling Efficiency Figure 4-33. Time-based treatment (T4) resu lts over the fall 2007 season for cumulative theoretical irrigation depth applied, daily effective rainfall, and 30-day moving totals of irrigation adequacy an d scheduling efficiency.

PAGE 192

192 0 2 4 6 8 10 12 14 16 18 20 0 50 100 150 200 250 300 350 400 9/19/159/2910/1310/2711/1011/24Daily Effective Rainfall (mm) Irrigation Applied (mm)Date (2007) Irrigation Required Irrigation Applied 0 20 40 60 80 100Percentage (%) Irrigation Adequacy Scheduling Efficiency Figure 4-34. Reduced time-based treatment (T5) results over the fall 2007 season for cumulative theoretical irrigation depth applied, daily effective rainfall, and 30-day moving totals of irrigation adequacy an d scheduling efficiency.

PAGE 193

193 Figure 4-35. Volumetric soil moisture content for the fall 2007 season for T2, the Toro controller. 61% 16% 11% 12% 6% 13% 16% 65% 77% 88% 100% 19% 35% 0% 20% 40% 60% 80% 100% > 90% 70% 90% 50% 70% < 50%Frequency Percentage (%)Ranges Irrigation Adequacy Scheduling Efficiency Figure 4-36. Weathermatic controller, T1, pe rcent frequency of i rrigation adequacy and scheduling efficiency scores for the followi ng performance ranges: greater than 90%, 70% to 90%, 50% to 70%, and less than 50%.

PAGE 194

194 31% 11% 7% 51% 35% 44% 17% 4% 42% 49% 100% 79% 96% 0% 20% 40% 60% 80% 100% > 90% 70% 90% 50% 70% < 50%Frequency Percentage (%)Ranges Irrigation Adequacy Scheduling Efficiency Figure 4-37. Toro controller, T2, percent fre quency of irrigation adequacy and scheduling efficiency scores for the following perfor mance ranges: greater than 90%, 70% to 90%, 50% to 70%, and less than 50%. 33% 17% 17% 33% 85% 10% 5% 1% 50% 67% 95% 99% 100% 0% 20% 40% 60% 80% 100% > 90% 70% 90% 50% 70% < 50%Frequency Percentage (%)Ranges Irrigation Adequacy Scheduling Efficiency Figure 4-38. ET Water controller, T3, percent frequency of irri gation adequacy and scheduling efficiency scores for the following perfor mance ranges: greater than 90%, 70% to 90%, 50% to 70%, and less than 50%.

PAGE 195

195 41% 20% 7% 32% 12% 16% 39% 33% 61% 68% 28% 67% 100% 0% 20% 40% 60% 80% 100% > 90% 70% 90% 50% 70% < 50%Frequency Percentage (%)Ranges Irrigation Adequacy Scheduling Efficiency Figure 4-39. Time-based treatment, T4, percent frequency of irrigation adequacy and scheduling efficiency scores for the following perfor mance ranges: greater than 90%, 70% to 90%, 50% to 70%, and less than 50%. 25% 22% 9% 44% 36% 37% 23% 4% 47% 56% 100% 73% 96% 0% 20% 40% 60% 80% 100% > 90% 70% 90% 50% 70% < 50%Frequency Percentage (%)Ranges Irrigation Adequacy Scheduling Efficiency Figure 4-40. Reduced time-based treatment, T5, percent frequency of ir rigation adequacy and scheduling efficiency scores for the followi ng performance ranges: greater than 90%, 70% to 90%, 50% to 70%, and less than 50%.

PAGE 196

196 Figure 4-41. Total daily rainfall and 30-day moving totals of i rrigation adequacy and scheduling efficiency for the theoretical irrigation requirement over the entire study period.

PAGE 197

197 CHAPTER 5 CONCLUSIONS AND FUTURE WORK Conclusions The goal of this research was to determine whether ET-based irriga tion controllers would be able to conserve water in Florida climate conditions. The primary objec tives of this research were to evaluate the ability of three brands of ET-based controllers to A) schedule irrigation by comparing irrigation application to a time clock schedule intended to mimic homeowner irrigation schedules, while maintaining acceptabl e turfgrass quality, B) estimate reference evapotranspiration compared to the ASCE Standardized Reference Evapotranspiration methodology, and C) schedule irrigation compared to a theoretically derived soil water balance model. Secondary objectives included a) quan tifying the variation betw een controllers of the same brand, b) compare the performance of th e ET controllers based on approximate distance to a publically available weather data source, and c) measure the ET controller performance using the SWAT testing protocol. Rainfall over the study period, May 25, 2006 through November 30, 2007, was less than the thirty year historical rainfa ll averages for the same time pe riod. Although drier than normal, there were two distinct wet periods ranging from June thro ugh September for 2006 and June through October for 2007. The study period ex perienced dry conditions containing 69% dry days. It was found that using a rain sensor w ith a time-based irrigati on schedule conserved 21% of water despite the unusual dry conditions. The three brands of ET controllers studied were as follows: Weathermatic SL1600 controller with SLW15 weather monitor; Toro Intelli-se nse utilizing the WeatherTRAK ET Everywhere service; and ET Wate r Smart Controller 100. There we re three controllers, one of each brand, installed in southw est Florida at a project site where the controllers scheduled

PAGE 198

198 irrigation on mixed landscape plots. Controllers observed wateri ng restrictions of 2d/wk and were programmed with a system efficiency of 100% or 95%for the first half of the study period. Watering restrictions were lifted to allow 7d/wk irrigation and included a system efficiency of 80% based on distribution uniformity test ing for the last half of the study. Nine ET controllers, three replications of each brand being tested, were installed in addition to the main project to determine if there was variability between controllers concerning irrigation scheduling, refere nce evapotranspiration (ETo) estimation, and proximity to weather data source. There were no differences between the replications of the controllers for both irrigation scheduling and ETo estimation. The Weathermatic controllers were not affect ed by the proximity of a weather station as they used an on-site weather monitor for data collection. These controllers over-estimated ETo by 8% due to the combination of using Hargreaves equation for ETo calculations and overestimating maximum temperatures. It was found using data collected from the Weathermatic controllers that Hargreaves equation was de pendent on quality of temperature data and independent of the source for ex traterrestrial radiation data. The Toro controllers estimated ETo within 1% when the weather station was within 100 m of the controller, but had a 16% error when the nearest publically available weather station was 11 km away. Th e ET Water controllers over-estimated compared to the public weat her station by 8% and under-estimated by 12% compared to ETo calculated from the on-site weather station data. There was very little variability between controller replications for ETo estimation or water application which increases the validity of the results found using the controllers at the main project site. Water application by the ET controllers on mi xed landscape plots was compared to a timebased schedule without a rain sensor. This treatment represented a conservative homeowner

PAGE 199

199 because the schedule was based on the historical net irrigation requirement and was updated monthly to reflect historical ch anges in weather conditions. Av erage savings compared to the time-based schedule without rain sensor across all seasons ranged from 35% to 42% for the ET controllers. Reducing the time-based schedule by 40% and including a rain sensor resulted in 53% savings showing that updating the time cloc k settings throughout the year can result in substantial irrigation savings. However, time-b ased schedules do not fluctuate with changing weather conditions and typical homeowners will not manually adjust on a regular basis. Thus, the ET controllers are necessary for consistent water savings. Weekly water application showed that the ET controllers applied le ss irrigation per week than the time-based schedule without a rain sens or. Also, the ET controllers applied weekly irrigation similarly to the time-based schedule uti lizing a rain sensor during seasons where water requirements were high such as spring and summer Conversely, weekly water application was similar to the reduced time-based schedule dur ing seasons when water requirements were low such as fall and winter. There we re not treatment differences in turfgrass quality over the entire study period and ratings remained above the mi nimally acceptable level. Also, there was no correlation between turf grass quality and amount of irrigation. Water application by the ET controllers and th e time-based schedules at the main project site was compared to a theoretical water requi rement determined from a soil water balance model. The treatments were subjected to irrigation adequacy and scheduling efficiency testing using 30-day moving totals similar to the SWAT protocol testing. Contro llers were evaluated on the frequency of scores within the following ranges: greater than 90%, 70% to 90%, 50% to 70%, and less than 50%.

PAGE 200

200 It was determined that the Weathermatic c ontroller averaged perf ormed in the highest range 61% of the time in irrigation adequacy and the lowest range 65% in scheduling efficiency while under-irrigating by -4% compar ed to the theoretica l irrigation requirement This controller was designed to operate on wate ring restrictions and irrigate d based only on accumulated water loss between allowable watering days. This cont roller scored high on i rrigation adequacy during dry periods when irrigation was the primary plant water input. However, irrigation adequacy suffered when the rain sensor bypassed irrigation events when under watering restrictions due to a mandatory 48 hour rain delay. Scheduling efficiency performance was poor when under watering restrictions because irrigation occurr ed to refill the crop evapotranspiration (ETc) loss without consideration to the so il water holding capacity. Once the watering restrictions were lifted, the Weathermatic contro ller applied irrigation everyday in small amounts resulting in higher scheduling efficiency results. The Toro controller performed above the 70% to 90% range for 42% of the time and in the less than 50% range for 51% of the time for irrigation adequacy. Scheduling efficiency remained high with 79% in the 70% to 90% range or above and 4% in the less than 50% range. Under-irrigation averaged 13% compared to th e theoretical irrigation requirement. This controller applied small am ounts of irrigation per event resulti ng in high efficiency performance, but low irrigation adequacy when under 2 d/wk wa tering restrictions. The ET Water controller had the same amount of under-irrig ation as the Toro compared to the theoretical irrigation requirement, but scored higher in irrigation adequacy a nd scheduling efficiency more frequently with 50% and 95% of the scores above 70%, resp ectively. Scheduling efficiency was high for this controller because it applied small amounts of irrigation every day. This controller did not update the irrigation schedule from April 9 through May 23, 2007 causing the controller to not

PAGE 201

201 account for increased plant water needs. Also, because irrigation occurred everyday and this controller did not recogni ze the rain sensor, the average irrigation adequacy was higher than the Toro controller. Cumulative irrigation averaged 27% over a nd 23% less compared to the theoretical irrigation requirement for the time-based sc hedule and reduced time-based schedule, respectively; both treatments utili zed a rain sensor. The time-base d schedule resulted in scores of 61% above 70% for irrigation adequacy and 72% below 70% for scheduling efficiency. The reduced time-based schedule scored 47% and 73% of the time above 70% for irrigation adequacy and scheduling efficiency, respectively. Thes e treatments were subject to 2 d/wk watering restrictions for the entire time pe riod, resulting in lower adequacy scores, and irrigated more than the soil water holding capacity. The reduced time-based treatment applied less irrigation per event, resulting in higher efficiency scores than the time-based schedule, but suffered in adequacy due to less water in the root zone fo r depletion. The ET c ontrollers under-irrigated compared to the theoretical, on average, but fell w ithin results seen for the time-based schedules. Future Work These controllers were tested near public we ather stations in a re search setting and were shown to save water under these conditions. However, water savi ngs for residential homeowners tend to be less than controlled research experime nts. Therefore, these controllers should be tested in real homes with homeowner interacti on to see if they are still important to water conservation efforts in Florida. Only three brands of ET controllers were tested with this study, but there are approximately a dozen ET controllers currently commercially available for residential landscape use. Testing other controllers would be info rmative in understanding all of the options on the

PAGE 202

202 market and would help to determine the best type of ET controller for maximum water conservation in Floridas climate. The results found from this study were for relatively dry conditions. Rainfall was shown to impact the ability of the ET controllers to sc hedule irrigation. Studies should continue through rainy years to fully determine average irrigation savings by using these cont rollers in Florida. Also, observing these controllers in wet year s could assist in ma king a recommendation to manufacturers on how to better in corporate rainfall into ET-based scheduling so that they can fully penetrate the Florida market eventually leading to statewide water conservation. The variability between controller replications was determined from only four months of data collection. Variability could increase with a longer time period. The replications should be monitored for an extended period of time to de termine if and when the variability between controllers becomes significant. It was determined from this research that the programmed settings for ET controllers are important in scheduling the correct amount of wate r application to ensure efficient irrigation. However, the average homeowner cannot program a VCR let alone a new technology such as this. Future work should include the de velopment of recommendations for common ET controllers to aid in the pr ocess of integrating the controllers into Florida homes. The results were compared to a theoretical irrigation requirement developed from a daily soil water balance. However, there are many assumptions made that increased the error in the calculations. It would be interesting to explore the possibility of schedu ling irrigation based on a smaller timestep, such as an hourly soil water ba lance, to determine th e effects of the daily timestep assumptions. Also, the inputs to the so il water balance model could be refined with measurements of site conditions such as field capacity and permanent wilting point.

PAGE 203

203 APPENDIX A STATISTICAL ANALYSIS AND RESULTS FOR CHAPTER 2 /* Statistics for Chapter 2 */ /* This section was created to analyze water applied (mm) for fall 2006 assuming weeks ar e reps within the season. Treatment 6 refers to the th eoretical time-based treatment. */ options nodate nonumber center formdlim= "*" linesize=88; TITLE 'Fall 2006 Water Application' ; data Ch2.fal06; set Ch2.fall2006; julian=juldate7(date); week=week(date); proc sort; by plot week; data Ch2.fall06a; set Ch2.fal06; by plot week; if First.week then week_sum = 0; week_sum + depth; if Last.week; run; proc sort data =Ch2.fall06a; by tmt; proc glm data =Ch2.fall06a; class tmt; model week_sum = tmt rep tmt*rep; means tmt/ duncan ; run; proc mixed data =Ch2.fall06a; class tmt rep plot week; model week_sum = tmt; random rep plot week; lsmeans tmt/ adjust =tukey pdiff; run; /* This section was created to analyze water applied (mm)

PAGE 204

204 for winter 2006-2007 assuming week s are reps within the season. */ TITLE 'Winter 2006-2007 Wa ter Application' ; data Ch2.win06; set Ch2.win2006; julian=juldate7(date); week=week(date); proc sort; by plot week; data Ch2.win06a; set Ch2.win06; by plot week; if First.week then week_sum = 0; week_sum + depth; if Last.week; run; proc sort data =Ch2.win06a; by tmt; proc glm data =Ch2.win06a; class tmt; model week_sum = tmt rep tmt*rep; means tmt/ duncan ; run; proc mixed data =Ch2.win06a; class tmt rep plot week; model week_sum = tmt; random rep plot week; lsmeans tmt/ adjust =tukey pdiff; run; /* This section was created to analyze water applied (mm) for spring 2007 assuming weeks are reps within the season. */ TITLE 'Spring 2007 Water Application' ; data Ch2.spr07; set Ch2.spr2007;

PAGE 205

205 julian=juldate7(date); week=week(date); proc sort; by plot week; data Ch2.spr07a; set Ch2.spr07; by plot week; if First.week then week_sum = 0; week_sum + depth; if Last.week; run; proc sort data =Ch2.spr07a; by tmt; proc glm data =Ch2.spr07a; class tmt; model week_sum = tmt rep tmt*rep; means tmt/ duncan ; run; proc mixed data =Ch2.spr07a; class tmt rep plot week; model week_sum = tmt; random rep plot week; lsmeans tmt/ adjust =tukey pdiff; run; /* This section was created to analyze water applied (mm) for summer2007 assuming weeks are reps within the season. */ TITLE 'Summer 2007 Water Application' ; data Ch2.sum07; set Ch2.sum2007; julian=juldate7(date); week=week(date); proc sort; by plot week; data Ch2.sum07a; set Ch2.sum07; by plot week;

PAGE 206

206 if First.week then week_sum = 0; week_sum + depth; if Last.week; run; proc sort data =Ch2.sum07a; by tmt; proc glm data =Ch2.sum07a; class tmt; model week_sum = tmt rep tmt*rep; means tmt/ duncan ; run; proc mixed data =Ch2.sum07a; class tmt rep plot week; model week_sum = tmt; random rep plot week; lsmeans tmt/ adjust =tukey pdiff; run; /* This section was created to analyze water applied (mm) for fall 2007 assuming weeks ar e reps within the season. */ TITLE 'Fall 2007 Water Application' ; data Ch2.fal07; set Ch2.fall2007; julian=juldate7(date); week=week(date); proc sort; by plot week; data Ch2.fal07a; set Ch2.fal07; by plot week; if First.week then week_sum = 0; week_sum + depth; if Last.week; run; proc sort data =Ch2.fal07a; by tmt; proc glm data =Ch2.fal07a;

PAGE 207

207 class tmt; model week_sum = tmt rep tmt*rep; means tmt/ duncan ; run; proc mixed data =Ch2.fal07a; class tmt rep plot week; model week_sum = tmt; random rep plot week; lsmeans tmt/ adjust =tukey pdiff; run; /* This section was created to analyze water applied (mm) and turfgrass quality ratings for every season assuming that the length of time affecting the quality rating would be for the two weeks prior to the rating day (2wk). */ TITLE 'Chapter 2 Water Applicat ion and Turfgrass Quality' ; data Ch2.new_water_qual; set Ch2.water_qual; proc sort data =Ch2.new_water_qual; by season; proc corr; by season; var two_wk tq; run; proc glm data =Ch2.new_water_qual; by season; class tmt rep; model two_wk tq = tmt rep tmt*rep; means tmt/ duncan ; run; proc mixed data =Ch2.new_water_qual; by season; class season tmt rep plot; model two_wk = tmt; random season rep plot; lsmeans tmt/ adjust =tukey pdiff; run;

PAGE 208

208 /* Chapter 2 Campus Plots Analysis */ options nodate nonumber center; data Campus_water; input date $ tmt $ rep $ depth; cards ; REMOVED DUE TO LENGTH ; proc sort; by tmt; proc glm; by tmt; class rep; model depth = rep; means rep/ duncan ; run; proc glm; class tmt rep; model depth = tmt rep; means tmt/ duncan ; run; proc mixed; class tmt date rep; model depth = tmt; random date rep; lsmeans tmt/ adjust =tukey pdiff; run;

PAGE 209

209 APPENDIX B TURFGRASS QUALITY RATINGS Table B-1. Turfgrass quality ra tings by the graduate research assistant for the summ er 2006 season. Plot July 10 August 10 1 7 6 2 7 6 3 6 6 4 6 6 5 6 6 6 6 6 7 5 6 8 7 6 9 5 6 10 5 5 11 5 5 12 5 5 13 5 5 14 7 6 15 6 5 16 6 5 17 5 5 18 5 5 19 7 5 20 6 5

PAGE 210

210 Table B-2. Turfgrass quality ratings by the graduate research a ssistant for the fall 2006 season. Plot September 12 October 12 November 2 November 16 1 4 5 5 5 2 5 4 5 5 3 5 4 4 5 4 5 4 6 7 5 5 5 4 5 6 5 5 5 5 7 6 5 4 6 8 5 6 4 7 9 5 4 5 6 10 4 5 5 5 11 5 5 5 5 12 4 5 5 7 13 5 5 5 5 14 5 4 5 7 15 4 5 5 6 16 5 5 5 5 17 3 6 5 6 18 4 5 4 5 19 5 5 5 7 20 5 4 5 5

PAGE 211

211 Table B-3. Turfgrass quality ra tings by the graduate research assistant for the winter 2006-2007 season. Plot December 12 December 20 February 1 1 6 5 6 2 6 8 6 3 5 6 6 4 7 7 6 5 5 6 5 6 5 7 5 7 6 6 5 8 7 6 7 9 5 4 5 10 6 5 4 11 7 6 5 12 7 5 6 13 5 6 5 14 6 6 5 15 7 6 6 16 6 5 7 17 6 5 6 18 6 6 6 19 8 6 7 20 6 6 6

PAGE 212

212 Table B-4. Turfgrass quality ra tings by the graduate research assistant for the spring 2007 season. Plot April 2 May 2 May 29 1 7 8 5 2 7 7 5 3 6 5 5 4 7 8 6 5 6 8 5 6 6 7 5 7 7 8 5 8 7 8 5 9 7 7 5 10 6 7 5 11 6 7 5 12 8 8 5 13 7 7 5 14 6 8 6 15 7 7 6 16 7 7 5 17 6 7 5 18 7 7 5 19 8 8 5 20 7 7 5

PAGE 213

213 Table B-5. Turfgrass quality ra tings by the graduate research assistant for the summer 2007 season. Plot June 26 July 27 August 28 1 5 7 5 2 5 7 6 3 4 6 5 4 7 8 6 5 5 7 5 6 5 7 5 7 5 8 7 8 6 8 7 9 5 7 5 10 5 7 5 11 6 8 6 12 5 8 7 13 5 7 5 14 5 8 6 15 5 8 7 16 5 7 7 17 5 7 7 18 5 7 6 19 6 8 7 20 5 7 7

PAGE 214

214 Table B-6. Turfgrass quality ratings by the graduate research a ssistant for the fall 2007 season. Plot September 27 October 30 December 4 1 5 7 8 2 8 7 8 3 5 6 8 4 8 7 6 5 6 6 5 6 7 7 6 7 6 7 7 8 8 7 7 9 6 6 4 10 6 8 8 11 6 7 8 12 8 8 8 13 5 7 4 14 6 7 8 15 6 6 6 16 8 7 8 17 6 8 5 18 6 7 8 19 8 8 8 20 7 8 8

PAGE 215

215 Table B-7. Turfgrass qualit y ratings by Master Gardener 1 over the study period. Plot Dec 20, 2006 Feb 1, 2007 Apr 2, 2007 Jun 6, 2007 Aug 3, 2007 Oct 30, 2007 1 7 NA 6 6 7 7 2 6 NA 6 6 7 7 3 6 NA 6 6 7 7 4 7 NA 6 7 7 7 5 6 NA 5 7 7 7 6 5 NA 5 6 7 7 7 6 NA 6 7 7 7 8 6 NA 7 7 7 7 9 6 NA 6 7 7 7 10 7 NA 6 7 7 7 11 7 NA 6 7 7 7 12 6 NA 7 7 7 7 13 6 NA 7 7 7 7 14 6 NA 7 7 7 7 15 6 NA 7 7 7 7 16 6 NA 7 7 7 7 17 6 NA 7 7 7 7 18 6 NA 7 7 7 7 19 6 NA 7 7 7 7 20 6 NA 7 7 7 7 *NA occurred when rater was not available to evaluate turfgrass quality.

PAGE 216

216 Table B-8. Turfgrass qualit y ratings by Master Gardener 2 over the study period. Plot Dec 20, 2006 Feb 1, 2007 Apr 2, 2007 Jun 6, 2007 Aug 3, 2007 Oct 30, 2007 1 5 6 NA 7 8 8 2 7 7 NA 7 8 7 3 7 6 NA 6 8 8 4 7 7 NA 8 8 8 5 6 6 NA 7 8 8 6 6 6 NA 7 8 8 7 7 7 NA 7 8 8 8 7 7 NA 7 8 8 9 6 6 NA 7 8 7 10 7 6 NA 7 8 7 11 7 7 NA 7 8 7 12 6 6 NA 7 8 8 13 7 6 NA 6 8 8 14 6 6 NA 7 8 8 15 7 7 NA 7 8 7 16 6 7 NA 8 8 7 17 6 7 NA 7 8 8 18 7 7 NA 7 8 7 19 7 7 NA 8 8 8 20 7 6 NA 7 8 8 *NA occurred when rater was not available to evaluate turfgrass quality.

PAGE 217

217 Table B-9. Turfgrass qualit y ratings by Master Gardener 3 over the study period. Plot Dec 20, 2006 Feb 1, 2007 Apr 2, 2007 Jun 6, 2007 Aug 3, 2007 Oct 30, 2007 1 8 7 7 7 8 8 2 8 7 7 8 8 7 3 7 6 7 7 8 6 4 8 7 7 8 8 7 5 8 6 7 8 8 8 6 7 6 6 8 8 7 7 8 7 6 7 8 8 8 8 7 7 8 8 7 9 8 6 7 8 8 7 10 8 6 6 8 8 8 11 8 7 6 8 8 8 12 8 6 7 8 8 8 13 8 7 7 7 8 8 14 7 6 7 7 8 8 15 7 7 7 8 8 8 16 8 6 7 7 8 7 17 8 7 7 8 8 8 18 8 6 7 8 8 8 19 8 7 7 8 8 8 20 8 6 7 8 8 8 *NA occurred when rater was not available to evaluate turfgrass quality.

PAGE 218

218 Table B-10. Turfgrass qualit y ratings by Master Gardener 4 over the study period. Plot Dec 20, 2006 Feb 1, 2007 Apr 2, 2007 Jun 6, 2007 Aug 3, 2007 Oct 30, 2007 1 6 NA NA NA 7 NA 2 7 NA NA NA 6 NA 3 6 NA NA NA 6 NA 4 8 NA NA NA 6 NA 5 7 NA NA NA 6 NA 6 7 NA NA NA 7 NA 7 7 NA NA NA 7 NA 8 8 NA NA NA 7 NA 9 5 NA NA NA 7 NA 10 6 NA NA NA 7 NA 11 7 NA NA NA 7 NA 12 8 NA NA NA 8 NA 13 7 NA NA NA 7 NA 14 7 NA NA NA 7 NA 15 7 NA NA NA 7 NA 16 8 NA NA NA 7 NA 17 8 NA NA NA 6 NA 18 7 NA NA NA 6 NA 19 8 NA NA NA 7 NA 20 6 NA NA NA 6 NA *NA occurred when rater was not available to evaluate turfgrass quality.

PAGE 219

219 Table B-11. Turfgrass qualit y ratings by Master Gardener 5 over the study period. Plot Dec 20, 2006 Feb 1, 2007 Apr 2, 2007 Jun 6, 2007 Aug 3, 2007 Oct 30, 2007 1 7 NA 7 7 8 8 2 7 NA 7 7 8 8 3 7 NA 7 7 7 8 4 8 NA 8 7 8 7 5 7 NA 6 7 8 8 6 7 NA 7 7 8 8 7 7 NA 7 7 8 8 8 8 NA 7 7 8 8 9 7 NA 7 6 8 8 10 7 NA 7 7 8 8 11 7 NA 7 7 8 8 12 7 NA 7 7 8 8 13 6 NA 7 7 8 8 14 7 NA 6 7 8 8 15 7 NA 7 7 8 7 16 8 NA 7 7 8 8 17 7 NA 7 7 8 8 18 7 NA 7 7 8 8 19 8 NA 7 7 8 8 20 7 NA 6 7 8 8 *NA occurred when rater was not available to evaluate turfgrass quality.

PAGE 220

220 Table B-12. Turfgrass qualit y ratings by Master Gardener 6 over the study period. Plot Dec 20, 2006 Feb 1, 2007 Apr 2, 2007 Jun 6, 2007 Aug 3, 2007 Oct 30, 2007 1 NA 7 7 NA NA 7 2 NA 8 7 NA NA 7 3 NA 6 7 NA NA 7 4 NA 7 7 NA NA 7 5 NA 6 7 NA NA 7 6 NA 7 7 NA NA 7 7 NA 7 7 NA NA 7 8 NA 7 7 NA NA 7 9 NA 7 7 NA NA 7 10 NA 7 6 NA NA 7 11 NA 7 7 NA NA 7 12 NA 7 7 NA NA 7 13 NA 7 7 NA NA 7 14 NA 7 7 NA NA 7 15 NA 7 7 NA NA 7 16 NA 7 7 NA NA 7 17 NA 7 7 NA NA 7 18 NA 7 7 NA NA 7 19 NA 7 7 NA NA 7 20 NA 6 6 NA NA 7 *NA occurred when rater was not available to evaluate turfgrass quality.

PAGE 221

221 APPENDIX C STATISTICAL ANALYSIS AND RESULTS FOR CHAPTER 3 /* Statistics for Chapter 3 of ET Controller Thesis */ options nodate nonumber center formdlim= "*" linesize=88; /* This section compares temperat ure and extraterrest rial radiation methods through calculated ETo values using Hargreaves Equation for the Gainesville turfgrass plot s. Temperature and extraterrestrial radiation values were from th e Weathermatic controllers and the on-site weather station. */ data Ch3.gnv_tra_et; set Ch3.gnv_tra; Title 'Parameter sensitivity for Hargreav es Equation at GNV Turfgrass Plots' ; proc sort data =Ch3.gnv_tra_et; by rep; proc glm data =Ch3.gnv_tra_et; by rep; class T_meth Ra_meth; model ET = Ra_meth T_meth T_meth*Ra_meth; means T_meth/ duncan ; means Ra_meth/ duncan ; run; /* This section compares temperat ure and extraterrest rial radiation methods through calculated ETo values using Hargreaves Equation for the GCREC. Temperature and extraterrestrial radiation values were from the Weathermatic cont roller and the FAWN weather station. */ data Ch3.gcrec_tra_et; set Ch3.gcrec_tra; Title 'Parameter sensitivity for Hargreaves Equation at GCREC' ; proc glm data =Ch3.gcrec_tra_et; class T_meth Ra_meth; model ET = Ra_meth T_meth T_meth*Ra_meth; means T_meth/ duncan ; means Ra_meth/ duncan ; run;

PAGE 222

222 /* This section will analyze differences in maximum and minimum temperatures for the Weatherma tic replications at the Gainesville turfgrass plots compared to the temperature measured from the on-site weather station. */ data Ch3.gnv_t; set Ch3.gnv_temp; TITLE 'Chapter 3 Temperature comparisons for the GNV Weathermatic Replications' ; proc glm data =Ch3.gnv_t; class season tmt; model tmax tmin = season tmt; means tmt/ duncan ; run; /* This section will analyze differences in maximum and minimum temperatures for the Weathermatic at the GCREC compared to the temperature measured from the on-site weather station. */ data Ch3.gcrec_t; set Ch3.gcrec_temp; TITLE 'Chapter 3 Temperature comparisons for the GCREC Weathermatic' ; proc glm data =Ch3.gcrec_t; class season tmt; model tmax tmin = season tmt; means tmt/ duncan ; run; /* This section was created to analyze ETo calculated/collected from the controller replicatio ns at the Gainesville turfgrass plots. */ data Ch3.campus_et; set Ch3.campus; TITLE 'Chapter 3 ET for Campus Controllers' ;

PAGE 223

223 proc sort data =Ch3.campus_et; by tmt; proc glm data =Ch3.campus_et; by tmt; class rep; model et = rep; means rep/ duncan ; run; proc glm data =Ch3.campus_et; class season tmt rep; model et = season tmt rep; means tmt/ duncan ; run; proc mixed data =Ch3.campus_et; class tmt month year season rep; model et = tmt; random month year season rep; lsmeans tmt/ adjust =tukey pdiff; run; /* This section was created to analyze ETo calculated/collected from the controllers at the GCREC. */ data Ch3.GCREC_et; set Ch3.GCREC; TITLE 'Chapter 3 ET for GCREC Controllers' ; proc glm data =Ch3.GCREC_et; class season tmt; model et = season tmt; means tmt/ duncan ; run; proc mixed data =Ch3.GCREC_et; class tmt month year season; model et = tmt; random month year season; lsmeans tmt/ adjust =tukey pdiff; run; /*

PAGE 224

224 This section will analyze the ET Wa ter rolling 7 day ETo totals compared to the on-site weather station a nd the GNV airport weat her station for the ET Water replications at the Gainesville turfgrass plots. */ data Ch3.gnv_etw_wket; set Ch3.gnv_etw; TITLE 'Chapter 3 ET Water 7 day rolling ETo for GNV Replications' ; proc glm data =Ch3.gnv_etw_wket; class season tmt; model wk_et = season tmt; means tmt/ duncan ; run; proc mixed data =Ch3.gnv_etw_wket; class tmt season; model wk_et = tmt; random season; lsmeans tmt/ adjust =tukey pdiff; run; /* This section will analyze the ET Wa ter rolling 7 day ETo totals compared to the other ET controllers where data was available for the controllers at the Gainesville turfgrass plots. */ data Ch3.gnv_7day_wket; set Ch3.gnv_7day; TITLE 'Chapter 3 7 day rolling ETo for GNV Controllers' ; proc glm data =Ch3.gnv_7day_wket; class tmt; model wk_et = tmt; means tmt/ duncan ; run; proc mixed data =Ch3.gnv_7day_wket; class tmt date; model wk_et = tmt; random date; lsmeans tmt/ adjust =tukey pdiff; run;

PAGE 225

225 LIST OF REFERENCES Allen, R.G., Pereira, L.S., Raes, D., and Sm ith M. (1998). Crop evapotranspiration: Guidelines for computing crop requirements. Irrigation and Drainage Paper No. 56, FAO, Rome, Italy. Allen, R.G., Walter, I.A., Elliot, R., Howell, T ., Itenfisu, D., and Jensen, M. (eds). (2005). The ASCE standardized reference evapotranspira tion equation. American Society of Civil Engineers Environmental and Water Resource Institute (ASCE-EWRI). 59 pp. Available at: http://www.kimberly.uidaho.edu/water/asceewri/ascestzdetmain2005.pdf Accessed 2 Decem ber, 2005. American Society of Civil E ngineers [ASCE]. (1978). Descri bing irrigation efficiency and uniformity. Journal of Irrigation an d Drainage Engineering 104(1): 35-41. Aquacraft, Inc. (2002). Performance evaluati on of WeatherTRAK irri gation controllers in Colorado. Aquacraft, Inc., Boulder, CO. Available at: www.aquacraft.com Accessed 21 October, 2005. Aquacraft, Inc. (2003). Report on perform ance of ET based irrigation controller: Analysis of operation of WeatherTRAK controller in fiel d conditions during 2002. Aquacraft, Inc., Boulder, CO. Available at: www.aquacraft.com Accessed 21 October, 2005. Bam ezai, A. (2004). LADWP weather-based irrigation cont roller pilot study. Western Policy Research, Santa Monica, CA. Available at: http://www.cuwcc.org/uploads/product/LAD W P-Irrigation-Controller-Pilot-Study.pdf Accessed 31 October, 2005. Black, R.J., and Ruppert, K. (1998). Your Flor ida Landscape: A Complete Guide to Planting and Maintenance. UF-IFAS publication. Univ ersity Press of Florida. Gainesville, FL. Burt, C.M., Clemmens, A.J., and Strelkoff, K.H. (1997). Irrigation perf ormance measurements: efficiency and uniformity. Journal of Irrigation and Drainage Engineering 123(6): 423442. Buss, E.A. (1993). Southern chinch bug manageme nt on St. Augustinegrass. ENY-325, Institute of Food and Agricultural Sciences. University of Florida, Gainesville, FL. Available at: http://edis.ifas.ufl.edu Accessed 15 March, 2008. Cardenas-Lailhacar, B., and Dukes, M.D. (2008). Expanding disk rain sensor perform ance and potential water savings. Journal of Irrigation an d Drainage Engineering 134(1): 67-73. Carriker, R.R. (2000). Florida's water: supply, use, and public policy. Department of Food and Resource Economics, Report FE 207, Univer sity of Florida, Gainesville, FL. Dewey, C. (2003). Sensors at work. Irrigation and green industry, September 2003. Irrigation Association, Falls Church, VA. Available at: http://www.igin.com/Irrigation/0903sensors.html Accessed 7 January, 2006.

PAGE 226

226 Diamond, R.A. (2003). Project review of Irvine ET controller residential runoff reduction study. Irvine Ranch Water District, Irvine, CA. Available at: http://www.irrigation.org/swat/im ages/irvine_runoff_reduction.pdf Accessed 1 November, 2005. Dukes, M.D. and Haman, D.Z. (2002a). Operation of residential irrigation controllers. CIR1421, Institute of Food and Agricultural Sciences, Un iversity of Florida, Gainesville, FL. Available at: http://edis.ifas.ufl.edu Accessed 7 January, 2006. Dukes, M.D. and Haman, D.Z. (2002b). Resident ial irrigation system rainfall shutoff devices. ABE325, Institute of Food and Agricultural Scie nces, University of Florida, Gainesville, FL. Available at: http://edis.ifas.ufl.edu Accessed 7 January, 2006. Fangmeier, D.D., Elliot, W.J., Workman, S. R., Huffman, R.L., and Schwab, G.O. (2006). Soil and Water Conservation Engineering. Fifth edition. Thomson Delmar Learning, Clifton Park, NY. Florida Department of Environmental Protecti on [FDEP]. (2002). Flor ida Water Conservation Initiative. Section 62-40.412(1), F.A.C. State of Florida, Tallahass ee, FL. Available at: http://www.dep.state.fl.us Accessed 10 January, 2008. Florida Statutes, Part VI, Chapter 373.62. n.d. W ater Conservation: automatic sprinkler systems. State of Florida, Tallahassee, FL. Available at: http://www.floridadep.org/water/stormwater/npdes/docs/ch373.pdf Accessed 16 February, 2006. Florida Statutes. (2001). Changes to Chapter 373. State of Florid a, Tallahassee, FL. Available at: http://www.dep.state.fl.us/cm p/federal/files/373ana01.pdf Accessed 16 February, 2006. Haley, M.B., Dukes, M.D., and Miller, G.L. ( 2007). Residential irrigatio n water use in Central Florida. Journal of Irrigation and Drainage Engineering 133(5): 427-434. Haman, D.Z., Clark, G.A., Smajstrla, A.G. ( 1989). Irrigation of lawns and gardens. CIR825, Institute of Food and Agricultural Sciences. University of Florida, Gainesville, FL. Available at: http://edis.ifas.ufl.edu Accessed 31 October, 2005. Hargreaves, G.H. and Sa mani, Z.A. (1982) Estimating potential evapotranspiration. Journal of Irrigation and Drainage Engineering, 108(3):223-230. Harivandi, M.A. (1984). Turfgr ass irrigation efficiency. California Turfgrass Culture, 34(4):2123. Hunt, T., Lessick, D., Berg, J., and Wiedmann, J. (2001). Residential weather-based irrigation scheduling: Evidence from the Irvine ET Contro ller study. Irvine Ra nch Water District, Irvine, CA. Available at: http://www.irrigation.org/swat/im ages/irvine.pdf Accessed 30 October, 2005.

PAGE 227

227 Irmak, S. and Haman, D.Z. (2003). Evapotrans piration: Potential or reference? ABE 343, Institute of Food and Agricultural Sciences. University of Florida, Gainesville, FL. Available at: http://edis.ifas.ufl.edu Accessed 31 October, 2005. Irrigation Association [IA]. (2005). Landscape irrigation scheduling and water m anagement. Irrigation Association Wa ter Management Committee. Falls Church, VA. IA. (2006a). Smart Water Application Tec hnology (SWAT) Performance Report: ET Water. Irrigation Association, Falls Church, VA. Available at: www.irrigation.org Accessed 26 January, 2006. IA. (2006b). Smart Water Application Technology (SWAT) Performance Report: WeatherTRAK. Irrigation Association, Falls Church, VA. Available at: www.irrigation.org Accessed 26 January, 2006. IA. (2006c). Sm art Water A pplication Technology (SWAT) Turf and Landscape Irrigation Equipment Testing Protocol for Clim atologically Based Controllers: 7th Draft. Irrigation Association, Falls Church, VA. Available at: www.irrigation.org Accessed 15 March, 2007. IA. (2007). Sm art Water App lication Technology (SWAT) Perf ormance Report: Weathermatic. Irrigation Association, Falls Church, VA. Available at: www.irrigation.org Accessed 26 January, 2006. Jenson, M.E., Burman, R.D., and Allen, R.G. (1990). Evapotranspiration and irrigation water requirements. ASCE manuals and reports on engineering practices No. 70. American Society of Civil Engineers. New York, NY. Jia, X., Dukes, M.D., Jacobs, J.M., and Haley, M. (2007). Impact of weather station fetch distance on reference evapotranspiration calculations. American Society of Civil Engineers Environmental and Water Resources Institut e Conference Paper. Reston, VA: ASCEEWRI. Mayer, P.W., DeOreo, W.B., Opitz, E.M., Kiefer, J.C., Davis, W.Y., Dziegielewski, B., and Nelson, J.O. (1999). Residential end uses of water. American Water Works Association Research Foundation. Denver, CO. Metropolitan Water District of Southern California [MWDSC]. (2004). Weather based controller bench test report. MWDS C, Los Angeles, CA. Available at: http://www.cuwcc.org/irri gation_controllers/MWDWeat her-Based-Controller.pdf Accessed 1 November, 2005. Natural Resources Conservation Service [NRCS]. (2006). Florida supplements to National Engineering Handbook: irrigation guide. Flor ida NRCS, Gainesville, FL. Available at: http://www.fl.nrcs.usda.gov/technical/irrigation.htm l. Accessed 12 January, 2008.

PAGE 228

228 National Oceanic and Atmospheric Administ ration [NOAA]. (2005). Monthly precipitation 1975-2005 for Parrish, FL. United States Department of Commerce National Climatic Data Center, Asheville, NC. Available at: http://cdo.ncdc.noaa.gov/pls/plclimprod/ poem ain.cdobystn?dataset=DS3220&StnList=086 880NNNNN Accessed 1 November, 2006. Pittenger, D.R., Shaw, D.A., and Richie, W.E. (2004). Evaluation of weather-sensing landscape irrigation controllers. Riverside, CA: Univ ersity of California Cooperative Extension. Riley, M. (2005). The cutting edge of residential smart irrigation technology. California Landscaping. July/August pp 19-26. Shearman. R. C. and Morris, K. N. (1998). NTEP Turfgrass Evaluation Workbook. NTEP Turfgrass Evaluation Workshop, October 17, 1998, Beltsville, MD. Solley, W.B., Pierce, R.R., and Perlman, H.A. (1998). Estimated use of water in the United States in 1995. United States Geolog ical Survey Circular 1200. 78 p. Solomon, K.H., Kissinger, J. A., Farrens, G. P., and Borneman, J. (2006). Performance and water conservation potential of multi-stream, multi-trajectory rotating sprinklers for landscape irrigation. Applied Engineering in Agriculture 23(2): 153-163. Southwest Florida Water Management Distri ct [SWFWMD]. (2006). Regional water supply plan. SWFWMD, Brooksville, FL. Available at: www.watermatters.org Accessed 30 June, 2007. Tichenor, J., Dukes, M.D., and Trenholm, L.E. (2003). Using the irri gation controller for a better lawn on less water. ENH978, Institute of Food and Agricultural Sciences. University of Florida, Gainesville, FL. Available at: http://edis.ifas.ufl.edu Accessed 31 October, 2005. Trajkovic, S. (2007). Hargreav es versus penman-monteith under humid conditions. Journal of Irrigation and Drainage Engineering 133(1): 38-42. United States Census Bureau [USCB]. (2005). Population estimates. US CB, Washington, DC. Available at: http://www.census.gov/popest/estimates.php Accessed 19 January, 2006. USCB. (2007). C40 Residential Housing Units by State. USCB, W ashington, DC. Available at: http://www.census.gov/const/C40/Table2/tb2u2005.txt Accessed 13 January, 2008. United States Departm ent of In terior [USDOI]. (2004). Reclamation: Managing water in the west. USDOI, Washington, DC. Available at: http://www.usbr.gov/waterconservation/docs/ET%20controller%20report.pdf Accessed: 1 Nove mber, 2005.

PAGE 229

229 United States Environmental Protection Agency [USEPA]. (2004). Market conditions report: High-potential water using produc ts (Draft). USEPA, Washington, DC. Available at: http://www.epa.gov/owm/water-efficien cy/pdf/m arket_conditions_7-5-05.pdf Accessed 2 November, 2005. Wade, G. and Waltz, C. (2004). Emerging and existing technologies for landscape water conservation (Draft). University of Georgias Center for Urban Agriculture, Athens, GA. Available at: http://apps.caes.uga.edu/urbanag/BookBMPS/Chapter6Technologies.pdf Accessed 31 October, 2005. Wong, F., Harivandi, M.A., and Hartin, J. ( 2005). UC IPM Pest Managem ent Guidelines: Turfgrass. University of California Agriculture and Natural Res ources Publication 3365T, Davis, CA. Available at: http://www.ipm.ucdavis.edu/PMG/r785100311.html Ac cessed 15 March, 2008.

PAGE 230

BIOGRAPHICAL SKETCH Stacia L. Davis began her educ a tional career at Colfax Elementary School in Fairmont, WV. Colfax Elementary was a modest school, serving approximately 60 children, first through sixth grade, in a split classroom structure wher e her education was personalized and creativity was high. She was uprooted from Colfax Elemen tary toward the end of sixth grade to attend Wendover Middle School in Greensburg, PA. Th is change sparked a need to define her strengths and abilities while choo sing advanced courses in all subject areas. Hempfield Area Senior High School became an im portant stepping stone in determining her place in academia because Stacia chose to concentrate on the advanced math and sciences offered. It was clear to everyone, including Stacia, that engineering was her niche. Ms. Davis chose to attend the University of Pittsburgh because of their diverse engineering program. Freshman engineering brought about many opportunities; some opportunities included presenting on wetlands at a sustainability confer ence and gaining tools st ill inadvertently used (such as HTML and C programming). Her next 3 years were used studying civil engineering with a concentration in environm ental engineering while serving as president of Chi Epsilon and treasurer of the National Society of Collegiate Sc holars. She also worked in the engineering field for companies such as CDM, Allegheny Energy, and Rhea Engineers and obtained her Engineer-In-Training status before comp leting her Bachelor of Science. Ms. Davis narrowed her focus when leaving PITT and chose to pursue a higher degree in the land and water resources engineering section of the Agricultu ral and Biological Engineering department at the University of Florida. Here, she researched water conservation through residential irrigation focusing on ET-based irrigation controllers a nd turfgrass. She enjoyed her time spent in this department and will continue on with her academic goals by seeking a doctorate in the same area.