Stream-gaging procedure


Material Information

Stream-gaging procedure a manual describing methods and practices of the Geological Survey
Series Title:
Water-supply paper (Washington, D.C.) ;
Physical Description:
xvi, 245 p. : ill., plates, tables (1 folded) forms (1 folded) diagrs. (some folded) ; 23 cm.
Corbett, Don Melvin, 1900-
U.S. G.P.O.
Place of Publication:
Publication Date:


Subjects / Keywords:
Stream measurements -- Handbooks, manuals, etc -- United States   ( lcsh )
Gaging -- Handbooks, manuals, etc -- United States   ( lcsh )
federal government publication   ( marcgt )
handbook   ( marcgt )
non-fiction   ( marcgt )


Statement of Responsibility:
by Don M. Corbett and others.
General Note:
Reprint. Originally published: 1943.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 030834338
oclc - 45906815
System ID:

Full Text

To tou





I It4

mvl AK

itr t. 1 t "Y'


P 4-

STEWART L. UDALL, Secretary GEOLOGICAL SURVEY Thomas B. Nolan, Director Water-Supply Paper 888









rch 3,

Reprinted 1962


GOVERNMENT PRINTING OFFICE WASHINGTON : 1962 For sale by the Superintendent of Documents, U.S. Government Printing Office
Washington 25, D.C. Price $2.00 (paper cover)

Digitized by the Internet Archive
in 2013



Foreword, by Nathan C. Grover ------------------------------------- xiii
Introduction ------------------------------------------------------ 1
Administration, personnel, and acknowledgments. ---------------------- 2
Organization for water-resources investigations ------------------------ 3
Water Resources Branch ---------------------------------------- 3
Division of Surface Water ----------------------------------- 3
Division of Ground Water --------------------------------- 4
Division of Quality of Water -------------------------------- 4
Division of Power Resources -------------------------------- 4
Division of Water Utilization -------------------------------- 4
Administration and operation ----------------------------------- 4
Personnel ----------------------------------------------------- 5
Recruiting of personnel ------------------------------------ 5
Training t- personnel -------------------------------------- 6
General procedure ------------------------------------------------- 7
Records of stage --------------------------------------------------- 8
Methods of obtaining gage-height record ------------------------- 9
Observed gage heights -------------------------------------- 9
Recorded gage heights -------------------------------------- 10
Observers ----------------------------------------------------- 10
Availability ----------------------------------------------- 10
Trustworthiness ------------------------------------------- 10
Duties --------------------------------------------------- 10
Compensation --------------------------------------------- 13
Measurement of discharge ------------------------------------------- 13
Current-meter measurements ------------------------------------- 13
Types of discharge measurements ---------------------------- 14
Wading measurement ---------------------------------- 14
Cableway measurement -------------------------------- 16
Bridge measurement ------------------------------------ 19
Boat measurement ------------------------------------- 21
Measurement through ice cover -------------------------- 24
General precautions ----------------------------------------- 28
Suspension of current meter and measurement of depth -------- 28
Rod suspension ---------------------------------------- 29
Hand-line suspension ----------------------------------- 30
Reel suspension --------------------------------------- 31
Single-conductor system with hand-line or reel suspensions --- 31 Electrical hook-up with dry cell eliminated --------------- 33
Measurement of velocity ----------------------------------- 34
Vertical velocity-curve method -------------------------- 35
Two-point method ---------------------- --------------- 37
Six-tenths-depth method -------------------------------- 38
Two-tenths-depth method ------------------------------ 39


Measurement of discharge-Continued.
Current-meter measurements-Continued.
Measurement of velocity-Continued. Page
Three-point method ------------------------------------ 41
Subsurface method ------------------------------------ 41
Integration method ------------------------------------ 42
One-point continuous method --------------------------- 42
Measurements of streanis that are deep and swift -------------- 43
Measurement of depth ---------------------------------- 44
Placement of current meter ----------------------------- 52
Procedure -------------------------------------------- 54
Tag-line method --------------------------------------- 56
Stay-line equipi-nent ----------------------------------- 57
Horizontal angles -------------------------------------- 58
Recording of data ----------------------------------------- 59
Form 9-27 ------------------------------------------- 59
Form 9-275a ------------------------------------------ 61
Form 9 275b ------------------------------------------ 62
Form 9-275c ------------------------------------------ 62
Form 9-275d --------------------------------- ---------- 62
Factors affecting the accuracy of measurements --------------- 65
Use and care of equipment ------------------------------ 65
Selection of measuring section --------------------------- 66
Spacing of verticals ------------------------------------ 68
Measurement of depth --------------------------------- 69
Ice in measuring section -------------------------------- 69
Freezing temperatures ---------------------------------- 70
Turbulent flow and turbulence -------------------------- 70
Angle of current --------------------------------------- 71
Effects of piers, piling, and eddies ----------------------- 73
Vertical and horizontal motion of meter ------------------ 75
Insufficient weight on current meter --------------------- 75
Wind ------------------------------------------------ 75
Drift and aquatic growth in channel --------------------- 76
Mean gage height for a discharge measurement -------------------- 76
Special methods ----------------------------------------------- 80
Weirs ---------------------------------------------------- 80
Slope area ------------------------------------------------ 81
Floats ---------------------------------------------------- 84
Venturi meter --------------------------------------------- 85
Marshall measuring flum, ------------------------------------- 86
Submerged orifice ------------------------------------------ 87
Chemical methods ----------------------------------------- 88
Salt velocity ------------------------------------------ 88
Salt dilution ------------------------------------------ 89
Contracted openings --------------------------------------- 90
Velocity-area method -------------------------------------- 91
Computation of discharge over dams ------------------------- 93
Calibration of water wheels and valves, gates, and sluices- - - - 94 Volumetric method ---------------------------------------- 96
Determination of discharge by adjustments for gain or loss in
storage ------------------------------------------------- 97
Laboratory models ----------------------------------------- 97


Measurement of discharge-Continued. Pap
Determination of flood discharge -------------------------------- 98
Extension of rating curve ----------------------------------- 98
Extension on logarithmic paper -------------------------- 99
Extension by d-Q -------------------------------------- 99
Extension bystudies of areas and velocity ---------------- 99
Extension of discharge as a function of A d -------------- 100
Slope-area method ----------------------------------------- 100
Selection of reach -------------------------------------- 100
Survey ----------------------------------------------- 101
Office procedure --------------------------------------- 101
Assumptions and computations -------------------------- 102
Contracted-opening method --------------------------------- 102
Selection of the contracted opening ---------------------- 103
Survey ----------------------------------------------- 103
Office procedure --------------------------------------- 103
Assumptions and computations -------------------------- 104
Flow over dam -------------------------------------------- 104
Selection of dam --------------------------------------- 104
Survey ----------------------------------------------- 105
Office procedure --------------------------------------- 106
Assumptions and computations -------------------------- 106
Other methods -------------------------------------------- 107
Critical-depth method ---------------------------------- 107
Timing of flood drift ----------------------------------- 108
Flood formulas ---------------------------------------- 108
Flow characteristics in open channels -------------------------------- 109
Stage-discharge relation ---------------------------------------- 109
Permanent controls ---------------------------------------- ill
Low-water controls ---------------------------------------- 112
Sensitiveness of low-water controls ----------------------- 115
Submergence of low-water controls ----------------------- 115
Artificial controls ------------------------------------------ 117
Design of artificial controls ----------------------------- 117
High-water controls ---------------------------------------- 118
Dams ------------------------------------------------ 120
Contractions ------------------------------------------ 121
Bends in channel -------------------------------------- 122
Channel controls for high stages ------------------------- 122
Shifting controls ------------------------------------------- 125
Effect of scour ---------------------------------------- 126
Effect of fill ------------------------------------------- 126
Shifts in the channel above the cont rol ------------------- 127
Overflow --------------------------------------------- 127
Effect of ice on the stage-discharge relation --------------- 128
Vegetal and aquatic growths ---------------------------- 130
Slope-stage-discharge relations at gaging stations affected by variable
slopes ------------------------------------------------------ 130
Variable slopes caused by backwater ------------------------- 132
Velocity-head increment to slope ------------------------ 135
Relation between discharge and fall ---------------------- 135
Curves of relation ------------------------------------- 137
Constant-fall stage-discharge relation ---------------- 139


Flow characteristics in open channels-Continued.
Slope-stage-discharge relations-Continued.
Variable slopes caused by backwater-Continued.
Curves of relation-Continued. Page
Normal-fall stage-discharge relation ------------------ 139
Discharge-fall relation --------------------------- -- 139
Effects of channel conditions ------------------------ 140
Effects of difference in datum ----------------------- 140
Effects of fluid turbulence -------------------------- 140
Practical application of the curves of relation 141
Tennessee River at Chattanooga, Tenn --------------- 141
Tennessee River at Guntersville, Ala ----------------- 143
Ohio River at Metropolis, III ------------------------ 143
Kootenai River near Copeland, Idaho ---------------- 145
Determination of discharge ----------------------------- 146
Arithmetical method ------------------------------- 146
Graphical method --------------------------------- 147
Development of discharge hydrograph -------------------- 148
Normal-fall method ------------------------------------ 149
Stage-ratio method ------------------------------------- 150
Selection of gaging stations on streams affected by variable
slope ----------------------------------------------- 151
Length of reach ----------------------------------- 151
Type of reach ------------------------------------- 151
Selection of reach ---------------------------------- 151
Position of gages ----------------------------------- 152
Variable slopes caused by changing discharge ------------------ 152
Adjustments using slope as a factor ---------------------- 152
Velocity of the flood wave -------------------------- 154
Adjustments using channel storage as a factor ------------- 156
Return of overbank flow -------------------------------- 157
116sum6 of equations ----------------------------------- 157
Gaging stations with two gages ---------------------- 157
Gaging stations with one gage ----------------------- 157
Adjustment of discharge measurements ------------------- 158
Jones method ------------------------------------- 159
Wiggins method ----------------------------------- 160
Boyer method ------------------------------------- 161
Lewis method ------------------------------------- 162
Determination of discharge ----------------------------- 163
Jones method ------------------------------------- 164
Wiggins method ----------------------------------- 164
Boyer method ------------------------------------- 164
Lewis method ------------------------------------- 165
Development of discharge, hydrograph -------------------- 165
Variable slopes caused by backwater in conjunction with changing
discharge ------------------------------------------------ 166
Instruments and miscellaneous equipment ----------------------------- 167
Current meters ------------------------------------------------ 168
General requirements in design of current meters -------------- 169
Types of current meters ------------------------------------ 170
Price -current meter ------------------------------------ 171
Price acoustic current meter ---------------------------- .171
Small Price current meter ------------------------------- 172


Instruments and miscellaneot's equipment -Continued.
Current meters-Continued. Jlttge
Principal parts of the type-A small Price current jrwt(-r 173
Yoke ------------------------- 175

Bucket wheel ------------- 175
Bucket-wheel hub ----- -------------------------------- 17,5
Shaft--- ------ 17 .5
Pivot ------------------------- 176
Pivot bearing ------------ 17( 1 P
Penta gear -------- 176
Contact chamber ------------ ----------------------- 176
Binding posts---- 176
Changes for special low-velocity work- 177
Assembly and adjustment of the type- A current meter- 177
Operation and care of current meters- 178
Balance with cable suspension ---------- 178
Mounting on wading rod ------------------------- ------ 179
Shaft alinement ----------------------------------------- 179
Care of pivot and bearings ------------------------------- 179
Insertion of new parts ---------------------------------- 180
Spin test ----------------------------------------------- 180
Pygmy current meter --------------------------------------- 181
Rating of current meters ------------------------------------ 182
Current-meter rating station at the National Bureau of
Standards ------------------------------------------- 182
Current-meter ratings ---------------------------------- 183
Water-stage recorders ---------------------------------------------- 184
Development of water-stage recorders- 184
Advantages ----------------------------------------------- 187
General requirements --------------------------------------- 188
Height element ----------------------------------------- 188
Timing element ---------------------------------------- 189
Recording mechanism ---------------------------------- 189
Gage-height ratio- - - -_ - ----------------------------- 190
Time scale --------------------------------------------- 190
Instrument case --------------------------------------- 190
Precautions --------------------------- : --------------------- 191
Specifications -------------------------- ---------------------- 191
Nonrecording gages -------------------------------------------- 191
Vertical-staff gage ----------------------------------------- 193
Chain gage ----------------------------------------------- 193
Inclined staff gage ----------------------------------------- 194
Hook gage ------------------------------------------------ 194
Float gage ------------------------------------------------- 194
Electric tape gage ----------------------------------------- 195
Canfield wire-weight gage ----------------------------------- 196
Type-A wire-weight gage ----------------------------------- 196
Sounding equipment ------------------------------------------- 197
Wading rods ---------------------------------------------- 198
Sliding support ---------------------------------------- 198
Meter cables ---------------------------------------------- 198
Hand cable 199


Instruments and miscellaneous equipment-Continued.
Sounding equipment-Continued.
Meter cables-Continued. rage
Direct-lay cable --------------------------------------- 199
Reverse-lay cable -------------------------------------- 199
Single-condu'ctor cable ---------------------------------- 200
Hand reels ------------------------------------------------- 200
Lee-Au hand reel -------------------------------------- 200
Morgan reel ------------------------------------------ 200
Connectors ----------------------------------------------- 201
Weights -------------------------------------------------- 201
Hangers for suspending weight and meter --------------------- 202
Cranes and reels ----------------------------------------- 203
Type-A and type-D cranes ------------------------------ 204
Types A, B, D, and E reels ----------------------------- 205
Sounding protractor ---------------------------------------- 206
Miscellaneous equipment --------------------------------------- 207
Headphones ----------------------------------------------- 207
Gaging cars ----------------------------------------------- 207
Lee-Au tag-line reel ---------------------------------------- 208
Veatch horizontal-angle coefficient indicator- ------------------ 208
Reference-mark tablet -------------------------------------- 208
Leveling equipment ---------------------------------------- 209
Timers --------------------------------------------------- 209
Use of field equipment ----------------------------------------- 209
Nonexpendable field equipment ------------------------------ 209
Equipment in stock ---------------------------------------- 210
Maintenance of gage datum ----------------------------------------- 210
Necessity for a single and permanent gage datum ------------------ 211
Checking of reference marks, reference points, and gages-----------. 212
Precautions to be taken in checking gages -------------------- 212
Procedure in checking gages --------------------------------- 213
Reference marks --------------------------------------- 215
Reference points --------------------------------------- 215
Vertical-staff gage ------------------------------------- 215
Inclined-staff gage ------------------------------------- 215
Chain gage ------------------------------------------- 216
Wire-weight gage -------------------------------------- 217
Electric tape gage ------------------------------------- 218
Float gage -------------------------------------------- 218
Hook gage -------------------------------------------- 218
Reference to mean sea level --------------------------------- 219
Routine field work ------------------------------------------------- 220
Regulations and instructions governing travel --------------------- 220
Preparation for field trip --------------------------------------- 220
Equipment for current-meter measurements. ------------------ 220
Wading measurements --------------------------------- 220
Measurements from cableways and bridges --------------- 221
Boat measurements ------------------------------------ 221
Measurements through ice cover ------------------------- 221
Supplementary articles ------------------------------------- 222
Supplies -------------------------------------------------- 222
Data concerning river-measurement stations ------------------ 223
Itinerary ------------------------------------------------- 224


Rot ije field work Coiit iii tied. Page
Wor k t o be dtone at, the st atio 11 224
Inspect ion of gages and wat er-s tage recorder, 22-1
Inspection of float and intake pipes. 226
Cleaniing of stilling well and initake- 22 7
Referencee of flood marks-. 228
Necessity for discharge ineasurenint s 228
Inspect ion of the control .228 Determination of stage of zero flow. 229j
Measurement of chain length .-229 Checking and correct ion of wire-weight gages-. 230
Contacting the observer ----------230 Collection of data for supplying missing records of stage- 231 Flood heights --------------------- --------- 232
Repairs at station - - - -- - -- - - - - -- 232
Levels ----------------------------------------------- 233
Preparation for winter operation. 233
Preparation of report of field work at station ----------------- 234
Trespassing------------------------------------------------ 235
Reports to the district office ------------------------------------- 235
Hazards of field work ------------------------------------------- 236
Accidents and poisoning------------------------------------ 236
Poison oak and poison ivy -- - --- - - - -- 237
Spotted fever - - - -- - - -- - - - -- 237
Snake bite.. -- -----------------------------237
Frostbite --------------------------------------------- 237
First-aid manual and equipment ------------------------- 237
Automobile operation- -------------------------------------- 237
Swimming ------------------------------------------------ 238
Cableways and gage wells ----------------------------------- 238
Construction ---------------------------------------------- 239
Measurements through ice ----------------------------------- 239
Bridge traffic ---------------------------------------------- 239
Floods --------------------------------------------------- 240
Index ------------------------------------------------------------ 241


PLATE 1. Stream-flow measurement station on Ant ietam Creek near
Sharpsburg, Md ----------------------------------Facing 8
2. A, Type-A wire-weight gage; B, Water-stage recorder in operation on Rahway River at Springfield, N. J ----------Facing 16
3. A, Wading measurement on Patuxent River near Burtonsville,
Md.; B, Measurement from cableway on Licking River at
Toboso, Ohio ---------------------------------- Facing 17
4. A, Bridge measurement with type-A crane and type-A reel on
Rillito Creek near Tucson, Ariz.; B, Heavy crane and hand
reel for bridge measurement ---------------------Facing 24
5. A, Power-driven truck carryirig crane and reel for bridge measurement, used on Mississippi River at Memphis, Tenn.; B,
Equipment used in boat measurements ------------ Facing 25

PLATE 6. Discharge measurement through ice cover: A, Current meter supported on a rod; B, Reel and sled equipment used on Mississippi River at St. Paul, Mini -... ...... ...... Facing 32
7. A, Natural control oIn Smith River near Bristol, N. It B,
Trenton-type artificial control on Rahway River at Springfield, N. J_ Facing 120
8. Changes in stage-discharge relation at station on Merrimack
River at Lowell, Mass -------. ID pocket
9. Curves showing relations of stage to discharge, Tennessee
River at Chattanooga, Tenn ------------- ------ In pocket
10. Curves showing relations of stage to constant-fall discharge
and discharge ratios to fall ratios, Tennessee River at Chattanooga, Tenn ------------------------------ In pocket
11. Curves showing relations of stage to discharge and discharge ratios
to fall ratios, Tennessee River at Guntersville, Ala- -In pocket
12. Plan and profile of Ohio River in the vicinity of Metropolis,
Ill------- I.------------------------ In pocket
13. Curves showing relations of stage to discharge and stage to fall,
Ohio River at Metropolis, Ill ------------------ In pocket
14. Curves showing relations of stage to normal discharge and
discharge ratios to fall ratios, Ohio River at Metropolis, Ill. --------------...---------- In pocket
15. Curves showing relations of stage to constant-fall discharge
and discharge ratios to fall ratios, Kootenai River near
Copeland, Idaho --...........--------------- In pocket
16. Curves of equal fall, Ohio River at Metropolis, Ill__ In pocket 17. Adjustment of discharge measurements for changing discharge,
Ohio River at Wheeling, W. Va., during the period March 14-27, 1905 -----.-------------------------- In pocket
18. Adjustment of discharge measurements for changing discharge,
Tennessee River near Scottsboro, Ala ----------- In pocket
19. Horizontal-axis propeller-type current meters --------Facing 176
20. .4, Original Price current meter; B, Type-A small Price current
meter ---------------------------------------- Facing 177
21. A, Pygmy current meter; B, Discharge ntegrator - Facing 92
22. Typical rating table for current meter ------------In pocket
23. A, Graduated sections for vertical-staff gag s; B, Vertical-staff
gage on concrete gage well ---------------------- Facing 200
24. A, Chain gage; B, Hook gage -------------------Following 200
25. A, Float gage; B, Electric-tape gage --------------Following 200
26. A, Canfield wire-weight gage; B, Flat wading rods. Following 200 27. A, Hand reels; B, Connectors ----------------------Facing 208
28. A, Columbus-type sounding weights; B, Meter hangers, weight
pins, and other miscellaneous equipment --------Following 208
29. A, Type-D reel mounted on type-D crane; B, Type-D reel
dismounted--------- -------------- Following 208
30. A, Type-A reel; B, Type-B reel--_, -----------Following 208
31. A, Type-E reel; B, Tag-line reels ------------------- Facing 216
32. Reference-mark tablet. -------------------------Following 216
33. Form for level notes ---------------- --------- -..In pocket
FIGURE 1. Typical vertical velocity curves ----------------------------36
2. Position :of a sounding weight and line in deep, swift water. -- 45 3. Discharge-measurement notes, form 9-275 -------------------60
4. Current-meter notesi ice cover, form 9-275a_-61


FIGURE 5. Discharge-measurement notes, two-I able method, form ()'75 63
6. First sheet of dchreeauentnote-s, form 9- 275c---.
7. Supplementary (lisci arge measiiremnit note-s, form 9- 275d__ '
8. Typical design of the 'Trenton It vpe of[ artificial corntrol 119
9. Rating curves for Mlerrimack River below Concord Rive-r at
Lowell, Mass-- 12
10. Slope and energy relations in an open chan nel ------ 132 11. Gages on Tennessee River at Chattanooga, TPenn ----142 12. Assembly diagram of type-A small Price current meter-. 174 13. Water-stage recorder used in 1876~ 185
14. Inspection of recording gage, form 9-1761 192


TABLE 1. Air-correction table giving difference, in feet, between vertical
length and slant length of sounding line above water surface
for vertical angles between 4' and 360--------------------- -46
2. Wet-line table, giving difference, in feet, between wet-litie
length and vertical depth for vertical angles between 40
and 36 ---------------------------------------- ---- 50
3. Summary of effect on air and water corrections caused by
raising the meter from the sounding position to the 0.8-depth
position ------------------------ ----------------------- 54
4. Amounts to be added to observed vertical angles to obtain
actual vertical angles ------------------------------------59
5. Values of n for canals and ditches and for natural streams- 83
6. Slope of the energy gradient, Tennessee River at Chattanooga,
Tennm------------------------------------------------ 135
7. Relation between stage, fall, and discharge adjustment factor,
Ohio River at Metropolis, Ill ----------------------------147
8. Discharge measurements of the Tennessee River at Scottsboro,
Ala------------------------------ --------------------- 159
9. Adjusted discharge for measurements of the Tennessee River
near Scottsboro, Ala------------------------------------- 163
10. Types of reels used in sounding ---------- ------------------206



The Sundry Civil Appropriation Act approved October 2, 1888, contained an item 0 2
For the purpose of investigating the extent to which the arid region of the United States can be redeemed by irrigation and the segregation of irrigable lands in such arid region, and for the selection of sites for reservoirs and other hydraulic works necessary for the storage and utilization of water for irrigation and for ascertaining the cost thereof and the prevention of floods and overflows, * the work to be performed by the Geological Survey under the direction of the Secretary of the Interior * *
In order to carry out this mandate, a knowledge of the quantities of water available for storage, diversion, and utilization in irrigation was needed. At that time, there were no systematic records of the flow of the streams and little knowledge of the methods that would best serve in obtaining such records; and no adequate instruments, apparatus, or equipment for collecting records of stage and discharge of streams were available. As a first and essential step in the investigation, Maj. J. W. Powell, Director of the Geological Survey established, in December 1888, a camp at Embudo, N. Mex., on the Rio Grande, where instruments and methods were studied and vounsy men were instructed in the undeveloped art of stream gaging.
With the establishment of the Embudo camp, the Geological Survey began systematic work in collecting records of stream flow and in studying the problems related to the utilization of water for irrigation and other purposes, and this work has continued uninterrupted to the present time. Specific appropriations for stream gaging, carried first in the Sundry Civil Appropriation Act for the fiscal year 18951 3 have been made annually. More than 4,000 stream-gaging stations are now being operated by the Geological Survey in cooperation with other Federal agencies, and with States, counties, and cities. These stations are distributed throughout the 48 States and the Terri' Chief Hydraulic Engineer (retired), Geological Survey. 225 Stat. L. 960-961,
28 Stat. L. 398.


tory of Hawaii. The field operations are conducted through 38 district offices, each of -which has a district engineer in charge and a group of assistants proportional in number to the amount of work to be done.
Study of the instruments, equipment, methods, and technique involved in systematic stream gaging, which was started at Embudo, has been continued by the hundreds of engineers engaged in it, and their findings have contributed both directly and indirectly to this report. Among the reports issued by the Geological Survey that have presented various aspects of the development of the art of stream gaging are the following Water-Supply Papers: 64. Accuracy of stream measurements, by E. C. Murphy. 1(.02. 99 pp., 4 pls. 94. Hydrographic manual of the United States Geological Survey, by E. C. Murphy,
J. C. Hoyt and G. B. Hlollister. 1904. 76*pp., 3 pis.
95. Accuracy of stream measurements (2(1, enlarged edition), by E. C. Murphy.
1904. 169 pp., 6 pls.
150. Weir experiments, coefficients, and formulas, by R. E. Horton. 1906. 189
pp., 38 PIS.,
180. Turbine water-wheel tests and power tables, by R. E. Horton. 1906. 134
PP., 2 pbs.
187. Determination of stream flow during the frozen season, by H. K. Barrows
and R. E. Horton. 1907. 93 pp., 1 p1.
200. Weir experiments, coefficients, and formulas (revised), by'R. E. Horton,
.1907. 195 pp., 38 pis.
337. The effects of ice on stream flow, by W. G. Hoyt. 1913. 77 pp., 7 PIS. 371., Equipment for current-meter gaging stations, by G. J. Lyon. 1915. 64 pp.,
37 pis.
868--A. Investigations of methods and equipment used in stream gaging, part 1,
Performance of current meters in water of shallow depth, by C. H. Pierce.
19-41. pp. 1-35, -pls. 1-27.
868-B. Investigations of methods and equipment used in stream gaging, part 2,
Intakes for gage wells, by C. H. Pierce. 1941. pp. 37-75, pls. 28-31.
In recent years, reports on special phases of stream-gaging activities have been issued as mimeographed circulars under the general title "Equipment for river measurements." These circulars have been issued in small editions and, like the earlier water-supply, papers, most of them are no longer available for distribution. Developments in -instruments, equipment, and practices are reported and discussed currently in the 'Water Resources Bulletin, a mimeographed pamphlet that is. issued periodically for the- official use of engineers of the, Geological Survey.
At a conference of district engineers held in January 1930, the development of a field manual was discussed, and its preparation was subsequently authorized by the Director. Attempts were made over a period of several years to obtain a draf t through the efforts of individuals who were fully occupied with other work. Finally, in 1934, steps were taken to make Don M. Corbett, then an assistant


hydraulic engineer, available for part-time service in preparing the manuscript for the manual. He drew ijpon all sources of inforrmition, including published and iinpubli-siied reports and suggestions by district engineers and others. He bT'Ought to the task 10 years of personal experience in field and office in four Geological Survey districts. 'In 1935 he completed a draft of the manual, which was mimeographed and distributed in that year to all. engineers of the Water Resources Branch of the Geological Survey with a request for criticism and suggestions. As Mr. Corbett could no longer be made available, C. H. Pierce, senior hydraulic engineer, was assigned to the responsible and exacting task of revising the manuscript, utilizing the many suggestions received.
This manual is far more than a compilation and adaptation of information previously published in scattered reports, as it embodies the results of the work of many engineers who have been active in all sections of the country over a period of many years. It presents the technique that has stood the test of experience under a wide variety of conditions as it has been applied by skilled engineers who have been constructively critical and have not hesitated to modify, abandon, or propose substitutions for any procedure that seemed to be unsatisfactory.
The report has been prepared primarily, for use in the training of young engineers for work in the Geological Survey. ]During 1938, 1939, and 1940 the number of new engineers added to the organization averaged 85 a year. The work of training these young men represents an undertaking that warrants the furnishing of the best possible facilities. The report will serve also to systematize, stabilize, and improve the work of stream-gaging as a whole. In a farflung field organization it is not easy to obtain consistency in methods and cesults among many groups that perform their work without frequent contacts with each other. The report will be useful also in connection with the training of students in the engineering colleges-an activity in which the Geological Survey is much interested because it must recruit its personnel from such students. It will also serve as an aid to practicing engineers who may be called upon to measure and record the flow of streams, as it contains much' new and valuable information not to be found elsewhere in engineering literature.
The information given relates to both the science of the flow of water in open channels and to the art of measuring and recording river discharge. Because the technique followed is perhaps as important as the instruments and equipment utilized, much attention is given to the details of the field procedures that have been found to yield the best records of river flow.


Iii stream gaging, as in all other engineering activities, over-all costs are .important, especially so when the funds available are inadequate for the work to be done, as has always been the situation in connection with the systematic recording of river discharge by the Geological Survey. The instrunients, equipment, and procedures have been developed, therefore, with a view to obtaining reliable records at a minimum cost. Since the greatest elements of cost are those of salaries and traveling expenses, including subsistence, all field instruments and equipment are designed for operation by one man or, at most, by two men.
The report is limited in its scope t o the field side of stream gaging, and except for the office work related directly thereto, it does not discuss the many office practices that are fully as important as the field activities in their relation to the collection, computation, and publication of records of stream flow.
In an activity that is current, it must be understood that there will be continuing improvements, and even as this report goes to press it is probably not strictly up to date, because of changes made in instruments or procedures since the latest revisions were made in the manuscript. It can be said, however, that the methods and practices described herein represented, in general, the best used by the Geological Survey at the time the report was prepared.



By DoN M. CORBETTand others

Water is a requisite of both plant and animal life. Its availability in usable form defines the limits within which human activities can be carried on, because the growth of cities and towns, the production of food supplies, the maintenance of transportation facilities and other public utilities, and the operation of many industries depend upon the availability of suitable supplies of water.
In its unending cycle between the clouds and the surface of the earth, water follows many courses. Precipitation, if in sufficient amount, produces surface runoff in stream channels, but not all of the total precipitation appears in surface runoff because a large part enters the ground directly or returns to the atmosphere either by evaporation or by transpiration. In its journey to the sea, water may follow courses below the surface of the ground or it may flow in surface channels. In general, there is much intermingling and interchange of surface and subsurface waters.
Systematic studies of the water resources of the United-States have been made by the Geological Survey, United States Department of the Interior, for more than 50 years, and a vast amount of valuable information has been collected and made available by publication in more than 900 water-supply papers. The investigations have included measurements of the flow of streams, studies of the quantity and availability of ground water, chemical analyses to determine the quality of water with respect to its use in agriculture and industry, surveys of river channels and valleys, studies of power production, and collection of other data, needed in determining the best methods of utilizing water resources. With the increase in growth and population of the country, the problems of providing adequate water supplies and of controlling the flow of streams to prevent damage and to promote utility have become increasingly difficult; and the del


inand for reliable information has be comel more insistent with each succeeding year.
Measurements of the flow of streams, investigations of underground currents and artesian wells, and determinations of the available water supplies of the United States were begun by the Geological Survey in 1888 in connection with special studies relating to the irrigation of the public lands. Systematic records of stream flow at more than Air800 places in the United States, including Alaska and Hawaii, have been obtained by the Geological Survey Thsrcodoftea flow% usually extend over long periods of time and include determinations of the (daily flow of the stream at the place of measurement. The data this obtained are used in the, administration of water rights, in studies in hydraulics and hydrology and in engineering studies related to the design, construction, and operation of hydraulic structures; to domestic, industrial, livestock, and irrigation water supplies; to the design of hydroelect ic, power plants; to litigat ion involving the determination of water rights; and to stream pollution, flood control, navigation channels and blocks, drainage of agricultural lands, sanitary and storm sewers for urban areas, railway and highway bridges, road drainage, and erosion -control structures and practices.
In July 1941 more than 4,000 river -measurem en t stations were being maintained by the Geological Survey in cooperation with Federal bureaus, with nearly all of the States, with the Territory of Hawaii, and with many counties, municipalities, and other organizations.
The greatly increased usefulness in recent years of the results of the water-resources investigations of the Geological Survey and its resulting increased personnel have led to a pressing need for a manual on stream-gaging procedure. In recognition of this need, the district engineers during their conference in 1930 recommended that the committee on field methods prepare a, preliminary draft of a field manual. Tin the conference in 1931 further consideration and approval was given to the project. Preparation of the complete text of the field manual extended over a period of several years. As the manual received the attention of many of the engineers of the Geological Survey, the descriptions of methods and procedure are representative of the general practice of the Geological Survey in measuring and recording the flow of streams.


The entire project of writing the manual on stream-gaging procedure was under the administrative direction of Nathan C. Grover, Chief Hydraulic Engineer, and C. G. Paulsen, Chief, Division of Surface Water.


The committee on field methods, which prepared the preliminary draft of the manual in accordance with the recommendations of the district engineers in conference in 1930, consistedl of E. D. Burchard, chairman, M. H. Carson, J. J. Dirzulaitis, H. E. Grosbach, A. B. Purton, and M. R. Stackpole.
In 1934 the assignment of expanding the preliminary draft into the complete manual was given to Don M. Corbett who carried the work forward until the spring of 1938, when he was assigned to other duties. C. H. Pierce then undertook the completion of the manuscript.
Parts of the manuscript were prepared, or reviewed and revised, by other members of the Water Resources Branch, as follows: The section on "Measurements of rivers that are deep and swift" by G. C. Stevens; the sections on "Slope -stage -d isch arge relations at gaging stations affected by variable slopes," "Variable slopes due to backwater," "Variable slopes caused by changing discharge," and "Variable slopes caused by backwater in conjunction with changing discharge" by M. C. Boyer; and the section on "Instruments and miscellaneous equipment" by A. H. Frazier.
At various stages in the preparation of the manual, the manuscript had the benefit of review and revision by the district engineers of the Water Resources Branch.
The Geological Survey is divided into five branches-eologic, Topographic, Water Resources, Conservation, and Alaskan. Water resources are investigated by the Water Resources Branch, which is subdivided into five divisions. The part of the work relating to measurements of river discharge comes within the activities of the Division of Surface Water, whose work is in many ways closely connected with the work of the other divisions of the branch. The various types of investigations conducted by the divisions of the Water Resources Branch are discussed below. DiviMSion of Surface Water.-The Division of Surface Water measures the flow of streams at selected stations in the 48 States, the District of Columbia, and the Territory of Hawaii. At these stations not only is the flow of "water measured, but records of stages and other data are collected. A stream-flow measurement station on Antietam Creek near Sharpsburg, Md., is shown in plate 1. The Division of Surface Water operates through 38 district offices with a district engineer in charge of the work in each district. In general, the districts are limited to single States, although a few districts


include two or more States or parts of States. In river basins that cross State boundaries, the work is efficiently and economically coordinated among the di st ricts concerned.
Division of Ground JVater.-The Division of Ground Water investigates those waters that lie below the surface in the zone of saturation, from which wells and springs aire supp~liedl. That division studies the source, occurrence, quantity, and head of these waters; their conservation; their availability and adequacy for domestic, industrial, irrigation, and public supplies and for watering places for livestock and desert travelers; and devises methods of constructing wells and recovering water from them and of improving springs. Each year surveys are made of selected areas where problems of water supply are acute, and the results of the surveys are made available to the public.
Division of Quality of lVater.-The Division of Qualit 'yof Water makes, chemical analyses of water from surface and underground sources with i'eferenc( to its suitability for both industrial and agricultural uses and for such domestic use as is not related to questions of health, so far as such use is affected by the dissolved mineral matter.
Division of Powrer Resources.-The Division of Power Resources studies and reports on water-power resources, collects and compiles information regarding not only the capacities of water wheels in water-power plants in the United States of 100 horsepower or more, but also the records of p)ower production.
Division of Water Ut iization.-The Division of Water Utilization investigates problems that are related to the utilization and control of the waters of streams and issues reports on floods and water utilization; interprets records of stream flow: andc performs such administrative work as relates to supervision and investigTation of those activities of the field organization of the branch that pertain to power projects of the Federal Power Commission and of the Department of the Interior.


Under the Director of the Geological Suirvey, the Chief Hydraulic Engineer is the administrative officer of the 'Water Resources Branch. The operation of the district and field offices of the Division of Surface Water is supervised by the chief of that division, who is responsible for the procedure used in collecting, records of stream flow and who assists the Chief Hydraulic Engineer in administrative matters pertaining to that work. The Division of Surface Water cooperates with the other divisions of the Water 'Resources Branch and with many bureaus of other departments in obtaining stream-flow information needed in connection with the work of those divisions and


bureaus. The work in each district is under the direction of a district engineer, who has charge of the details of the work in his district and is responsible for the selection of the sites; and for the establishment, construction, and operation of river-measurement stations; and also for the analysis and compilation of records for publication. The district engineer is the contact officer between the Geological Survey and the Federal, State, and municipal officials who cooperate in obtaining records of stream flow at stations in his district.

Appointments to the staff are made from lists of eligibles certified by the Civil Service Commission. Vacancies in the higher grades are nearl y alwa ys filled by promotion of experienced persons selected from the lower grades; therefore new appointments, especially in the professional and scientific services, are usually made in the lowest, or entrance, grades. A position may be filled by transfer from another Government agency if the applicant's request for transfer is approved by such agency and by the Civil Service Commission and if the qualifications of the applicant meet the requirements of the work. A vacancy may be filled by the reinstatement of a former employee if the person is eligible for reinstatement under Civil Service rules and regulations. A v.-cancy may occur because of promotion, resignation, transfer, or death, or because of an increase in the work of any district or branch division.
College and university graduates who have established Civil Service eligibility provide the normal sources of supply for professional and scientific personnel in the Geological Survey. These graduates may acquire experience in other kinds of work before receiving an appointment in the Geological Survey. If this is not extended for too long a time it is helpful in broadening their outlook on general problems.
Some educational institutions provide training in water-resources investigations so that men preparing for this type of work have an opportunity there to apply theoretical hydraulics to practical problems. Instructors in such colleges welcome the opportunity to have a representative of the Geological Survey address their students on the subject of hydrography. These personal contacts, which engage the interest of the students in the work of the Water Resources Branch, may result in an increased number of high-grade eligibles on the Civil Service lists from which the Survey replenishes itig personnel.

New appointees to engineering positions in the Geolo ical Survey 91
serve a probational period of 1 year, during which time their services may be terminated if the y are not found satisfactory. New appointments must be made at the lowest salary in the grade, which is usually the junior grade. The new appointee is normally assigned first as an assistant to an experienced man, under whose close supervision he obtains s practical experience in the highly specialized technique of stream gaging.
Much of the routine field work is done by engineers traveling alone, with no assistance except such as they may- obtain locally. The newly appointed junior engineer is preferably given instruction by more experienced engineers and acquires some field experience while acting as an assistant before he is given the responsibility of -independent field work. 'In order to insure the best possible instruction for each junior engineer upon his introduction to the work to which he ha& been assigned, it is desirable that the, district engineer supervise the instruction personally and that he arrange for one of his ablest assistants to.give such technical advice and instruction as may be most helpful to the appointee in becoming acquainted with the details of his work. This instruction should include details in connection with the work to be performed at gaging stations. Instruction received by an assistant while observing the methods used by an experienced man, supplemented by his own experience when doing similar work under the direction of his superior, will make a deep and lasting impression. This personal instruction might well continue throughout a field trip covering all the gaging stations likely to be visited by him on his first unsupervised trip. If the trip does not include the different conditions which prevail at low. and high water, for which different methods might be used, it should be repeated when those conditions do prevail.
An engineer engaged in this work should be able to adapt himself readily to the various local conditions that may prevail as he moves from place to place. These conditions vary not only in different districts but also in different parts of the same district. Some of the districts are so large that it is expedient to divide them into two or more areas with a group of engineers assigned to each area. After an engineer becomes thoroughly acquainted with the conditions in one area, he may be assigned to another area and there obtain experience of value in supplementing that previously obtained. These changes in assignment within a district ordinarily can be made by the district engineer to the mutual advantage of all concerned, as they not only provide wider experience for the men assigned to the work, thereby developing their resourcefulness and initiative, but they


also bring the accumulated experience of several men to each problem of an area.
Although essentially the same methods are used in all disti cfts the application of those methods to the great variety of condition in different sections of the country results in variations in the u,k that require engineers able to meet new situations and new probloilts as they arise. Therefore a broad and well-rounded experienceH that can best be acquired by service in several districts is always beneficial in developing well-trained and efficient engineers.
Many helpful suggestions in regard to contacts with the public can be given to a junior engineer by those who are familiar with the local customs. As the Geological Survey is a service organization it should not be necessary to enlarge upon the importance of courteous conduct. A mere suggestion that each engineer has his share of responsibility in maintaining the good will of the public toward the organization he represents should serve to direct his attention to the importance of creating a favorable impression.
The scientific aspects of the work should never be overlooked. The engineer employed in obtaining basic stream-flow data should realize that he is also a scientist engaged in research and that these data, or deductions from them, may be used in establishing fundamental laws in hydraulics and hydrology as well as in more immediate engineering problems.
Above all else, the young engineer should be imbued with loyalty to his work and to the organization of which he is a part. He is one unit among many units engaged in the same undertaking. Confidence in his superiors and loyalty to their programs will carry him through many difficulties and enable him to use his talents to the best advantage.
In a program for obtaining systematic records of stream flow, the technical work may be classed under three major heads: (1) establishing and constructing stream-flow measurement stations; (2) operating and maintaining those stations; and (3) computing, compiling, and preparing stream-flow data for publication.
Before a stream-flow measurement station is constructed a general reconnaissance is made in order that a suitable site for the gage may be selected. This reconnaissance is facilitated by an examination of topographic maps or such other maps of the area as may be available. Tentative sites for the gaging stations may be indicated on the maps, each site being subject to critical examination, and to rejection if the physical characteristics of the stream channel at and near it are unfavorable.


Consideration should be given to the channel characteristics in the vicinity of each proposed site with particular reference to the hydraulic conditions necessary for maintaining a fixed and permanent relation between stage and discharge at the gage. The selection of a suitable cross section of the stream for use in making discharge measurements and the proper placing of the gage with respect to the measuring section and to that part of the channel which controls the stage-discharge relation are of special importance. As the suitability of a particular site with respect to the hydraulic requirements for a gaging station may vary with different stages of the stream, it is desirable that information be obtained in regard to the hydraulic conditions prevailing at times of low, medium, and high stages. If time and opportunity do not permit examinations at various stages, ond if the engineer must make his selection of sites for gages from in -formation obtained at one stage, he should give consideration to the probable changes in conditions that might prevail at other stages before making his selection.
The construction of stream-flow measurement stations includes all the work pertaining to installation of the staff gages, building of the gage wells and shelters, erection of cableways and structures from which discharge measurements are made, improvement of the channels and control sections, and placing of reference marks and other tennis of permanent equipment. Structures typical of those generally used at stream-flow measurement stations are shown in plate 1.
Operation and maintenance work begins with the completion of the construction of the station. It involves collecting basic streamflow data and comprises most of the routine activities of the field organization when it is away from field headquarters.
The data are computed and compiled in the district offices. This work includes the analysis of all discharge measurements; the development of rating curves defining the relation between stage and discharge; the computation of mean daily, monthly, and annual rates of flow; and the preparation of descriptions of the stations to accompany the tables of discharge. Reviewing and assembling the data for publication in water-supply papers of the Geological Survey are functions of the reviewing section of the Division of Surface Water in the office at Washington, D. C.

The stage or gage height of a stream is the height of the water surface above a chosen datum corresponding to the zero of the gage. An accurate record of stage is one of the essential factors in determining river discharge, and it provides basic data necessary for computing the rates and quantities of flow for any period of time. An





engineer assigned to field work should recognize the obvious f act that a record of discharge can be no more accurate than the record of stage upon which it is largely based.

Gage heights may be obtained either f rom direct observations on a nonrecording gage or by the use of a mechanically operated recording instrument and are classed either as observed gacre heights, which consist of gage readings made by individuals, or recorded gage heights, which are those obtained by means of a recording instrument.
A record of observed stages consists of one or more daily readings of the height of the water surface as indicated by a nonrecording gage. Where such a record is used in determining mean daily discharge a sufficient number of observations should be made each day to assure a satisfactory record of stage f rom which the mean daily discharge can be computed. A record of this kind usually consists of two daily readings with additional observations made during floods and periods of large or rapid variations in stage, although for a stream not affected by diurnal regulation and -%Nrhere changes in stage occur very slowly one daily reading may suffice. The frequency with which the gage can be read depends to a large extent on the availability of a crage observer, the distance he must travel to reach the crage, and the compensation that can be made for the work. A Geological Survey type-A wire-weight gage used in obtaining observed records of stage is shown in plate 2, A.
A record of observed stages is kept in a gage-height book (Geological Survey form 9-175) provided for that purpose. The book contains sufficient space for a 3-month's record of stage observed twice daily and such additional information as may be obtained by the observer. The name of the gaoring station and the date for each day during the 3 months' period is entered in the proper places in the book before it is issued to the observer. For each day of the period, spaces are provided for two observations of stage, for the time of each observation, and for remarks pertinent to the accuracy and significance of the observations.
The accuracy and sufficiency of an observed record of stage depends largely upon the capability and faithfulness of the observer as well as upon the number of readings each day. The ever-present personal or human element is probably the greatest single factor with which the field engineer must contend in obtaining an accurate and reliable observed record of stage, because the observations and


records are made by persons of varying intelligence, carefulness, and faithfulness and not by mechanical instruments.
A record of recorded stages is obtained by the use of a mechanically operated recording instrument known as a water-stage recorder. This instrument may be designed either to produce a graphic record of the rise and fall of the water surface with respect to time or to print or otherwise indicate the stage at definite time intervals. Of these two types of records, the graphic record is generally accepted tis being both more accurate and more usable.
The record produced by a water-stagp recorder largely eliminates the personal element that always prevails in an observed record and provides the refinement that is necessary for a high degree of accuracy. A water-stage recorder in operation in a concrete shelter over a 5-foot square gage well at the gaging station on the Rahway River at Springfield, N. J., is shown in plate 2, B.

Gage observers are employed for obtaining records of stage either by directly observing nonrecording gages or by attending to the operation of water-stage recorders.
Availability.-In many places little or no choice in the selection and employment of gage observers may make it necessary to employ the person living nearest the gage, as he may be the only individual reasonably available. If a choice is possible, it is good practice to interview first the person on whose property the station is situated. If that person, or someone else connected with the property, is interested in the work and appears to be competent, it is better to employ him than another person who must trespass in order to perform the duties, thus avoiding potential trouble with the property owner as well as possible embarrassment to the Survey.
The necessity for daily observations at definite hours, eliminates as gage observers many persons who are profitably employed otherwise. Therefore the selection may be limited to comparatively few persons even where the gaging station is in a thickly populated area.
Trustworthiness.-The observed record of stage, should be accurate and reliable, which is possible only if the observer is competent, faithful, trustworthy, and has a full understanding of the need for accurate observations. These characteristics are not always obvious when the individual is first interviewed but are usually encouraged by proper contact and instruction. On the other band, a capable and honest observer may become careless and unreliable if the engineer fails to show proper interest in his work. Some observers


whose capability and honesty were doubted at t1v time of their employment have rendered excellent and faithful service as a result of tactful and diplomatic efforts of the field engineer.
Discontent and laxity of the observer can generally be avoided if, at the time of his employment, he is given definite instructions as to his duties and the compensation he is to receive. It is a good policy to issue instructions, either personally or by correspondence, with consideration and diplomacy, and thus avoid as much as possible any inference that the supervision of the work is to be conducted in tin arbitrary manner.
It is desirable for the observer to realize that there are ways of detecting errors in gage readings. If he is shown that inconsistencies in gage readings may be discovered by comparison with records at other gaging stations on the same stream or on other streams in the vicinity, he is likely to be more careful in his work and less apt to think that he can "estimate" the gage height without actually reading the gage.
Duties.-The character of the observer's duties varies with the type of the gage-height record required. For an observed record of stage the duties are specific. They consist mainly of the careful reading of the gage and the proper recording of the readings. In addition the observer is required to assist in keeping the gage in good condition and to report promptly to the office any trouble that he cannot correct. For example, at times he may need to remove such small amounts of mud, debris, or ice as interfere with his accurate reading of the gage.
The instructions in each gage-height book must be followed by the observer if his duties are to be performed most satisfactorily. The gage-height book should either be kept in a compartment in the gage box at the gaging station or taken by the observer to the gage at the time of each reading so that the record may be entered at the moment the observation is made. The observer should be careful to record each gage reading, with the time, date, and remarks related thereto, in the proper columns. Recording these data elsewhere for later entry, or trusting to memory, should always be avoided. If for any reason the observer is unable to read the gage, he should leave the spaces blank in the gage-height book except for a note stating that no reading was obtained. If he employs someone to attend to his duties, a note to that effect should be made under "remarks."
If a rise and fall in stage has occurred between visits and if the highest stage reached by the river can be seen from the high-water mark on the gage., the gage height of such a high stage with the probable time of its occurrence should be recorded under "remarks" for the day. Information on conditions that affect the height of water at

the gage, such ac the forming or breaking of log jams, the growth of grass in the channel, or the collection of logs and debris on the control, is of value to an engineer using the record and should be noted by the observer. Statements regarding the weather, the approximate amount of precipitation (whether rain or snow), and the trend of river stage at the time of the observation (whether rising, falling, or stationary) should be made. During winter the observer should describe the condition of the river below the gage, or at the riffle or rapids controlling the ordinary stage-discharge relation, by stating whether the channel is open, partly open, or frozen over; whether there are ice jains above, below, or on the riffle; and whether ice is anchored to the bed at the riffle or floating in the stream. If the stream at the gage is covered with ice the observer should record the stage of the water surface and not the height of the top or bottom of the ice.
At the end of each week the observer should copy on a card (Geological Survey form 9-176) the week's data from his gage-height book and mail the card to the district office, thus providing the office with an up-to-date record of the stage of the stream. At the end of the 3-months' period for which the gage-height book was issued the observer should sign the book and mail it to the district office.
The duties of the observer at a station equipped with a water-stage recorder are so exacting that they must be performed by the field engineer if a capable observer is not available. In several districts where the stations are so situated that they can be visited frequently by an engineer, it has been found economical and practicable to operate well-equipped gaging stations without the use of observers. In other districts experience has shown that every reasonably accessible gaging station should be visited by an observer at least once a week in order to insure continuity of the records. Improved highways and transportation facilities have made it possible in many sections for a field engineer to visit gaging stations frequently and to perforni the essential duties more efficiently than the ordinary observe er. Modern water-stage recorders installed with adequate gage wells and shelters require less frequent attention than older types of recorders.
Manipulation of a water-stage recorder by an observer who has no knowledge of its operations may result not only in additional work for the field engineer but also in the loss of valuable records. Instructions to the field engineer in regard to his duties at a gaging station also apply to the observer for those duties that he must perf orm.
Upon the completion by the observer of those duties that are directly related to the operation of the water-stage recorder, a summary of the results obtained should be recorded on the inspection card


(Geological Survey form 9-176-C), which outlines specifically those duties that should be performed on each visit to the aging station. The card is then signed and sent to the district engineer.
Compensatio.-Observers, like all other indiviuals earlig a livelihood, expect compensation for the work they do, even though it may require but a few minutes of their time each day. The conpensatioi of each observ( r should be based on the amount of work he has to do and the distance he must travel to reach the g'aging station.

The discharge or rate of flow of a stream is the quantity of water flowing past a cross section of the stream in a unit of time. Tho unit in which discharge is usually expressed is cubic foot per second, which is contracted into second-foot. A second-foot of water is defined as the quantity flowing through a cross section 1 square foot in area at a velocity of 1 foot per second. The procedure of measuring the area of a cross section of a stream and the velocity of flow past the section is known as the velocity-area method of measuring discharge. The product obtained by multiplying the area of the crosssection by the velocity constitutes a discharge measurement for that area.
The velocity of water may be measured either directly or indirectly, depending on the method employed. A direct method consists in observing the rate of travel of a float or a chemical placed in the stream. An indirect method consists either in the measurement of the slope of the water surface from which the velocity is computed by means of a slope-velocity formula or in the use of an instrument to measure the velocity of flow within a selected section. Discharge measurements are classified according to the method used in measuring velocity.
The detailed collection of stream-flow records needed in a wellarranged investigation of water resources has become practicable through the development of methods and equipment for accurately measuring the velocity of water in open channels. In the following discussion the merits of several of the more generally accepted methods are reviewed and considered with respect to the adaptability and application to general stream-gaging procedure.
A velocity-area measurement of discharge in which the velocity is measured by an instrument known as a current meter is a currentmeter measurement.
In making a current-meter measurement, the total area of the cross section at the place of measurement is divided into small or


partial sections and the area and the mean velocity of each is determined separately. The small sections are each bounded by the water surface, the stream bed, and two imaginary vertical lines, called verticals. Each vertical, therefore, being a common dimension for two adjoining sections, fixes the point at which observations of depth and velocity are made. Sufficient, velocity observations are made to establish the mean velocity in each of the two verticals forming the side boundaries of a section, and the velocities in the two verticals are then averaged to determine the mean velocity in the section. The product of the mean velocity thus obtained and the area of the section, which in turn is the product of the distance between the two verticals and the mean of their depths, is the discharge in the section. The sum of the discharges in all the partial sections is the discharge of the stream.
The specific procedure followed in making a current-meter measurement depends on the method selected for use in determining the depth and the mean velocity in the vertical, the manner in which the meter is suspended, and the type of structure or support from which the measurement is made.
Discharge measurements are made in various ways and therefore may be said to be of different types, depending on the kind of support used by the engineer in crossing the stream and the manner in which the current meter is held in the desired position in the water. The specific procedure for each type of discharge measurement is described below. Factors affecting the accuracy of measurements of the different types are discussed on pages 65 to 76. The meter and sounding equipment are described on pages 168 and 197. Stop watches, note forms, angle charts, and other items of equipment are assumed to be available.
Before making a wading measurement, the engineer should examine various cross sections in the vicinity of the gage to find the one most suitable for this type of measurement. With the measuring section selected and the equipment assembled. the next step is to span the measuring section with a tag line at right angles to the direction of the current. As this cannot always be done with respect to angularity throughout the width of the stream, it may become necessary to measure and record the amounts by which the angles formed by the current with the tag line at some of the measuring points deviate from 90'. Any unnecessary deviation should be avoided by careful placing of the line. While placing the tag line, the engineer should obtain a general idea of the proper spacing of


verticals by observing the total -width of the section and the character of the stream bed.
After the tag line is placed, the actual discharge measurern(,nt is begun. A wading measurement in progress at the gaging A,,6011 I)II Patuxent River near Burtonsville, Md., is shown in plate 3, A. The edge of the water in reference to a marker on the line tand ,dso the bank from which the measurement is started (-NNhether left or right bank, looking downstream) are recorded. The rod is then placed in a perpendicular position in the first selected vertical and flie df-pth observed. If the depth in the vertical is 1.5 feet or greater, the two-point method should be used. If the depth is less than 1.5 feet the method used will depend largely upon the type of current meter, the depth of the water, and the roughness of the stream bed.
The performance of current meters in water of shallow depths has been investigated by the Geological Survey at the National Hydraulic Laboratory of the National Bureau of Standards. The results indicate that the 0.6-depth method should be used for depths between 0.5 foot and 1.5 feet and that the 0.5-depth method should be used for depths less than 0.5 foot.
It appears that coefficients other than unity may be necessary for current-meter measurements in very shallow depths for two reasons: First, the distribution of velocity in a vertical may be such that the actual velocity at the point of observation is not the mean for the vertical as, for instance, an observation at 0.5 of the depth; and second, the registration of the current meter may be affected by its proximity to the water surface or the stream bed. Sometimes the errors from those two sources may be of opposite sign and therefore compensating. Under other circumstances the errors may be of the same sign., or may be predominantly of one sign, and therefore not compensating. As a result of the investigations mentioned above, coefficients have been determined for use with observations of velocities in shallow depths.'
In wading measurements the engineer should stand in a position that least affects the velocity of the water passing the current meter. Field and laboratory studies conducted by the Hydraulic Laboratory Committee of the Geological Survey indicate that the position of the engineer least affecting the accuracy of a discharge measurement by wading may be described as follows: With the meter rod at the tag-line and facing the bank with the water flowing against the side of his leg, the engineer should stand from 1 to 3 inches downstream from the tag-line and 18 inches or more from the meter rod. If facing the left bank, he will naturally hold the meter rod with his
'Pierce, C. H., "Performance of current meters in water of shallow depth": U. S. Geol. Survey Water-Supply Paper 86&-A, pp. 1-35,1941.


left hand if facing the right bank hie will hold it with his right hand. The results of the investigat ion show that no coefficient for position need be appliedd if this position is used. The engineer can maintain a standing position 18 inches or more from the meter rod with a reasonable degree of comfort and at that distance can also give proper' attecnt ion to the current meter.
Care should be taken to keep the rod in a vertical position and the meter parallel to the direction of flow while the velocity is being observed. If the flow is not at right angles to the tag-line, the amount lby whIich the angle deviates from 90' or the angle coefficient for that difference should be recorded. This angle coeflicient, which is the cosine of the angle of difference, may be determined by the use of an angle chart or a protractor held in proper alinement with the tag-line while the "angle-coefficient line" that corresponds most nearly to the direction of the flow is being observed on the chart or protractor. Upon completion of the necessary observations at the first measuring point, a similar procedure is followed successively at each of the remaining verticals.
If the velocity at the edge of the water is not zero, it is customary to estimate this velocity as a percentage of the velocity measured at the first vertical or measuring point. In order that no appreciable error may be introduced into the total measurement-as a result of such estimates, care should be taken to space the verticals so that the flow in the section bordering the edge of water is extremely small in comparison with the total flow. FurthermoreI it should be kept in mind that the vertical-shaft cup-type meter tends either to underregister or to overregister when used close to a vertical wall or bank where the velocity is nonuniform, the direction of deviation depending on whether the bank or wall is to the right or left of the meter, looking downstream.

A measurement of discharge may be made from a cableway with the meter and weight suspended either from a hand-line or from a reel assembly. The use of a reel assembly with a type-B gaging car on a cableway at the gaging station on the Licking River at Toboso, Ohio, is shown in plate 3, B. A hand-line used from a type-A gaging car is generally operated over a sheave placed on the side of the gaging car near the end opposite the operator (see pl. 1). A scale, marked in feet and tenths, attached to the side of the gaging car and in direct alinement with the sheave, materially assists in measuring the depths and in placing the meter in position in the vertical. If a type-B gaging car is use, the hand-line may be operated over the end heave and the depths measured with a graduated- scale.






P4 00-0 it
LO E-4




A relassembly may be used from either the type-Ao the type-B gaging cars. In the type-A gaging car, in which the operator works from a sitting position, the reel assembly is operated most satisfactorily if the rack on which it is mounted is placed at right anglesto the sides of the car. In this arrangement the axis of the drum is at right angles to the cableway. In the older type-B gaging car, III which the operator works from a standing position, the reel asseinIbly is mounted on the upstream side of the car with the axis of the drum parallel to the cableway. The improved type-B car has an endl sheave, and the reel assembly is mounted so that the axis of the drum is at right angles to the cableway. Provision for easily mounting the reel assembly is essential in the construction of either type of gaging car.
If the observations of depth and velocity do not necessitate the consideration of the vertical angle (see p. 45), the specific procedure in measuring from. a, cableway may, in general, be considered the same for both the hand-line and the reel assemblies. This procedure is as follows: Identify the edge of the water in relation to the initial point or zero marking, which is usually at one of the structures supporting the cableway, and the record the bank (whether left or right, looking downstream) where the m-easuremnent is started. The meter and weight are lowered at the first vertical until the horizontal axis of the meter is flush with the water surface; they are then lowered until the weight touches the bed of the stream, during which procedure the amount of line let out is measured. If the bed of the stream is composed of stable material, the weight may be again raised and lowered a short distance to determine irregularities in the profile of the section at that point. If there is a difference in the two observations, the mean is usually accepted. If, however, the stream bed consists of mud or shifting sand and silt, the depth should be noted the instant the weight touches the bed, as repeated churning with the weight may cause a scour that will result in an unduly large observation of depth. To the part of the depth thus measured, the distance from the bottom of the weight to the horizontal axis of the meter must be added to obtain the total depth.
Some engineers prefer to use the bottom of the weight instead of the axis of the meter as an index for determinations of depth. Where the suspension cable of the current meter makes a considerable angle with the vertical line from the point of suspension to the water surface, commonly called the vertical ang~le, and where floating drift is present in the measuring section, there is a recognized advantage in this method; but where a vertical angle and floating drift are not present, the swerving and swinging of the meter will be materially


reduced if the weight is submerged, and consequently a more accurate measurement of depth will be obtained, particularly if the cableway is far from the water surface.
After the depth is measured and recorded, the observations of velocity are made. With the weight at the stream bed after the measurement of depth, the most convenient procedure is to raise the current meter first to the 0.8-depth position for an observation there, and then to bring it to the 0.2-depth position for the observation at that place. In computing the distance that the current meter is to be raised, the distance from the meter to the bottom of the weight must be taken into consideration. In working from a cableway, the velocity observation should not be started until the cable and car have ceased oscillating.
The two-point method, in which velocities are measured at 0.2depth and at 0.8-depth, should be employed wherever the depths will permit. The minimum depth at which this method may be used in cableway measurements depends on the position of the meter with respect to the bottom of the weight. If the horizontal axis of the meter is placed 0.5 foot above the bottom of the weight, the minimum depth in which this method can be used is 2.5 feet. For the 0.7-f oot position of the meter, the minimum depth is 3.5 feet; for the 1.0-foot position, 5 feet. The 0.6-depth method should be used where the depth in any vertical will not permit the use of the two-point method; except that in very shallow water where depths are less than 0.5 foot the
0.5-depth method may be preferable.
If the angle made by the current with the cableway deviates f rom 900, the correction f or this angle may be determined from an angle chart, a protractor, or an indicator operated from the side of the gaging car by observing the "angle-coefficient line" which coincides with the direction of flow. The direction of flow may be detected by observing particles of drift that pass the measuring section near the measuring line or by noting the position of the meter when it is~ placed just beneath the surface of the water. The latter practice is preferable, and is reliable provided the velocities are great enough to overcome any torsional force that the twist. of the measuring line may exert on the meter. When observing passing drift, note carefully that the particles under observation are not affected by action of the wind. In the absence of floating drift when the surface is fairly smooth, the angle coefficient may be observed by bisecting the wake produced by the meter cable. The wake is usually identified by two small ripples that make an acute angle with the measuring line at the apex.
After the completion of the above procedure in the first vertical, each successive vertical is treated in like manner until the stream has been crossed.


If there is floating drift at or below the surf-,tc4, it may be- necessary to raise the meter for inspection or cleaning after the depth is measured and before the velocity observations are completed. After snclian interruption of the routine procedure it may be desirable to measure down from the water surface when placing the meter for the velocity observations. With this change in procedure from that usdin placing the meter from the bottom upward, care should be taken to see that the meter is lowered the computed distance below the water surfa ce.
An emergency may arise in using a hand line where the available weight is insufficient to obtain reliable measurements of depth in the usual manner. Under those circumstances the line and weight may be cast upstream and a f airly accurate measurement of depth obtained with the line pulled taut when the weight is on the bottom of the stream directly beneath the cableway.
A measurement is sometimes required in a section of shallow depth that cannot be waded because of a soft stream bed. In making a measurement from a cableway under such conditions, it may be necessary to place the meter below the weight in order to obtain a 0.5-depth or 0.6-depth observation. If the meter is used in that position great care must be used in placing it so that it will not come in contact with the bed. Under these circumstances the measurements of depth should be made either with the meter removed and before the observations of velocity are begun or by means of an auxiliary weight and line.
Although it is desirable to adhere to specific procedure in measuring depths and velocities, departures are sometimes permissible. For example, in streams that are comparatively free of drift and in which it vertical angle is not involved, the tag-line method may be used to advantage in measuring the depth and in placing the current meter. The tag-line method may be used also when the vertical angle must be considered, provided the angle does not change when the index tag is brought to the water surface (see p. 56). However, before adopting any method that apparently tends to simplify a generally accepted practice, the engineer should first assure himself that his proposed departure from customary methods will not affect, the accuracy of his measurement.

On some rivers subject to overflow, the only places where the total flow at flood stages can be measured are at highway or railroad bridges where the roadway embankments restrain the flow to channels under the bridges. During flood stages, when the use of a cableway is impracticable, it may be necessary to utilize the bridge in measur-


ing the discharge. It is, of course, desirable that the rilge should be normal to the current and that the cross section of ti channel beneath the bridge should be reasonably uniformi. Old pier footings, piling, and other obstructions left in the channel after the coinpletion of the bridge may sometimes seriously impair the accuracy of measurements made at bridge sections. Bridge piers in the mteasuring section are particularly objectionable where the piers make large angles with the direction of flow.
A discharge nieasure'iiwit is rarely made from a bridge as quickly and conveniently as frown a cableway. The necessity of dividing the area into a larger number of sections because of bridge piers and greater irregularity in the stream bed, th,. inabiity 1") observN e floating drift if working from the downstream side of the bridge, the greater possibility of damage to the meter, and the interference of bridge members are some of the factors that make a bridge measurement difficult and undesirable.
The downstream side is generally preferred in measuring from a bridge, as the direction of flow on that side often appears to be more nearly normal to the structure. The tendency for the meter line to move away from the lower bridge members affords a better opportunity for measuring the vertical angle of the meter line, if any is present. This tendency also largely eliminates the continual contact of the line with the bridge members, which cannot be avoided on the upstream side.
Either a hand line or a reel assembly may be used from a bridge in essentially the same manner as from a cableway. In the use of a hand line assembly from a bridge, it is generally more convenient to lower the sounding weight first to the bed of the stream and measure the depth by raising the meter and weight from the bottom to the surface This procedure reduces to a minimum the contact of the meter line with the bridge members.
If a weight heavier than 30 pounds is required, it is customary to use a crane-and-reel assembly, the reel being mounted on a crane designed to clear the handrail of the bridge and to guide the meter line beyond any interference with bridge members. The crane is attached to a movable base for convenience in transferring the equipment from one measuring point to another. The use of such equipment in making a bridge measurement at the gaging station on Rillito Creek near Tucson, Ariz., is slhown in plate 4, A. Another somewhat heavier type of crane with hand reel that is sometimes used for measurements of the overflow sections of the Mississippi River at Vicksburg, Miss., is .shown in plate 4, B. If the depths and velocities require weights greater than 200 pounds, the crane-and-reet equipment is usually mounted on a power-driven truck, the motor


of which is arranged to operate the reel as well as the fl~.A power-driven truck carrying crane-and- reel efIiiipIment tiie(l at the, gaging station on the Mississippi River at Mfeminpbis Tenn.. i lO~1 in plate 5, A. If the current deviates from the 900" :tnre with the bridge, the angle correction may be best obtained by p~lacinig the( angle chart or protractor against the handrail and reading the aiqgle coefficient in the same manner as for cableway mecas uremnents.
For measurements of canals, t ailraces, and small stremsifo
bridges are sometimes designed and built to provide facilities for making discharge measurements by use of the rod suspension. Where the depths and velocities do not permit the use of a, standard wading rod, a special rod may be designed, or a stay-line and wire arrangements may be employed to hold the rod and meter in place. Although for low velocities the procedure may be the same as for a wading measurement, it is often advisable at higher velocities to measure the depth in the following manner: For each selected vertical a point is established on the bridge. With this point ais an index, the distance to the water surface is measured by lowering the rod until the base plate touches the water. The rod is then lowered to the bottom of the channel, and the rod reading is again noted at the index point. The difference in readings is the depth of water in the vertical. Measuring the depth in this manner tends to eliminate errors that may be caused by the piling up of water on the upstream f ace of the rod.


Measurements of discharge from a boat are generally made by the Geological Survey only on those rivers that may be readily spanned by a temporary cable of sufficient strength to hold the boat in position while the observations are being made. The equipment used will depend largely on the size and character of the stream, the frequency with which discharge measurements are to be made, and the availability of a suitable boat. The boat should have a flat bottom, should be at least 15 feet long, and should be heavy enough to dampen the effects of the action of waves. The temporary cable used to span the measuring section must serve as a means of holding the boat in position and also of measuring the width of the stream. This cable should be wound on a reel that can be conveniently operated from the stern of a boat to pay out the cable as the boat is propelled across the stream.
A hand-line or reel assembly operated from a boom is most adaptable for measuring the depth and velocity from a boat, although where the depths are shallow and the bed of the stream is uniform and smooth, the depths may be measured with a rod. The boom should extend upstream beyond the bow of the boat a sufficient dis-


tance to eliminate, any possible effect of the boat, on the velo(.ity of the water at the meter. It is preferable that the boom be so (I(,,signed as to permit the use of a Geological Survey type of sounding reel, although reels of special design may be developed for use on individual streams where boat measurements become general practice. In general, the equipment should be of such nature that a boat measurement may bo made by one engineer with the aid of not more, than one helper. Equipment used in the St. Paul district of the Geological Survey in making boat measurements of the Mississippi River is shown in plate 5, B.
Although one particular assembly of boat equipment may not be adaptable for all types of streams and conditions, the equipment specified in the following list may be considered as typical for use with river spans of 1,000 feet or less and for velocities up to 5 feet per second. Deviations from this list of equipment may be necessary because of specific local conditions. Assuming that a boat is available and that an engineer has the usual current-meter outfit, the following additional equipment is necessary:
1. A demountable tag-line reel and support carrying about 1,300 feet Of %2-inch 19 wire galvanized aircraft-strand boat cable graduated into 10-, 50-, and 100-foot intervals by solder marks using one, two, and three marks, respectively.
2. A pair of lightweight double blocks and tackle with a parallel jaw or a winch to be used in taking up the slack in the boat cable. With the latest design of tag-line reel, the slack can be taken up by means of the reel crank.
3. A hand ax, sledge, stakes, and a short length of rope for use in anchoring the boat cable.
4. A boom consisting of a single beam constructed either of an oak timber reinforced along the sides by angle irons or of two structural aluminum channel shapes, one telescoped within the other to permit adjustments in length. This boom is equipped with a bronze sheave at the far end and must be of sufficient length to reach from a cross beam attached to the gunwales of the boat to a point at least
3 feet beyond the bow.
5. A cross beam, preferably constructed of structural aluminum angle shapes or channels, for attachment to the gunwales of the boat near the center. This cross beam should be provided with a base to permit the ready mounting of the measuring reel. It may be provided also with a guide sheave at each end under which the boat cable is passed. A combination boom and cross beam has been designed as part of the standard equipment for boat measurements.
6. Sufficient J-bolts or C-clamps to attach the cross beam to the gunwales of the boat and the boom to the cross beam.


7. A sounding reel of the type described on pages 205-206 or a specially designed reel equipped with the, necessary length of measuring line.
The recommended procedure in making a boat measurement is as follows: After the measuring section has been selected and the necessary equipment taken to the site, substantial anchorage to which the boat cable can be temporarily attached must be provided. If a tree or other object is not available near the edge of the water for an anchorage, an iron pin of the proper length and size may be driven into the, ground. To stretch the cable, the free end is fastened to the anchorage on one bank. The boat reel assembly is placed in the stern of the boat with the cable leading off the top of the reel. As the boat is propelled across the stream the cable is paid out. As the cable is paid out, the boat should be headed somewhat upstream so that the opposite bank may be reached at the desired point or slightly above it. If the cable is allowed to unreel easily it will fall to the river bottom practically at right angles to the current. When the bank is reached, the reel and support are taken ashore and the cable tightened by means of a block and tackle, or by the use of the reel and reel crank, until the cable clears the water by 6 to 12 inches. In preparation for the measurement the boat is placed under the boat cable with the bow upstream. The cross beam is then attached to the boat, and the boat cable is passed through the guide sheaves on the cross beam. These guide sheaves may be two J-bolts installed in an inverted position at each end of the cross beam, or two C-clamps attached directly to the gunwales opposite each other. Usually there is sufficient friction between the cable and guides to hold the boat in position while the observations are taken. By placing the boom over the forward gunwale slightly to one side of the stem at the bow, the operator may more easily observe the meter.
With the equipment assembled in this manner the procedure followed in making the measurement is essentially the same as for cableway measurements, and sufficient weight can generally be used to eliminate the vertical angle. The two most common factors contributing to the inaccuracy of a boat measurement are the upstream and downstream movement of the boat, which is caused by variable wind or nonuniform velocity ofthe water, and. the vertical movement of the boat, which is caused by wave action. Inaccuracies in measurements of velocity resulting from the upstream and downstream oscillation, which is usually rather uniform, may be overcome by taking velocity observations over a longer period, or by starting and stopping the velocity observations at the same point in the oscillation. Upstream or downstream wind may destroy the uniformity (if this oscillation and also create enough wave action to produce vertical movement. This vertical movement, in addition to making observations of depth


more difficult and less reliftble, may also affect the accuracy of the velocity observations, especially those taken with types of meters affected by vertical motion. For this reason, boat measurements should not be attempted when the velocity of the wind is more than 15 miles an hour.
On navigable streams the boat cable should be fastened so that it may be readily lowered to the bed of the stream. This is usually accomplished by releasing the tension through either the block and tackle or the reel crank. The boat cable should be well tagged so that approaching boats can see it. Where these precautions have to be taken an additional helper may be desirable or necessary to operate the cable.
A measurement of discharge may be made through ice cover if the ice is of sufficient thickness to support the engineer safely and if no open water measuring section is available. In some situations it may be necessary to measure the flow for a part of the distance across the channel in open water and the remainder through ice cover. This procedure should be limited to those channels that can be waded as the surface ice may be thin and unsafe where such conditions exist.
The equipment used for a measurement through ice cover must include, in addition to that used for open-water measurements, tools for cutting holes through the ice and for measuring its thickness and hangers to permit such changes in the suspension of the meter as may be found necessary. A shovel and an ice chisel or an axe are most commonly used for cutting holes. The shovel is used principally for clearing away surface snow and for removing chips of ice from the hole during the process of cutting. The use of either an ice chisel or an axe is a matter of preference where the surface ice is thin, but for a thickness of ice greater than 1 foot an ice chisel is indispensable. If a suitable ice chisel cannot be purchased, one can be constructed by a blacksmith from an ordinary steel bar. It should be at least 4 feet long, should be looped at one end to permit easy gripping, and should weigh at least 14 pounds. The blade should be wedge-shaped and drawn out to a section about 1/4 inch in thickness and 3 inches in
idth. The blade must be tempered so that it can be ground to a keen edge and readily sharpened with a file or stone. The edge should be sharpened on one side only and the cutting surface kelpt straight. To assist in keeping the lower part of the hole free from ice chips, the lower 6 inches of the chisel may be so bent back that the cutting edge is not more than%6of an inch to the rear of the chisel above.
A scale about 4 feet long, graduated in feet and tenths and having an L-shaped projection at the lower end, is generally used for measuring the thickness of the ice and the distance from the water surface


A4 >4 04 jr



00 'tat




Ilk k w



U u Z


to the bottom of the ice. For use in a newly cut hole, the horizontal part of the L should be at least 4 inches long so that it may vxlelld beyond any irregularities on the under side of the ice. If the holvs are used for stibsequent measurements, the lower cd(_-rcs may beconle so rounded that the horizontal part of the L should b( about I foot long for reliable determinations of thickness of ice.
The holes cut throtigh the ice should be elliptical, with the inajor axis of the ellipse parallel to the current, and should be no larger than is necessary to perinit the raising and lowering of the meter. For most efficient Cutting, the worker should strike with the chisel so that as many of the chips as possible will fly clear of the hole. This can be best accomplished if the worker cuts with the blade slightly inclined toward the center of the hole as he makes the first cuttings around its circumference, and with the blade inclined toward the outer edge of the hole as he makes the second round of cuttings. The groove thus completed will be several inches deep and coniparatively free from chips of ice. The solid central part can then be arred loose usually in one piece, by striking it a wedging blow with the chisel. This procedure is repeated until the bottom of the ice is reached, although special care is required to complete the cutting before an appreciable amount of water is admitted into the hole. The entrance of the water may be largely prevented by cutting the last groove to a uniform depth just above the bottom of the ice and striking the center part with a sharp blow to loosen it. The remaining pieces of ice will Ifloat to the surface and may be easily pushed aside with the chisel or shovel.
It is usually advisable to chop the first hole through the ice in the middle of the measuring section and then proceed to the quarter points. This practice may lead to the detection of slush ice and so enable the engineer to investigate another section if necessary before much time and effort have been expended. On small streams where measurements of velocity are desired at many points close together, the ice may be removed from the entire measuring section, thus materially reducing the vertical pulsation of water that sometimes occurs at holes cut through ice cover.
A measurement through ice cover may be made by either a rod, a hand-line, or a reel suspension. A meter supported on a rod may be used satisfactorily where depths are shallow and temperatures such that ice will not readily form in the meter bearings when the meter is removed from the water (see pl. 67 A). If the temperatures are subfreezing and if the regular wading assembly is used, it is better practice to keep the meter at a fixed position on the rod throughout the measurement, say at a point 0.5 foot above the base plate. If the surface of the ice or water is used as an index point, the meter


may be placed at the proper position in the vertical for velocity observations without removing it from the water until it is transfe ired to the next vertical. If the meter is retained in the fixed position on he rod, the assembly may, in the absence of the L-shaped scale, be used for determining the thickness of the ice and the distance from I he water surface to the bottom of the ice.
A small portable reel assembly mounted on a tripod, stand, oi- sled1 in such a manner that the weight and meter can be raised and loweredl directly over the hole in the ice is useful in making this type of measurement. Reel and sled equipment used in making discharge measurements of the Mississippi River at St. Paul, Minn., through ice cover is shown in plate 6, R. The tripod or stand should be collapsible and have adjustable legs by which its height and spread ma~y he regulated. These features of design assist in making the. equipment readily transportable. A hand-line assembly may be conveniently used if the hand-line consists wholly of rubber covered cable and if the velocities are such that reliable soundings are obtainable with a 15- or 30-pound soundling weight. When either the hand-line or rel suspension is used during subfreezing temperatures, the observations of depth and velocity in a vertical may be obtained without exposing the meter to the atmosphere once it has been placed in the water. This may be accomplished by tagging the measuring line at a known distance above the horizontal axis of the meter and by using this point as an index for measuring the depth and for placing the meter.
The two-conductor system as used in making open-water measurements of discharge is generally employed with satisfactory results in measuring through ice cover with either a hand line or a reel assembly. Satisfactory results may also be obtained with the single-conductor system if the ground wire is placed in running water while the observations of velocity are being made. Some difficulty may be experienced with the single -conductor system if there is a. layer of frazil or anchor ice (p). 69) under the surface ice, unless the ground wire extends into the running water.
After holes have been cut and positions of the verticals established, observations of depth and velocity are started, The thickness of the ice and the distance from the water surface to the bottom of the ice are measured at both the upstream and the downstream ends of the hole. If the measurements differ, the mean of the two measurements is recorded. The total depth of the water is then measured, and from this depth the distance from the water surface to the bottom of the ice is subtracted to obtain the effective depth. If there is a layer of frazil or anchor ice below the surface ice, its thickness must also be determined. The lower limits of the layer are found by raising


the meter upward f romi the bottom of the stream until it f ails to register any velocity. TJhe distance from the meter in this position to the top of the water surface is measured and subt)1racted froin the total dep)thI to obtain the effective depth.
Th~le method selected for measuring the velocity will depend~ on the type of meter suspension used and the effective depth. If the effective dIepth is 2.0 feet or mfore and the meter is supported on a rod, the twopoint method should be used. If the meter is suspended fr-om a hute and is placed 0.5 foot above the bottom of the sounding weight, the two-point method should be used for all depths of 2.5 feet or greater. Where the two-point method is not p~racticalble, the velocity observalion may be taken at either 0.5-depth or 0.6-depth, and coefficients may be applied to reduce the observed velocity to the mean in the vertical, these coefficients being less than unity. Tfhe results of experiments by the Geological Survey 2 Show that for velocities under ice cover, the average coefficient for a 0.5-depth observation is about 0.88 and for a 0.6-depth observation about 0.92. Wherever practicable it is advisable to establish the coefficient by defining one or two vertical velocity curves in sections where there is sufficient depth below the ice.
The procedure of placing the meter for velocity observations will be simplified if the percentage depth is determined from the bottom upward regardless of the type of suspension used. The distance that the meter must be raised when the weight or the base plate of the rod is on the bottom is computed by multiplying the effective depth by the desired percentage depth above the bottom and by subtracting from the figure thus obtained the distance from the horizontal axis of the meter to the bottom of the rod or weight.
In subfreezing temperatures where the stage is likely to remain reasonably constant during the measurement, it is advisable to complete all correlated steps in the procedure, such as the determination of distances between verticals, the measurement of depths, and the preparation of forms f or notes, before making the velocity observations. In this manner, inconveniences caused by the formation of ice in the meter bearings and the collection of f razil in the meter cups will be appreciably reduced.
The vertical pulsation of water in the holes cut in thie ice must be given careful attention in determining depths and velocities. Pulsations of half a foot or more are not uncommon and may cause appreciable error if they are not carefully observed and averaged. Because of pulsations the meter should be held as far upstream as possible when the velocity is being observed. Where pulsations are large,
2 Barrows, H. K., and Horton, R. E., Determination of stream flow during the frozen season. U. S. Geol. Survey Water-Supply Paper 187, 92 pp., 1907.


inaccuracies in depths and placement of meter may be reduced by using, as an index, the top of the ice or a rod placed across the hiole;or, if the depths are shallow, the L-shaped scale may be Used to measure the effective depths directly. It is desirable to observe( thje revolutions of the current meter over a longer period of time thlan is usual in open water measurements.

Experience has shown that the following precautions will increase the accuracy of discharge measurements by eliminating sources of error.
Test the meter immediately before and after each discham-re measurement, by submitting it to a spin test in quiet air. Eor use in velocities of less than 1, foot per second, the meter should have an initial spin test of at least 2 minutes and should slow down gradually in coming to a stop. Only in streams of high velocity or in streams heavily laden with sand and silt is it advisable to use a meter with a pmi test of less than 1 minute.
The current meter should be inspected at frequent intervals. As meters hold to their ratings only when operating freely, it is essential that~ they be kept in good operating condition at all times.
The cross section of the stream at the place of measurement should be divided into a sufficient number of partial sections so that the measuseqMenits of depth will develop an accurate profile of the bed and so lint the measurements of velocity may be made close enough to each other for an accurate determination of velocity in each of the partial sections. Where the velocity is not uniformly distributed, a larger number of sections will be needed than where uniformity prevails. The number of necessary verticals will depend therefore upon the roughness of the bed and the variations in the velocities (see p. 68).
The stop watch used in timing velocity observations should be checked frequently, as accurate timing is obviously essential.
In slow and irregular velocities, if the stage is not changing rapidly, observe the revolutions of the current meter over a longer period of time than usual.


Measurement of depth at each of the selected verticals in the measuring section is essential, both in determining the area of the cross section and in the correct placing of the meter. The method used in measuring depth depends largely on the manner in which the meter is suspended or supported. The depth must be the true vertical distance from the surf ace of the water to the bed of the stream.


Measurements of depth and velocity at each vertical are us,,ually made iii successive operations. For this reason the following (1Scussion not only includes the various methods of suspendling theC Meter and measuring the depths but also describes briefly the assembily of the equipment for each type of meter suspension.


The determination of depths and velocities by a current meter supported on a graduated rod is restricted generally to measurements made by wading; to measurements made from complete ice cover where depths and velocities are not excessive;- to measurements made from a boat; and to measurements made in canals, flumes, and narrow streams that are spanned by footbridges placed just above the surface of the water. The rod suspension cannot be used where the velocity is so great that the rod and meter cannot be held in position by the engineer. A rod 4 or 5 feet long attached to a base plate is used for the ordinary wading measurement. The base plate largely prevents the rod from sinking into the stream bed or from entering narrow crevices in the bed that are not representative of the general profile of the measuring section, and thus it eliminates certain tendencies to obtain soundings -that are too large. The rod may be round, half an inch in diameter, in one piece or made up of 1-foot or 1 /-foot sections joined together. Flat rods of various designs are used also (see pl. 2,B). The current meter, 'with a sliding support inserted between the tailpiece and the meter, is mounted on the rod. Care must be taken to observe that the bucket wheel rotates in a horizontal plane at right angles to the rod. This sliding support is so designed that when unclamped it can be easily moved along the rod, and when clamped it will hold the meter firmly at the desired position.
Occasionally there are places where mud or shifting sand at the measuring section makes it impossible to rest the base plate of the rod on the bed of the stream, or where the collection of ice on the rod is sufficient to interfere with the placement of the meter. For these conditions it may be necessary to dispense with the base plate and to hold the rod in position by hand, first clamping the meter at a selected position on the rod.
Revolutions of the current-meter rotor are electrically indicated through a telephone head set (p. 207) connected to two small insulated wires of suitable length, one leading to a binding post on the head of the meter and the other to the attachment cap on top of the rod or to the hanger screw on the meter. A small dry battery is introduced into the circuit to provide current for actuating the telephone receiver. but s5,tisfa.ctory results can be obtained without this battery.


Depths are measured by reading the position of the water surface on the rod when the base -plate rests on the bed. In high velocities the water will tend to pile up on the upstream face of a rod and to draw down on the downstream face, so that the observations of depths may be in error if this fact is not given proper consideration. Under this condition, the accuracy of the measurement of depth mtay be increased by standing at one side of the meter and viewing the rod across and at right angles to the direction of flow and by observing the position of the water surface on the side of the rod. For very small depths it is advisable to slide the meter clear of the water before the sounding is made in order to eliminate surf ace disturbance caused by the current-meter bucket wheel. Although the smallest graduation on the rod is one-tenth of a foot, shallow depths should be read to half -tenths or hundredths of a foot if the smoothness of the measuring section will permit. A folding rule graduated to hundredths of a foot is convenient for m-eastiring shallow depths.

The current meter and sounding weight are generally suspended from a hand-linie if accurate soundings are obtainable with a 15- or a 30-pound weight. This procedure is adaptable for measurements made from a bridge, a cablewlay, a boat, or ice; but its use is limited to small depths and comparatively low velocities. If vertical angles are too large to be neglected and if the required weight is more than can be conveniently manipulated by hand, hand-line suspension should be replaced by reel suspension.
A hand-line consists of joined sections of two types of cable, one section for use above the surface of the water and the other sections for use below. For ease in handling, the upper part is a rubber-covered cable about 3/8-inch in diameter, which contains two insulated copper strands. The underwater part is usually a strong, flexible wire cable about '/1-inch in diameter, which is used to reduce to a minimum the resistance offered to the moving water. The cable has an insulated copper core for use with the two-way electrical circuit., A small reel where the two sections are joined permits adjustment of the underwater part to the depth of the stream. This reel also provides a means of electrically connecting the two cables. Occasionally a square knot is used to join the cables, but it provides no means for readily adjusting the underwater length. At the upper end of the hand-line the copper strands are attached to the terminals of a connection plug, which permits ready connection to the telephone head set. The hanger, to which the meter and sounding weight are fastened, is discussed below, as are also the small connector that joins the measuring line to the hanger and the manner in which the meter is electrically connected to the measuring line.

The practicable assembly for measuring high velocities and great depth is a current meter and sounding weight suspended from a reel and line. This type of suspension is convenient and simple and utilizes sounding weights that are too heavy for use on a hand-line.
A reel suspension is most commonly used from a cableway or bridge, although it is adaptable for measurements made from a boat or an ice cover. Suitable apparatus for supportimg the reel and for guiding the measuring line must be provided for each of these types of measurements.
A typical reel assembly consists essentially of a drum of known diameter, around which the measuring line is wound; of a counter calibrated to the diameters of the drum and cable for measuring the depths; and of a crank and brake with which to operate the reel. These parts may be assembled and supported on a base in a portable unit. The measuring line is a small metal cable, the diameter of which depends on the strength needed. It has an insulated copperstrand core, which makes possible the completion of a two-way electrical circuit. A small cAmnector fastened to the lower end of the cable provides for its attachment to the weight hanger. Before the line is fastened to the connector the end of the copper-strand core is freed so that it may be readily joined to another small insulated wire strand of sufficient length to reach the binding post on the meter.
A typical hanger from which the sounding weight and meter are suspended is designed to permit the placement of the current meter in certain definite positions relative to the bottom of the sounding weight. These positions, depending on the type of hanger used, are such that when a 15-pound sounding weight is used the horizontal axis of the meter is 0.5, 0.7, 0.9, or 1.0 foot above the bottom of the weight; when a 30-or 50-pound weight is used the axis of the meter is 0.7, 0.9, or 1.0 foot above the bottom of the weight; and when a 75-, 100-, or 150-pound sounding weight is used the axis of the meter is 1.0 foot above the bottom of the weight.
Although a two-conductor electrical system is generally used with either a hand-line or reel assembly, a single-conductor system that includes a ground return may be adaptable for emergency or general practice. In an emergency the two-conductor system can be converted into a single-conductor system by electrically insulating the weight hanger from the measuring line and electrically connecting the measuring line to the binding post on the meter head. If the single-conductor system is used in general practice, the measuring line may be the same


as used for the two-conductor system, or it may be a smaller wire cable or strand without the- insulated copper core.
If tho single-conductor system is used f rom a cableway the electrical circuit may be completed through the cable, provided the anchorages are well grounded. The battery and head-set assembly must then have an extension consisting of two insulated wires, one leading from the measuring line to the battery and head-set assembly and the other from this assembly to the cableway. A battery clip or small thin sheet of lead or zinc attached to the end of the wire that is to contact the cableway facilitates the completion of the circuit. In some situations a short piece of copper wire permanently attached to a cable-car hanger bolt will suffice if the bolt is insulated from the reel. The battery clip or sheet of metal, when gripped about the cableway by hand, will make a satisfactory contact if the cableway is free f rom paint and rust; otherwise the wire used to ground the circuit should be attached by means of a sharp instrumenCtliat will insure a good contact. A small radio ground clamp with a pointed setscrew is convenient for this purpose. If the rehirn circuit cannot be completed. through the cableway, a ground attached to the battery and head-set assembly and extending beneath the surface of the water may be substituted. A piece of copper or zinc fastened to the submerged end of this line is usually needed to improve the ground.
In using a sing] e-conduct or system from a bridge the return circuit is commonly made with the ground line as mentioned above, although sometimes it is possible to complete the circuit through a metal handrail or drain pipe effectively grounded. Bridge measurements by this system should be made from the downstream side to avoid possible short circuits from interfering bridge members. The wire used to contact the handrail or drain pipe should also be attached to a sharp instrument so that any rust or paint on the surface may be easily penetrated. If the metal, particularly on the handrail, cannot be reached without marring its surface, a wire the length of the bridge and grounded at the ends can be used to .complete the circuit.
The successful application of the single-conductor system when using a battery as a source of electrical current depends largely on the depth of the stream and the chemical composition of the water. As the leakage in the electrical circuit increases with the depth of water, the identification of the revolutions made by the rotor of the current meter becomes'more difficult with increasing depth. Basic or acidic water assures better conductivity than relatively pure water. At times thissystem may yield favorable results without a dry cell; but a dry cell in the circuit is recommended, with the polarity reversed when necessary, as explained below.





Without a dIry cell in the cicut the source of electromotive force may be considered as a wet volfai( cell in which the meter :usse(Iiily, or a part of it, acts as one electrode; a trailing piece of metal, grotind cable, or insulated part of the meter assembly acts as the other- (leetrode; and the river water acts as the electrolyte. This lnethiod was developed by the Geological Survey in its Montana district andl was later used in its California district, where B. C. Colby and( HI. M. Orent (om.ducted extensive experiments on its adaptability, with the meter sii)ported on a rod, by a hand line or by a reel, using either the twvo-condluctor or the single-conductor system. It was found that the current in the wet voltaic cell, for a given hookup and depth, has a definite direction, which may be the same or the opposite of that supplied by the dry-cell battery. As the condition of a wet voltaic cell is also in effect when a dry cell is used in a single-conductor system, it is desirable that the dry cell be connected so as to aid and not oppose the wet call action. In some single-conductor hook-ups there is a definite but narrow zone in which no meter signals will be heard. This may occur either with or without a dry cell in the circuit.
In studying the action of the wet voltaic cell alone with various hook-ups it was found by Colby and Orem that the current through the telephone receiver reversed as the meter was lowered from the top to the bottom of the stream, and around the point of reversal was a zone of weak signals. A strength of signal that does not vary appreciably with depth can be had by increasing the size of the zinc electrode by attaching an insulated piece of zinc to the back of the sliding support or by using an insulated meter tailpiece constructed of zinc, these additions to the zinc electrode being connected to the galvanized suspension cable above the insulator on the meter hanger. The hook-up for this arrangement, except the head-set assembly and the use of an additional wire to connect the insulated hanger or tailpiece with the measuring line, is the same as for the single- conductor systems
A convenient head-set assembly for this hook-up, which is adaptable to rod, hand-line, or reel suspension, is as follows: A 7-foot length of small rubber-covered two-strand extension cable is connected with the head-set and thel insulation removed from the free ends for about 6 inches. These free ends are made into terminals by coating one end with solder and attaching the other to a small galvanized or zinc battery clip.
When this telephone assembly is used with rod suspension one of the loose terminals is attached to the binding post of the meter


contact chamber andl the other is clipped to a9n insulatedtapic or to the Sli(Iifg support. The electrolytic action, between, the bjattery clip and the meter asset mbly will usually produce sufficient current to give satisfactory signals. If at times it, is foundA necessary to strengthen the signals, this can be done by r~esetting thie battery clip or by fixing a strip of zinc around the iwsulatedl slidling support to provide additional electrode area. The battery-clip electrode and insulation are placed behind the rod on the meter assembly so that there will be negligible interference with the flow past the meter.
When this assembly is used with hand-line or reel suspension from a bridge or cableway, the loose terminal of the telephone cable
-is fastened to the terminals of the hand-line or to the frame of the sounding reel. If the usual two-conductor rubber-covered cord is used for a hand-line, it is convenient to twist the two strands together and attach them to a battery clip, which in turn may be clamped to the loose terminal of the head-set cable. 'The batteryclip terminal of the head-set cable is then fastened to the best available ground, which is usually the cableway or handrail. For this purpose it is desirable to have about 3 feet extra of insulated cable, stripped of insulation at one end and attached to a battery clip at the other, for use as an extension from the sounding reel or as a ground wire when a direct contact is made with the cableway (or metal bridge rail.

In determining the mean velocity in a section used in making a discharge measurement, the mean velocity normal to the measuring section must be measured at each vertical. This is done with the current meter by measuring the velocity at specific points in the vertical. The number and position of the selected points with reference to the total depth identifies the method used in measuring the velocity.
After the current meter is placed at a selected point in the vertical it should be permitted to become adjusted to the current before the observation of velocity is started. The time required for such adjustment is usually only a few seconds if velocities are reasonably high. In low velocities, particularly if the current meter is suspended by a cable, this adjustment may require a minute or more. The failure of the operator to wait for full adjustment ma~y lead to inaccurate results.
The number of revolutions made by the rotor of the current meter is ordinarily observed over a period of 40 to 70 seconds except for velocities less than 1 foot per second, in which a longer run is usually taken. Observation of time to the nearest half-second is


usually sufficient. The observed number of revolutions and the responding time are converted into velocity in feet per second by use of a rating table for the current meter.
In the vertical velocity-curve method a series of velocity observations at points well distributed between the surf ace of the water and the bed of the stream is made at each of the selected verticals. If there is considerable curvature in the lower part of the vertical velocity curve as indicated by the intervals between revolutions of the currentmeter rotor it is advisable to space more closely the observations in that part of the depth. The velocities measured in each vertical will, when plotted against depth, define the vertical velocity curve from which the mean velocity in that vertical can be determined. In order that vertical velocity curves at different stations may be readily Comparable, it is customary to plot the curves with proportional parts of the depth as ordinates. Typical vertical velocity curves at stream-flow measurement stations are shown in figure 1.
Studies of vertical velocity curves made under widely different hydraulic conditions show that their shapes usually correspond to part of a parabola, the axis of which is parallel to the surface of the water, coincides in general with the filament of maximum velocity, and is generally between the surface and one-third of the depth. The velocity decreases gradually upward from the axis to the surface and downward nearly to the bottom, where the change becomes more rapid as a result of the friction and turbulence produced by the bed. For the same mean velocity and different depths of water, the curvature of the vertical velocity curve decreases as the depth increases. For the same depth and different velocities, the curvature increases as the velocities increase.
In making observations of velocity for constructing vertical velocity curves, three of the points selected for observations should be at distances below the surface equal to 0.2, 0.6, and 0.8 of the total depth, so that the results obtained by the vertical velocity curve method may be compared directly with those obtained by the two-point and 0.6-depth methods. Observations made near the surface should always be checked, because irregularities in the movement of the, water are likely to occur in that part of the vertical and because the current meter is erratic in its action if it is only partly submerged. Velocity observations at points in the vertical frequently give erratic results, owing to
pulsations and surges affecting the distribution of velocities. These pulsations or surges vary the velocity of the water in cycles ranging
$Murphy, E. C., Accuracy of stream measurements: U. S. Geol. Survey Water-Supply Paper 95, pp. 28-32, 1904.


V4M 4J 00) V


c 0 g : 0-c

c V)o

.2 L6 "ll

w u*) 3 .?colo CC cc Ic 0.00-- CL 04.,.- 0 W
0-wo 8to

< "141 Zr'14 w cd Z w
0 u &
QD 0
(r _j
W j 4J to -to

%0 IAJ 0 cr C-c

120 .T

> t3 v

0 U FE-4
w N"
.................. > b i


4 < L

C-c 2 ,


.............. ..

> L

L -10
(:) 3.
tni EN Cc

0 0 1 10
ci 0 0
H.Ld3a W101.40 18Vd WN0118W08d


from a few seconds to several minutes, depending on the nature of the stream. To obtain a reliable vertical velocity curve under such conditions, it is necessary either to run each velocity observation long enough to averaLe the effects of the pulsation and surge or to determine the velocity limits of the cycles. These limits may be approximated either by taking several short velocity observations while the meter is at one position in the vertical or by increasing the number of observation points in the vertical while decreasing the time required for each observation, so that two or more velocity readings are obtained within the same phase of the cycle at two adjacent depths. Although either of these procedures may result in reducing the probable accuracy of an individual velocity observation, the information thus obtained is useful in systematically weighting the individual velocity determination when constructing the vertical velocity curve.
In constructing the vertical velocity curve the velocity observations are plotted with the velocities as abscissas and depths below the water surface as ordinates. The mean velocity is determined from a smooth curve drawn as nearly as possible through the plotted points. With a sufficient number of velocity observations available to define a velocity curve for each vertical, the discharge may be determined by plotting on a graph of the cross section the velocities at different, depths as indicated by the vertical velocity curves. Lines of equal velocity are then drawn and the areas between them are measured by means of a planimeter. The sum of the products of the area and the average velocity for each area is the discharge in the section. This method of determining discharge is described in textbooks on hydraulics.
The vertical velocity-curve method is valuable in determining coefficients for application to the results obtained by other methods. The method is not generally adapted to the making of routine discharge measurements, chiefly because the apparently increased precision thus obtainable may be more than offset by errors resulting from changes in stage during the longer period of time needed for making such measurements.
In the two-point method of measuring velocities, o1w-ervations are made in each vertical at 0.2 and 0.8 of the depth below the surface. The average of these two observations is taken as the mean velocity in the vertical.
The two-point method is based on many studies of actual observations and also on the theory that the vertical velocity curve corresponds to part of a parabola with axis horizontal at the point of highest velocity, for which it may be mathematically demon-


strated that the average of the velocities at 0.2114 and 0.7886 of the depth is equivalent to the mean velocity.4 Studies of many vertical velocity curves made for different depths, velocities, and condlitions of stream bed support this theory. Experience has shown that this method gives more consistent and accurate results than any of the other methods, except the vertical velocity curve method when it is used under measuring conditions of constant stage and steady flow. Because of the support given the two-point method by both theory and practice, it is generally used by the Geological Survey in current-mieter measu renments of discharge.
There aire, however, a few situations where correct results are not obtainable by the use of the two-point method. One situation relates to the use of the vertical-shaft cup-type current meter, which und~erregYisters near the surface and near the bed of the stream, so that for shallow depths a coefficient greater than unity may be required, as shown by Pierce.5 Trhe coefficient may vary with both depth and velocity. With the cup-type meter it is generally not advisable to use the two-point method in depths of less than about 2.0 feet unless a coefficient is applied. Occasionally, conditions, may necessitate the application of a coefficient less than unity to obtain the correct discharge, as for example in deep water immediately above a dain where the measuring section is sloping upward. Before any coefficient is applied, however, its applicability should be thoroughly established by vertical velocity curves for the entire range of conditions covered by the measurement or by such other data as are available.
In the 0.6-depth method an observation of velocity is made in each of the selected verticals at 0.6 of the depth below the surface. This method is based on the theory that the vertical velocity curve corresponds to part of a parabola with the maximum abscissa within the upper third of the ordinate representing the depth. On this basis, the mean abscissa lies between 0.58 and 0.67 of the depth below the surface. If the maximum abscissa is in the upper fourth of the measured depth, the 0.6-depth ordinate is very nearly the mean. Although a large percentage of velocity curves that have been studied indicate that the mean velocity in the vertical is at approximately 0.6 of the depth below the surface, experience on
4 Pardoe, W. S., Methods of stream gaging: Eng. News, vol. 75 p. 889, 1916; Liddell, W. A., Stream gaging, pp. 39-40, 1927; H~oyt, John C.. The use and care of the current meter as practiced by the United States Geological Survey : Am. Soc. Civil Eng. Trans., vol. 66 (1910), p. 9T and 110 (discussion by HI. K. Barrows).
SPierce, C. H., Performance of current meters in water of shallow depth : U. S. Geol. Survey Water-Supply Paper 868-A, pp. 3.5,'1941.


certain streams, particularly those of great depths or with smooth beds, has shown that the results obtained by this method tend to be slightly greater than those obtained by the two-point method. Under those circumstances it is possible that the maximum abscissa for several of the selected verticals in the measuring section may be more than one-fourth of the depth below the surface. Laboratory investigations of performance of current meters in water of shallow depth indicate that vertical-axis cup-type current meters when used at 0.6 of the depth give results that are too small (requiring coefficients greater than unity) for velocities of 0.3 foot per second and less and also where depths are 0.5 foot or less.
Although the 0.6-depth method generally gives fairly satisfactory results, nevertheless as the variations of individual observations may be somewhat greater than those shown by the two-point method it is used by the Geological Survey only if the two-point method is found impracticable because of insufficient depth or for other reasons such as a rough stream bed or aquatic growth.
The 0.2-depth method consists of an observation of velocity in each of the selected verticals at 0.2 of the depth below the surface and is used only when the depths, velocities, or other causes do not permit the use of the methods previously described. The depth is obtained from a standard profile of the measuring section or, if the cross section is not permanent, from soundings made before the velocity observations are taken or as soon thereafter as possible. Discharges are determined in the following manner: The measurement is computed by using the 0.2-depth velocity observations without coefficients as though each was a mean in a vertical. The approximate discharge thus obtained (called for convenience the 0.2depth discharge) divided by the area of the measuring section gives the weighted mean value of the 0.2-depth velocity. Studies of many measurements made by the two-point method show that for a given measuring section the relation between the mean 0.2-depth velocity and the true mean velocity either remains constant or varies uniformly with the stage. In either case this relation may be determined for a particular 0.2-depth measurement by recomputing measurements made at the site by the two-point method to obtain the observation values of the 0.2-depth discharge and the mean 0.2depth velocity. The plotting of the true discharge and the 0.2-depth discharge as coordinates for each measurement will give a dischargerelation curve, and the plotting of the mean velocity and the mean 0.2-depth velocity as coordinates will give a velocity-relation curve.


These curves may bc extended to determine the true discharges coiresponding to the 0.2-depth discharge either by reading directly from the discharge-relation curve or by multiplying the true mean velocity obtained from the velocity-relation curve by the area of the measuring section.
It has been found that many of these relation curves are practically straight lines passing through the point of origin, thus showing an approximately constant relation between the 0.2-depth values and the true values of mean velocity and discharge. However, where backwater or overflow conditions affect the flow or where the distribution of the flow changes decidedly with the change in stage, the discharge-relation curve may give a more accurate extension. If the measuring section shifts materially and if the velocity distribution is not seriously disturbed by change in stage, an extension of the velocity-relation curve may be the more reliable. It is therefore advisable to plot both these relation curves and select for extension the one more nearly approaching a straight line.
Results obtained by this method are reliable provided the discharge relation and the velocity relation do not change materially between the time of definition of the curves and the time when the measurement is made. The relation curves for shifting channels should be checked frequently by data obtained from measurements made by the two-point method.
The 0.2-depth method may be advantageously used under conditions where high velocity, floating ice or debris, or inadequate measuring equipment makes it impracticable to obtain reliable velocity observations at 0.8 of the depth; also at times when the stage is changing so rapidly that it is desirable to complete a measurement in as short a time as possible. This method is generally considered superior to the subsurface method because the relation between the mean 0.2-depth velocity and the mean velocity is usually more nearly constant and more easily determined than the relation between the mean subsurface velocity and the mean velocity. It is also useful in studying the accuracy of a measurement made by the two-point method, especially if the uncertainty relates to the 0.8-depth observations. The method should not be used if the depths are so small as to bring the meter close to the surface of the water. Appreciable errors in velocity because of uncertainty in regard to placing the meter at the exact 0.2-depth position are not likely to occur, even if soundings are roughly determined, because most vertical velocity curves for depths where this method might be required show little curvature in the vicinity of the 0.2-depth point. The results obtained by this method are considered more reliable than those based on coefficients determined by vertical velocity curves occasionally made at the station.


In the three-point inethod the velocity observati1o11, which are madt~e at 0., 0., and 0.8 of the depth below the surface, combine the twopoint and the 0.6-depth methods. In this method the 0.2- and 0.8depth observations may first be averaged and the result 'averaged with the 0.6-depth observation or, if it is desired to give more wNeight to tile mean of the 0.2-depth and 0.8-depth observations, the arithmetical mean of tile three observations may be used. The two-1)oift method is considered more reliable than the 0.6-depth method; thlerefore, in using the three-point method it is probable that additional weight should be given the mean of the 0.2-depthi and 0.8-depth observations to the extent of using the arithmetical mnean of tile observations at the three points. The method is used principally for- comparison withl other methods, rdthougih its use may be desirali fvlcte appear' to be abnormally distributed in a vertical or if the 0.2-depth observation is near tile surface and the 0.8-depth observation is made ill that part of the depth where tile velocity is seriously affected by friction or by turbulence produced by the stream bed. The method is based on the assumlpt ion that tile mean velocity obtained by the twopoint method alone is too small and by the 0.6-depth method alone is too great, and that an average of the results obtained by the two methods represents more nearly the true value. Its use is recommended only for unusual conditions where other methods are not entirely applicable.

In the subsurface method a velocity observation is made in each of the selected verticals at a uniform depth below the surface. This depth should be at least 2 feet below the surface and preferably at a greater depth for deep, swift streams in order to avoid the effect of surface disturbances. The measured velocity must be multiplied by a coefficient to reduce it to the mean velocity in the vertical. Whether this coefficient should be applied to each velocity observation or to the computed discharge for the measurement would depend upon information obtained from vertical velocity curves. If the vertical velocity curves are well distributed across the section and show variable coefficients, a higher degree of accuracy will be obtained if the coefficients are applied separately to each measured velocity by using the coefficient deemed most applicable to each vertical. If the coefficients are fairly uniform, an average coefficient can be applied to the total discharge. It should be kept in mind that these coefficients are likely to vary with the stage, depth, and position in the measuring section.
As it is generally difficult to determine accurately the exact coefficients for use with the subsurface method, results of high accuracy


cannot be expected by this method. It should be employed only where it is impracticable to obtain reliable soundings or where the 0.2-depli method is not adaptable.

In the integration method the current meter is lowered and raised at a uniform rate in each of the selected verticals in the measuring section. The number of revolutions is timed over two or more complete cycles, and the result when converted into velocity is the mean velocity for the vertical. This method may give accurate results if sufficient care is used in its application. However, as vertical movement of a vertical-shaft cup-type meter affects the motion of the rotor to some extent, the method is not recommended for general use. In order that, the vertical movement of the meter shall not seriously affect its rotation, the rate at which the meter is raised and lowered must be limited to a small percentage of the average velocity of the water. The integration method requires additional precautions on the part of the engineer and possibly additional assistance, which are added reasons why it is not particularly adaptable to routine stream gaging. Its only use by the Geological Survey is for purposes of comparison.
In the one-point continuous method the velocity is measured continuously at a point in a selected vertical with a current meter which is provided with a recording apparatus. The position of this point, which varies with the depth, should be in the region of the higher velocities and should be between 0.3 and 0.6 of the depth below the surface to insure the best results. This measured velocity is averaged, usually for 24-hour periods, and a coefficient is applied to obtain the average mean velocity for the period.
The coefficient to be used in obtaining the mean velocity is determined from discharge measurements made at different rates of flow at the section in which the recording current meter is placed. The revolutions registered by the recording current meter are observed for the actual time of each discharge measurement and converted into average velocity in feet per second for the period of observation. The coefficient is then computed by dividing the mean velocity obtained from the discharge measurement by the average velocity measured by the recording meter in the selected position. The coefficient may vary for different rates of flow if the distribution of velocity in the measuring section changes materially. Where such a condition prevails, a velocity-relation curve is developed by plotting the mean velocity for the discharge measurement as the ordinate and the average velocity measured at the selected point as the abscissa.


The accuracy obtained by this method may be of a high order, but its adaptability is so limited that its use is restricted to phices 'NN'Ilere the discharge is largely a function of the slope and where the variatioll in stage is relatively small, such as in some canals, tailraces, flumesawl lake outlets. The coefficients or velocity-relation curves used in this method can generally be defined easily and accurately, and a, record of the revolutions made by the recording current meter can be obtaiiied either by an elect really -control led counter or by graphs made on a time chart. In the application of this method the operator must be sure that the recording current meter is always placed at the selected depth, that it is kept free of algae and floating debris, and that its mechanical operation is unif ormly maintained. This method has been used by the Boston district of the Survey in determining daily discharge at the outlet of Lake Whinepesaukee at Lakeport, N. H.
Measurements of depth and velocity of streams that are both deep and swift require additional precautions and special equipment. Accurate determinations of depth and of mean velocity in the verticals are of primary importance in obtaining an accurate measurement of discharge because the discharge is the product of the area and the velocity; so an error in the measurement of either would result in a corresponding error in the discharge measurement. Those parts of the cross section that have the greatest depths and velocities usually carry the major part of the total discharge, and it is in those sections that accurate soundings are most difficult to obtain.
Improvements in cable cars, booms, reels, sounding weights, and suspension lines have increased the accuracy and efficiency of streammeasurement work, particularly on streams of high velocity and great depth. There is, however, a practical limit to the extent to which the difficulties caused by high velocities and great depths can be overcome by the use of the available equipment, and when that limit has been reached under the usual methods of operation, other methods must be used to increase the accuracy of soundings and the precision of placement of the current meter.
In rivers of high velocities and great depths where the current meter and weight drift downstream from the vertical, the accuracy of the measurement may be increased by measuring the angle made between the line and the vertical and by applying corrections to the indicated depth to offset the effect of the inclination of the line above the water surface and the effect of curvature of the line between the surface and the weight. Thus it is possible not only to obtain a more accurate determination of depth but also to place the meter more nearly in the desired position. The use, of tags or markers


placed at known intervals on the measuring line will expedite measurements in comparatively deep water under certain conditions.
In streams of high velocity and comparatively shallow depth the use of stay-line equipment will be of assistance in measuring the depth and in placing the meter.
Measurements of depth made by the usual methods are too large if the depth and velocity are such as to cause the weight and line to drift downstream from the vertical. The downstream drift of the weight and line will place the weight downstream from the vertical when it reaches the river bed, causing the sounding line to be curved from the water surface to the weight and to be inclined above the surface. The length of line under water is therefore greater than the vertical depth, and the measured line includes the effect of the inclination of the line above the water. The excess in length of the curved line over the vertical depth is indicated by the vertical angle made by the line at or above the water surface, and the excess in apparent depth caused by inclination of the line above the water is a function of the same angle and the vertical distance between the surface and the point of suspension of the line. Therefore the vertical angle of the line at or above the surface and the vertical distance from the surface to the point of suspension of the line must be measured as additional observations in the process of obtainiiia a correct measurement of depth.
The error that may occur in such a measurement of depth is illustrated in figure 2, which shows the position of a sounding weight and line where the depth and velocity are such as to cause the equipment to drift downstream. The error consists of two parts if the structure supporting the equipment is not at the water surface. If the index on the measuring reel is read when the sounding weight is at the surface (b) and then read again when the weight is at the river bottom (e), the distance (ce) represents the amount of line let out by the reel during the process of lowering the weight from the surface to the river bottom. This distance (ce) may be called the observed depth. The error in the observed depth consists of (1) the distance (ed) above the water, and (2) the difference between the wet line depth (de) and the vertical depth.
The correction above the water surface (ed on the diagram), commonly called the air correction, depends upon the vertical angle of the line and the height (ab on the diagram) of the apex of the vertical angle above the surface. This correction is obtained by using the exsecant of the vertical angle. Computed figures for this correction for even-niunbered angles between 41 and 36' and vertical


heights between 10 and 100 feet are shown in table 1. To apply this table, the vertical angle and the height of the apex of the angle abo- the water surface at different stations along the cableway

,Apex of vertical any/C
in oundinyln

Vertical ang/e

1*~ ~C
bI ae ufc


RIe e

hudedh are gie od Wnitroatio s rdres n lengthsof line


ocqrr=oo ONV ooo c>Nllt=00 ccqt ocr t"INNNN mmmmm

tc w

lf' 1!7,
m 00 m t- CIA f o c xf* 0 xf^ c ""4 00 M -r
ci C4 v c,6 4 Z t.: rz 06 06 06 ci ci 6 c, I m -5

q6 -V CC t- CA
ci ci C4 C6 C6 4 4 4 L6 L6 t6 6 t, 1,, 1 -,6 06 ci ci ci C ,-;

c lf t- M 00 tl4 0 CC N oc M c tf t- m 00 11, 0 to N 00 m 0 t- oc c1l if) 0 M c 0 M t- 0,14 00 lf 00 N "c c m c 0 m
. . . . . . . .
cq cq CIA M m m r tn. u: c r cc r- t- 1- 00 00 00 cl 0 0

1.0 tc t- oc 00 C C-Nm N M .t -t U- rc tL- 00 -,I, t- 0 It t- = m c c, N U- X X "'r t- 0 m z =
. . . . . . . . . .
N N N m m m It Irv "ir 11, tr xf: xf t- r- r-xwwx

c C9 c lf cq X L- oc -f, t- '14 t- m c z m c c cq c
o M Lf 00 M c c r- vt t- c N lr t- c M Lf', 00 M tc X r- 'et =
oo cli cli C4 c C15 C6 C6 C6 4 4 4 Uz tri L6 "i 6 C6 t6 CZ5 t_ V.: t-:

oc 0 00 c m ul 00 0 m 00 0 m c 00 co
oc C N 414 t- c".1 f oc C N Lf'., t- rr 00 0 m ul
cq C4 ci C4 C4 cli CI'S cli cl cle$ 4 4 4 4 4 16 L6 An. L6 C6 C6 6

kf qt N -4 0 0 W t- cc -14 M N C c t- c-, LO It M N
0 M X0 t- X C N 11 c W 0 N ,I* rc t- 0 M ko t- c M t
'14 cli C i C4 cli C4 cl cyi cl C115 cy cl 4 4 4 4 4 vi t6 X6

t- M Xktt= c t- M W ll C c t- M W,4!t C X0
o c r- C9 "14 LL, t- X C cq M V: f c W c M,14 c t- N "4 V*
"?. cq c cli t1i cli C4 cli cli Cos C6 06 C6 vi cl 4 4 4 4

M c oc rt, t- 0 M CC I- C N kf: 00 "14 t- N

d c'j C4 c'j C 4 C4 C4 cq C4 c"s C'6 C6 C6 V6 C'i

D t
c c C4 cj ci C4 C'i C'i C'i C'i

0 x cc .,I- m U- m t- in, m c t- U' M N 0 00 c 114
o LO CZ t- X C r-4 N M
C4 N, N* cq* N*

L: t- M 0 z C9 X, 14 0 c N 0* e) t- m c lll 00
o V CC 1-c r- oc 00 c C N N m Irr ul: U_ t- tN I- IC 0 Cl rt oc C#'Z t- t;:) U W t- N fc
o N N u

o N N N I'd, lf xcl LO tc fZ tc t- t- t- cr w w w

o -.4 r..4 N N N N N Nmmmm m 'At '"q 114 ul UO in to

N Mtt U' t- ooc 0 N M 't lf X -4 cq
c c 0 C4 cq N N N N N NNmmm
. . . . . . .

cl m m .!V It X0 xf: rZ fc t- t- 00 00 m C> (z C4 cq m m l4tt lt4
o r.
. . . . . .

W 'W

0 cq It c 00 0 cq 1114 c 00 0 cq c 00 0 N mt tc 00 0 eq -.14 o 00
r..4 ".4 r.* r..4 P-4 N cq N N cq m m m co co mw '"q It 1.14 in ul X0 Ul L*


0 N Itiq cz 00 0 cq 114 o 00 0 N I", o 00 c N It to 00 0
t- r- I- r- t- 00 00 00 00 00 0) m 0 0 0 0

cc "'t 00 tn C,4 0 t-,", c Mot- kocq mtcm r.-q
c Lo 0 Lo 0 It c Irv 00 m 00 m t- N U- 1 c tc 4 tt t6 c6 c6 t- t- t-: o6 o6 ci ci 6 c5 4 c4 6 c6 c6
cq cq N N cq cq C9 cq

t- M 0 cq "t U' tc t- M 0 N M U" CC t- 00 0 cq
M t- N c 0 t:V W cq c 0 U' M m t- r-q ul m m w N c
C-4 C4 vs 06 4 4 41.6 C95 C6 C15 r- tl- 06 od 06 ai ci 6 c
r--o cq N

W -4 ( 4 C6 6i t,- m 00 1-t 0 c cq
t- "ll 00 kc C) N tc m m c 0 1114 t- 00 N to (M
c5 -. -; ci c4 ci c6 oi v 4 4 L6 t6 u 6 c.6 6 t-.: t,: r.:
".4 "..4 V-4 "..4 r..q r..q r" r.-4 r-4 r-4 V-0 r-4 r" r..4 V-..i r-4 r--i "..4 r-.4 "..4 "..1

00 0 0 N m 1 v to CC t- 00 0 c c cq m 't &A.,) c tN to (M Cl xf oo t- 0 M cmm c cr, N U'* X -I Ite
(. 6 6 6 6 6 r-4 cli cli cli cli cl cys ces 4 4 4 16 U-S
r-.4 r-.4 "..0 r-.4 r-.4 "..0 P" r-.4 r..4 ".4 "..4 r-.4 "..4
U- W 14 t- ct 0 t- mo cmm c
t- 0 Nlowom o 00 -,Tr c M N r- (M N
v o6 o6 o6 6 ci ci 6 6 6 6 6 -; -4 ci c ci ci c6
".4 -.4 r..4 r.-4 -.4 r..q "..q -4 r..q r-4 ".0 V-4

c 00 M "o 00 M CC 00 co cc 00 P- co tc 00 M cc
r- M N 't ,c W M U: t- C N 11 to M -Mtoxo cq
C45 C95 t-1: t--: t-: 06 06 06 06 ci 6 6 6 oi 6 c 6 6
"..4 "..4 ".4

w t- I* to r N 0 M w I- c Lo -,t:V M N 0 c t- PIC
to X 0 N cwo-m U" t- M V-4 M to r- = 0 N
u t6 c.6 c6 c6 6 6 t- t-4 r-: t-Z t-Z t-: o6 o6 o6 o6 o6 ci oi ci

- t- M 00,ft 0 ki-) t- M 00 "t 0 UD -4 t- N W -14 0 kt: t- 00 0 r.-q M u* C c 00 M "..4 N 'It c t- = o N m ul t- w 4 4 L6 ui U- U U 16 L6 C6 6 C 6 C6 C6 t r4 t4 r14 t- t-:

U' 00 V- It CC = N 10 00 T- c a N Lo womcc= N
oo M --4 cq M ,I, o t- W 0 N M U' ro t- c 0 N '"q
CY3 cles 4 4 4 4 4 4 4 U 16 X6 ui L6 L6 L6 L6 C6 C6 C6 C6

0 mmoo 0 - NNNMM m m Itt "
0 N -:! U' o t- X c 0 N M -4v Lo o t- W = CVS C6 cl cl C6 C16 cl C6 cl 4 44444 444416 U

cq 0 w cc Rt N 0 w f c .1t N 0 m t- to m "-4 m t- ul m
"tt U* U") = t- W 0) M 0 --4 Nmm"to = r- t- w M 0
cli cli cli 6 cli c c cli cvi cfs cys cl$ 06 C6 vi C6 V6 C6 cys cys 4

'", 0 cc N X -tv 0 1- M O 10 t- m = u N w -rv 0 CIO
WMMOO -NNMM 1-t Lo U" co c -t- 00 00 c 0 0
tli ci cli cli cli ci cli cli cli ci ci ci cli cli cli cli C6 C16

o Lo 0 M M w N t- ec 0 't C m
m m Itil Lo to I- t- t-WW=M 0 0 ".., '.4 '.4 N
-. 14 r.; 1; 14 14 ci cli cli c c cli

cc mr-om c M N W r--4 "cr
NNMMM m Itr 1TV to to
. . . .
T--4 "-4 "..4 ".4 ".4 ".4 r-.4 r-.4 ".4 ".4 r-.4 ".4 ".4 ".0 V--4

M co to I- M --( M UD t- M"-4 M 0 1- 00 0 N "do o 00
to o cc = o PIC t- t- I- r- t- 00 00 00 00 00 0) 05 m M . . . . . .

MRt 10 ro t- 0 N m Lo Cc t- 00 0 N M -14 Lo
commmm m tv it It M4 to to Lo in Lo ko
. . . . . . .

Lo 10 c c t- 1- 00 00 = 0 0 0 0 --q N cq m m P-4 V-4 ".4 V-4 "-4 r-q ".4 ".4 ".4 cq cq N N cq cq cq N cq eq . . . . . . .

0 c 4 c 08 6 cq, 4 c c c; c 4 e 06 c to 00 0 w o to cc 4;0 1,-- 1- t- I- t- wwwww (Mc === V-4


The correction for excess in length of line below the water surface is obtained by the method for determining the true depth of a sounding from the wet-line depth and vertical angle of the line at the 'Wat er surface, described by F. C. Scheneh on16 who was formerly assistant engineer in charge of discharge measurements of Niagara River. The method depends on an elementary principle of mechanics: if a known horizontal force is applied to a weight suspended on a cord, the cord takes a position of rest at some angle with the vertical, and the tangent of the vertical angle of the cord is equal to the horizontal force divided by the vertical force due to the weight. If several additional horizontal and vertical forces are applied to the cord, the tangent of the angle in the cord above any point is equal to a summation of the horizontal forces below that point, divided by the summation of the vertical forces below the point.
In applying the above principle to conditions of measurements of depths of flowing water it is assumed that with a properly designed sounding weight the horizontal pressure on the weight in the comparatively still water near the bottom can be neglected. The distribution of total horizontal water pressure along the sounding line is made in accordance with the variation of velocity from surface to bottom. The excess in length of the curved line over the vertical depth is the sum of the products of each tenth of depth and the exsecants of the corresponding angles derived for each tenth of depth by means of the tangent relation of the forces acting below any point.
Table 2, which is taken from Shenehon's report,7 shows corrections for even-numbered angles between 4' and 360 and wet-line depths between 10 and 100 feet. Corrections for wet-line depth of 100 feet are numerically equal to percentage correction for any depth.
The corrections from table 2. which may 6e called the water-correction, cannot be ascertained until the air correction has been deducted from the observed depth and the wet-line depth obtained by means of table 1. The air correction is zero wh-Ien the apex of the vertical angle is at the water surface. The air correction may be nearly eliminated by using tags or markers at selected intervals on the sounding line. This practice is almost equivalent to moving the reel to a position just above the surface and gives a measurement of wet-line depth with small error. The vertical angle of the line at or above the surface must be measured so that the wet-line depth may be reduced to vertical depth by use of table 2.
8Shenehon, F. C., In Lydecker, G. J., Survey of northern and northwestern lakes: Ann. Rep. Chief of Engineers, 1900 U. S. Army, pt. 8, Appendix III, pp 5329-5330, 1900.
'7 Shenehon, F. C., op. cit., p. 5330.


The following points concerning the method for deternii mug,. the vertical depth of the water from. the wt-line depth and verticalt a I le of the line at or above the surface should be kept iii mind by' wsers ot the method:
1. The weight. and line are stich that the weight will o to the bottom (despite the force of the current.
2. The sounding is made when the weight is at the bottom bitt entirely supported by the line.
3. Horizontal pressure on the weight when in the sounding positi101 neglected.
4. The table is general, not for any particular line oi- ANiF t or sounding weight, provided they are designed so as to offer little resistance to the current, as the vertical angle is a. f unction of the resistance offered by the line and weight.
The correction tables show the amount and rate of variation of the corrections. In each table, for a given 'vertical angle, the correct ions vary directly as the distance-either the distance above water or the wvet-line distance. Also in each table. for a given distance, the covrections vary approximately as the squares of the vertical angles. The correction tables show also that, for a given angle, the air and water corrections are about equal when the distance above water is about one-third of the vertical depth. For many gaging stations at which the distance above water is equal to or greater than the depth. air corrections will be necessary for depths and velocities not requiring water corrections.
Soundings made in deep, swift water sometimes have been reduced to vertical depths by applying the cosine of the vertical angle to the measured depth. If the measured depth is the wet-line depth the method will give results that are too small. The reason for this statement is evident from the diagram in figure 2. If the meter line under water continued to the bottom at the same angle as at the surface, the cosine method would give true depth if applied to the wet-line depth, but the wet-line depth (de) is shorter than the straight continuation of the line; therefore, application of the cosine of the angle to the wet-line depth gives too small a depth. The error is compensated to some extent if the cosine is applied not to the actual wet -line depth but to the observed depth, which is greater than the wet-line depth by the amount cd. As the compensating effect increases with the distance between reel and water surface, it follows that for a certain distance between reel and water surface an application of the cosine to the observed depth (measured at the reel) will give the true depth. For a greater distance, the cosine method will indicate depths that are too large; for a lesser distance, depths that are too small.


rz C40 c C-4 --T- -f :: z jr C, I
N C4 C, I C'I C,

OC C'I C Tll I
C.C 00 C M to CI. c C14 C-, r z t- C.
6 Ci Ci C,4 Ci 6 r,J C-- v C,

C9 -,tl t- m C 4 -14 Cz Cj -4 rt CC, Go m r 00 c c N 1 1
IC t- COO C! C4 M 4 C I- or- C C4 C -t C O t- C4
. . . .
C'l C*4 C4 C-1 C4 Cq C4 C-1 m C6 ce cl cl

"t to -14 to cc t- CC CC C*, f
t- or_ C C C14 C^ -t t- 00 C"I
Ci Ci Ci 6 Ci C'i
4 Cli Cli cl

C'4 -j-, t, t,- r= kf' lx n*
-14 to e c t- 00 0 -r tn C r-- v- coo C C) -* Cq m TV
C 6 C'i 6 6 C C4 Cj

1,* cv. C14 C J-' CC C-4 0 00 p
00 C C C11 Cv: m rz U- (X OC C.
6 Cli 6 Cli Cli

k- :Y
cy Cq M
. . . .
C14 C

OC C C) Cq M C^

4t C14
to C) 10 C o lr c --IW C -t, C, (m
0 OC Dr, C C C C C'I Cki m
C11 C5

rt M t- r--O to m m I-- C M X N C,.: C -t. OC N C.o rt4 v '14 Lo to I C C rc t- t- X x
C ) . . . .

Lo C .4 to
N N C14 M m m cll -t -14 to le-, llz I=

M kf X 0 CY) I C X 00
-V t- C C11 -14
o 1-4 N C4 C14 M M m -t --v zr rr I- lf. to I- cc C t- tCz . . . . .

CZ) X -41, io u- M m krz t- M t- C7, M tr., 1,
it N N C9 N Nmmmm m rr jl, to If"
C . . . .

o C-1 N N N N N N M M M cf M M M !t -r

lf 00 C 0 C-1 m C t- 00 0 M TV to Cz t X (Z
o ococo r--O -" r N N N N N N N
0 . . . . .

"q 14 kf tc cc t- 00 00 M Cq C4 m m to Cm r.- r 00 0)
o C 0 0 C 0 0 0 0 0
N N N M M m -tt ",-q, ko io to Cc to C t,. t.- t,- CC 00 00
0 0000c 0 C C o c cocco occoc 00000
. . . . . . . .
C,4 C9 Cq N N N N o4mmmm mmmmm
o C) C) d o o 0 C) 6 0 0 0


Cq't O x
MMMMM 't v q:r .tr 14 to Ln


"4 tc t- 00 0 C4 -t

C4 IR14 r- m N It 1 0 --4 00
x (7) c N m -14 t- X 0, 0 C4 ^0, -t t- 00 C C C,
1. L 'f

m cli cli co C6 cl cl

M M N N C C C CS x rr C14
oo M 0 --1 C14 m "I" "Rlt kf z t- or- C 0 cl m Cq C C'i cy C'$ C's C6 ces C's C, C, C, C, 4

c -I- C-1 c cc cc U' m (m f- -t N C
t- 00 c C
cli cli N cli C cli cli C6 cl cl-S C6 C-- C-13 cl cy m C',

00 kc cq (M c M N C c IM
N M CYZ "dq to Lr c t- 00 x (7 C C c", N m -11
ci C4 C C'i C C C'i ci -i c"i Ci C ci C, C" z C' C,

00,-t C 10 t- m C In. -4 t- m C It-: C r cl) oc -V 0 1- 00 oc C 0 Cl M cf -t, -fl in. t- 11C t- t- GC C
4 C4 C'i C4 C'i C C Ci ci c4 C 4 Ci C4 ci ci Ci Ci Ci

05 -,t (m -,t x m XMXMX mxmxm x
'Ri4 lf X0 cc cc I- t- x x (7) C4 N m m -t
4 r4 r-4 4 4 C cli ci C'i C'i 6 Ci 04 C'i C'i

C14 c lf C M I- -t oc Cl C.5 5 "tm Cc cc i, i, x coc x m 5 5 c
r-4 r- 1-4 --; 4 Ci C4

00 Cq LO 00 cq llll OC kf 00 vt OC "kt CC r.-4,.tt r- r -t,
0 C 0 r-4.cq cq r-l m m m It -tl It 1. If-, CC o

w C C (m M 0 coo- P-- C14 N N C11
. . . . .
r-q r..q V-0 r-.4 r-q r -4 r-.4 r-.4 r-14

(m V-4 CVD lf: r- CVD vt CC 00 C'l -t cc 00 c C'l -14 00
to CC to o CD cc t r t t- t- 00 or-, OC 00 00 C C C C
. . . . .

kf 00 = (=) cq m 1.0 c OC C C N M kl CZ 00 cli
44 .14 ,d4 U- Ix to to Lf- kf lf c c z (= c rc t. . . . . .

0 P-4 N M *t4 X0 cc 1 00 C. C N M -V 1 1 r- 00 = C
mmmmm mmmmm -t -14 -. r 14 -1" -t -tl 'T -f, It
. . . . .

cq cq ct 1-tt "q 00 cc C
N N cq N N N N Cl C,-l N N Cl N NNMMM m
. . .
(D C) Cl eq N N m m m -" I!t
-.4 r-.4 r---o r--o V-4 r--4 r--O 7-0
'10 ""1 1,,
. . . I .
-qt -.iJ4 -RIO It In lf L('j lf IM ll" Lf LC = C CC C Z
0 C) 0 C) C> 0 0 0 0 C 0 0
. . . . .

The accuracy of a measurement of discharge depends upon the( correct determination of the mean velocity iin the vertical, and this determination requires the p~lacemlenlt of the ciirremit meter at- selected positions in the vertical depth. The conditions that cause a measurement of depth to be in error when made by the usual methods wold Ialso affect the placing of the meter at selected positions in the vertical. A current meter placed in deep, swift water by the usual methods for observations at selected positions in the depth will be too high in the water, and the amount of error in placement increases rapidly with increases in the vertical angle. That part of the error due to inclination of the line above the water surface may be corrected by use of the .4i r- correction table. The water-correction table for use in measurements of depth is not strictly applicable, however, to the determination of the vertical distance to an intermediate position.
The assumptions upon which the water- correction table is based are listed oni page 49. The assumption that horizontal pressure on the weight when in the sounding position is neglected is based both on the fact that the sounding weight* is in the zone of minimum velocity and on the supposition, that it is so shaped as to offer very little resistance to the current. If the weight is raised above the bottom it is subjected t~o higher velocities, the pressure on the weight is increased, and the vertical angle of the line at the surface will change with any change in the total pressure on the equipment that is under water. Thie position and shape of the sounding line, if the weight is above the bottom will therefore not be identical with those shown in figure 2 for the sounding position. As the water-correction table was computed for a weight and line, it is not strictly applicable for a sounding made with a current meter and weight. As the meter is placed about a foot above the weight, it is in a zone of higher velocity than the weight, and because of its shape it offers more resistance to the current.
Correction tables for placement of the current meter for velocity observations have not been prepared, as it is apparent that they would be specific and not general in their application. It is evident, however, that the use of the air- and water-correction tables prepared for use in measurements of depth will tend to reduce the error in placement of the meter for velocity observations, and, al though they are not strictly applicable, their use for this purpose has become general in measurements of rivers that are deep and swift.
The two-point method for determination of mean velocity in the vertical, which is in general use, requires observations at 0.2 and 0.8 of the depth. If the use of correction tfjbles is needed, the nmter is usually placed by measuring down from the water surface for the


0.2-depth position and by iieasiInug up froiii the bed of the stt' for the 0.8-depth position. Inaccurate I)lnceillent of the iieter f(,t the 0.2-depth position will result in less error than for the 0.8-depth position, as the velocity near the upper position usually vwlries 1es rapidly. It is obvious that letting out a length of line equal to 2U percent of the vertical depth for the 0.2-depth position and reeling up the same amount for the 0.8-depth position will not phlce the meter at the desired positions unless consideration is given to the amount of line involved in the air and water corrections.
For the 0.2-depth position the water correction is negligible as the wet line is practically a straight line. The amount of line to be let out, however, in addition to 20 percent of the vertical depth, is somewhat more than the air correction corresponding to the vertical angle and a distance equal to the height above water. This amount of line, including the air correction, may be determined from the air-correction table by using the vertical angle, observed when line equal to 20 percent of the vertical depth has been let out, and a distance equal to the sum of the height above water and 20 percent of the vertical depth. If the angle increases appreciably when the additional line is let out, more line must be let out until the total additional line, the angle, and the vertical distance are in agreement with figures in the air-correction table. For the 0.2-depth position the wet line depth instead of the vertical depth is sometimes used as the basis for the setting. Its use is assumed to result in placing the meter at the desired depth without additional corrections for change in vertical angle.
In the placing of the meter for the 0.8-depth position, a correction. to the amount of line reeled in must be made for the difference, if any, between the air correction for the sounding position and the air correction for the 0.8-depth position. This difference is designated as c in the summary below. If the angle increases for the 0.8-depth position, the meter must be lowered; if it decreases, the meter must be raised. The angle is an indication of the total pressure on the equipment. If the angle does not change when the meter and weight are moved from the sounding position to the 0.8-depth position, the increased pressure on the meter and weight has been compensated by the removal of pressure on the line, because a shorter length of line is subjected to the action of the current when the meter is moved to the 0.8-depth position.
For the 0.8-depth position of the meter the water correction may require consideration if the depths are more than 40 feet and if the change in vertical angle is more than 5 percent. As the wet-line depth when the meter is in the 0.8-depth position is less than when it is in the sounding position, it is obvious from table 2 that, if the


vertical angle remains the same or decreases, the water correctioJn for the 0.8-depth position is less than the water correction for the sounding position by some difference designated as c' in the summary below. If the vertical angle increases, the difference c1 between the water corrections diminishes until the increase in angle is about 10 percent; for greater increase in angle the difference between ( oriections increases also.
The effect on the air and water corrections caused by raising the meter from the sounding position to the 0.8-depth position, and the resulting effect on the corrections for the 0.8-depth position are summarized below.
"lABLE 3.--Summary of effect on air and water corrections caused by raising the meter front the so ending position to the 0.8 depth position
Air correction Water correction
Change in
angle Direction Corrc io to meter ositi )ir c tiof t(f (C'orrect ion to meter posiof change change t ion
None. None. None. )ecrease, Raise mn ter the diftferv ll(t CI.
Decrease. Decrease. Raise meter the difference c. creases. lie meter the diitfer( I we C1.
In(r ase. Inretae. Iower ueier the dlfference c D)ccree)-r then l in- Raise meter Ihe dillercre A. ell(e C1,

'The difference c will decrease with increase in angle and will change sign when increase in angle hals become about 10 percent. Beyond that point the correction to the meter position (because of water correction) requires lowering the meter the difference (c1) between the sounding and the 0.8-depth water corrections.
For slight changes in the vertical angle the adjustments that must be made to the wet-line length because of the differences c and c', in the air and water corrections, in order to obtain the correct position of the meter for the 0.8-depth observation, are small and usually may be ignored. Inspection of the summary table given above in'!icates that the meter may be placed a little too low in the water if the adjustments are not made. Because of this possibility, the wetline depth instead of the vertical depth is sometimes used as the basis for the 0.8-depth position with no adjustments for the differences c and c in the air and water corrections, respectively.
For all positions of the current meter, the distance between the center of the meter and the bottom of the weight must not be ignored if the bottom of the weight is used as the index in the measurement of depths.
The procedure generally followed in making observations and corrections for depth and for velocity positions in discharge meas-


ureuieits of rivers that are deep and swift has been developed fr, ol experience acquired at various river-measurement station) The equipment for such work includes reels, booms, depth indicators or counters, and protractors for indicating the vertical angles. After the equipment has been assembled and tested, the routine procedure is as follows:
1. Place the gaging car or the reel and boom equipment in )osItion at the station on the cableway or bridge at which the observations are to be made. Measure and record the vertical distance from the water surface to the apex of the vertical angle.
2. Place the bottom of the sounding weight at the water surface and set the counter or depth indicator to read zero. Lower the sounding weight to the bed of the stream. Read and record the observed depth and vertical angle when the sounding weight is at the bed of the stream but entirely supported by the sounding line. With the aid of the correction tables compute and record (1) air correction, (2) wet-line depth, (3) water correction, (4) vertical depth, (5) 0.2 of vertical depth.
3. Raise the meter from the sounding position a distance equal to 0.2 of the vertical depth minus the distance from the meter to the bottom of the weight. This places the meter approximately at the 0.8-depth position if the vertical angle has not changed and if the depth and angles are not sufficient to cause an appreciable change in the water correction. If the vertical angle has increased. the meter must be lowered the difference between the air corrections for the 0.8-depth position and the sounding position; if it has decreased, the meter must be raised the difference between the air corrections. If the depth and angles are sufficient to cause an appreciable change in the water correction, an adjustment on this account will be necessary.
4. Observe and record data for velocity at the 0.8-depth position.
5. Raise the meter from the 0.8-depth position until the counter reads 0.2 of the vertical depth plus the distance from the meter to the bottom of the weight. Note the new angle for this position. From the air-correction table, find the proper correction for this angle and a vertical distance equal to the height above the water plus 20 percent of the vertical depth, and lower the meter an amount equal to this correction. This places the meter approximately at the 0.2-depth position unless the vertical angle increases appreciably. If the angle increases, additional line should be let out until the amount of line in excess of 20 percent of the vertical depth, the angle, and the vertical distance are in agreement with the air-correction table.


'. Observe and record (dt :i for velocity at the 0.2-depth position.
IL work on streams carrying (rift and fine debris in suspension, the procedure may need modification. Accumulation of debris on the line increases the angle progressively as the accumulation increases, and it requires that the sounding be made as rapidly as possibile and that the meter and weight be brought to the surface immediately after the sounding for inspection and cleaning and protection from large drift. During tle time required for a velocity observation at a given point, the accumulation of debris on the line may cause the meter to rise above the selected depth of observation. To compensate for this condition, the vertical angle must be observed constantly and the necessary additional amounts of line must be let out.
The tag-line method of measuring depth and placing the meter Consists essentially in using index tags or markers fastened to the meter line at known distances from the nieter or bottom of the weight. 'Flie tags, which may be streamers of different colors fastened on the neter line by solder beads, small cable clips, or adhesive tape, should be easily identified and should remains fixed in position. A narrow strip of inner tube tightly stretched in a double turt around the cable makes a satisfactory marker. After the weight touches tle stream bed the depth of water is determine( bI drawing up the meter line enough to measure the depth of tli(' first tag below the water surface ald adding to that depth the known distances of this tag above the bottom of the weight. In the placing of the meter for a velocity observation, the tag nearest the selected depth is used as a reference point in raising or lowering the meter to the desired depth.
The tag-line method is considered advantageous in the measurement of deep water with low or moderate velocities, as it eliminates the necessity for measuring a large part of the wet line and thereby avoids the necessity for raising the weight of the surface at each sounding. It permits the placing of the meter in the correct position for a velocity observation with less effort and greater speed. The current meter remains below the surface, where it is not so likely to encounter floating debris or ice. The method is used more frequently with hand-line equipment than with a reel assembly, although it may be used with either; and it is particularly useful in freezing temperatures, as it avoids exposing the meter to the cold air.
The tag-line method may be used where there are vertical angles in the sounding line, although certain precautions must be observed. The air correction for sounding and placing of meter is eliminated if the measuring equipment is supported at the wafer surface. The use of the tag-line method will reduce the air correction to a negligible


amount if the vertical angle does not cllanlge when an ifldex tag is raised to tle water surface. The method, however, cannot be considered as equivalent to moving the nieter-line support to the water surface, as the support, if at the water surface, would be in a fixed position; whereas with the support above the water the intersection of the line with the surface is free to move upstream and downstreaii" and therefore the length of line in the water may increase or decr(ease. If the vertical angle increases when the index tag is brought to the surface the difference in the air correction for the two angles must be added to the observed depth to obtain the wet-line depth. If the angle decreases, the difference must be subtracted. The wet-line depth thus obtained must be corrected to vertical depth by use of th e watercorrection table. These corrections shou]l(1 all bIe taken into consideration, especially if the stream is deep and swift. The use of adequate sounding weights will ordinarily eliminate the necessity for corrections.
Differences between air-corrections due to moveinent upstreain or downstream of the intersection of the line with the water surface are of consequence only when they are such percentage of the vertical depth as to affect the accuracy of the sounding or the placing of the meter. It is therefore essential that users of this method become familiar with the magnitude of possible er-rors in order to know its limitations. It should be noted that the height above water of the support of the equipIm ent is a significant factor, as the air correction varies directly with the height and approximately with the square of the vertical angle. The procedure for making discharge measurements by the tag-line method with corrections for vertical angles is essentially similar to that described on page 54.
Measurements of shallow streams with high velocities may be made with increased accuracy by the use of a st'ay-line. The stayline wire or cable is erected parallel to the measuring section and should be about the same height above water as the cableway or bridge from which the measurements are made. It should be fari' enough upstream from the measuring section so thait the angle made by the stay-line with the water surface will not be greater than about 30'. The stay-line cable carries a traveling pulley, to which is attached a swivel pulley. The stay-line, which is attached to the meter hanger above the meter, passes through the swivel pulley and then to the operator or the bridge or cableway. A sm-all wire of high strength should be used for the part of the stay-line in the water. A strong cord or small rope that will pass over the pulley readily should be used above the water.


To obtain the best results with the least effort and ill the shortest time, make the soundings by measuring down from the water surface, and place the meter for velocity observations by mne:isurinlg up from the bottom.
A stay-line is most useful at measuring sections where the struc ture from wlichl measurements are made is at a considerable height above the vateir, where I he water is comparatively shallow, and where the velocities are high. Under these conditions its use eliminates the necessity for measuring the vertical angles made by the meter line and the applications of air corrections. It eliminates also error that might be caused by the downstream drift of the meter alid weight into a sect ion not equivalent to that at the measuring section.
The principal error in th le ise of a .tay-line occurs when depthl and velocity are sufficient to cause curvature in the meter line when the sounding weight and mleter are brought into the vertical plane of the mleasulriig sectoloi by ten isionl in the stay-line. If the use of a larger sounding weight will not reduce the error in the observed depth to a negligible amonmt,- the use of a stayline is not recoilmended. Errors due to clurvature in the meter line, where a sounding is made with tihe use )f a stay-line equliplment, cannot be removed by mieais of the water-correction table as the assumlptio1s on which that table is based tdo not apply.

If the direction of flow is not perpendicular to the measuring section, the angle in the mineasuring lii-e as indicated by the verticalangle protractor will be less than the actual angle in the line. The corrections to air line and wet line based on the observed angle will be too smal: consequently incorrect measurement of depth and incorrect placement of meter will re.;iult. The horizontal angle between the direction of flow and a perpendicular to the measuring section may be measured from the gaging car or from the bridge with angle charts or protractors prepared for that purpose.
If the horizontal angle of the direction of flow may be called H, the measured vertical angle P, and the actual vertical angle X, the relation between the angles may be expressed by the formula tan. P
tan. Xcos. H
Angles P and H are measured, and therefore angle X may be
derived from the equation.
Table 4, which gives the amounts, in tenths of degrees, to be added to observed vertical angles to obtain the actual vertical angles for a range of horizontal angles between 8o and 280, has been computed from the relation between the angles H. P. atid X.


I"ABLE 4.-Amounts to be added to observed vertical angles, to obtai actolu twrtiral angles
Horizontal angles
Observed vertical angles
80 120 160 200 240 280

80 0. 1 0.2 0.3 0.5 0.8 1.0
120- .1 .3 .5 .8 1. 1 1.5
16 0. .. .1 .4 .6 1.0 1.4 2.0
200.. .2 .4 .7 1. 2 1.7 2.4
240 ...... ...2 .5 .8 1.4 2.0 2.8
28--- .2 .5 1.0 1.5 2.2 3.0
320 ..2 .6 1.0 1.6 2.4 3.3
36 .2 .6 1. 1 1.7 2.5 3.4
40 ---- ...2 .6 1. 1 1.8 2.6 3.5

Applicability of the table will depend on the degree of refinement with which the vertical angle is observed. No additions to the observed vertical angle are necessary for any vertical angles below 20' if the refinement is 10 and if the horizontal angles are below 120. If the refinement is 20, the limits are 240 iii the vertical angle and 200 in the horizontal angle.
To systematize the recording and coinputiug of data pertaining to current-meter measurements, the Geological Survey uses a series of standard forms on which data are recorded and computed.
Form 9-2?75.-Form 9-275, shown in figure 3, is used for all openwater discharge measurements where the observations are not corrected for vertical angles. It contains 13 columns, each with a heading explanatory of the data to be inserted. On the left margin are the angle coefficients referred to on page 72. The position of the vertex of the angle is indicated by the small circle on the right margin. When this form is used it is common practice to utilize two horizontal lines for recording the observations and calculations made at each vertical, regardless of whether the two-point or the 0.6-depth method is used. Such procedure permits the recording of data for 12 verticals on one sheet, avoids possible confusion of notes, and allows sufficient space for recording any miscellaneous observations that may be necessary. Data in the following order are recorded on the upper of the two horizontal lines, beginning at the left margin of the sheet: (1) Distance from the initial point to the vertical, (2) angle coefficient or cosine of the angle between the direction of flow and the normal, (3) depth in the vertical, (4) actual depth at which the first observation of velocity is taken, (5) revolutions of the rotor of the current meter,
(6) time over which these revolutions are observed, (7) calculated velocity at the point of observation, and (8) mean velocity in the ver-


Be 75



River, at ______ ____________Dist. Angle Depth Rev- Time VE-LOCITY
92- from Co0-Det of ob- ol in Mean Mean Area Meanith Dicag
initial fi. Dphserva- U LosSec- At 1vr in death ihag
point cient tion tnsonds point i 1secin

98. __ _- __-_98. __ _92. -_ _ _ _


No. of Sheeta. Comp. by -Ck. by
FIG.u liE 3.-Discharge-measurement notes, form 9-275.


tical. If a second observation of velocity is taken in the same vertical the actual depth, revolutions, and time observations are recorded in ther proper spaces on the second line, followed by calculations of velocity at the second point of observation as follows: (9) Mean velocity in the section, (10) area, (11) mean depth, (12) width, and (13) discharge. If but one observation of velocity is taken, the second line is used only for recording the calculations of mean velocity in the section. area, mean (lepth, width, and discharge. The time at which an observation in a specific vertical is completed is noted in the first
Date .19- No. of Me&#.
River, at
Creek, near
Thick- Total VELOCITY
nes of depth of Depth
Dist. from ice water oF meter Rev- Time Mean
initial below olu- in Area dept Width Dischtge
point water tons seconds At Me Me a r depth Wh
W. S. to Effective surface point vertical section
bet. ice depth

TOTALS. ______No. of Sheets. Comp. by Chk. by Make notes on back.
FIGURE 4.-Current-meter notes, ice cover, form 9-275a.

column directly below the figure designating the distance from the initial point. To identfy the bank at which the measurement is started or finished, the letters "L. E. W." or "R. E. W.", signifying the left edge of water or the right edge of water and followed by its distance from the initial point, are written in the first column at the beginning and end of the measurement. Each sheet should be dated and identified to avoid confusion of notes. To obtain consistency in computatons, calculations should be carried to three significant figures except below 1, where hundredths is sufficient. If the figure following the last significant figure is 5, it should be either dropped or added to make the preceding figure even.
Form 9-75a.-All observations pertaining to the measurement of discharge under ice cover are recorded on form 9-275a (see fig. 4).


This form is essentially the same as form 9-275 except that the angle coefficient column is omitted and the "Depth" and "Depth of observation" columns are replaced by three columns headed "Thickness of ice," "Total depth of water," and "Depth of meter below water surface." The system of note keeping is also the same except for the data tabulated in the three columns mentioned. In the "Thickness of ice" column, the thickness of ice is recorded in the upper space and the distance from the water surface to the bottom of the ice in the lower space. In the "Total depth of water" column, the total depth is recorded in the upper space and the effective depth, which is the depth of the water beneath the ice, is recorded in the lower space. In the column headed "Depth of meter below water surface," the actual computed figure showing the position of the meter is recorded.
If the top of the ice, or a rod placed across the hole, is used as an index for measuring the depths, the distance from the index to the bottom of the channel may replace the total depth of water in the "Total depth of water" column, and the distance from the index to the bottom of the ice may replace the thickness of ice in the "Thickness of ice" column. The figures recorded in the column "Depth of meter below water surface" should, it' this case. identify the position of the meter with respect to effective depth.
Form. 9-75b.-Where an open-water measurement requires consideration of the vertical angle, form 9-275b is used. This form, shown in figure 5, is an expansion of form 9-275, with seven columns replacing the two columns headed "Angle coefficient" and "Depth." These additional columns provide space for tabulating the following data: Distance above water surface, vertical angle, observed depth. air correction, wet-line depth, water correction, and vertical depth.
Form 9-275c.-Form 9-275c, shown in figure 6, serves as the first sheet for each discharge measurement and contains a complete list of headings showing necessary data that must be collected and tabulated if the measurement is to be of maximum value. The column headings are self-explanatory, and no measurement of discharge should be considered complete until all information indicated by the column headings has been obtained and entered in the spaces provided. All entries of data should be made directly on this form immediately upon their observation and not recorded elsewhere to be transferred at a later date. Those headings that do not apply to the particular measurement should be deleted so as not to leave any inference that the information may have been overlooked. The substitution of initials for names, the abbreviations of names of places, and the record of the date by figures should be avoided. Form 9-275d.-Form 9-275d (see fig. 7) is used for compiling general information indicated by the topics listed at the top of the sheet


C; C. cle


cn <

cc rz:

4-j cd

.0 v


w 4
w >

2 0 1 od
0: 1 U z
w :; Z
1.6 < w 0
U 2 3: ;T4
w 0 w
0 W 0.00
0 W J
0 0 W M
z (n w -Z ta
w W of
0 0
W w w a -00.
j- 0 3: c
< U) < of
%-3 ..% L w 0
0 W lu


9-275 0
(July 1935)UNTDSAE
(JU.V mLJNITEDE STATES Mews. No.............
DISCHARGE MEASUREMENT NOTES Date.......... .Date .-----------------., 19- Party .........-..-...-.........................................
11idth ..--------- Area -------.-- Vel. .-.-- G. H ...---------- Disch .--..........
Method ------- No. secs. .------- G. H. change ------ in ---- hrs. Susp ........
Method coef. ------. Hor. angle coef ----.... Meter No. .------- Susp. coef .......
GAGE READINGS Date rated --.....--------------- Used rating
Time Recorder Inside Outside
for rod ....--------- usp. Meter -----.. ft.
Slart above bottom of wt. Tags checked .......-..

Spin before meas. ...---------- after ........
S------ ---------- ---------Measplots- %diff.from -------.. rating
-. .- ..------- ---------------- Wading, cable, ice, boat, upstr.,downstr., side
-- -.------- ----------------- bridge .....----------- feet, mile, above, below

Finish gage, and ----...-.......-...-.....--...............
Chain, wire, found ...........------------------Weighted M. 0. H ..------------ ....... ---------- changed to -......-------------- at ..............
0. H. correction. .... ..-------.---------- ---------- Correct length -----.....---------------------Correct M. 0. H .... ..------.---------- ---------- Levels obtained -........---------------------.

Measurement rated excellent (2%), good (5%), fair (8%), poor (over 8%), based on following conditions: Cross section -..........-.....-......----------------------------------------Flow ------.. ........-------------------------------- Weather .... ......--------------------------.. . .
Other ...............................-------------------------------------------------------------...
Gage .-..-...-...-......--........-........................................................................
...................------------- Record removed .. ..------------ Intake flushed .....---------Observer ----.------......-........-..........-------------------------------------------------------------Control ---------..-...-.......----.-.----.-----..-.-...--..-.......----------------------------------Remarks .--.----.........-.--................-------------------------------------------------------------Sheet No ----- of ----- sheets. G. H. of zero flow ........-.------------------------- ft.
Fi(;uLa (;.---First sheet of discharge-measurement notes. form 9-275c.


and is the basis of the field report to be prepare t by Ib eni e r at the time of his visit to the station. If the data relate to a discharov measurement, the form is attached as the final sheet of notes oncern ing the measurement; if they have no reference to a disclhrg measurement they are filed with other data pertaining to the station.


Date ......................... ,---------------- 19 ..- No. of Mea......

Enter on this form ample notes in regard to the following:
1. Accuracy of measurement; 2, gage; 3, observer; 4, bench marks; 5, gageheight corrections; 6, adjustments to total discharge; 7, station equipment; 8, channel, control, and point of zero flow; 9, rating, backwater; 10, diversions, regulation; 11, records; 12, cooperation.

No. of sheets 6-4226 v. amS covn erae rewVo: 1s
FrGnRF 7.--Supplementary discharge measurement notes, form 9--275Id.


For accurate and reliable measurements of discharge, especially in natural stream channels, a knowledge of many factors is essential in addition to the specific procedure followed in making a discharge measurement. The wide variety in the character of streams, in climatic conditions, both seasonal and regional, and in behavior of the measuring equipment when used under various circumstances; gives rise to many problems which may affect the specific procedure of the work.

Accuracy in measurements of discharge can be expected only when the equipment is properly assembled, adjusted, and kept in good condition. The current meter and stop watch, in particular, must receive the best of care and protection, both when in use and when being


transported, as they are the two m-ost delicate ailnsiid elements o)f the measuring equipment.
The current meter necessarily receives a certain amtount of hard usage that may result in damage, such as a broken pivot, chipped bearing, or bent shaft, any one of which may cause the meter to underregister. Observations of velocities near bridge piers and abutments-. soundi n 0s taken at sections having irregular and uncertain profiles with the meter attached to the measuring line, and the presence of floating drift or ice probably include the greatest hazards to the meter equipment. Floating drift and ice, under careful observation, can usually be seen in time to allow the removal of the equipment from the water. Occasionally a measurement is so valuable that considerable hard usage of the equipment is justified. After such usage each part of the current meter should be thoroughly examined.
Damage to me-asuring equipment during transportation is generally clue to careless packing or negligence in protection. A standard case is provided for use in transporting the current meter and the several small andl more delicate articles of equipment. Pivots, bearings, contats ndote eta rts carried into the field should be carefully packed. The stop watch. should always be 3arriecl in a container that provides protection against dampness and sudden jarring. The headset assembly should be packed carefully to avoid accidental short circuits, which may discharge the battery. If several sounding weights are to he carried1 a box should be used with a separate compartment for each weight, and the compartments should be arranged and fitted so as to protect the weights from flattening and scarring. Other articles. such as sounding reels and waders, should always be carried in separate cases. The means for transporting more bulky equipment, such as cranes and r'eels, will vary with the type of conveyance used. Transportation of equipment in assembled form f rom, one gaging station to another is one of the most common sources of damage and too often results in a loss rather than a saving of time. An engineer who takes pride in the care and protection of his measuring equipment will find himself amply repaid for the extra, time and effort that may be required to maintain it in the best possible condition.

The selection of the section at which to measure discharge should be given ample consideration, when measuring either in open water or through ice cover. Although this phase may have been given careful study during reconnaissance prior to the establishment of the gaging station, nevertheless changes in channel conditions may occur from time to time which necessitate the select ion of other sections, especiallyN for low-water measurements.

If results of the discharge measurement are. to be accurately related to the flow at the gage during the timne of ineas urement, it is essential that the place of measurement should be near' the gage. Only where it is known that there is no appreciable difference between the discharge at the gage and at the measurIngr Section during the time of measurement, or wher-e there is some reliable means by which the difference may be determined, is it permissible to have the measuring section far from the gage. The characteristics of the measuring section may differ from those of the section controlling the stage at the gage, and they need to remain permanent only during the time required for making the measurement. Inasmuch as a measuring section must meet certain requirements, it is not uncommon to select several sections to be used for wading measurements at different stages.
The ideal open-water measuring section is perpendicular to the direction of the flow and should be in a stretch of the stream where the bed and banks are uniform. Accurate measurements are most easily made if the minimum velocity at any one vertical is not less than 0.5 foot per second and the depth of the water is sufficient to permit the use of the two-point method. Such a section is not always found in natural streams. but reliable and accurate results of measurement are obtainable at measuring sect ions showing considerable departures from the ideal if the engineer uses proper precautions and good judg(Yment in choosing the cross section and in operating the current meter and if hie makes sufficient observations of velocity to assure a representative mean velocity for each part of the section.
Sections for- measurements through ice cover should have the same gyen e i'a I ch aracteristices as good open -water sections. On. s-treams where ice cover is anticipated, these sections should be selected during the open-water season and their positions indicated by markers that will uot be obliterated by ice and snow. Prior selection of the section will enhance the accuracy of the determination of the discharge under the ice, as the distribution of velocities under ice cover and in open water is very similar. If the general distribution of velocity in the ice-covered section is different from that for openi-water conditions, it will have resulted from the pressure caused by the rigidity and weight of the ice. A previous knowledge of the cross section will assist in finding the edges of flowing water and in spacing the holes in the ice so that the discharge. measured between verticals in each part of the area will be in proportion to that measured in open water.
The section directly under a cable or bridge used for open-water measurements may be used at some stations for measuring through ice cover, although when measuring through the ice at a bridge it is generally advisable to select a cross section far enough upstream or downstream from the open-water section to avoid the disturbing influence


of the piers. Because the anchor ice and frail from which slush ice is formed collect more thickly in the upstream part of an ice-covered reach, the most favorable ice-covered section is just above a place of open water.
Appreciable errors may result in determining both the area and the mean velocity if the Verticals in the measuring section are not properly spaced. With regard to measurements of depth, these verticals should be so spaced that the error involved in determining the area becomes negligible if the part of the profile of the section between two successive verticals is assumed to be a straight line. With regard to velocities, the verticals should be so spaced that th6 average of the mean velocity in any vertical and that in the preceding or following vertical will represent the mean velocity in the section between the verticals. Usually where the profile of the measuring section is very irregular the velocities are irregular, and more verticals are necessary for an accurate measurement of discharge. Only where the velocity appears to be well distributed and where the profile of the cross section is reasonably regular and smooth is it desirable to space the verticals at equal intervals throughout the measuring section. In some deep channels the surface velocities may appear to be fairly uniform, but irregularities in the bed of the stream may require irregular spacing of verticals in order to define accurately the profile of the cross section. If the stream bed is irregular it is generally advisable to measure the velocity in each of the corresponding verticals, as the velocities near the bed inay be affected by the irregularities in the profile.
Besides the effect of distribution of velocity and the character of the bed at the measuring section upon the number of verticals required, the number varies with the depth and width of the stream. Because of these differences and because a few measuring sections are divided into two or more channels, no fixed rule can be made as to the exact number of verticals required for a measurement. Even where the stream is confined to one channel and the stream bed and velocities are reasonably uniform, no definite rule can be followed except that the verticals should be so spaced as to disclose the real shape of the bed and the true mean velocity of the flowing water. The verticals may be as much as 15 or even 20 feet apart in a smooth channel 400 feet wide and perhaps as much as 40 feet apart in a 1,000-foot channel. The channel should be decided into 20 or more parts except for very small streams, where a somewhat smaller number may be sufficient if the distance between verticals becomes less than 1 foot. The division is generally made so that there will be not more than 10 percent, and preferably not more than 5 percent, of the discharge between any two verticals.

The accuracy of a measurement of discharge in any section depends as much on the correct determination of depth as it does on the accuracy of measurements of velocity. Inasmuch as the discharge is a product of area and velocity, any error in the area as a result of an incorrect measurement of depth will produce a corresponding error in the discharge. This error would affect the determinations of areas and discharges in the two adjoining sections.
Inaccuracies in soundinLys are most likely to occur in those sections having great depths and high velocities where the engineer is unable to measure the vertical depths directly. As those sections generally carr the larger part of the total discharge, it is essential that methods of measuring be used which will either eliminate this source of error or reduce it to a negligible amount.
In regions of subfreezing temperatures, surface ice, frail, or anchor ice may form during each winter. Ice in the measuring section adds to the difficulty and inconvenience of making discharge measurements, and as frail or anchor ice it tends to reduce the accuracy of the measurements. The accuracy can be considered equal to that of openwater measurements only where there is surface ice alone and where it serves as a support from which to measure.
Surface ice ordinarily forms first at the edges of the stream and gradually proceeds toward midstream until the stream is spanned. The ice cover forms last in that part of the stream where the velocities are highest because agitation by the current breaks away many small crystals of ice as rapidly as they form.
The surface ice during the period when it is breaking up should not be used as a support for the engineer. Floating cakes of ice are as great a menace to meter equipment as any other type of drift, and accurate determinations of depth and velocities are always difficult when such ice is present.
Frazil is ice in fine elongated needles, cubical crystals, and thin sheets, formed at the surface of the stream when the disturbances of the water caused either by high velocities or by wind are too great to permit the formation of ordinary surface ice. It never forms under an ice cover but may be carried under it by the current as a mass of floating slush.8 Anchor ice, which resembles frail ice, forms between sunset and sunrise on the stream bed and attaches itself to boulders and other objects. It is rarely found at great depths but forms
9 Barnes, H. T., Formation of anchor ice and precise temperature measurements: Am. Soc. Me-ch. Eng. Trans., vol. 26, pp. 558-583,1905.


rapidly and in large quantities in shallow streams where the bed is rough. During the daytime, particularly when there is sunshine, anchor ice may break loose, rise to the surface, and merge in the frazil.
Measurements through ice cover at a section containing frazil or anchor ice should be made only as a last resort because of the una voidable inaccuracies in determining both the velocities and the effective depths. The fine particles of ice are likely to collect in and about the rotor of the meter and thereby retard it enough to cause underregistrati on. If the thickness of the frazil or anchor ice is taken as that part of the depth in which the current meter indicates no velocity, the assumed effective depth may be too small because of the effect of the ice on the free operation of the rotor'

Freezing temperatures not only add to the discomfort of the enigineer while operating the meter but may cause serious errors in velocity determinations if the meter is used after ice has formed in the bearings and contact chamber. In making a discharge measurement at subfreezing temperatures, the engineer, once he has put the meter into the water, should expose it to the air only when absolutely necessary. If the measurement is made through ice cover, the meter should be moved from one vertical to another in the minimum time and should be kept under water as much as possible during the measurement. The formation of ice on the meter equipment may be minimized by making the soundings without attaching the meter and by computing all meter settings before the meter is put into the water. When the meter must be removed temporarily from the water, the rotor should be permitted to revolve for a short time after it is again submerged before the revolutions are recorded, as the river water with a temperature above freezing will tend to thaw any ice that may have formed in the bearings. If the rotor is so sealed with ice that it ceases to turn, the meter should be thoroughly dried by a fire or heated with warm water before it is again used.

Turbulent flow is that type of flow in which any particle of water may move in any direction with respect to any other particle and in which the loss in head is approximately proportional to the second power of the velocity. The range in depth, velocity, and viscosity of the water in natural streams is generally within those limits in which turbulence occurs. Where the condition of flow is such that the water

9Hoyt, W. G., The effect of ice on stream flow: U. S. Geol. Survey Water-Supply Paper 337, pp. 24-30, 1913.


is considerably agitated by cross-currents, boils, and eddies, the accuracy of the velocity observations obtained by a current meter, regardless of the type used, may be affected to some extent. However, many comparisons and tests have shown that unless these disturbances of flow are pronounced, the errors will be small and mostly negligible. Because of the inaccuracies due to excessive turbulence, engineers of the Geological Survey generally specify that sites for stream-gaging stations shall be so selected that current-meter measurements can always be made in water comparatively free from cross-currents, boils, and eddies.
The nature of errors caused by excessive turbulence depends largely upon the type of current meter used and to sonie extent upon the method of suspension. lit is generally accepted that a vertical-shaft cup-type current meter if used with rod suspension will overregister in water of excessive turbulence and that a horizontal-shaft propellertype meter will underregister under those conditions. Some experiments have shown that the over-registration of vertical-shaft cup-type meters in turbulent water approximately equals the underregistration of certain horizontal-shaft propeller-type meters. Although the available information on this subject is far from conclusive, it suggests that the degree of accuracy is enhanced if two measurements are made under the same conditions--one with a cup-type meter known to overregister and the other with a propeller-type meter known to underregister by about the same amount-and the results averaged. Such a procedure has been practiced in several districts in the measurement of flow in flumes, tailraces, and other channels where the effects of excessive turbulence cannot be avoided. The results thus obtained have been found to compare favorably with results obtained by other methods.
The angle of current, as applied to stream-gaging procedure, is the difference between 900 and the angle made by the current with the bridge or cableway. To eliminate errors -introduced by such angles it is necessary either to convert the width between two measuring points into the width normal to the current, or to obtain the component of the velocity normal to the measuring section. The method by which this angle is corrected depends on the type of current meter used. A vertical-shaft, cup-type meter, if supported on a rod, will tend to measure the full velocity in the direction in which the water is moving and not in the direction represented by the horizontal axis of the meter. Experiments have shown that the angle between the current and the axis of the meter must exceed 250 before any appreciable error is introduced in the meter registration as a result of the disturbance created by the meter yoke. This characteristic of the cup-


type imeter p)ermiits some departure of the horizontal axis from the direction of flow without serious error. The horizoital-shaft propeller-type meter when supported on a rod tends to measure only the c ('ompolient of the current parallel to the axis of the meter; therefore the meter must be held in line with the direction of flow if correc tions are to be made for angularity. The horizontal axis of either type of meter suspended from a cable will take the direction of the current if no torsional effect is caused by tie suspension cable and if the flow is (omI par:atively free from excessive turbulence. If it is necessary either to correct the width between two measuring points to the width normal to thle current or to obtain the (component of the velocityV normal to t.le nieasuring section, this width or velocity uniist he muultiplied by the cosiiie of the angle made by the current with the nornlal to the section. Several methods may be used in correcting for the 4,,ngle of current, such as the use of an angle chart or ecially revised eagle idicator. on which are shown directly the cosines of the algles: or a direct measurement of the angle may be imade by a l)pro'ttractor. To eliminate the use of a separate angle chart, angle coefficients have been printed on (lisch arge-measuremei it form 9) 275 (see fig. 3). ihe points representing these coefficients are spaced along the left margin of the sheet with the vertex of the angle at the right margin and are so iarr':nged that the form itself can be used as an angle-correct ion chart by holding the vertex parallel to the measuring section and ldoting the coefficient that corresponds most satisfactorilv with the direction of current. The angle prot ravtor is of notebook size and is constructed on transparent celluloid. Litvs representing the angle coefficients al)p)ear symmetrically along both margins of the protractor and are drawn so that the lines representing the lesser coefficients are nearest the base of the protractor. When this protractor is used to measure a counterclockwise angle, the right corner is held in position on an object parallel with the measuring section and the protractor rotated until the right edge coincides with the direction of flow. The angle made by the protractor with the parallel object is the angle of current and can be read directly on the left margin. For a clockwise angle, the left corner of the protractor is held in position on the object parallel with the measuring section, the left edge is brought to coincide with the direction of flow, and the angle of current is read along the right margin of the protractor. A special device known as the Veatch angle-of-current coefficient-indicator, described on page 208, is especially adaptable for use when measuring from a cal)le car or with a crane and reel.
If no angle chart, protractor, or indicator is available, or if their use is inconvenient, the angle of current may be determined in the


follow i ng 1 uier:" Hol a tape parallel to the direction of flow witl a foot mark on a lie (,' object parallel to tle measuring section and with t(he irext foo mar ;dk at a lixcd point. 'IlTe tape is the pivoted about the fixed( point untit it is at right angle.g the ineasuringl section and itle new distancee bet'eenl the fixed poinlit anId the line or object parallel to thle measlurinJg section Is obse r I 1f the tape is graduated in tenths and hundredths of a foot, the new observe(d distance will be the co.-i n of tihe angle (of current. The procedu(res used inII these methods of correct ioul for the angle of (current are based on the assumption that the angle observed on the surface of the water is representative of thle one prevailing throughout the vertical. A further assumption is that vertical angles cause no effect( on tlhe observation of the horizontal angle.

Whei-'e tl here are piers, piling, or teddlies in the mleasurig section, the procedure of measuring becomes more difficult. a nl the possibilities of error are increased in determiining both the cross-section area and the mean velocity. The volume of water that would ordinarily flow through the space occupied by a pier or pile must pass the ineasuring section in adjacent openings. In order that these adjacent openings may carry the additional flow, either the crosssection area or the mean velocity in the open section, or both, must be increased. If the bed of the channel is subject to scour, it is possible that the area in the open section may be sufficiently increased by scour of the bed so that the additional flow may pass without any material increase inl mean velocity. In any event, the distribution of the velocity will be affected for some distance out from the obstruction. The distance from the pier or pile to the point in the open section where the distribution of the velocity is least affected is dependent upon the character of the stream bed, the depth and velocity of the water, the size and shape of the piers and piling, and the distance between them.
Where a pier is free of drift and where the velocities are sufficiently slow to permit depth and velocity observations close to the structure, there is every reason to expect accurate results if additional verticals are used and properly spaced. If turbulence about the pier is so pronounced that -making reliable velocity measurements within several feet of the pier is impracticable, estimates must be made to complete the measurement. Every effort should be made to define the profile of the cross section up to the pier by sounding without the meter as a part of the assembly, as changes in depth because of scour of the channel bed may be greater near the pier than in any other part of the channel. In velocity measurements that are made close to a pier where the turbulence is excessive, the


meter, if suspended by a cable, will tend to drift from the vertical toward the thread of greatest velocity. Consequently, large errors may result if such velocity determinations are used without consideration of the adjacent velocities. A common method of estimating velocities near a pier is to measure the velocity up to a point on each side of the structure where there is no visible effect on the velocity distribution, then to plot the mean velocity in each vertical with respect to its position in the measuring section, and to base the estimates of velocity at the pier on the normal trend of the meanvelocity line extended to the pier. A study of the trend of meanvelocity lines developed for several measurements made at different stages is helpful in increasing the accuracy of estimates made in this planner, I ii these estimates the engineer should assure himself that (lie -,l'ected ar'ea about the )ier contains no back eddy or, negative floW.
The effects froin piles, wlich are usually cylindrical, should be considered when neasurinlg the dischat-ge ini overflow areas where the velocities are usually lower than in the main channel and where shifts in the bed are less likely to occur. With extremely low velocities and in the absence of drift lodged against the piling, the water will flow around each pile in streamline fashion with little or no effect on the distribution of velocity in the open section. Consequently, if the piles are spaced several feet apart and if the velocity observations are made from the upstream side and midway between the piles accurate results will probably be obtained by including the area of the piling in the area of the cross section. If the distribution of velocity at a point midway between the piles is in any way affected by their presence, the area of the piles should be taken into consideration and the total or a part of such area deducted from the area of the cross section. Whether to include or to deduct the area of the piling is therefore a problem requiring careful consideration for each measurement in which piles are involved. An eddy causing a reversal of the direction of flow within the measuring section may, for practical purposes, be likened to a cylindrical roller having a vertical axis. If it may be assumed that the eddy is a rotating cylinder having concentric or excentric rotation with the upstream flow equal to the downstream flow within the revolution, it may be treated as if it were dead water. It is possible, however, that the downstream velocity on one side of the cylinder may be greater than the upstream velocity on the other side, and an effort should be made to determine the net downstream velocity within the area containing the eddy. Soundings should be taken in the eddy for use with the estimates of the net downstream velocity in completing the estimates of discharge.