v. .34 y _. .r .. a a 5 v ‘ x n. a . 4.“ .o. .. _. \ ....Iu :5... o .. .- U “a. w "th ..Ob .\ a ‘ a .c . w . ..- .. v ... r .l‘ . ,5 AK Q3 ...” «'V - . . 4 V: R. ... o\ .~ fit, 9' a. ... ”.1 .‘ ~ ...r u 5 1“ .0 d (-6.. ‘90. I. \ e . p. ...‘Qu. I w. 9 L o. . b m” a 5 H w .5. y a... . a «.1. ... 3 . . ..In. ..... .. . .z . ..., £7. ., . my. .. . w? w m. m m. A... we F . \ w . .. l ... MI.“ a... I Q . u ‘ Q: - ‘ u’L'fi ...3 ...; , .. .. .v. 5..“ .... C. .m... .. . v on v r. . o. ‘1 ,4 n. .. ‘5 a .5 61s» ‘W o M.... . .... {IR P .5 TI.’\. .p 0 ””6 New.“ If. . H 35 ... L. r. «E ”"34.- rflfl ‘ . . _ .. .n . L :2 . z . ..-.g a ..-- ...; f s _ ...» .. a . :1 a £3.— 7 . .\ o El.“ .. M .. u o‘o . a." I g (r .. J. ... v. ...“ 1!.“ ”96. a... . ‘ i. ,... Z z a... .. 3 .... a e 5 .i . . . ... u . . _ um)...» I“... . . U. N w‘. . til a 6.4“ 1.0.3.. . as...» x ‘ . .1. HI. 6. 1 1 ‘ O’mf‘ ‘l. ”‘5. .10 .351 at.“ «(1‘ ,. r of! ”Maw J v& 0!}: .r. .* .kvrr m r .0 "'0!- . r t t. 2". no...“ a ..w In!” 10.0..“ am “cw EM'5 "v n a. ..\ 9 Q ”Mal“ - ”W. H mm . I 1.. Av i s o .4 u «x z; :0 PE m «a. «v 1 .w “All. ... c t o . . ... U .... ........,. E g m... -... a O. «I...» at” . fL '5. r..\ ‘ A: . g, i j Z , e d, a, ,3 ., ,7, , A, f I". -v u . - I I' \ This is to certify that the thesis entitled ”Recommended Yields of Surface Runoff from Small Watersheds on the Hillsdale Soil Complex and Rates of Surface Runoff for Use in the.Design of Farm Ponds in Southern Michigan“ presented by David Franklin Witherspoon has been accepted towards fulfillment of the requirements for Master of Wag“ in Agricultural. Engineering Ma or professor pm March 10, 1952 - I l ‘7‘ w—r—c'- -‘_._——.-_~ “... “Faun-am u '-' u“. a... ... . RECOWDED YIELDS OF SURFACE RUNOFF FROM SMALL UATERSHEDS ON THE HILISDAIE SOIL COMPLEX AND RATES OF SURFACE RUNOFF FOR USE Ill THE DESIGN OF FARM PONIB IN SOUTHERN MICHIGAN by . David Franklin Eitherspoon AN ABSTRACT Submitted to the School of Graduate Studies of Michigan State College of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Agricultural Engineering W Approved - aw 9i .1375" 2 David Franklin iitherspoon ABSTRACT Recommended Yields of Surface Runoff from Small Watersheds on the Hillsdale Soil Complex and Rates of Surface Runoff for Use in the Design of Farm Ponds in Southern Michigan The purpose of this thesis is to make recomendations for yields of surface runoff to be used in the hydrologic design of farm ponds. A comparison is made of the peak rates of runoff occurring ones in twenty-five years for use in spillvay design, modified for Southern . Michigan fron those recommended for the north Appalachian Region and the Claypan Prairies as well as those recommended for Michigan in the United States Department of Agriculture Parners' Bulletin lumber 1859. The main factors governing the design of farm ponds are dis- cussed. 1. These are as follows : Evaporation from free water surfaces Precipitation falling on the reservoir Surface runoff (a) Total yield from the watershed (1:) Rates of runoff for spillway design Subsurface runoff (seepage to the pond) Demand use (the required quantity of water to be taken from the pond for livestock and other uses) Beepage (away from the pond) Silting or sedimentation David Franklin Uitherspoon The recommendations for yields of surface runoff are only applicable to small watersheds of soils of the lillsdale Soil Complex as it is found in southern Michigan. These soils should have similar textures and profile characteristics to those described in this thesis. Since the rates of runoff are derived from general recommendations for other areas these may have wider application. The recommended yields of runoff were estimated from a synthesis of fifty years of runoff records. A relationship between rainfall and runoff for the BMI‘ and winter seasons was established on the basis of the ten years of record from the cultivated watersheds of the liichigan Hydrologic Research Project. A frequency analysis was nade of the fifty years of rainfall: records of the United States Weather Bureau at Lansing, Michigan. From the results of the frequency analysis and the rainfall-runoff relationships the runoff was found that could be depended upon seventy-five percent of the time, and ninety-six percent of the time. ‘ The amount of runoff for an eighteen nonth period as found in this study is as follows: 2.69 inches of runoff can be expected seventy-five percent of the tile . 1.52 inches of runoff can be expected ninety-six percent of the tine . The degree of safety desired and sound Judgment should govern the use of these recommendations. The reconendations contained in this thesis should be considered tentative subject to revision when David Franklin Uitherspoon more complete records are available. A simplified method of design is given in the Appendix to shov the use of the recommendations made here. RECC'IEIDED YIELIB OF SURFACE RUIOFF FROM SIMLL WATERSHEIB O! mm mm SOIL comm AND RATES OF SURFACE HURON FOR 083 II THE DESIGN OF FARM PCUIB II SMHERH MICHIGAN by David Franklin Witherspoon A THESIS Submitted to the School of Graduate Studies of Michigan State College of Agriculture and Applied Science in partial fulfillment of the requirements ' for the degree of MASTER OF SCIHCE Department of Agricultural Engineering 1952 ACKNWIEDGEHEFIS The author wishes to express his sincere thanks to Mr. E.R. Kidder under whose inspiration, supervision, and unfailing interest this investigation was undertaken. He is also indebted to Dr. E.P. Whiteside for his guidance and help in the soil descriptions on the watersheds, as well as their definition for the specified areas of Michigan. Grateful acknowledgement is also due to Mr. J.O. Veatch for his map of the Soil Associations of Michigan and to Mr. I..I.. Rarrold and Mr.l.E. Minshall and others of the United States Soil Conservation Service for their helpful suggestions and assistance. Acknowledgement is also made for the data and negatives of maps of the Michigan Hydrologic Research Project Operated by the United States Department of Agriculture, Soil Conservation Service in cooperation with the Michigan Agricultural Experiment Stat ion . TABLE OF CONTENTS INTRODUCTION ......OOIOOOOOOOOOOOOOO000............OOOOOOOIOOOOO m or ”MATURE 00.0.0000......OOOOOOOOO......OOOOOOOOOOOOO “nerd 0.00..........0.0.0.0000.........OOOOOOOOOOOOOOOOOO Evaporation ............................................... Surface Runoff ............................................ Subsurface Runoff ......................................... Demand U86 ................................................ Seopage ................................................... Silting ................................................... METHOD OF STUDY ............................................... Description of watersheds ................................ Area of Application of Iield.Recommendations ............. Runoff Yield Design Value Determination .................. Rates or leOff for Spill-'8’ Dead-En eeeeeeeeeeeeeeeeeeeeee DISCUSSION .................................................... General .................................................. Evaporation .............................................. Precipitation ............................................ Surface Runoff’- Yield ................................... Surface Runoff - Rates ................................... subsurface Rumff 00.0.0...O......OOOOOOOOOO00.00.00...... W U89 .....OOOOOOOOOO0.0...0.0.0.000...0.00.00.00.00. somooooqosqu—i F4 F1 c: c: 5151 26 26 27 30 3O 3O seepage ......OOOOOO......O.......OOOOOOOOO00.000.00.00. 31 Silting .0.......OOOOCOOOOOOOOOOOOO0......00.00.000.000. 31 sum-ARI 0.00.000.00.000...........OOOOOOOOOOOOOOO.....OOOOOOC 32 APPmIX ......OOOOOOO0.00.000.000.000.........OOOOOOCOOOO... 34 umm CITE ....OOOOOOOOOOOOOOO0.0.0.000...00.00.0000... 38 LIST OF FIGURES Figure l. Rainfall - Runoff Relationship, Sumner ........... 18 Figure 2. Rainfall - Runoff Relationship, Winter ........... 19 Figure 3. B31968 Of Runoff eeeeeeeeeseeeeeeeeeeeeeeeeeeeeeese 2.3 LIST OF PLATES Plate 1. lap of Area of Application in Michigan ........... 3 Plat. 2. lap Of cultivated Watersheds eeeeeeeeeeeeeeeeeeee 11 Plate 3. Slope Map of Cultivated Watersheds .............. 12 LIST OF TABLES Table I. Minimm Dependable W311 eeeeeeeeeeeeeeeeeeeee 17 Table II. Estimated Dependable Runoff Yields .............. 21 RQIIIJYI care-Ava . l a g m I: I v- A a a I a F A I e I D " l D '- ‘I F n .. I I C a l o C I q ‘ p INTROIIIC'I‘ ION Farm ponds are a useful surface water conservation practice. In specialized farming areas supplemental irrigation is rapidly becoming an essential practice during the critical periods of crap growth. Farm ponds offer a water supply for this purpose . Water for spraying purposes in orchards , where underground or other surface water supplies are inadequate, may be obtained from a farm pond. In the ”grassland“ type of agriculture advocated by some con- servationists , farm ponds fulfill the need for additional stock- watering facilities. As well as the uses already mentioned farm ponds also provide water supplies for fire protection, wildlife and recreation. The demand for information concerning farm ponds by the farmer is quite apparent from the volume of extension literature available from state and federal agencies. Very little basic hydrologic data necessary for the design of a farm pond has been published. This is particularly true in Michigan. At the present time the United States Soil Conservation Service is using generalized data adapted to this area frqs regional information. It is the purpose of this thesis to discuss in detail surface . runoff as it affects the design of a farm pond in southern Michigan with regard to available data and literature . ‘ The other factors governing the design of a farm pond are also discussed. - 2 - In farm pond design the following factors are taken into considerations 1. 2. 3. 4. 5. 6. 7. Evaporation from free water surfaces Precipitation falling on the reservoir Surface runoff: (a) Total yield from the watershed (b) Rates of runoff for spillway design Subsurface runoff (seepage to the pond) Demand use (the required quantity of water to be taken from the pond for livestock and other uses) Seepage (away from the pond) Silting or sedimentation The yields of surface runoff recommended in this thesis are only applicable to the Hillsdale Soil Complex. The area of’Michigan to which these recommendations apply is designated in black on.Plate 1, page 30 The soil series making up the associations designated are Hillsdale, Coloma, Bellefontaine, and.Miami. The recommended yields for this area are based on synthesized fifty year runoff records. These were determined from relationships established on the basis of the ten years of record of the cultivated watersheds of the Michigan Hydrologic Research.ProJect. The rates of runoff are derived by extrapolation of recommendations made for the North.Appalachian.Region and the Claypan.Prairies as well as recommendations fer Michigan from the United States Department of Agriculture Parmers' Bulletin Number 1859. N IICHIGAN MAP W ”IL Assoclmous Plate 1. Area of Application designated in Black - a - In most cases the site of pond construction is predetermined by economics and feasibility of construction. The recommendations made in this thesis will aid in checking whether the size of watershed available will yield a water supply which can be depended upon. ’ The recommended rates of runoff will aid in the design of spillways. In the science of hydrology rigorous proofs are seldom possible . Therefore, the reconendations contained in this thesis should be considered tentative, subject to revision when more complete records are available . -5- REVIEW OF LITERATURE General .Most extension.publications which comprise the volume of the literature on farm.ponds make only general statements concerning farm pond design and construction. These apply only in a specific area. Calkins (1) lists the essential requirements of the basic structural types of farm ponds as: l. The pond should meet a definite need, that is, a water supply for livestock, spraying, irrigation, fire protection, wildlife or recreation. Source of water for the pond.should.be free from.barnlot drainage and.undesirable industrial wastes. Pond.must have an impervious dam.and floor. Borings or test holes should be drilled.to determine soil conditions. Adequate spillway capacity - ponds through which surface runoff flows require spillways designed for flood runoff. All ponds should have an Open auxiliary spillway to prevent water overt0pping the earth fill in case the drop inlet becomes clogged or its capacity is exceeded. Inflow should.be regulated.to the need by adding to or subtracting from.natural watersheds by diversion ditches and.terraces. Ponds should.be constructed on topography which.permits construction at reasonable cost, that is, on watercourses of feur percent or less slepe and side hills of less than eight percent slope. -6- 7. Ponds should have a drain, especially fish ponds. Evaporation Estimates of evaporation in Michigan are somewhat variable. Thornthwaite and Holzman (14) state that bodies of free water make available a continuous supply of moisture for evaporation and actual losses are dependent directly upon meteorologic factors. For this reason it has been possible to develop emperical formulae which permit the computation, with reasonable accuracy, the anticipated losses from lakes and reservoirs in terms of meteorologic data alone. Harrold (7) estimates, based on his knowledge of hydrologic phenomena, that, evaporation from, and precipitation on a pond area in Michigan might be expected to balance over critical dry periods of six months or more. Follansbee (A) in his study of evaporation notes that the area of lowest evaporation from a free water surface in the United States is the Great Lakes Region where it ranges from fifteen to twenty inches per year. Kimball (9) estimated the evaporation in the area under consideration as a maximum of thirty-five inches a year. Horton (8) interpolated values using data collected at Germfask in the upper peninsula of Michigan. He estimated, by converting the data to a base for the continental United States, the evaporation over this area of southern Michigan under consideration, ranged between thirty-seven and forty-three inches per annum. This was for a Class A.Weather Bureau pan. .. 7 .. Thornthwaite (15) estimates the moisture deficiency, that is, the difference between precipitation and water losses in Michigan, as two to three inches during the dry season of the year and surplus as ten inches for the wet seasons of the year. He estimates the potential evapo-transpiration from plants as twenty-four to twenty-six inches a year . , Meyer (11) using his evaporation formula on United States Weather Bureau data found the evaporation over the area under consideration varied from twenty-five to thirty-five inches annually. He shows that there is a surplus of two to seven inches of precipitation on a free water surface over the amount lost by evaporation. At Lansing, Michigan, he calculated that the evaporation has been greater. than precipitation only two years in the thirty years of record from 1910 to 19%. This deficiency was six and seven inches during two years in the early thirties. Surface Runoff In this particular phase of farm pond design most of the work has been done with reference to large watersheds . Considerable study has been done in areas where the most of the flood peaks are caused by thunderstorm activity instead of the spring freshet as is the case in Hichigan. Hamilton and. Jepson (5) state that the watershed characteristics (slope, shape, size, cover and soil) and storm characteristics (amount, duration and intensity, of rainfall) have a direct effect on the annual yield and peak flow of surface runoff from any area. nan-old (6) .- 8 - observes that profile drainage differences are of less importance in causing flood peaks than they are in causing differences in annual yield. Himshall (12) found that as the soil became less permeable the relation of rainfall to runoff became more consistent. Subsurface Runoff Very little material is available for use in the estimation of this factor of farm pond design. Cook (3) states that the weakest point in most procedures for the design of a farm pond is the assumption that surface runoff constitutes the only source of inflow. He cites as an example the many farm ponds in central Missouri with very smll grassed contributing areas. He states that these often receive considerable inflow from rains which produced no surface runoff. Parsons (13) in Alabama found the groundwater flow to a pond which he studied was 1L6 inches where the surface runoff was 5.22 inches from twenty-seven acres of terraced land. / Demand Use 5 Hamilton and Jepson (3) developed a chart for the estimation of livestock water requirements . These are general over-all recomendations m the continental United States . Seepage Parsons (1k) in his study in Alabam found that the loss due to seepage from the pond he studied was 0.35 acre-inches per day on a mean-annual basis. The pond area was 1.5 acres at spillway elevation. The soil on the floor of this pond was a sandy topsoil of twenty-five - 9 .. per cent non-capillary porosity with a clay subsoil of 10.5 per cent non-capillary porosity. There has been considerable work done on losses due to seepage from irrigation canals. However, these investigations for the most part are concerned with lower heads of water than are normally encountered in the design of a farm pond. Carpenter (2) in his early investigations in Colorado found that the success of the silting process in sealing the reservoir against seepage may be expected to be greater in small reservoirs than in large reservoirs . Silting Harrold (6) states that a watershed cover of ninety per cent in grass or woods will reduce silting or sedimentation to a negligible amount. However, the results presented by Harrold in this publication show that under conditions of less cover silting may be serious in reducing the reservoir storage capacity , -10- METHOD OF STUDY Descriptions of Watersheds Three watersheds are included in the Michigan Mmlogic Research ProJect. The two cultivated watersheds, the data from which is used in this stuw, are located on the Michigan State College Farms about two miles south of East Lansing, Michigan. The third watershed which has a permanent wooded cover is located on the Rose Lake Conservation Farm about nine miles east and north of East Lansing. The data rom this watershed was not used in this study. This hydrologic project was set up in 1911-1 for the purpose of studying the effect of cover on soil loss and surface runoff. The two cultivated watersheds are managed in a manner similar to that of an average farm using conservation practices. Across the slope cultivation and four year rotations of corn, grain, hay, hay, are the main conservation practices used. Climatic instrumentation on the cultivated watersheds is by standard United States Weather Bureau instruments . The runoff measuring installations are the standard Soil Conservation Service Research Type E flunes used on small watersheds. Silting basins and Bamser divisors are used in conjunction with the flames for the measurement. of sediment load in runoff. A detailed instrumentation and soils map of the cultivated watersheds is shown on Plate 2, page 11. A map showing the tapography of the cultivated watersheds in detail on Plate 3, page 12. ‘1‘ a. 3.300.. gases-30.5 oval . Calls-s ass-...... 0.“ .....h.’ It ccccc as... assesses ....s (I Sean-o...”- Hn..:ql. . s... a... use... 0...:qu assess. as... partial... 0.. ' ti... ...-l. asses-s so 0!-..0l its: so! ass. s as... sass ll.l 3 .... 31:... ton 93s.. on ..(u 0 h: sass Is. sec!!- sls-Ipsso s run-.3- uaapacuaa: I: . 0 000000000... 33:31:... \ z/I pauses ... 313 “s. e ' III o 11.0]de . . o . Io... use! and no a 3:. . 03.. .00.. 38¢ .31. El OP I‘I sass-El. 05.340. 3.00.0 034 9.0338 ® 22:; .3233 32:: ® . g 82.50 4.0. to: of loci (use. ... e! .- ...: .2 «Issue cusses 36. .832; as... :83. c noses- S: .s oases-.....o. nuts... 8 No.02 a gm- ;usn .- x 3.6.... :3 33.3» a. If. .. 0 4.1.2.! 30* .0 Icidaaliuud zo_momu 00.8096. >523: 'C 30:00.3 0!... mp Fla-ta 20 aim: 52:35... 32:31:: no US$58: uaapaauaa: 4.00 ale sinus: sates:- a not. {so use: .so losses-sauces... I . I ,/ . + -._ ‘ ----.. Is. so... I. ...... caisson-...... 83....- . aiihmulmawz— Gad audHfl in! dais: 3‘- 3 s33 Jan: 8343 51.23... 1104 ciao-3.8.140; 8.4.10.7: :38 office .840; rot: as; java? . .3 mac. .ozaa as: .4409}... .03 lacs 52: as: :3:st .nuo or: 1... 2.40.. sap-l .... a; suds a use! assis- uacpo ...-:35 as .33... ‘03.; ’10.: 0.0308... social I; .i; g _ as a: .8! seas-loco a. ‘ UO_>¢U. 20.h(>¢Uh8¢0 4.00 5% mouzmmubxs) ouk<>rfizu #41 .3. 332 ‘ .. .0 :4 2a.... as... 2 ToPodnApH‘cAL wnvav or Ruuvovv may» ’ART or aw. 74 or bu.no,vanflnw. 0W co, Mus-0 ‘3‘” "7”.me fillldflfl Plum HUDOON VARM. mac. M5. 0. “pm“ '0'“ Sada" \ man - so 3-.0. “I‘d” WLW"‘ ‘I't'lfitw "A?!” Undeth'nad figures on Lknahoo‘s. 0.419.500“ Sine. No 3336 cmmm r50 mrznnsns m3. m. 6454. lea-m “q, $Irvu‘ov. / I g , [1 ”73¢ $1... + H.— m PL q): _ W [mm m... ~ - . I I. / 1" 1 ... m I 1 ,> . .-.. n ' 3-4 as 17 . ‘ 3, [fits u t, mrmaraav cuss I 2 W“ "‘ rams! m c-s 12 at F ‘ i imp “7...; " ..n u C 0.: mm OW ‘ - . “*- " “'r , ’ - sou. couaenvanou seawcs r—trn on as? "a .L‘. a 0am a. l final lass Isa-o ”m Mm . m m "C ”M ,m_____‘ ‘ W Au. Wu Sun 'A' . £01 ‘ "-5., A- Urns-rust” '3' - 5.: z s-u-ea s s on. s-u-«l I ... . Qt. R48 L L l Plate 3. - 13 - The data from.the wooded watershed was considered inadequate and is not used here. A.probable reason for the lack of surface runoff under wooded cover is the increased interception of rainfall, evapo- transpiration and absorption of water'by the hwmus of the forest floor. From the ten years of record of surface runoff from.the wooded watershed it was Observed that the total surface runoff in all years was less than five per cent of the annual rainfall. In four out of the ten years of record the runoff was less than one per cent of the rainfall. On the average the records showed that the runoff from the cultivated.watersheds was eight times that of the wooded watershed. An area producing little or no runoff cannot be considered a satisfactory source of supply of surface runoff suitable for filling farm.ponds. Since Plate 2 page 11 was made the soils of the cultivated watersheds have been re-mapped. The boundaries between the soil types are essentially the same as on the original map. The code numbers in the soil descriptions refer to the code numbers on the soil map on Plate 2 page 11. The soil descriptions as made by Dr. E.P.‘Whiteside of the soils on the cultivated.watersheds of the Michigan.Eydrologic Research Project are as follows: - 1h - Code No. 515 Soil Type: Spinks loamy fine sand Horizon Depth Texture Remarks AP 0-9" loamy fine sand Bp 9-13” loamy fine sand B is mottled A2 13-27" loamy fine sand Occasional pebble B 27"- heavy loamy fine found. sand to light loamy fine sand Profile drainage fair to good Topography - Rolling, depressional Original vegetative cover - hardwood forest - beech, maple. Code No.819 and 511 Soil Type: Coloma loamy fine sand . Horizon Depth Texture Remarks Ap 0-10" heavy loamy fine sand A2 10-27" loamy fine sand 32 27-37" loamy fine sand Code No.510 Soil Type: Hillsdale fine sandy loam Horizon Depth Texture Remarks Ap 0-7" ‘ fine sandy loan 32 7-27” clay loam D1 27"- fine sand to loamy Dl has streaks fine sand of texture difference. - 15 - Area of Application of Yield Recommendations The area of application is shown on Plate 1, page 3, and is designated in black on the Map of the Soil Associations of Michigan. The soil associations of the area are made up of Bellefontaine, Hillsdale, Coloma, and Miami series. These associations cover a total area of 1,099,100 acres of southern Michigan. General descriptions of these soil series are as follows: Bellefontaine - Sandy loans and light loams, moderately stony. Reddish sandy and stony friable subsoil and coarse pervious substratum. gigsdale - Sandy loans and light loans; light brownish and yellowish surface soil underlain by yellowish friable but moderately retentive sandy loam and gritty clay. Ma - Sands or light sandy loans, underlain by yellowish dry sand to three feet or more, then by pervious heterogenous sand, clay, and stones. ling; - Light brownish loan and silt loam over brownish compact and retentive but granular gritty clay. 01w substratum extends to several feet. It can be readily seen from these general descriptions that these soil associations contain soil types which can have wide textural differences and consequently profile differences. Therefore, it is imperative that before the yield recommendations made in this thesis can be safely used for any watershed, the soils of the watershed should be mapped in detail with complete profile descriptions. These descriptions should be compared with those given in this thesis for - 16 .. the experimental watersheds of the Michigan Hydrologic Research Project. It can be safely assumed that the recommendations made here can be safely applied to the more impermeable soils of these associations, that is, Miami and Hillsdale series. however, care should be taken when applying these recommendations to the soils of lighter texture than those found on the cultivated watersheds at East Lansing, Michigan. These soils would be made up of the lighter textm'ed soils (sands) of the Coloma series. Runoff Yield Design Value Determination In order to synthesize fifty years of runoff records which would have some degree of validity and accuracy, as well as be safe for use in design, it was decided to divide the year into two periods. These were the non-growing season (winter), that is, October to May (8 months) or October to April (7 months), and the growing season (stunner) May to September (5 months) or June to September (h months). Two different length periods were chosen for each season. This was done because May and June are the months of highest rainfall in this area. In May newly planted vegetation may not be well established. Therefore, the month of May was included in both seasons to minimize any variation in runoff caused by this condition. For the above chosen periods the rainfall records of the United States Weather Bureau at Lansing, Michigan were subjected to a frequency analysis to find the minimum amounts of rainfall which could be depended upon seventy-five and ninety-six per cent of the time. The results of this frequency analysis are shown in Table I, page 17. -17- TABLE I MINIMUM DEPENDABLE RAINEALL Percentage of the Time Dependable Period I7§per cent 96 per cent ‘Winter 8 months 17.62 inches 1h.28 inches ‘Winter 7'months 13.9h inches ’ 10.21 inches Summer 5 months 13.10 inches 9.30 inches Summer h months 7.98-1nches 5.59 inches The runoff records of the Michigan Hydrologic Research Project were plotted against the rainfall for the winter periods of eight and seven months and the summer periods of five and four months. This was done on log-log paper as is shown in Figures 1 and 2 pages 18 and 19 respectively. I The rainfall-runoff records for these periods were then analyzed by the method of least squares. Various combinations of the data were used to Obtain the maximum.number of relationships, that is, each watershed separately, both watersheds together to find an over-all relationship. Relationships were then chosen for the summer season and the winter season. These are as follows: -5 3.61 Summer R 1.18 X 10 P Figure 1 page 18 -3 2.hl Winter R 2.h3 X 10 P in blue on Figure 2 page 19 Where R is runoff and P is precipitation in Indus Ram/o/l 6* i 9 . j 0 d0/ a) ' 70 Runoff in Inches M/IYMZL “RUNOFF ATM 770/V5fl/P Summer 4 0/25 Noni/15 fly. / 70 in lac/Its ” ” (Jae/x 00/ -19- ~'cz/ ' ' "to Runoff m MCI": RAM/fill Z *fl’U/VOFF Rf [47/0/15 fi/PS Mn/‘z/ 7 or é’ Mon #25 fig. 2 /0 - 20 - For the summer relationship the one which gave the lowest value of runoff was chosen. As can.be seen in Figure 1, page 18, the runoff during the summer months is extremely variable. Therefore, fer safety the previously mentioned relationship was chosen. Since the majority of the runoff yield occurs during the winter months the relationship chosen as reliable fer this period is very important. In Figure 2, page 19, two relationships are shown. The one shown in.black R = 0.227 P00977 is the relationship which gives the most consistent values of runoff as compared with the actual record. The relationship shown in.blue R ; 2.h3 x 10 -3 P ZOkl is the relationship actually used fer the recommendations made here since it gave values of runoff which.more nearly approximated those which could.be depended upon seventy-five per cent and.ninety-six per cent of the time. This relationship also gave the lowest values of runoff for the winter period. The runoff yield values obtained by applying the minimum dependable rainfall in the chosen relationships are given in Table II, page 21. - 21 - TABIE II ESTIMATED DEPENDABIE RUNOFF YIELDS Percentage of the Time Dependable Period 75 per cent 96 per cent Winter 8 months 2.h3 inches l.h6 inches Winter 7 months 1.37 inches 0.67 inches Summer 5 months 0.13 inches 0.03 inches Summer h months 0.02 inches 0.00 inches The most desirable time of construction of a farm.pond is in the spring or early summer. Under conditions of runoff found in.Michigan the newly constructed pond would have very little water in it the first summer and would probably be filled by the spring snow melt of the following year. If the demand use is to be taken into account in the design of a pond, enough water will have to be stored which will adequately supply the demand for the critical dry periods in eighteen months, that is, the demands of the following spring and summer and any possible winter demands. Therefore, to Obtain an eighteen month period for design from.Tab1e II, page 21, it is necessary to use one eight month winter period and two five month summer periods. The eighteen month period of design for demand use is used.by both.Herrold (6) and Krimgold and Minshall (10). For use of the design recommendations for small farm ponds, a method of design is given in the Appendix. - 22 - Rates of Runoff for Spillway Design The experimental watersheds at East Lansing, Michigan are approximately 1.9 acres in size and data is not available from.1arger areas in Michigan except from the very large watersheds gaged.by the United States Geological Survey. Therefore, estimates of rates of runoff for use in the design of pond spillways cannot be satisfactorily made on the basis of recorded data fer Michigan. However, Earrold (6) has made recommendations for rates of runoff for a frequency occurrence in twenty-five years for the lorth.Appalechian Region based.on records of the‘United States Geological Survey and data obtained from the Soil Conservation Service experimental watersheds at Csshocton and Zanesville, Ohio. He gives location factors to be applied to the rates of runoff for the variation in rainfall intensities. Krimgold and Minshall (10) using records obtained from the United States Geological Survey and experimental watersheds within the Claypan Prairies of lower Illinois made recommendations for rates-of runoff occurring once in twenty-five years which also can be modified by location factors. In Farmers' Bulletin Number 1859 general recommendations for rates of runoff are made for the continenta1.United States. These rates of runoff are also modified for various locations by recommended factors. Figure 3, page 23, was Obtained for southern Michigan by plotting the recennendations of Earrold (6) and Krimgold and Minshall (10) using their recommended location factors extrapolated.parallel to those recommended in Farmers' Bulletin Number 1859. The curve for a watershed with high runoff producing characteristics as given in Farmers' Bulletin ,OH'O n cubic fafpsr sscom’ ..."... . £unofl' § rub-n x I Cyc It. I \ i O 09". SIC-IIO U B ’0‘ I n a s’ .‘ I}: l 8 5 I19 . Wakrs/ud Ilsa in Jena C arm Der/Jed from: Rift-5 CO/I— RWf-f xr/mxszpmbalx : fig, 5 as.” Mamas/som— - 2:; - Number 1859 and.modified for the area under consideration is also shown for comparison. Harrold (6) makes the following recommendations fer use with the flood.peaks given in his publication: 1. These flood.peak runoff rates are for watersheds having good permanent vegetal cover (50 to 75 per cent or more impasse or woods) 2. Where the entire watershed is cultivated in a 3- or h-year rotation (1 or 2 years in grass), multiply the peak.values by 1.7. 3. Where the entire watershed is in woods or good grass cover, multiply the peak values by 0.6. The recommendations of Krimgold and.linshall (10) are made for mixed cover. In both cases the authors specify that no safety factor need.be applied to their recommended.values of peak flows occurring once in twenty-five years. By a comparison of the recommendations made in other areas and extrapolated for hichigan an extremely good agreement is Obtained as is shown in Figure 3, page 23. Therefore, it can be assumed.with a reasonable degree of safety that these values are valid for Michigan. Since the recommendations of'fiarrold (6) and.Krimgold (10) are based on actual watershed records, any value used within the range of the two curves would.be reasonably satisfactory fer a specific area unless the watershed in question has extreme characteristics which might cause greater peak flows than anticipated in these recommendations. - 25 - Ho specific recommendations are made for the area of Michigan under consideration. 'when designing a pond.spillway from this data the rates of runoff used should.be governed.by the runoff characteristics of the watershed and sound Judgment. - 26 - DISCUSSION General Since the majority of the runoff in this region occurs during the spring runoff, the size of watershed does not necessarily need to be changed by the use of diversion ditches and terraces if found.too small for the size of pond. .Additional water.might be obtained'by the use of trees and other obstructions on the watershed which could cause appreciably increased snow accumulation during the winter season. Snow ridging has been used for this purpose in North Dakota. The essential structural requirements of the basic types of farm ponds (refer to page 5) as listed by Calkins (1) should be strictly adhered to if a pond is to perfbrm.in the manner for which.it was designed. Evaporation According to the literature it would seem that evaporation need not be taken into consideration in the design of a farm.pond in southern.Michigan. However, some observers have estimated evaporation from a free water surface through a wide range of fifteen to ferty- three inches per annum.fOr the area of’Michigan.under consideration. Meyer (11) whose work is generally accepted in estimation of this hydrologic factor verifies the assumption that the evaporation equals the precipitation on an annual basis. However, when further - study of this factor of farm.pond design has been made and additional records from the newly installed.Weather Bureau Evaporation Pan at the - 27 - Michigan Hydrologic Research ProJect, are available, this factor could be more accurately estimated for the area. When this has been done evaporation might be taken into consideration in design for the critical summer months and the size of watershed could.be more accurately determined. Precipitation Over the area designated on Plate I, page 3, the average annual precipitation varies from.thirty to thirty-six inches. The mean annual precipitation at Lansing, Michigan is 31.3% inches. Therefore, this analysis is on the safe side for the area under consideration. Surface Runoff - Yield The method used in the determination of the estimated runoff yields was derived from suggestions by Minshall (12). The author realizes it does not satisfy the true scientific approach and is not a rigid analysis governed by mathematical laws. However, it does offer a temporary solution to the problem.of runoff estimation in the absence of better methods of analysis and.more complete data. The results that were obtained show that h.l+6 acres of watershed are required to supply one acre-foot of surface water. This amount would be supplied seventy-five per cent of the time in eighteen months. To supply one acre-foot of water in eighteen.months ninety-six per cent of the time it would.require 7.89 acres. Hamilton and Jepson (5) recommend.five acres of watershed in the area under consideration to supply one acre-foot of surface water. no mention is made of the - 28 - frequency these recommendations are made for. Krimgold and Minshall (10) estimate that for a seventeen month period on the Claypan Prairies 2.55 inches of runoff could.be expected seventy-five per cent of the time and 0.30 inches could.be expected ninety-six per cent of the time. These recommendations were for northwestern Illinois and.Iowa based on data from watersheds at Edwardsville, Illinois and.McCredie,Missouri. Harrold.(6) recommends O.h8 inches of runoff could.be expected seventy-five per cent of the time and 0.2h inches of runoff ninety-six per cent of the time for well-drained areas having no seeps or springs. These recommendations are made for an eighteen month.period in southeastern Ohio. In Michigan as a result of this study the estimates for an eighteen.month.period froa.an area on the Hillsdale Soil Complex are 1.52 inches of runoff ninety-six per cent of the time and 2.69 inches of runoff seventy-five per cent of the time. The recommendations made here do not agree very closely with those made in Illinois and Ohio. .A.possible reason for this is that the recommendations made in Illinois and Ohio do not require a safety factor in use whereas those made in this thesis should.be modified ‘by field experience and sound Judgment. Another possible reason for the variation in the recommendations is the fact that in Illinois and Ohio a larger portion of the runoff is caused by summer storme. From observation of the records of the Michigan Hydrologie Research ProJect it was feund.that for nine out of the ten years of record, over eighty per cent of the total annual runoff occurs during - 29 - the first three months of the year. This shows that very little runoff is caused.by summer storms in the area of Michigan under consideration. To further verify the estimates made here for eighteen months the actual records for these periods were examined. The lowest runoff value during the ten years of record is 1.6505 inches fromUWatershed.A and 1.7309 inches from Watershed B. These occurred during the eighteen months covering the growing periods for l9hh and l9h5. The precipitation during this period was 12.78 inches for one growing season of five months in 19%, 26.09 inches for the other growing season of five months in 19h5, and lh.6h inches for the winter season of eight months. Referring to Table I, page 17, the value 12.78 inches could be depended upon slightly more than seventy-five per cent of the time; 26.09 inches could be depended upon less than seventy-five per cent of the time, and lh.6h inches could be depended.upon slightly less than ninety-six per cent of the time. Since the precipitation for the winter months is very near the minimum which could be depended upon ninety-six per cent of the time, the runoff is correspondingly low and within the estimated.mintmmm for this period, This occurrence shows that the relationship chosen is valid for the minimum rainfall Obtained from the frequency analysis. These recommendations are lhaited in their use. They should only be used as recommended.where surface water is the only source of inflow to the pond and for ponds of a capacity of one to twenty-five acre feet. s 30 - Surface Runoff - Rates In Figure 3, page 23, a comparison of the rates of runoff from other areas, modified for Michigan, is shown. Since these curves were derived from.genera1 recommendations based on actual occurrence, the rates of runoff for the range of watershed sizes up to two hundred acres should be reliable. For watersheds of larger size Ithan two hundred acres special methods of flood.peak flow estimation should be used. Subsurface Runoff This hydrologic factor should.be estimated for the individual site of pond construction. Seeps and springs are usually not apparent on very small watersheds, but as the size of watershed increases the flow derived from these sources becomes increasingly important. The surface runoff estimates given in this paper are made far areas where there is no subsurface flow to the pond. If there is an estimable amount of subsurface flow to the pond the yields of surface water should.be modified accordingly. under certain conditions of high water-table on the pond site, subsurface flow constitutes a maJor portion of the water supply. Demand'Use This factor can be estimated‘by simple calculation. Irrigation and livestock needs can‘be estimated from experience and hy consultation with the farmer. -31-‘ Seepage This factor is prObably the most difficult to estimate. Any underestimation of this factor will result in failure of the pond. The ideal site for construction is on a heavy clay soil where seepage is a minimums However, this type of site is not always available. Silting after the pond is in operation or puddling of the soil on the pond site may decrease seepage. The best practice is a complete investigation of the pond site with deep borings and where possible a study of its glacio-geology. In this way any permeable material may be avoided and seepage kept at a minimum. Silting or Sedimentation Reservoir storage capacity can be severely reduced by silting. More data is required.before reliable estimates of this factor can be made. However, silting is a known prOblem.when a large part of the watershed area is cultivated. Therefore, the most advisable practice is to keep as much of the contributing area as possible in permanent vegetation. In particular, the area immediately surrounding the pond should.be kept in permanent vegetation to keep the reservoir silting at a minimum. - 32 - SUMMARY The factors governing farm pond design are complex. Those discussed in this thesis with regard to their estimation for conditions in southern.Michigan are as follows: 1. Evaporation 2. Precipitation 3. Surface runoff (a) Yield of surface runoff (b) Rates of surface runoff h subsurface runoff 5. Demand use 6 Seepage 7. Silting Surface runoff is discussed in detail. .A tentative method of analysis is presented for the determination of yield estimates from short period records. These recommendations for determining watershed sizes fer ponds of one to twenty-five acre-feet capacity Obtained.by this method are: . 2.69 inches of runoff can be expected seventy-five per cent of the time over an eighteen.month.period. 1.52 inches of runoff can be expected.ninetybsix per cent of the time over an eighteen.month period. These recommendations only apply to the soils of the Hillsdale Soil Complex having similar profile characteristics to those fOund.on - 33 - the cultivated watersheds of the Michigan Hydrologic Research Project. The recommended yields should be modified by sound Judgment in use. A comparison is made of the most reliable estimates of peak rates of runoff modified for use in southern Michigan. The curves shown in Figure 3, page 23 , will serve as a guide for design values for spillways. When additional data is available the recommendations made in this thesis should be reviewed and modified accordingly. A method of design of farm ponds is presented in the Appendix. - 3h - APPENDH A.method for the hydrologic design of farm.ponds using the recommendations of runoff yields made in this thesis is given here. This.method is adapted from.a simplified.method developed by Harrold (6) for use in the design of small ponds in the North Appalachian Region, the cost of which does not exceed five hundred dollars. To determine the size of drainage area proceed as follows: Step 1 For the selected.pond site, determine the surface area and depth of the pond at the elevation of the principal spillway. Check the geology of the pond site for possible seepage losses. Step 2 Determine for the watershed, the predominant type of soil and compare the profile characteristics with those given in the descriptions on page IQ. If the soils are lighter than those described the runoff yield.values should be decreased according to the Judgment of the designer. SteP 3 Calculate the volume of storage using the depth and the mean surface area of the pond (O.h of the pond area at spillway elevation). Allow for the amount of water contributed.by seeps and springs. Using the value of runoff ydeld dependable seventy-five per cent of the -35— . . time (2.69 inches) calculate the size of watershed required to fill the pond in an eighteen month period which includes two growing seasons. Step h Estimate the water seepage loss plus the water use demands for the pond for an eighteen.month.period which includes two growing seasons. If this total exceeds the value of estimated storage in Step 3 the spillway elevation of the pond will have to be raised to increase the storage and the size of watershed will have to be increased. If this amount is less than that found in Step 3 the size of drainage area should be checked.using the value of runoff yield.dependab1e ninety-six per cent of the time (1.52 inches) that it will supply the use demands and seepage losses ninety-six per cent of the time. Step 5 ‘ For excessive use demands, determine the additional drainage area required using the value of runoff for ninety-six per cent of the time expectancy (1.52 inches). Example: Given a pond site having a surface area at spillway elevation of 2.35 acres and a depth of 7 feet. Steps 1 and 3: The mean surface is 2.35 x 0.h e 0.9M acres. - 36 - The watershed is a loamy sand.with profile characteristics similar to those mapped on the experimental watersheds. The pond site is a slowly permeable clay soil. Therefore, seepage would be a minimum and the values recommended.here would apply. The volume of storage is 0.9h x 7 g 6.58 acre-feet. ‘Watershed.required to fill the pond a 6.58 x 12 m 29.3 acres 2. 9 The watershed contains no seeps or springs. Step h: ' Seepage is negligible, the demand use is estimated at 7 acre- feet. Therefore, the size of drainage area will have to be increased and the spillway elevation increased if possible. Additional storage required 7 - 6.58 : O.h2 acre-feet .Additional drainage area required 0.h2 x 12 z 3.32 acres 1.52 Total drainage area required is 32.6 acres. If additional drainage area was not required, for example, when the use demand is 6.0 acre-feet, this should.be checked to find if additional watershed area is needed to supply this demand ninety-six per cent of the time. Check 6.0 x 12 : h7.h acres ‘ ‘l.52 Therefbre, the size of watershed would have to be increased 18.1 acres to provide a watershed which will supply the demand use ninety- six per cent of the time. When the demand use is lower the watershed .. 37 - found which will fill the pond in eighteen months, seventy-five per cent of the time , will often be sufficient to supply a dependable source of water ninety-six per cent of the time. - 3g - LITERATURE CITED Calkins, R.S. Essential Reguirements of Basic Structural Types of Farm Ponds. Agric. Eng.28: l$89492, (November, 191W). Carpenter, L.G. ,The Loss of Water from Reservoirs by Seepage and Evaporation. Colorado Agricultural Experiment Station. Colorado Publication No.15. (1898). Cook, H.L. written communication. Follansbee , Robert. Evaporation from Water Surfaces, A Symposium. Am. Soc. Civil Eng. Trans. 99:671-7h7, (19310»). Hamilton, C.L., and Hans G.Jepson. Stock-mterinLDevelgpmnts: Wells, Spring; and Ponds. United States Department of Agriculture, Washington, D.C., Farmers' Bulletin 110.1859, (July, 19%). Harrold, L.L. Hydrologic Design of Farm Ponds and Rates of Runoff for De_sigp of Conservation Structures in the North Appalachian Region. United States Department of Agriculture, Soil Conservation Service, Research, Washington 25, D.C. , scs-rr 6h (December,19h7) . Harrold, L.L. written communication. 8. 10. 13. 11+. .. 39 - Horton, R.E. Evpporation Maps of the United States. American Geophysical Union, Transactions, Part II, National Research Council, Washington, D.C. (January, 19%). Kimball, H.H. Evaporation Observations in the United States. Eng. News, (April 6, 1905). Krimgold, D.B. , and H.E. Minshall. Hydrologic Desifl of Farm Ponds and Rates of Runoff for Design of Conservation Structures in the Claypan Prairies. United States Department of Agriculture, Soil Conservation Service, Research, Washington 25, D.C., SCS-TP 56, (May, 1915). Meyer, A.F. Evaporation from Lakes and Reservoirs. Minnesota Resources Commission, St. Paul, Minnesota, (June, 1912) . Minshall, R.E. written communication. Parsons, D.A. The gagglogy of a Small Area near Auburn, Alabama. United States Department of Agriculture, Soil Conservation Service, Research, Washington 25 , D.C. , SCS-TP 85, (September, l9’+9). Thornthwaite , C.W. , and B. Holmn. Measurement of Evaporation from Land and Water Surfaces . United States Department of Agriculture, Washington, D.C. Technical Bulletin Ho. 817 (19h2). - ho - 15. . We]: toward a Rational Classification of Climate. Geographical Review. 38:1 pp 55-9“ (January, 19h8). IIIII IIIII IIIIIIIIIIII 31329