COMPARISON o; CALCULATED AND MEASURED EVAPOTRANSPIRATION or POTATOES AND TOMATOES rem suoarpemoos or TIME Thais {a the. on... or M. s. Mia-new 31m umvmsn‘v Cortland Ennis Young, Jr. ' 19.60 0-169 Illllltfljfllllflfllllfllfllflll 1/ This is to certify that the thesis entitled Comparison of Calculated and Measured Evapotranspiration of Potatoes and Tomatoes for Short Periods of Time. presented by Cortland Ennis Young, Jr. has been accepted towards fulfillment of the requirements for _M.3;_ degree in Agriculmral Engine e ring 5/ . Major professor Date May 20: 1960 LIBRARY Michigan State University Am": gm. COMPARISON OF CALCULATED AND MEASURED EVAPOTRANSPIRATION' OF POTATOES AND TOMATOES FOR SHORT PERIODS OF TIME By Cortland Ennis Young, Jr.‘ A THESIS Submitted to the College of Agriculture of Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirements for the degres;of MASTER OF SCIENCE Department of Agricultural Engineering‘ 1960 \ Approved by 42§Z;// ( (3/0577 / .r‘ ACKNOWLEDGMENTS The author wishes to express his appreciation to Professor E. H. Kidder of the Agricultural Engineering Department for his helpful suggestions and guidance as major professor during the course of this study. Appreciation is also expressed to Dr. A. E. Erickson of the Soil Science Department for his interest and constructive suggestions. The author's wife, Judy, deserves much credit for her patience and encouragement through the entire study, and helpful suggestions while reviewing the manu- script. II III IV . VI VII VIII Ix TABLE OF CONTENTS INTRODUCTION. . . . . . . . . . . . . . . . . . REVIEW OF LITERATURE. . . . . . . . . . . . . . A. Methods of Estimating Evapotranspiration . B. Methods of Determining Evapotranspiration. C. Practical Uses of Evapotranspiration Data . . . . . . . . . . . . . . . . . r . PROCEDURE 0 . . . . . . . . . . o . . . . . . . A. EXperiment Procedure . . .‘. . . . . . . . E. Calculation.Procedure . . . . . . . . . . C. Comparison Procedure . . . . . . . . . . . DISCUSSION OF RESULTS . . . . . . . . . . . . . SUMMARY . . . . . . . . . . . . . . . . . . . . CONCLUSIONS . . . .I. . . . . . . . . . . . . SUGGESTIONS FOR FURTHER STUDY .‘. A . . . . . . REFERENCES. . . . . . . . . . . . . . . . . . . APPENDICES. . . . . . . . . . . . . . . . . . . Appendix I . . . . . .I. . . . . . . . . . . . Appendix II. . . . . . . . . . . . . . . . . . Appendix III . . . . . . . . . . . . . . . . . Appendix IV. . . . . . . . . . . . . . . . . . Appendix V . . . . . . . . . . . . . . . . . . Appendix VI. . . . . . . . . . . . . . . . . . Appendix v11 . .'. . . .'. . . . . . . . . . . 3 l2 12 15 15 18 22 25 57 58 59 6O 62 62 63 64 65 66 67 68 Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure 10. Figure 11. Figure 12. l. 2. 6. 7. 9. LIST OF FIGURES Page Experiment Site. . . . . . . . . . . . . . 24 Comparison of Measured Evapotranspiration and 5 Day Evaporation Pan Averages for Potatoes, 1958 (Entire Season) . . . . . . 36 Comparison of Measured Evapotranspiration and 5 Day Evaporation Pan Averages for Tomatoes, 1959 (Entire Season) . . . . . . 37 Comparison of Measured Evapotranspiration and 3 Day Evaporation Pan Averages for Potatoes, 1958 O O O O O O O O O O O O O O 38 Comparison of Measured Evapotranspiration and 5 Day Evaporation Pan Averages for Petatoes’ 1958 O O O O O 0 O O 0 O O O O O 39 Comparison of Measured Evapotranspiration and 7 Day Evaporation Pan Averages for 13013813068, 1958 o o o o o o o o o o c o o 0 40 Comparison of Measured Evapotranspiration and 3 Day Evaporation Pan Averages for Tomatoes, 1959 . . . . . . . . . . . . . . 41 Comparison of Measured Evapotranspiration and 5 Day Evaporation Pan Averages for Tomatoes, 1959 . . . . . . . . . . . . . . 42 Comparison of Measured Evapotranspiration and 7 Day Evaporation Pan Averages for Tomatoes, 1959 O O O O 0 O O O O O O 0 O O 43 Comparison of Measured Evapotranspiration and 5 Day Blaney-Criddle Averages for Potatoes, 1958. . . . . . . . . . . . . . 44 Comparison of Measured Evapotranspiration and 5 Day Blaney-Criddle Averages for Tomatoes, I959. . . . . . . . . . . . . . 45 Comparison of Measured Evapotranspiration and 5 Day Thornthwaite Averages for Potatoes, I958. . . . . . . . . . . . . . 46 ' LIST OF FIGURES (continued) Page Figure 13. Comparison of Measured Evapotranspiration and 5 Day Thornthwaite Averages for Tomatoes, 1959. o o o o o o o o o o o o o 47 Figure 14. Comparison of Measured Evapotranspiration and 5 Day van Bavel (simplified) Averages for Potatoes, 1958. O O O O O O O O O O O 48 Figure 15. Comparison of Measured Evapotranspiration. . and 5 Day van Bavel (simplified) Averages for Tomatoes, 1959. . . . . . . . . . . . 49 Figure 16. Comparison of Measured Evapotranspiration and 5 Day van Bavel (detailed) Averages for Potatoes, 1958. . . . . . . . . . . . 50 Figure 17. Comparison of Measured Evapotranspiration and 5 Day van Bavel (detailed) Averages for Tomatoes, 1959. . . . . . . . . . . . 51 Figure 18. Comparison of van Bavel (detailed) and Evaporation Pan 3 Day Averages for Potatoes, 1958. . . . . . . . . . . . . . 52 Figure 19. Comparison of van Bavel (detailed) and Evaporation Pan 3 Day Averages for Tomatoes, 1959. . . . . . . . . . . . . . 53 Figure 20. Comparison of Measured Evapotranspiration and van Bavel (detailed), Evaporation Pan, and Penman 5 Day Averages, for Tomatoes, 1959. O O 0 O O O O O I O O O O O O O O 0 54 Figure 21. Thornthwaite's Nomograph for Determining Evapotranspiration. . . . . . . . . . . . 62 Figure 22. van Bavel's Nomograph for Determining Evapotranspiration. . . . . . . . . . . . 63 LIST OF TABLES Table 1. summary Of R88u1t8 “'" 1958 o o o o o o o o o 55 Table 2. Summary of Results - 1959 . . . . . . . . . 56 INTRODUCTION Vegetables grown in Michigan will undergo periods of drought during the growing season. Farmers, realizing this fact, have purchased irrigation equipment to use during drought periods. ' For maximum crap production it is important to irrigate at the correct time and with adequate amounts of water. Proper use of evapotranspiration data can aid in determining the correct time to irrigate. Evapotranspiration for any plant-soil system is dependent on climatic conditions, soil moisture con- ditions, and type and stage of plant growth. Extensive work has been done to determine the effect each variable, and their interactions, has on evapotranspiration. As a result of these studies, several methods of estimating evapotranspiration have been developed, which are usually based on climatic variables. These estimates have been compared with measured evapotranspiration in many loca- tions, in order to determine their validity under differ- ing climatic conditions. In an eXperiment at Michigan State University, the evapotranspiration rate of two crOps, potatoes in 1958 and tomatoes in 1959, was determined by the use of Bouyoucous moisture blocks. This thesis compares this measured evapotranspiration to evapotranspiration esti- mates computed by the methods of Blaney-Criddle, Thorn- thwaite, Penman, and van Bavel. Evaporation pan data is also compared to the measured evapotranspiration. These comparisons cover periods of five days in order to determine the validity of each estimate during measured maximum and minimum evapotranspiration periods. In comparisons covering more than five days, maximum and minimum evapotranspiration periods are minimized and thus make conclusions inaccurate. Special emphasis is placed on determining the relationship between evaporation pan data and measured evapotranSpiration. REVIEW OF LITERATURE There are two ways of determining evapotranspir- ation values: (1) estimates of evapotranspiration based on climatic conditions, soil moisture conditions, and physiology of plants; (2) direct measurement of the soil moisture condition and the change in soil moisture. Ac- curate evapotranspiration values have a practical signif- icance for horticulturalists, soil scientists, agronomists, engineers, and ultimately for the farmers interested in irrigation and drainage. Methods of EstimatinggEvapotranspiration The three basic methods of estimating evapo- transpiration, either directly or indirectly, can be grouped into the following broad categories: (8) (1) Empirical methods, in which existing clima- tological data are fitted into certain arbitrary factors. (2) Theoretical methods based on the removal of vapor that must take place in the evaporation process. (3) Theoretical methods based on the balance of energies available and used in the evaporation process. In this section, each of the above methods will be discussed. Blaney-Griddlequuation Blaney and Criddle (3) developed one of the first workable empirical equations for determining evapo- transpiration. Their equation is expressed as: we: monthly crop coefficient per cent of annual daytime hours occurring that month monthly evapotranspiration in inches average monthly temperature in ° F. dc: ESN' "fl They based their equation on the fact that ‘ evapotranspiration rates change proportionately with tem- perature changes. Recognition is given to radiation in- fluence by the inclusion of day length. Seasonal crop factors have been evaluated for many crops. Monthly crap factors have been determined for a few craps. These monthly crop factors show the fluctuation of water usage during a growing season. Phelon (13) proposed that the k values vary proportionally as the temperatUre varies over a growing season. His work on alfalfa showed that the water use requirement varied considerably during the course of a season. Thornthwaite's Equation An empirical approach has been developed by Thornthwaite (20). His equation, like Blaney-Criddle's, placed maximum climatic emphasis on temperature and was expressed as: E = (c) (t)a = monthly potential evapotranspiration in em. = mean monthly temperature c&a = constants for a given year a = 0.000000675 (I)3 - 0.0000771 (1)2 + 0.01792 (I) + 0.49239 I = i 1 (monthly heat indexes) 1) 1.514 5) ( ( Al I The equation for evapotranspiration then resolves itself to: E = 1.6 (3519-5)8 The computation of evapotranspiration has been made relatively easy with a nomograph constructed from this equation. ‘Thornthwaite did not attempt to distinguish between crops. The equation was designed for use primar- ily in central and eastern United States. However, it has been tested against values from evapotranspirometers all over the world. Agreement has been good in areas of similar climate, but wide discrepancies have deve10ped in some areas (9). Dalton's Equation ‘ Dalton (8) was one of the first to work on the rate of evaporation from a water surface. His theory has been the basis for most of the later work in this field. He derived an equation, based on the effect of vapor pres- sure and wind movement, for estimating evaporation from a water surface. His equation is expressed as: E 3 (es ‘ ed) f (u) E = evaporation in a unit time e8 = vapor pressure at the evaporating surface ed = vapor pressure at some height above the water surface - f (u) = a function of the horizontal wind velo- city Rohwer (18) evaluated constants for the equa- tion to give: U2 = wind speed at two meters in miles per hour Since Rohwerls equation represents water loss over a water surface, evapotranSpiration cannot be obtained directly from the equation. Thornthwaite-Holzman Equation Thornthwaite and Holzman (l9) Pioneered work in the theory of vapor diffusion as applied to evaporation. They developed an equation of the form: E = P K2 (q1 ' Q2) (U2 "U1)~. (ln737zl) E.= Rate of evaporation specific humidity 2 mean wind velocities* - two different heights gman's constant ql&q U1 & U Z1 & Z K = a K The difficulty in applying this equation lies in the necessity of making exact measurements of specific humidity and wind speed at the two different heights. The lack of available equipment and data reduces the practical application and testing of this technique. Pasquill (11) has deve10ped a modification of the above equation: E = BU2 (el - 62) KZM (1 - U2/U1) 2 B: RT (1n ZZ/zl) M R T e & e U1 a U2 z1 a 22 1 2 This equation is limited in use to a site of a molecular weight of water gas constant absolute temperature vapor pressures wind speed heights 1 & 2 fixed roughness. B will vary as the roughness of an area changes. Other Equations Mankkunk, Turc, Haude, and Uhlig (15) have de- ve10ped equations that estimate evapotranspiration. They are theoretical and make use of climatic conditions as the variables for determining evapotranspiration. They have been tested in specific locations in Europe. Penman's Equation Penman (12) has made use of the theories of both energy balance and diffusion of vapor in arriving at his method of determining evapotranspiration. His estimate represents evaporation from a water surface and is classi- fied as potential evapotranspiration. Two basic steps are involved in solving his equation; first, the evaluation of net radiation energy and second, determining how this energy is divided in heating the air and evaporation. He assumed that: (1) the gain in heat by the soil is relatively small over a short period of time, and (2) advective heating is neg- ligible. Penman's most recent equation is expressed as: E = 45H + X Ea .A + 8’/SD E = potential evapotranspiration in mm per day A = slope of saturated vapor pressure curve 8 = 0027 . H = net radiation E8 = auxiliary quantity (measure of the diffu- sion of vapor) S = factor denoting influence of diffusion resistance D = factor denoting influence of length of day H = Ra (l-r) (.18 + .55 n/N) + (Ta4 (.56 + .092433) (.1 + .9 n/N) R8 = mean monthly extraterrestrial radiation in mm per day r = radiation reflection coefficient n/N = ratio of actual to possible hours of sun- shine g = Stefan Boltzman constant 2.01 x 10'9 mm per day Ta = absolute air temperature in deg Kelvin e = actual vapor pressure of the air in mm of mercury F1 l - 0.35 (ea-ed) (l + 0.0098U2) mm per day e = saturation vapor pressure at mean air tem- perature 2 = wind speed at two meters U S = La/ (La + 0.16) L = 0.65 (1 + 0.0098U2) effective diffusion a length of air U ll N/24 + 1/7! sin N 11724 2 ll hours from sunrise to sunset Penman suggeSts that a factor of 0.80 will modify the potential evapotranspiration to actual evapotranspira— tion for England. This factor will vary from area to area. An extensive effort has not been made to find other fac- tors. van Bavel (23) suggests a factor of 0.75 for the southeastern United States. Penman has not incorporated a crop factor into his equation. He has sdggested that a crop factor would increase the accuracy of the equation. The assumption of no advective heat losses will cause serious discrepancies in Texas and other plains areas. The estimates are made on the basis of two dimen- sional analysis and plants have a third dimension. The Penman equation has been checked and found accurate in many parts of the world. The difficulty of computation and of measuring climatic conditions has limited its use. 10 van Bavel's Method The approach by van Bavel (23) is based on the theory of energy balance as expressed by Penman's equa- tion. A nomograph,based upon the measurement of radia- tion and air temperature, has been prepared to simplify computation. van Bavel's procedure is based on the theory of a conservation process going on in nature. Evapotrans- piration intensity is bounded by limits of 0 and 0.35 inches of water per day. Evapotranspiration is principally dependent upon incident radiative energy. Radiation is determined by latitude, time of year, and cloud-cover. The first two variables are easily predicted. Cloud-cover is unpredictable but can be measured. Temperature is significant to this process since warm air can retain more water than cold air. Windspeed and relative humidity are of minor significance for evapotranspiration from large homogenous areas. Two practical methods of arriving at daily evapotranspiration estimates have been devised. A choice of which to use depends upon the precision desired. Both methods are based on the Penman equation and have given reasonable predictions for the eastern part of the United States. These methods are illustrated in Appendix IV. ll Evaporation Pan Method The question of whether evaporation from a water surface is in preportion with evapotranspiration from plants in the surrounding area, has provoked numerous experiments in the past decade. Harrold (6) compared evaporation pan values against measured evapotranspiration from monolith lysi- meters and found an encouraging relationship. Pruitt (l4), Bouwer (4), and others have also obtained promising re— sults. The theory behind the validity of evaporation pan values as estimates of measured evapotranspiration is that an evaporation pan in the same habitat as a grow- ing crap will experience the same effect from climatic conditions as the plant. (6) Any change or equalizing effects from climatic factors will affect pan evaporation and crop evapotranspiration pr0portionally, if not equally. In order to achieve a good relationship, the soil on which the crOp is produced should always be kept at a high level of available moisture. Thus lack of soil moisture will not be a limiting factor in evapotranspira- tion. Even though the soil is kept at a high moisture level, the total evapotranspiration for a set period of time will not usually be as high as the evaporation pan 12 loss. Plants cease transpiration almost completely at night, while the evaporation pan continues to lose water. It is important that the difference between the two be fairly constant. This difference can be compensated for by multiplying evaporation pan values by a factor to ob- tain values at or near the rate of measured evapotrans- piration. Methods of DeterminingEvapotranspiration Two methods for measuring the evapotranspiration rate of a plant-soil system are: (l) a direct measure of the soil moisture condition, and (2) a measure of the change in weight, due to moistUre loss, of a plant-soil system. Methods proposed to measure the soil moisture condition directly include gravemetric (2), tension (2), electrical (2), and neutron scattering measurements (17). Lysimeters have been used to determine the change in weight of a plant-soil system. Harrold (7). Thornthwaite (9), Visser (15), and Tanner (16) have de- veloped different types of lysimeters to measure evapo- transpiration. Practical Uses of Evapotranspiration Data van Bavel (24) points out that at the present time evapotranspiration data can be used in conjunction 13 with past and current weather records for practical ap- plications such as: determining drought and flood prob- abilities, and irrigation requirements of crops. The practical applications listed are based upon the soil moisture condition of an area at a given time. To know this condition, several variables must be measured or estimated. The time and amount of effective rainfall must be known as it increases soil moisture. Evapotranspiration and percolation rates must be obtain- able as they represent moisture loss from the root zone of the soil. vnn Bavel (21), Allred and Chen (1), and Mor- anic (10) have used evapotranspiration estimates in con- jUnction with past rainfall data to predict drought prob- abilities. van Bavel (24) points out that evapotranspira- tion estimates can be used in conjunction with past rain- fall data to predict flood probabilities. The foremost use of evapotranspiration data in recent years has been for determining time for irrigation. The common procedure has been to test the reliability of various methods of estimating measured evapotranspiration. The best evapotranspiration estimating method is used to determine the time for irrigation. Harrold (6) found that data obtained from a BPI - evaporation pan could be correlated to measured 14 evapotranspiration. He compared measured evapotranspira- tion with estimates computed by the methods of van Bavel, Thornthwaite, and Blaney-Griddle (5). The results were encouraging, both from the standpoint of relative accuracy of present methods, and the possibility of developing ad- justments where desirable and practical. Pruitt and Jensen (14) compared evaporation pan data, Thornthwaite values, and Blaney-Griddle values to measured evapotranspiration. During periods of good crop cover, tank-evaporation rates gave a much closer estimate of actual evapotranspiration rates than either the Blaney-Griddle or Thornthwaite procedures. van Bavel and Wilson (22) compared Thornthwaite estimates with measured evapotranspiration. The Thorn- thwaite values were accumulated to estimate time for irri- gation. They found that these accumulated estimates pre- dicted a time to irrigate that was very close to the actual time when irrigation should have occurred according to measured soil conditions. Bouwer (4) has made use of the relationship between evaporation pan data and measured evapotranspira— tion in developing an integrating rainfall-evaporation. recorder, which is a modification of a standard evapora- tion pan. This device could make possible fully automatic irrigation with solid sprinkler irrigation systems. PROCEDURE Bouyoucous moisture blocks were used to obtain soil moisture data from a controlled irrigation experiment conducted at the Michigan State University horticultural farm. This data was used to compute actual evapotrans- piration for the plant-soil system in the experiment. Measured evapotranspiration was compared to evapotrans- piration estimates computed by the methods of Blaney- Criddle, Thornthwaite, van Bavel, and Penman. The com- parisons were made of five-day averages to determine the correlation during short periods of time. The maximum and minimum water use periods and the cycling of evapo- transpiration values showed up very clearly in this length of time when plotted as comparison curves. Experiment Procedure The problem of determining the minimum avail- able moiSture level at which crops should be irrigated in order to produce the best yields was the basis for the experiment, of which this study was an outgrowth. In order to establish an experiment which would give an answer to this problem, certain variables had to be controlled. (1) A uniform soil type was selected for the plots. 15 l6 (2) Tile drains were used to keep the water table below the reach of plant roots. (3) Roofs with an automatic control system were constructed on rails so that when rain began, the roofs would come out and cover the plots. (4) Cultivation and fertilization were identical on all plots. (5) All plots were planted at the same time, with the same variety of plants, and with standard spac- ing throughout. (6) All plots had the Same number of guard rows. (7) All plots were irrigated and harvested by the same procedure. The general plan of the experiment site is shown in Figure l. The portable irrigation apparatus was oper- ated at the same application rate on all plots. Bouyoucous gypsum moisture blocks were used to measure soil moisture. Four blocks were installed at 0, 1, 3, 6, 9, 12, 15 and 18 inch depths in each plot. Daily readings were taken on all moisture blocks with a Bouyoucous moisture meter, to measure per cent of available moisture. The moisture content of the soil in each plot was expressed as the average of the available moisture readings from the blocks located in the three to eighteen inch depth in the soil. It was assumed that 17 the crop extracted its water from the surface eighteen inches. In 1958, with potatoes as the crop, plots B-2 and D-l were kept above 70 per cent of available water. Plots B-l.and C-2 were kept above 40 per cent for the first part of the season and above 70 per cent for the remaining part. Plots A-1 and D-2 were kept above 40 per cent. Plots A-2 and 0-1 were kept above 10 per cent, and check plots A-3, B-3, C-3, and D-3 received natural rainfall. The percentages represent the minimum level of available moisture in the root zone to which the plots were allowed to fall. When the soil moisture level reached this minimum value they were irrigated to bring the soil back to field capacity. In 1959, with tomatoes as the crop, plots B-1 and D-l were kept above 70 per cent of available water. Plots A-1 and D—2 were kept above 70 per cent for the first part of the season and above 40 per cent for the remaining part. Plots B-2 and C-2 were kept above 40 per cent. Plots A-2 and C-1 were kept above 20 per cent. Plots A-3 and C-3 were natural rainfall check plots. The eight tomato plants in the middle of each plot were harvested each week. The potatoes were har- vested after a blight killed the vines in late August, 1958. An analysis of variance test was conducted on the crop yields data. 18 A weather sub-station was located at the experi- ment site, where temperature, humidity, rainfall, wind, and evaporation pan data were obtained. These data were used in subsequent calculations of evapotranspiration. In order to obtain net radiation values to use in conjunction with Penman's equation, a net radiometer was operated on the plots during a period from September 1 - 28! 19590 Calculation Procedure It was necessary to average the measured evapo- tranSpiration data for five to eight day periods due to the apparent non-uniformity in moisture block readings after an irrigation. It took from three to four days before the soil moisture readings in a plot would begin dropping after an irrigation. A five to eight day aver- age was needed to give a realistic picture. In order to obtain consistent results, moisture data taken between irrigations was averaged for periods of five to eight days. The magnitude of evapotranspiration losses be- tween plots, for the same period of time, was not always the same. At times values for certain plots seemed un- reasonable. Instead of disregarding periods of question- able data, a daily average was made of the six plots kept above the 40 per cent available moisture level. This 19 method of averaging was equivalent to averaging all the blocks for a five to eight day period. Another reason for-a11 plots being averaged together was that, due to the staggered irrigations on the plots, values for the six plots represented rates nearer potential evapotranspiration. The average avail- able moisture level was well above fifty per cent during the whole growing season. An example of the procedure used in obtaining the measured evapotranspiration rates is illustrated in Appendix V. Blaney-Criddle's Method The Blaney-Griddle equation was designed to estimate evapotranspiration on a monthly or seasonal basis. Therefore, only seasonal crop factors (K) and a few monthly crop factors (k) have been deve10ped for various crops. With no available way of determining k factors for short periods of time, the seasonal factor of 0.65 for potatoes and 0.70 for tomatoes was used. Each factor in the equation was'converted to a daily basis, as is illustrated in a sample calculation, Appendix IV. Thornthggite's Method Potential evapotranspiration as computed by Thornthwaite's method uses the following equation. e = 1.6 (IOt/I)a 20 In order to simplify working with this equation, Thornthwaite developed tables and a nomograph. The com- putation was reasonably easy with mean air temperatures and the latitude of the location known. The equation gave unadjusted rates of potential evapotranspiration. It became necessary to reduce or increase the unadjusted rates by a factor that varied with the day and with lati— tude. The first step in calculating evapotranspiration was to obtain the heat index, I. Thornthwaite's table 1 (9) provided monthly values of 1 corresponding to monthly mean temperatures. Summation of the twelve monthly values gave the index I. The value of I was used to construct the nomo- graph, Figure 21. Since there is a linear relationship between the logarithm of temperature and the logarithm of unadjusted potential evapotranSpiration, straight lines on the nomograph define the relationship. All lines pass through the point of convergence at t = 26.5° C and po- tential evapotranspiration = 13.5 cm. The slope of the line is determined by the heat index (I). Values of evapo- transpiration were then found for any temperature. These values of potential evapotranSpiration were adjusted for latitude and day length. Tables (9) provided correction factors by which the unadjusted 21 potential evapotranspiration of each day was multiplied. A sample calculation is found in Appendix IV. zgn Bavel's (simplified) Method This method of estimating evapotranspiration was very simple. All the information that was needed to make an estimate was van Bavel's table, Appendix III, and the per cent of possible sunshine that reached the earth each day. The United States Weather Bureau at Capitol City Airport, Lansing, Michigan, makes a daily reading of the per cent of possible sunshine that reaches the earth. These values were used in conjunction with the table to arrive at daily estimates. This method is illus— trated in Appendix IV. van Bavel's (detgiledlMethod The detailed van Bavel method yielded evapotrans- piration values readily with the aid of a nomograph, Fig- ure 22. It was necessary to know the per cent of possible sunshine for a given day and the mean daily air tempera— ture. Also, the amount of radiation intensity that will reach our atmosphere at a given latitude on a certain day was needed, Table 2 (23). This method is illustrated in Appendix IV. 22 Penman's Method Calculation by Penman's equation was made much easier by using the net radiation readings taken on the plots of growing plants. It was necessary to have mean daily air temperatures, mean daily dew point temperatures (obtained from the United States Weather Bureau), wind speed (measured at the experiment site), and number of hours from sunrise to sunset (obtained from the United States Weather Bureau). A sample calculation appears in Appendix VI. Eggporation Pan Evaporation pan readings were taken daily at 5:00 p.m. from a United States Weather Bureau Class A, Pan. A standard weather bureau eight inch rain gauge was located near the pan to provide data for days when rainfall made it necessary to adjust evaporation pan values. Comparison Procedure The measured evapotranspiration data was accurate for average periods of from five to eight days. A standard period of comparison of five days was chosen. Comparisons for 1958 and 1959, for the average periods of time indicated, consisted of measured evapo- tranSpiration compared to evaporation pan three, five, and seven day averages, van Bavel five day averages, both 23 simplified and detailed, Blaney-Griddle five day averages, and Thornthwaite five day averages. Comparison was also made between van Bavel (detailed) three day averages and evaporation pan three day averages. For a period during 1959 comparisons were made between measured, evaporation pan, van Bavel (detailed), and Penman evapotranSpiration values. ‘ These comparisons appear as curves in Figures 2-20. Plotted points represent an average of the daily values within the stated period of comparison. Exact values of maximum differences between curves has been tabulated in Tables 1 and 2. 7a. 'I‘II'III'I'I 0-2 C-I 8-2 A-l IRRIGATION. +— 20'-—° APPARATUS D-3 A-3 TEMPERATURE AND HUMIDITY FIGURE 0 o RAIN GAUGES O SITE EXPERIMENT EVAPORATION PAN DISCUSSION OF RESULTS Comparisons were made between measured and com- puted evapotranspiration in order to determine which com- putation method would give the best estimate of evapotrans— piration. These comparisons covered short periods of time during the growing season so that maximum and mini- mum water use periods would be apparent. A summary table for each year, Tables 1 and 2, shows the results taken from the comparison curves of computed and measured,evapotranspiration. Data of par- ticular interest are measured and computed seasonal totals, and largest total over and under estimations. Evaporation Pan Values Compared to Measured Evapotranspira- tion As shown in Figures 2 and 3, in periods leading up to July 8, 1958, and July 1, 1959, the evapotranspira- tion was well below evaporation pan values. The reason for this great difference was that the plants in an early stage of growth were transpiring at a rate less than that of a mature plant. The other methods of estimating evapo- transpiration were not compared to measured evapotranspira- tion during this period because their estimates are based on mature transpiring plants. ~25 26 Around August 18, 1958, a blight killed the potato vines in the plot area. The blight resulted from extremely moist atmospheric conditions existing while the crops were covered. Figure 2 illustrates the immedi- ate reduction in crop transpiration and the resulting differences in the two curves. A 1:1 ratio existed between the curves from September 9 - 16, 1959. This could have been an accident or an indication of the actual relationship during the latter part of the growing season for tomatoes. A frost occurring on September 16, 1959, killed 50 per cent of the plants. The tranSpiration rate was greatly reduced and the evapotranspiration curve remained low when the evaporation pan rate increased, Figure 3. To be of value in comparison with measured evapotranspiration, the evaporation pan curves, Figures 4-9, for each year had to be adjusted by a factor. A possible reason for the difference between measured evapo- transpiration and evaporation pan data is that plants almost completely cease transpiration at night while the evaporation pan can lose water at any time that there is a moisture deficit in the air. The evaporation pan correction factors were obtained by dividing the total evapotranspiration for a certain period of time by the evaporation pan values for 27_ the same period. Factors were determined for periods of approximately a month. Only one factor was determined for 1958 because there was relatively the same magnitude of difference between the two curves through the entire period, Figure 5. During 1959, the relative difference between the two curves changed around August 5. The fac- tor for 1958 was 0.707. The factors were 0.72 and 0.80 respectively for the first and last half of 1959. Adjusted evaporation pan values gave very good estimations of measured evapotranSpiration losses for both years, Figures 4-9. This was true regardless of whether the pan average that was chosen for comparison was three, five, or seven days. The largest single over- estimation was 0.28 inches during a six day period in 1958. The largest single underestimation was 0.24 inches covering a nine day period in 1959. These resulted from a comparison of three day pan averages to measured evapo— transpiration. The close agreement can be explained by the hypothesis that evaporation from water and evapotranspira- tion from a plant-soil system should react similarly to various climatic changes. This similar reaction should result in approximately the same cycling for each set of data. The same cycling can best be observed from the 28 comparisons of five and seven day evaporation pan data and measured evapotranspiration, Figures 5, 6, 8, 9. The three day evaporation pan averages curve fluctuated quite irregularly back and forth across the measured evapo- transpiration curve, but this was because the measured values were computed in five to eight day increments, Figures 4 and 7. The two curves were often out of phase. As has been pointed out, even though this did happen, the periods of over or under estimation were not significant. The plant-soil system probably reacted more slowly to climatic change than did evaporation from a water surface. The moisture blocks would have been the last part of the whole system to feel the effect of any changes. Care must be taken in using evaporation pan data as a source of estimating actual evapotranspiration loss from a plant-soil system. The accuracy is greatly dependent on using the correct factor to adjust the evapo- ration pan values. Within two year periods the factors varied depending on (1) the type of crOp, (2) the stage of develOpment of crop, and (3) the portion of the growing season concerned. Blaney-Griddle Estimates Compared to Measured Evapotrans— piration Blaney-Griddle total seasonal estimates ran 29 somewhat below total measured evapotranspiration for both 1958 (0.50 inches), and 1959 (0.75 inches), Figures 10' and 11. For 1958 the underestimation was fairly constant and consistent for the whole season. For 1959, the major- ity, (0.55 inches), of the underestimation occurred in the period of maximum water usage, July 1 - July 17. For the remainder of the season the agreement was good. The accuracy of computed values obtained by the Blaney-Griddle equation in 1958 was dependent on the crOp factor used. An increase of the cr0p factor from 0.65 to 0.72 would have brought the two curves into their best agreement. Since temperature is the climatic variable that determines evapotranspiration in the Blaney-Griddle equation, the good agreement during 1958 seems to give support to their hypothesis that evapotranspiration is directly proportional to temperature. In 1959, an adjustment in the crop factor was not needed over the whole season, but just during periods of maximum water usage. Since high water use periods cannot be predicted, an additional crop adjustment factor cannot be used. A combination of temperature and other climatic conditions caused the period of high water usage from July 1 - July 17, 1959. Therefore, Blaney-Criddle's equation, based only on temperature, is unreliable during these periods. 3O Thornthwaite's Estimates Compared to Measured Evapotrans- piration For both years, the total of the values found by using Thornthwaite's method of estimating evapotrans— piration was within 0.10 inches of measured totals, Fig- ures 12 and 15. For 1958, due to the good agreement, the maximum over or under estimation did not exceed 0.15 inches for any period. In 1959, an extreme period of underestima- tion (0.80 inches) occurred from July 1 - July 15, and an extreme period of overestimation (0.67 inches) occurred' from August 5 - September 1., I In spite of the good agreement during 1958, there was a definite phase shift between the two curves. According to Thornthwaite's hypothesis, potential evapo- transpiration is dependent on temperature at a given time. Therefore, the values computed by his method showed an immediate response to temperature changes. The moisture conditions measured by the blocks reflected a slower re- sponse to temperature changes, due to the buffering effect of the soil and the physiological adaptations of the plants to meet climatic changes. ‘ In 1959 this method proved unreliable for esti- mating evapotranspiration during a period of high water use, and also during a different period of extremely high temperatures. The underestimation during July 1 - July 17, was probably due to the failure of the equation, based 31 primarily on temperature, to estimate properly extended periods'of high evapotranspiration rates resulting from all climatic variables. The period of overestimation, August 5 - September 1, was probably due to Thornthwaite's hypothesis that the logarithm of evapotranspiration in- creases with the logarithm of temperature. Because of this, evapotranspiration estimates are too high during extended periods of high temperature; August, 1959, was the fourth warmest August on record. van Bavel (detailed) Estimates Compared to Measured Evapo- transpiration van Bavel's (detailed) method provided values too low for both 1958 and 1959, Figures 16 and 17. The magnitude of difference between the two curves was fairly consistent for each year. A factor times the van Bavel values would produce a reasonable estimate. A factor of 1.52 for 1959 and 1.19 for 1958 would produce the best results for estimation purposes. I A The large emphasis van Bavel places on the ef- fect of daily amounts of radiation was the probable reason for the consistently low values. He pr0poses that on days with low effective radiation due to cloud cover, the evapo- tranSpiration values will also bellow -- regardless of the temperature. During the summer, there are enough days with reduced radiation in this area to make computed 32 evapotranspiration estimates consistently lower than meas— ured evapotranspiration. It can be seen from the compari- son curves that values computed using van Bavel's equation will always be low for this area. Adjusting the nomograph would better relate the effects of radiation. van Bavel (simplified) Estimates Compared to Measured Evapotranspiration The van Bavel (simplified) method of estimating evapotranspiration provided high values for both 1958 and 1959, Figures 14 and 15. In July, 1958, the estimates were considerably above the measured evapotranspiration for that period and made up the majority of the 1.09 inches of overestimation for the season. In August, 1958, the two curves were in close agreement. In 1959 the method provided good results as an estimate of measured evapo- transpiration even though there were some periods of over- estimation, which reached a maximum of 0.43 inches for one extended period. It is difficult to eXplain why the van Bavel simplified estimated values approximated measured evapo- transpiration so well for July, 1959, and so poorly for July, 1958. The estimated values are based on the effect of temperature and radiation, but during July of both years these values were approximately the same. July, 1959, had higher wind movement than July, 1958; and this 33 could have accounted for the difference by producing higher evapotranspiration rates in 1959. Ian Bavel (detailed),Estimates Compared to Evaporation Pan Values , A comparison of evaporation pan values and van Bavel (detailed) values is shown in Figures 18 and 19. van Bavel places maximum emphasis on radiation and tem- perature changes but practically no emphasis on wind speed and vapor pressure. If the losses from an evaporation pan fluctuate due to the same variables and in the same proportion as the van Bavel estimates, the plotted data should show a relationship. A three day average was chosen so that any maximum and minimum periods would be readily apparent. I The two curves fluctuated together with the same cycling frequency and with no phase shift for 1958. The agreement was not as good in 1959. The two curves still generally followed each other. For both years the magnitude between peaks and valleys was greater for the evaporation pan than for the van Bavel curve. The results are encouraging for assuming that radiation and temperature changes provide the greatest influence for fluctuations in evaporation from a pan. Other climatic factors may change the magnitude, but do not appreciably alter the established trend set by 34 radiation and temperature. Both evaporation pan data and van Bavel (detailed) values, when adjusted, have proven to be goéd estimates of measured evapotranspira- tion. ggmpgrison of Penman; van Bavel (detailed)L Evaporation PanL and Measured Evapotranspiration Values These four different values were compared be- cause all but the van Bavel values were affected by tem- perature, radiation, wind speed, and vapor pressure. The van Bavel values were affected only by temperature and radiation. All values would be expected to react similarly to the same climatic changes. The Penman and measured curves ran extremely close together until the frost killed the plants on Sep- tember 16, 1959, Figure 20. Evaporation pan and Penman curves were together between September 8 and 25. The van Bavel (detailed) curve ran proportionally below the other curves for the entire period. The results of comparing Penman values to meas- ured evapotranspiration indicated that actual evapotrans- piration was affected by a combination of temperature, radiation, wind speed, and vapor pressure. The Penman values provided a better two week estimate of measured evapotranspiration than the other methods used for both seasons. After the frost hit the tomatoes, the low 35 values for the measured evapotranspiration curve, as com- pared to the Penman curve, indicated the reduced trans-l piration by the crop. A correction factor was not needed for the Penman values computed during September, 1959. This was probably because measured net radiation readings were used to com- pute Penman values; also the plant-soil system was near to its potential evapotranspiration and equal to pan evaporation during this period. Net radiation readings were taken for only one month, thereby limiting the amount of computed Penman values for comparison with measured evapotranspiration. It would be impractical to predict the Penman -- measured evapotranspiration relationship for an entire season. PER DAY) (INCHES EVAPOTRANSPIRATION 36 N r m. —MEASURED RATE (5-8 DAY AVE.) ’ “”‘EVAPORATION PAN RATE (5 DAY AVE.) ¢ —_ ( /C7\ “9 ./O/ \ /)—— ‘-—O/ ‘- o o (O _.. o o P O o ,, . . m g 0. 0* o 1 A 20 28 6 l4 . 22. 30 7 I5 23 3| JUNE JULY AUG. FIGURE 2 COMPARISON OF MEASURED EVAPOTRANSPIRATION AND EVAPORATION PAN DATA FOR POTATOES 37 N - n ///0\\\I - , ' -— MEASURED RATE (5—8 DAY AVE.) A , -—-—EVAPDRATIDN PAN RATE (5 DAY AVE.) >- . <1 . /o\ /O\ O , . N . . . . l m \\\0/// ///y—~\. CL m .»-<\ . . g . . S .\\\y/// 0 . Z c 3 0 ° / . 2 ~ 4"“ . \O/. 2 . O o ,/O‘- C ' . 0 X \0 <1 0 (I E e U) e Z < E A - —————~ 0 ‘3 o- 0 <1 > LIJ 23 , 9 T :7 25 2 '10 IS 26 3 . ll l9 ‘ 27 AUG. SEPT. JUNE JULY . ‘FIGURE 3 COMPARISON OF MEASURED EVAPOTRANSPIRATION AND EVAPORATION PAN DATA FOR TOMATOES I959 38 mom. mwOHdfibn. mom 40.40 24d ZO_._.w 024 ZO_._.m 0mmDm42. 0. __ ¢ _ 0-4 3 ._I “III P. :(MK. A0wkm330m..n A.w>< >40 my whdm 24¢ zo_._.<¢oa<>w!- ..w>< >40 mlmv wkdm outamdwc‘l 02' 99' NOIiVUIdSNVHlOdVAB (AVG 83d SBHONII I) (3 mm mmm. mmo._.4._.on_ mom 40.40 240 20_._.4moa4>m 024 20_b4m_n_mz4m._.oa4>w 0m03m4w2 no 209040200 0 mxami .034 >43... 0. __ 4 mm _N 4. L iris; . KR. A0whm0504v w._.4m Z40 20_h4m004>m.u| A.m>4 >40 3 9.41 240 zo_._.4moa4>w!u A.m>4 >40 mlmv mk4m 0wmnm4w2I SI' or 90' , j NouvsIdSNvaiodVAa 03' 93' 92' (M10 83:! SBHONII AD 000. mw00.40.0n. 00.0 40.40 240 20.0.4m0n04>w 0Z4 zo_h4m_amz4m0.oa4>m 0mm3m4ms. .00 200340200 0 mm30_u_ .034 >43... mm m. __ ¢ 0N _N v. N w i i S i _ w _ m , 90° _ 83833 wt; 25. zoCemoengL 3:2 :3 t NEE zen. 29:335-.-. 0 A.m>4 >40 mlmv 00.40 0mm3m4m2lrll. Ql' NOIiVUldSNVUlOdVAB 09' Ol' 03' 93' 92' (AVG 83d SBHONII mom. mw00.45_00. «.00 40.40 Z40 20.0.4moa4>w 0Z4 20.0.4m_¢mz4m0.oa4>m 0wm3m4w2 00 200—040—200 h mm30_0 mem .034 >43... 41 m N mm m_ N. 0 mm mm m. m A0wk03504. wk4m 24¢ zo_._.4coa4>wII éu>< >3 2 we: 24.. 29:33:-.- wm>< >m 024 20.0.4m_amz4m0.004>m 0mm304w—2 .00 200.140200 0 w130_u_ .Hmmw .034 >43... la .0. mm m. N. m mm mm m. m _ T i - i 0 0 II . II b ISII IL. I 1 4| I II I _ i i -I \-.o..- I: .......... II C C I I I O I IO_O \ O O\h\.\IMu/.IdI O \InVI l \ o I O O! C _\A/ HI 0 s I’ s. O O 6“ .sOIII r. :. .-,-----..-. I\O\AI; I 0 I IQ. _ -I IOIAI I OK _ . _ \n/ j 11 o s0 . 0 0 . . . _ o o0.\ “owkm3304v whmll__ / . .A -II . I»; ~.w>4 >40 my mkdm 24a ZO_kmIiw _/O.. . ~.m>4 >40 mlmv mk4m 0mm3m4w2l~ 90' 03 SI' Ol' 92' 02' 92' (AVG 83d SBHDNII NOIiVUIdSNVHiOdVAB mom. mmohqzo» mo“. .455 24s . zofiemosgm ozq 20.0..4¢_n.mz4m.r0n_4>w 0wm3m4w—z .00 209140200 0 wm30_0 0.000 .034 >43.J m m mm m. m. m mm S m. m a A .I M _ _ , a a .., ; -I m _ 0 0 - I I II. II n a m _ I... _ i _I _ (“.0 0 IO.\ 0 o . ozzlfil I_ k 0 . OI \.\ o _.\O\.. “ll/0 I 0‘0“. an. ell . o T. I+l.\°'olx T10-..“ c I III \Nw: Ol\_ 0 83333 mesa 24.. 225523.. ..m>< >3 t 2.4: 24.. zofiemosgms- _ cm: :3 one 8.2. omcnmamsl OI' v 02' 9‘? 90' 9|° 02' 92° NOIiVHIdSNVHLOdVAB (AVG 83d SBHONII 000. mwo0.40.0n_ mo“. mw344> M40900 I>mz44m 0Z4 20.0.4m_amz4m0.oa4>w 0mm3m4w2 00 200.04.05.00 0. wm30_0 .034 >43... mm m. 2 4 0N _N 4. '0 44 90' OI' 9I' II 2. (AVG 83d saHoNI) ‘NOIivsIdSstiodVAz-I 9 2' A.w>4 >40 0. m0.4m 0wh30200II — A.w>4 >40 mlmv N040 0wm3m4m2l 09' 99' mmm. mwo0.4_200. mon— mm344> w400.m0l>m244m 0Z4 20.0.4m_n_mz4m0.0n_4>w 0mm3m4ws. 00 2003040200 __ wm30_0 Famm .034 m N 0N m. N_ m 0N NN >43... 0. m 45 A.m>4 >40 my A.m>4 >40 mlmv N0.4m 09030200 .... MP4”. 0mm3m4m2I 92' 03' 9|' Ol' 90' 09' 99' NOIiVUIdSNVULOdVAB (AVG 83d SSHONII 000. mm0h40.00 ”00....— mw344> N004; IIPZIOIH 024 20.0.4m_n_mz4m0.0n_4>m 0wm3m4w2 .00 200540.200 N. wm30_0 .034 >431 mN m. 2 4 0N _N 4. db 3:2 :3 3 up: 3398.. . ..m>< > m....4>>I...ZmOI0. 024 20.0.4m.n.mz4m0.0n.4>w 0wm3m4ws. ....0 209040200 0. wm30..... .kawm .034 >433 m N 0N m. N. m 0N NN m. m _ . 4 i I I _ 4 _ ..o. . _ IIIT. . . TQ‘ I O b ILI O Irll‘bl o :9 o _ \\\ Joell o \ II \\..0.IIII c To 1110. In.vll\04 o c all.- _ I i 4 o . I A.w>4 >40 00 wk4m 0N0.3n_200I.. A.w>4 >40 mlmv N041 0wm3m4w2I 03' 9|' 0|' 90' 92' 09' 99' (AVG 83d SBHONII NOILVUIdSNVHLOdVAB 0mm. mmo...40.0n. ~00... mw344> ”090302.00 4m>4m 24> 024 20.0.4m.n.m24m0.0n.4>w 0wm3m4ws. 00 200540.200 4. wm30.0 .034 >437 mN m. ._ 4 0N _N 4. 9I' NOILVUIdSNVHiOdVAB 92' 0w>4 >40 my N0.4m 0w...3n:200..n éw>4 >40 mlmv mh4m 0m¢3m4m2l 90' Ol' 03 09' 99' (AVG 83d SBHDNII 49 .000. 0N00.4s.0.—. 00.0 0N344> 5N.0.4n.$..0. 4N>4m 24> 024 20.0.40.0mz40._.0n.4>N 0N0304N2 00 200-040—200 N030..0 .Pmmm .034 >433 m N 0N m. N. m 0N NN m. 0 a . . .o: -10 O 0 II .IIQI II I 0‘ ILI d’ O O I x I AV! b\ o I 0'2' ..N>4 >40 0. ..N>4 >40 min. N040 0N0.30200 II N040 0N0304N2 I 93' 9|' Ol' 90' 09' 99' NOILVHIdSNVULOdVAB (AVG 83d SBHONI) mmm. 0N00.40.00 000 mN344> .0N4.40.N0. 4N>4m 24> 024 20..r40.002400.004>N 0N03m4Ns. 00 209040.200 0. N030.0 .034 >43. mm m. .. 4 mm .m 4. 4. . .o .1 9 h . . 0 . 0 I I. . . O \¢-l II In. H o .I ._ Z . . O we 9 ..m>< :3 3 0.21 35.123... W ..N>< >40 mlm. N040 00039402! .2 c. NOIlVHIdSNVUiOdVA 3 (AVG 83d SBHONI) ZO.P<0_0mZw DMIDmdwi 00 200—040—200 w030.0 .Pawm .034 >43... 0 N mm m. N. m mm mm m. m _ \\\diuod _ .0 II x .01: ss\ 0 III \\\ AV! Ipilo .0 '6'- O 0 II Ogilb\ O bl. s . _ IQ. . A.N>4 >40 0. N0.40 09.30200-.. ..N>4 >40 min. N0.40 0N0304N2I mmm. 0N0...4s.00. 000 0N344> .0N4_40.N0. 4N>4m 24> 024 03' 9I' Ol' 90' 93' 09' 99' NOIlVUIdSNVHLOdVAB (AVG 83d SBHONII wmm. mN00.40.00 000 0N344> .0N4.40.N0. 4N>4m 24> 024 40.40 240 20.0.40004>N 00 200.040.200 0. N030..0 .034 >433 DN 0. ._ 4 0N .N 4. h (. .. . . 4 . . . s 0.4 4 ‘0" s. . e I; ~. .0. \tp xs .0 . . . .0 I \‘Q (Q _ ...EI:.- .......II. -I q s . 3 s s OK ..N>4 >40 0. N040 0N._.30200I ..m>< :3 2 PE”. 24s 29.23.25: Ol' 90' NOIlVUIdSNVHIOdVA 3 9I' 93' 03' 09' 99' (AVG 83d SBHONII I 000. 0N0...45.00. 000 0N344> .0N4.40.N0. 4N>4m 24> 024 40.40 240 20.0.40004>N 00 209040.200 m. N030_0 .0.0Nm .034 >433 m N 0N m. N. m 0N NN m. m 90' F ___I_ o :- ‘L ‘p I "I ,6 d‘ ‘9 A... q. 1,) 25 l Pd ..N>4 >40 mv N0.40 0N0.30200I 9I' NOILVHIdSNVUlOdVAB S 2' -—I—,o . Is <1: 4' 4..m>< :5 m. 0.2;. 24.. 29.23.25..- 4 Ol' 03' 09' 99' (AVG 83d SBHONII ‘34 3...... 302200 mo“. $3.3 3.3.458 4m> 024. .240 zolqmoa<>m .252sz .20; I<0.0m2400.004>N 0N03m4N2 0o 209040.200 0N N030.0 0.0mm mm ON 0. o. m .. 3 x O HAv Of. 4 * 9 d d: O ’0’ II- . 8 V . N m m H V o. 10/ 0 "¢ . N m ..N>4 >40 0 444. N040 .0N4.40.N0. 4N>4m 24>.Ii mt; z<2zms--..l/ wt; 24.. zo_._.mIi /. _. PE... owmamqmzl b 03' 93' (AVG 8 3d SBHONII 55 .nmpmz 00 monocfl A.mfissa CH .00 .Hm>mm cm> w .m.> .mpflmsnpnnoca u .49 .mavofiponhmcwam u o.m .20flumA00m>m u .m .QOHumgfiqmsmApoam>m H .mm "hmx . .. m.®>w N06 I mfi ass< when He eo.H II I- I- me.m ee.s a. esflflsp OH H33 Isv .m.> I.D>m has .0: . . I- I- II -mA A e mass.me mo.A mm.e 04.3 n.esA0Aas m H55 Iandm .fl.> mm %H30 0%0 . 4H .%54 2% . 3. . \.U I mA Afise e e ms 0 I m .mse I es 0A IA 0 A. e es s . .MW mm” .0: I mH a I mass mm om.o II I- II mm.m 0e.e ..e>s ass I.D m haw Hm afise mm aAse , s mhm 9 EJ o c o o J Q. I AA aflse e NA A. o I mm sass m we m mo 0 0e e ee 0 .m..wme . .m>s awn e .ss< As . on sass ma . . . . 4 s I am mass m e m 0H 0 I em aflsw as 0 ma 0 mm 9 0e 0 .m..mm< V shade 0N .AHSHJ ARMMN WNW mam . as . .. . m cm. I on Ease s s Am 0 I em aAse s e s mm 0 mm s Ne w .u..ne< .em .as .pmmIAmocsjfithmcs: .pmm .pmmIAm>o .pmoupm>o cmuSQBOD ashammms .Bm msH pmmthH pmmwpma Ihmcc: pmmmpma pmmmpma .pmolpm>o.wo HNpOp mo Hm000.¢30800 mo mo coasom.ficcuwcmq pmmmpmq mo oofihmm mo numcmq pmmmhmq sommmm commom muonpmfl mmmfl I .00 umpmasoflwo Op nmASmNNE mo momflamqsoo mo mpazmmm Apmpwa mo mmnocfiv ca .em .mHIH .pqmmI .Hm>mm cm> u .m.> .mpfimsnpspona I .ma .maoofihOImmcme u .o.m .QOflpwHomm>m u .m .GOfipwHHmmcmhpomm>m u .Bm "mmx _ n.m>m II II II II II II rcm.m *mm.m awn mv qwafimm NI .anmm . n10>d >IMU I H zfism mama mo Hm m II II II mm.HH vo.m ovacmafiwp Imcv .m.> NH .ms< a . m .wzq (aaImnwni I m .ms< m mu oH mo I ¢H mama mama om m¢. mm.HH om.mH Acmflmflflm Iafimv.m.> % s . mm va mwsw mama VH pm. IH¢ mmsm mzwc mm om. mm.HH mm.HH A.m>a mac 4 3 .3 a s .m ”w INM awsw ammo pa mp. Iuw .mwfi mama OH mm. mm.HH NH.HH A.m>m mac n~.o.m mm .ms< . v .ms¢ . . . m.m>m mwu I «H .m:< mhwo ma ma I mm mash mama m om mm Ha mm Ha .WV.WMM mm .ms< » . m .msg . . . A.m>m may I m .ms< m mg fim mm I mm has“ mama 0 mm mm HH m» HH .mv.wmm mm .ms¢ . .m: .m>a mac I pH .ws< mama m «m I mw mas“ mamn m ma. mm.HH m~.HH my can .m .flo4 . «I ... Gama: .am .pmthmccs.pmthmccs Acmm .cfiv.pmthm>o .pmqum>o Aomm .CHV wwummwwafcmpsmeoo .em mnfl pmmMpmH pmmmhma.pmmInmccg pmmmhwa ummmnma .pmmlum>o mo Hmuow mo HmpOp.é:ano we we coahmm mo numcmq pmmmpm m6 ooapmm mo npmcmq pmmwpmq somMmm, somwmm moonpwz mmmfl I .em ompmHSOHmo ow consmwms mo nomHanSOO mo mpadmmm m mapwe SUMMARY In a controlled irrigation experiment at Michi- gan State University, measured evapotranspiration was compared to six different methods of estimating evapo- tranSpiration in an effort to determine the accuracy and reliability of each. The comparisons have been made for short periods of time during two growing seasons to deter- mine the behavior of each estimate during maximum and minimum water use periods. Curves drawn from the computed values were com— pared to the curves of measured evapotranspiration, and periods of over and under estimation were determined for each method. Agreement or lack of agreement between the curves was explained. In cases of poor agreement adjust- ments to produce improved estimates were suggested. I 57 CONCLUSIONS (1) Evaporation pan values, when used with an adjustment factor, provide an accurate estimate of actual evapotrans- piration. (2) Factors that will correctly modify evaporation pan values will vary due to difference in crops, stage of crop growth, and the portion of the season under consideration. (3) Blaney-Griddle and Thornthwaite estimates can be ex— pected to provide a fair estimate of measured evapotrans- piration during average water use periods. (4) Blaney-Griddle and Thornthwaite estimates will run considerably lower than actual evapotranspiration during periods of extended high water usage by the plant-soil system. (5) Thornthwaite estimates will overestimate actual evapo- transpiration during extended periods of high temperatures. (6) Actual evapotranspiration will experience a phase lag behind adjusted evapOration pan and Thornthwaite values. (7) van Bavel (simplified) values should be tested further before use. (8) van Bavel (detailed) values times a factor can be satisfactorily used to determine actual evapotranspiration. (9) Penman values, when computed from measured net radia- tion readings, give indication of providing an excellent means of estimating actual evapotranspiration. 58 SUGGESTIONS FOR FURTHER STUDY (1) More basic data and information is needed of the evapo- transpiration rates and habits of different crops. A measuring device such as.a lysimeter should be used as a check of the data now being obtained by moisture blocks. (2) A field experiment making use of various methods of estimating evapotranspiration to determine time for irri- gation, should be made. The test of each method's effect- iveness would be the final yield of the crops. (3) Additional basic comparisons between measured evapo- transpiration and evaporation pan data should be made in order to determine more precise adjustment factors. (4) Further investigation should be made of the validity of various methods of estimating actual evapotranspiration rates for this area. (5) Net radiation readings should be taken for a crop during a complete growing season. Penman values computed from net radiation data would check the validity of his equation during longer periods of time. 59 10. ll. 12. REFERENCES Allred, E. R., and Robert Chen (1953). Evaluating irrigation needs in humid areas. Agr. Engr. 34: 611-615, 619. Baver, L. D. (1956). Soil Physics. 3rd ed. John Wiley & Sons, Inc., New York, 489 PP. Blaney, Harry F., and Wayne D. Criddle (1950). Deter- mining water requirements in irrigated areas from climatological and irrigation data. USDA Soil Con- servation Ser., SCS-TP-96. Bouwer, Herman (1959). Integrating rainfall-evapora- tion recorder. Agr. Engr. 40:278-279. Harrold, L. L. (1958). Lysimeter checks on empirical evapotranspiration values. Agr. Engr. 39:94-97. 'Harrold, L. L. (1955). Evapotranspiration rates for various crops. Agr. Engr. 36:669-672. Harrold, L. L., and F. R. Dreibelbis (1951). Agri- cultural hydrology as evaluated by monolith lysimeters. USDA Tech. Bul. 1050. Levine, G. (1959). Method of estimating evapotrans- piration. Transactions of the ASAE. 2:32-34, 40. Mather, John R. (1954). The measurement of potential evapotranspiration. Publications in Climatology. Vol. VII, No. 1, 225 pp. Morranic, R. J. (1959). Frequency of moisture de- ficiencies in the lower peninsula of Michigan. Thesis for degree of M.S., Michigan State Univ., East Lan- sing (Unpublished). ‘ Pasquill, F. (1949). Some estimates of the amount of diurnal variation of evaporation from clay and gastur; in fair spring weather. Royal Meteorological oc., 5. Penman, H. L. (1948). Natural evaporation from open water, bare soil, and grass. Proceedings of the Royal Soc., Series A., 193. 60 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 61 REFERENCES (Continued) 'Phelon, John T. (1954). Improving estimates of irri- gation water requirements for short periods of time. Presented to the ASAE Summer Meeting, Minneapolis, Minnesota, June 22. Pruitt, W. 0., and M. C. Jensen (1955). Determining when to irrigate. Agr. Engr. 36:389-393. Rijtema, P. E. (1959). Calculation methods of poten- tial evapotranspiration. Institute for Land and Water Management Research, Tech. Bul. 7. Tanner, C. B., K. M. King, and V. E. Suomi (1956). A floating lysimeter and its evaporation recorder. Trans. Am. Geophys. Union. 37:738-742. Taylor, Sterling A. (1955). Field determinations of soil moisture. Agr. Engr. 36:654-659. Rohwer, C. (1931). Evaporation from free water sur- faces. USDA Tech. Bul. 271. Thornthwaite, C. W., and B. Holzman (1939). The de- termination of evaporation from land and water sur- faces. Monthly Weather Rev. 67. Thornthwaite, C. W. (1948). An approach toward a rational classification of climate. Geogr. Rev. 38 : 55-940 van Bavel, C. H. M. (1953). A drought criterion and its application in evaluating drought incidence and phazard. Agronomy Jour. 45:167-172. van Bavel, C. H. M., and T. V. Wilson (1952). Evapo- transpiration estimates as criteria for determining time of irrigation. Agr. Engr. 33:417-418. 420. van Bavel, C. H. M. (1956). Estimating soil moisture conditions and time for irrigation with the ET. method. USDA ARS 41-11. 16 pp. van Bavel, C. H. M. (1959). Practical use of know- ledge about evapotranspiration. Transactions of the O (.I Appendlx l 3O 4O 20 (°C) IO TEMPERATURE 1 IL JllllJ l 2 345 I0 20 3O POTENTIAL EVAPOTRANSPIRATION (cm) THORNTHWAITE'S NOMOGRAPH FOR DETERMINING POTENTIAL FIGURE 2| EVAPOTRANSPIRATION (J ,1 .—0 in .NO 1 m» I .uoII m .3 I. .00.! .mo II .40.| i Iu u d II. m L L IL L m .0 I1 .00 a. I mo I.“ .1 mo zozommppz -J «on oozpcjzm T. u m» . mpoijozczaa u m . mczmézm e. A o T 4mg... 1...... n me. me 22.53: I .0 _w N0 N0 u0 um .mO n.0C3m NN <>z m>