EFFECTS OF IRRIGATION INTERVAL ON PEAK

EVAPOTRANSPIRATION PER INTERVALbY

J. C. WilcoxMember C.S.A.E.

Canada Department of AgricultureResearch Station, Summerland, B.C.

Summerland orchards in 1959, at ninelocations in the southern interior ofthe province in 1959, and at 11orchards in the Oliver-Osoyoos area in1962. The type of bellani plate assembly used has already been described (3. 6).

Evaporation records were kept eachyear from May to September inclusive.In most cases readings were takendaily, but at some of the outlying stations they were taken only weekly.Volume of evaporation in cubic centimeters was transposed into inches ofdepth by use of charts that took intoaccount the evaporating area of thebellani plates.

The evaporation records thus takenat each location were examined each

year to determine which 5-day periodhad the greatest exaporation, what thetotal evaporation was during thisperiod, and what the average dailyevaporation was during this period.The same was done for periods of 7,10, 15, 20 and 30 days. Peak evaporation values were determined in thesame manner. Where evaporation records were taken weekly, as at the outlying location, the periods used were 7,14, 21 and 28 days.

Total and average peak evapotranspiration were plotted against length ofinterval in days, and the regressionequations relating evapotranspirationto length of interval were calculated.The same was done with evaporation.

RESULTS

Total Peak Evapotranspiration andEvaporation

In any one year, the peak intervalsusually centred around an especiallyhot spell for that year. In 1958, a veryhot year, the peak evapotranspirationand evaporation intervals on the weather station centered around July 15to 25.

When the interval was plottedagainst total peak evaporation andtotal peak evapotranspiration on log-paper, straight-line trends were alwaysobtained. The points on the chartswere usually quite close to the regression lines (figure 1). Thus theequation relating interval to evaporation or to evapotranspiration was ofthe following type:

logy = loga+ blogx (1)in which y = total peak evaporationor evapotranspiration in inches,

x = interval in days,a = total peak evaporation or eva

potranspiration at interval of oneday,

b = slope of curve.The regression lines for the weather

station data in 1958 (figure 1) hadthe following equations:

Evaporation: log y = log 0.885 -f-0.935 log x.

Evapotranspiration: log y = log0.468 + 0.909 log x.

Though widely separated, these tworegression lines were almost parallel.Total peak evaporation was approximately twice the total peak evapotranspiration. Even at that, evapotranspiration was much higher in the lysimetersthan in nearby orchards (1). Theregression lines of 1959 had slopessimilar to those of 1958 but the values

were lower.

Research workers have found (2, 5)that sandy soils require more irrigation water during the season than dosoils of finer texture. This has been attributed primarily to a lower application efficiency of water on sandy soils.The higher annual water requirementof a sandy soil has been found (4,5)to go hand in hand with a higher peakflow-per-acre requirement during theheat of the summer.

Factors other than a low applicationefficiency may also help to cause thegreater peak flow required by sandysoils. It is the purpose of this paper todiscuss one such factor — the averageevapotranspiration per day during thatirrigation interval having the peak rateof evapotranspiration. A search of theliterature has not revealed any previousstudies of this specific phase of theproblem.

In this paper, "irrigation interval"or "interval" means the time in daysbetween the start of any one irrigationand the start of the next irrigation.With any given length of interval, the"peak interval" is the one occurringwhen evapotranspiration is the hightest;"total peak evapotranspiration" is thetotal evapotranspiration in inches during this peak interval; "average peakevapotranspiration" is the averageevapotranspiration in inches per dayduring this peak interval. "Total peakevaporation" and "average peak evaporation" are used in similar manner, interms of evaporation in inches fromevaporimeters. The "average daily irrigation requirement" is the recommended irrigation application in inchesdivided by the peak interval in days.

EXPERIMENTAL PROCEDURE

In 1958 and 1959, evapotranspiration of alfalfa, ladino clover and bromegrass was determined in lysimeters onthe Summerland Research Station wea

ther station. The procedure used fordoing this has already been reported(7). The soil was kept wetted to 70%or more of available maisture.

Evapotranspiration was also determined indirectly by the measurement ofevaporation from black bellani plateevaporimeters. Close correlations havealready been reported (1, 3, 7) between evapotranspiration and evaporation. Bellani plates were installed onthe weather station in 1956, at six

5 7 10 15 20 30

LENGTH OF INTERVAL, IN DAYS

Figure I. Irrigation interval in days plotted againsttotal peak evaporation and evapotranspiration,weather station at Summerland, 1953; also plottedagainst recommended irrigation applications tor

orchards.

The net values obtained by deducting the rainfall from the evaporationor evapotranspiration are not shown infigure 1. In 1958 the rain was only0.27 inch at both 20 and 30 days. Thisproduced a slight curvature of thelog-log line. A similar effect has beenobtained in almost every case duringyears of high peak use. The curvaturehas, of course, been much more pronounced during the wetter years. Thewetter years, however, are of littleinterest in so far as designing an irrigation system is concerned. In the

CANADIAN AGRICULTURAL ENGINEERING, FEB. 1966

Okanagan Valley, the effects of rainon the peak flow requirements in ahot year have proved to be negligible.

The approximate curve relating theinterval to the irrigation application, asnow recommended (5) for sprinklerirrigation of orchards in the OkanaganValley, is shown in figure 1. Sprinklerirrigation requirements have beenfound to be lower than peak evapotranspiration as determined in lysimeters at relatively high soil moisture contents.

Variations in both position and slopeof the curves of total peak evaporation were found from year to year. Byway of example, the data for Orchard12 in Osoyoos are shown for threeconsecutive years in figure 2. Theequations for these three curves wereas follows:

In 1962: log y

0.924 log x.In 1963: log y

0.806 log x.

In 1964: log y0.814 log x.

log 1.062 +

log 1.242 +

log 1.000 +

The slope of the 1962 curve istypical of most of the slopes obtained.The slope in 1963 was lower than usualbecause of an exceptionally cool July.The slope in 1964 was also lower thanusual and the points were not as closeto the line. The weather was quitevariable in 1964 and there was noone focal period for all peak intervals.

a"*5 12

o .o. << >

9-

III I I I

0 5 7 10 15 20 30

LENGTH OF INTERVAL, IN DAYS

Figure 2. Irrigation interval in days plotted againsttotal peak evaporation, Orchard 12, 1962, 1963 and

1964.

Average Peak Evapotranspirationor Evaporation

The equation for the average peakevapotranspiration or evaporation perday was of the following type:

log i = log a + (b — 1) log x(2)

in which r = evapotranspiration orevaporation per day in inches, and x,a and b are the same as in equation 1.The equations for the five examples in

figures 1 and 2 were as follows:

Weather Station, 1958—

Evaporation: log r = log 0.885—0.065 log x

Evapotranspiration: log r = log0.468 — 0.091 log x.

Orchard 12 in Osoyoos, evaporationonly—

In 1962: log r = log 1.062 —0.076 log x

In 1963: log r = log 1.242 —0.194 log x

In 1964: log r = log 1.000 —0.186 logx.

In all cases studied, both averagepeak evaporation and average peakevapotranspiration decreased with anincrease in the length of the irrigationinterval. This is indicated by the negative values for the slopes of the curvesin the above equations. The percentages of reduction in evapotranspiration, evaporation and irrigation requirement obtained by lengthening theinterval from five to 30 days were asfollows:

Evapotranspiration, weather station1958 12.1%

Evaporation, weather station,1958 11.1%

Evaporation, Orchard, 12,1962 13.8%

Evaporation, Orchard 12,1963 29.7%

Evaporation, Orchard 12,1964 25.4%

Irrigation requirement 53.6%

In those years with especially highrates of peak evaporation, as in 1958,1960 and 1962, the reduction in peakrate per day between a 5-day intervaland a 30-day interval averaged about12.5% of the rate at the 5-day interval. The percentage difference wasmuch greater than this in years oflow peak evaporation, as in 1963 and1964. System capacity, however, depends on peak evapotranspiration inhot years rather than in cool years,hence the effects of lengthening theinterval in cool years can be ignored.The 12.5% reduction accounted forabout 23% of the 53.6% reductionin the daily irrigation requirement. Inother words, about 23% of the decrease in peak irrigation requirementaccompanying an increase in intervalcould be attributed to the effect of

length of interval on average peakevapotranspiration.

Differences between and withinDistricts

Wide differences in total peak evaporation have been found betweendistricts located within the semi-arid

parts of the southern interior of British Columbia (table 1). The peakvalues at Vernon in 1962 ranged from19 to 30% lower than those atOsoyoos, even though both locationsare in the Okanagan Valley; at Crestonthey were 39 to 44% lower than theOsoyoos values.

TABLE I. VARIATION IN TOTAL PEAK EVAPORATION FROM DISTRICT TODISTRICT, 1962*

Length of Interval

District 7 days 14 days 21 days 28 days

inches inches inches inches

Creston 3.32 5.96 8.30 10.58

Grand Forks 4.02 7.22 10.20 12.98

Osoyoos 5.40 10.18 14.77 19.05

Oliver 5.25 9.96 14.40 18.71

Keremeos 4.26 8.20 12.04 15.78

Summerland 4.72 9.02 13.20 17.25

Kelowna 4.74 9.04 13.20 17.25

Vernon 4.40 7.63 10.58 13.39

Armstrong 4.10 7.15 9.88 12.40

Salmon Arm 3.71 6.96 9.97 12.84

*Each value shown is the total peak evaporation in inches for the interval indicated.

TABLE II. VARIATION IN TOTAL PEAK EVAPORATION AMONG FIVE

ORCHARDS IN THE OLIVER DISTRICT, 1962

Length of Interval

Orchard No. 5 days 10 days 20 days 30 days

inches inches inches inches

7 4.12 7.43 13.50 18.90

8 4.90 8.88 16.17 23.23

9 4.04 7.42 13.60 19.45

10 4.32 7.90 14.58 20.60

11 2.90 5.24 10.28 14.85

CANADIAN AGRICULTURAL ENGINEERING, FEB. 1966

In two districts, Oliver and Osoyoos,evaporation records have been kept forthree years at several locations withineach district. Surprisingly wide differences in peak evaporation were obtained within each district. By way ofexample, 1962 data are shown forOliver in table 2. They illustrate thatwhen records were taken at only onelocation in a district, this location didnot necessarily represent accuratelythe district as a whole.

DISCUSSION

In designing an irrigation system, anengineer normally tries to incorporateinto it a peak capacity that equateswith the peak requirements of the landto be irrigated. The peak requirements,in turn, depend on such factors aspeak evapotranspiration, rain, application efficiency, and whether growersare irrigating steadily or intermittentlyduring the heat of the summer. At anyone location, the peak requirementsnormally occur during the hottest partof the hottest year. Fruit growers generally believe that system capacityshould be great enough to satisfy thetime of greatest need; and that in sofar as system design is concerned, timesof lesser need are unimportant.

This paper has dealt with two related parameters — peak evaporationand peak evapotranspiration. The dailypeak value of each has been found todecrease with an increase in the irri

gation interval. As reported elsewhere(1, 5), the safe irrigation intervalduring peak evapotranspiration hasbeen found to be closely dependent onthe soil texture: the smaller the sizeof the soil particles, the longer is thesafe interval. In this investigation, peakevaporation per day has varied widelyfrom year to year, from district todistrict, and even from orchard toorchard within the same district, solelybecause of climatic variations.

Irrigation investigations conductedon tree fruits in the Okanagan Valleyhave led to recommendations of required peak flow per acre based primarily on the texture of the soil (5).More than twice as much water hasbeen recommended for a coarse-

textured soil requiring an irrigationevery five days at peak use as for afine-textured soil requiring an irrigationonly every 30 days. At time of peakuse, the average peak rate of evaporation has been found to account forabout 23% of this difference in irrigation requirement. At times of lessthan peak use it has accounted formore than the 23%; but since systemcapacity recommendations are basedon peak requirements only, this is theonly value that has proved applicablethus far.

It should be stressed that the effectsof the irrigation interval on averagepeak evaporation have not applied tothe total irrigation requirement for theseason. It is true that a sandy soil hasneeded more water during the seasonthan a silt or clay soil; but there is noevidence at hand to indicate that thisis due to greater evapotranspiration onthe sandy soil. It is more likely to bedue to other factors such as a lowerapplication efficiency.

It was found that in very hot yearsthe amount of rain falling within peakintervals up to 30 days in length wassmall enough that it could be ignoredfor design purposes. This finding cannot, obviously, be automatically applied to other areas — especially morehumid areas where rain is heavier ormore frequent.

SUMMARY AND CONCLUSIONS

Evapotranspiration was determinedusing lysimeteres at Summerland. Evaporation was determined by use ofbellani plate evaporimeters, at variouslocations in the southern interior ofBritish Columbia.

The greatest total (peak) evaporation per irrigation interval during theseason correlated positively with thelength of irrigation interval. The regression between them was linear onlog-log paper. The same held true withevapotranspiration.

The average evaporation or evapotranspiration per day during the peakinterval correlated negatively with thelength of the interval; in other words,it was less with long intervals than withshort intervals, hence less with fine-

textured soils than with coarse-texturedsoils. In hot years the difference inaverage evaporation per day between5-day and 30-day intervals accountedfor about 23% of the difference inpeak flow per acre required for irrigation.

Peak evaporation varied widely fromyear to year, from district to district,and even from orchard to orchardwithin the one district. In hot years,the benefit from rain during intervalsof peak use was negligible.

REFERENCES

1. Korven, H. C. and Wilcox, J. C.Correlation Between Evaporationfrom Bellani Plates and Evapotranspiration from Orchards. Can.Jour. Plant Sci. 45: 132-138.1965.

2. Newell, R. J., Johnson, W. W.,Patrick, A., Schofield, C. S. andSingleton, H. P. Columbia BasinJoint Investigations: IrrigationWater Requirements, Problems 4and 5. U.S. Bur. Reel. Spec. Rep.1945.

3. Robertson, G. W. and Holmes, R.M. A New Concept of the Measurement of Evaporation for Climatic Purposes. Proc. Inter. Ass.Sci. Hydr. 3: 399-406. 1957.

4. Wilcox, J. C. Effect of IrrigationInterval on Peak Flow Requirements in Sprinkler Irrigated Orchards. Can. Jour. Soil Sci. 40:99-104. 1960.

5. Wilcox, J. C. and Brownlee, C. H.Sprinkler Irrigation Requirementsfor Tree Fruits in the OkanaganValley. Canada Dept. Agr. Pub1121. 1961.

6. Wilcox, J. C. Note on Effects ofShielding Bellani Plates on Rateof Evaporation. Can. Jour. PlantSci. 42: 400-401. 1962.

7. Wilcox, J. C. Effects of Weatheron Evaporation from BellaniPlates and Evapotranspirationfrom Lysimeters. Can. Jour. PlantSci. 43: 1-11. 1963.

CANADIAN AGRICULTURAL ENGINEERING, FEB. 1966