Irrig Sci (1986) 7:237-243
Irrigation : clence © Springer-Verlag 1986
Estimating Evapotranspiration from Midday Canopy Temperature P. L. Gupta I and P. S. N. Sastry 2 Physics Section Haryana Agricultural University, Hisar-125004, India 2 Division of Agricultural Physics, Indian Agricultural Research Institute, New Delhi 110012, India Received May 1, 1986 Summary. Measurements were made at New Delhi (India) on wheat (Triticum aestivum L.) and mungbean (Vigna radiata L.) crops growing during the winter and summer seasons respectively, to evaluate the use of the equation of Bartholic, Namken and Wiegand (1970) for estimating daytime evapotranspiration from a single measured canopy temperature. Measurements on eleven days indicated that a single observation of the canopy temperature taken at any time between noon and 2 p.m. could be used to compute daytime evapotranspiration with an error (underestimate) of less than 27% as compared with values measured by the Bowen ratio method. The regression equation ET0 = 0.618 + 0.96 ETI has a coefficient of determination of 0.927 and can be used to relate daytime evapotranspiration by the Bowen ratio method (ET0) to that estimated by the Bartholic-Namken-Wiegand equation (ET0 for wheat and mung bean crops grown under adequately watered conditions in the New Delhi region.
Estimates of evapotranspiration (ET) are needed in agriculture for scheduling irrigation, predicting crop yields and management of available water resources for better crop production. Well-established methods such as the energy balanceBowen ratio, aerodynamic, Penman and Van Bavel combination equation, used to estimate evapotranspiration, require several field measurements. The rapidly developing technique of remote sensing has made the task of estimating ET on a regional basis easier by providing a measure of canopy temperature. Equations such as those developed by Bartholic etal. (1970) and Brown and Rosenberg (1972) for ET estimation, using remotely sensed canopy temperature, provide hourly ET values which on integration yield a ET value for the whole day. Jackson et al. (1977), Sequin and Itier (1983) and Saha et al. (1984) have developed simplified forms of the Brown-Rosenberg equation and have estimated the daily ET using a single canopy temperature. However, the feasibility of using the Bartholic et al. equation for this purpose has not been tested so far although this equation has been reported to be well-suited for estimating regional ET (Stone and Horton 1974; Gupta and Sastry 1984). Therefore, the present investigation was conducted
238
P.L. Gupta and P. S. N. Sastry
to test the feasibility o f using a single m e a s u r e d c a n o p y t e m p e r a t u r e (at a g i v e n t i m e o f the day) to o b t a i n the daily ET by m e a n s o f the e q u a t i o n o f B a r t h o l i c et al. (1970).
Materials and Methods The present investigation was carried out on a wheat (Triticum aestivum L.) crop during the rabi season of 1982-83 and on a mung bean (Vigna radiata L.) crop during summer of 1983 on sandy loam soils of Indian Agricultural Research Institute (I.A.R.I.), New Delhi. Wheat (Var. Sonalika) was planted on 9th Nov. 1982 and Mungbean (Var. PS-16) was sown on 24th April, 1983. Irrigation for both the crops was planned to maintain ample water, so that moisture in the effective root zone (15-45 cm depth) was not depleted below 50% of the available water held between 0.03 and 1.5 mPa soil water potential. A Teletemp infrared thermometer was used to measure the hourly canopy temperature (Tc) and the canopy-air temperature difference (Tc-Ta) during the day time. Hourly net radiation (Rn) was measured with a portable net radiometer mounted 50 cm above the crop canopy. Soil heat flux (G) was measured with a heat flux plate at 5 cm depth in the soil between the rows below the crop canopy. A meter was used to measure the leaf area index. The experimental site for the wheat crop (rabi season 1982-83) was surrounded by a similar wheat crop with a fetch of more than one kilometer. There was fetch ratio of 20 : 1 for the summer mung crop which was surrounded by cotton, mung and other vegetable crops. To compute the Bowen ratio (fl), the temperature and the vapour pressure gradients above the crop canopy were measured using aspirated thermocouple psychrometer system. For this purpose, five sets of psychrorneters were installed at the different heights (15, 40, 65, 115, 140 cm) in case of wheat and (20, 45, 70, 95, 145 cm) in case of mungbean on a white painted wooden mast erected in the field. Air temperature and vapour pressure observations were recorded for a 30-min duration at hourly intervals during the day time. These 30-min observations were then averaged to obtain temperature and vapour pressure for that hour. To minimize the effect of measurement errors the measurements were confined to within two metre above the crop canopy, height intervals corresponding to maximum gradients were considered for the computation of Bowen ratio (fl) which hopefully minimize the measuremerit errors (Fuchs and Tanner 1970; Reimer and Dasmarais 1973). In addition the Bowen ratios (fl) were only computed for days characterized by - 0 . 2 5 ~ Ri <- + 0.025 and fl =< 0.3, where R i is the Richardson number. Assuming the soil heat flux to be negligible, the hourly day time ET values were computed using the Bowen ratio equation, given below. R, ET = - l+fl
(1)
where R, is net radiation (kWm -2) and the Bowen ratio (fl) is given by the following equation: AT
fl = 7 - -
Ae
(2)
where 7 is the psychrometer constant (millibar/°C), A T and de are the temperature and vapour pressure d111erences between the two laelghts above the crop canopy. The day time evapotranspiration (ET0) by this method was obtained by integrating these hourly ET values, and used as a reference for comparison. The following equation suggested by Bartholic et al. (1970) using canopy temperature could be used to compute hourly ET values. ET =
Rn
Tc-Ta 1 + 7 - -
ec - ea
(3)
Evapotranspiration from Midday Canopy Temperature
239
where T c and Ta are the canopy and air temperature (°C) and ec ea are the saturated vapour pressure (mbar) at canopy and air temperature respectively. The day time evapotranspiration (ET2) by this method was computed by integrating these hourly ET values. A separate day time evapotranspiration (ETx) was computed from Eq. (3) using the single midday canopy temperature and the integrated day time net radiation. Richardson number (Rg) was computed using the following expression: R,
g dT/dZ T (du/dZ) 2"
(4)
Where g is the acceleration due to gravity (m/s2), T is the temperature of the air layer (°K) and dT/dZ and du/dZ are the temperature and wind gradients respectively expressed in °K/m and s-1.
Results
The midday canopy and the air temperature and the range of canopy to air temperature difference on the days of observations are presented in Table 1. The eight days of observations during the rabi season include all the important stages of wheat growth, namely, heading, anthesis, soft dough and hard dough stages. Three days in summer correspond to the flowering, pod development and the maturity stages of the mungbean development. The leaf area index for the wheat crop was between 5.2 to 6.1 and between 2.3 to 5.7 for the mungbean during the days of observations. Soil heat flux as a percentage of net radiation is presented in Table 2. The integrated day time soil heat flux as a percentage of day time net radiation was found to be one to six %, in general, for both the crops. However, during the early stages of crop growth, it was 8 to 10% of net radiation. Even at mid-day the soil heat flux as a percentage o f net radiation was found to be of similar order.
Table 1. Midday canopy and air temperature on the days of observations Season and crop
Date
Crop stage
Leaf area index
Midday temperature (° C)
(rc)
Daytime range of canopy to Air near air temp. the canopy difference (To) (°C)
17.6 19.8 20.2 21.1 26.4 23.9 30.3 28.8 42.4 40.6
2.84.84.04.21.72.62.94.34.43.3-
33.3
1.8- 5.4
Canopy
Winter (wheat)
Summer (mungbean)
4. 2. 83 6. 2. 83 10. 2. 83 18.2. 83 22. 2. 83 5.3.83 11.3.83 20. 3. 83 1.6. 83 15. 6. 83 27. 6. 83
Heading
5.5
Anthesis
5.9
Soft dough
6.1
Hard dough Flowering Pod development Maturity
5.2 2.3 4.0
13.7 13.4 14.6 12.3 19.4 20.4 24.8 20.5 33.1 25.9
5.7
31.3
6.2 7.6 7.6 9.9 7.0 8.0 9.7 8.2 10.6 15.8
P. L. Gupta and P. S. N. Sastry
240 Table 2. Soil heat flux at different crop growth stages Date
Crop growth stage
-Integrated day time-] soil heat f l u x ]
Midday soil] heat flux | x 100
Integrated day time I net radiation _]
Jan 21 Feb 4 Feb 10 Feb 26 March 5 March 20
Wheat Boot leaf Heading Anthesis Milking Soft dough Hard dough
June 1 June 21 June 27
Summer Mung Flowering Pod development Maturity
M i d d a y - - [x 100 net radiationJ
9 10 6 2 2 1
10 6 5 2 3 3
8 4 1
8 6 3
Table 3. Comparison between daytime evapotranspiration ET2 (integrated hourly ET values), ETa (value computed using the Midday canopy temp.) by Bartholic et al. (1970) model and ET0 (integrated hourly values computed by Bowen ratio method) Day
Day time evapotranspiration kWhm -2
ET1 - ET0 -
-
ET2
ETI
ETo
Winter (Wheat) 4. 2. 83 6. 2. 83 10. 2. 83 18. 2. 83 22. 2. 83 5. 3. 83 11.3. 83 20. 3. 83
1.84 1.64 1.84 2.38 2.44 2.46 2.37 3.00
1.80 1.65 1.72 2.42 2.52 2.39 2.50 2.92
2.38 1.94 2.38 2.94 3.05 3.00 2.97 3.48
-
Sumnier (Mungbean) 1.6. 83 15. 6. 83 27. 6. 83
3.78 2.84 2.63
3.72 2.95 2.70
4.14 3.47 2.98
- 10 - 15 - 9
x
100
ETo
24 15 27 22 17 20 16 14
The d a y t i m e evapotranspiration ETI (computed using single m i d d a y measurement of canopy t e m p e r a t u r e ) and ET2 (daytime integrated hourly values) by Bartholic et al. (1970) equation is presented in Table 3. The difference (ETI - ET2) ranged between - 0 . 1 2 to + 0.11 k W h m -2. On applying the p a i r e d 't' test, the calculated 't' value (0.03) was found to be much less than the tabulated value (2.23) at 5% level indicating no significant difference between ET1 and E T 2 . This result clearly indicates that using a single m i d d a y m e a s u r e m e n t on canopy t e m p e r a t u r e
Evapotranspiration from Midday Canopy Temperature
241
Table 4. Comparison of hourly evapotranspiration from wheat and mungbean crops ET 4 (computed by Bartholic et al. equation and ET 3 (computed by Bowen ratio) Daytime Hourly evapotranspiration W / m 2 hours (IST) ET3 ET4
07.00 08.00 09.00 10.00 11.00 12.00 13.00 14.00 15.00 16.00 17,00
100 x (ET4 - ET3)/ET3
10/2
18/2
11/3
1/6 a
10/2
18/2
11/3
1/6 a
10/2
183 259 354 355 342 454 274 101 15
193 325 436 478 498 455 328 166 13
244 372 487 478 384 448 349 121 51
214 284 365 465 451 526 541 461 383 313 170
145 205 279 296 281 384 186 70 10
142 265 374 392 383 367 265 117 9
208 305 404 394 318 366 289 103 39
168 237 339 423 411 475 477 412 342 235 132
. . -
. . 21 21 20 17 18 15 32 31 33
18/2
11/3
1/6 a
. . - 26 - 19 - 15 - 18 -21 - 19 - 19 -29 - 31
. . - 15 - 18 - 17 - 18 - 16 - 18 - 17 - 15 -24
21 16 - 7 - 9 - 9 - 10 - 12 - 11 - 11 -25 -23
a Mungbean WHEAT
FEB,18
0.6
WHEAT
0.4 0.2
m
,".._
L,
0
c3
....
-0.2
WHEAT
HU N6 BEAN
0.2 O
-O.2~1
1
~
~
,
IlL'
,
, ~P"...e.-.41~ i
51100~1300- 1500 - '~0 ~ 0 ~5 900 0 1100 0 0031
i
,
{
"rIME h ( I S T )
Fig. 1. Daytime variation of bowen ratio fl=TA-~_T(e)
To-To ec ea
and its equivalent factor # e =
(o)
-
in t h e B a r t h o l i c et al. e q u a t i o n is as v a l i d as u s i n g h o u r l y d a y t i m e values a n d i n t e g r a t i n g to o b t a i n d a y t i m e e v a p o t r a n s p i r a t i o n . C o m p a r e d to ET0 ( d a y t i m e i n t e g r a t e d h o u r l y ET values c o m p u t e d b y B o w e n r a t i o m e t h o d ) , B a r t h o l i c et al. (1970) e q u a t i o n u n d e r e s t i m a t e d e v a p o t r a n s p i r a t i o n b y values r a n g i n g f r o m 14 to 27% in t h e case o f t h e w h e a t c r o p in t h e rabi season, a n d b e t w e e n 9 to 15% for t h e m u n g b e a n in s u m m e r s e a s o n ( T a b l e 3). T h e h o u r l y e v a p o t r a n s p i r a t i o n c o m p u t e d b y B o w e n r a t i o (ET3) a n d b y B a r t h o l i c e t a l . (1970) e q u a t i o n (ET4) are p r e s e n t e d in T a b l e 4 a n d c o m p a r e d .
242
P.L. Gupta and P. S. N. Sastry
Figure 1 compares the daytime hourly values of Bowen ratio (fl) and its equivalent
(
Tc-Ta)"Theflvaluesweref°undt°decreaseduring
factor (/~e) where fie = 7 e-~---ea
the afternoon hours becoming negative in the evening. Similar observations o n / ~ values were earlier reported by Nkemdirim and Yamashita (1972) and Reimer and Desmarais (1973). Compared to p, the fie exhibited greater seasonal variation as its value at midday decreased from 0.45 on February 10 to 0.22 on March 11.
Discussion
Daytime evapotranspiration calculated by the Bartholic etal. equation using midday canopy temperature ET1, and by the Bowen ratio method ET0, were computed and compared neglecting the soil heat flux. This should not seriously influence ET estimates as the measurements presented were for almost complete canopy cover and the soil heat was less than 10% of the net radiation during this period (Table 2). The Bowen ratio method, a basic approach based on sound physical principles, has been considered as the reference for comparison. Moreover the error due to the assumption of equality of transfer coefficient for heat and the water vapour in this method was minimized to less than five % by selecting days • characterized by - 0.25 N R i~-t- 0.025 and fl ~ 0.3 as suggested by Campbell (1973). The underestimate of ET (14-27% from wheat and 9-15% from the mungbean) in the present investigation are similar to those reported earlier by Stone and Horton (1974), and could be attributed to assumptions in the Bartholic et al. equation, i.e. that (i) the crop canopy is a wet surface whose vapour pressure can be given by the saturation vapour pressure at the prevailing canopy temperature, (ii) that the air near the surface is saturated and its vapour pressure is approximated by the saturation vapour pressure at air temperature. The present investigations were carried out under water non-limiting condition which hopefully may satisfy the first assumption. However, the second assumption never seems to be satisfied under the field conditions. To examine the consistency of the percentage deviation, the hourly estimates of evapotranspiration ET4 (Bartholic et al. equation) and ET3 (Bowen ratio method) were compared. Compared to ET3, Bartholic et al. equation underestimated ET values to a greater extent in the morning and evening hours as compared to the noon hours values (Table 4). The consistent values during the daytime between 12.00 to 14.00 h (IST) on the days of observations (Fig. 1) indicated that for the estimation of evapotranspiration by Bartholic et al. equation, single observation on canopy temperature can be taken at any time between 12.00 to 14.00 h (IST). The approach of the fie and the fl values with the advancing season suggested a greater suitability of the Bartholic etal. equation as the season advanced and became comparatively warmer. The daytime estimates of evapotranspiration (kWhm -2) from wheat in the rabi season and from the mungbean during summer were pooled and simple regression analysis yielded the following regression equation with a coefficient of determina-
fie
Evapotranspiration from Midday Canopy Temperature
243
tion of 0.927. ET0 = 0.618 + 0.96 ET1.
(5)
The Bartholic-Namken-Wiegand equation, though underestimating evapotranspiration by up to 27%, has the advantage of being simple in its application in that it needs fewer input parameters. Net radiation values could be estimated from incoming solar radiation and published values of the albedo for a particular crop, air temperature could be measured on the ground, and canopy temperature with a remote thermal sensor. The regression Eq. (5) could then be used to compute daytime evapotranspiration over larger area in which ET estimate from BartholicNamken-Wiegand equation is used as independent variable. However, it should be emphasized that such a regression method is location specific.
References Bartholic JP, Namken LR, Wiegand CL (1970) Combination equation used to calculate evaporation and potential evaporation. USDA-ARS-Bull 41-170:14 pp Brown KW, Rosenberg NJ (1972) A resistance model to predict evapotranspiration and its application to a sugar beat field. University of Nebraska, Lincoln. Horticulture Prog Rep 25:51-5.31 Campbell AP (1973) The effect of stability on evaporation rates measured by energy balance method. Agric Meteorol 11:261 Fuchs M, Tanner CB (1970) Error analysis of Bowen ratio measured by differential psychrometry. Agric Meteorol 7:329 Gupta PL, Sastry PSN (1984) Remotely sensed canopy temperature based models for estimating evapotranspiration. Proceedings of fifth Asian conference on remote sensing, Kathmandu (Nepal), pp P7.1 Jackson RD, Reginato RJ, Idso SB (1977) Wheat canopy temperature: A practical tool for evaluating water requirements. Water Resour Res 13:651 Nkemdirim LC, Yamashita S (1972) Energy balance over prairie gras. Can J Plant Sci 52:215 Reimer A, Desmarais R (1973) Micrometeorological energy budget methods and apparent difficulty for Boreal forest and grass sites at Pinawa, Minitiba, Canada. Agric Meteorol 11:419 Saha SK, Ajay, Kamat DS, Singh AK, Aggarwal PK, Chaturvedi GS, Sinha SK (1984) Remote sensing of crop evapotranspiration using plant canopy temperature. Proceedings of Seminar on crop growth condition and remote sensing, Indian Agricultural Research Institute, New Delhi, pp 5.1.2-5 Seguin B, Itier B (1983) Using midday surface temperature to estimate daily evaporation from the satellite thermal IR data. Int Remote Sensing J 4:371 Stone LR, Horton ML (1974) Estimating evapotranspiration using canopy temperatures. Field evaluation. Agron J 66:450