Evapotranspiration (ET) is one of the major components of the hydrological cycle and its accurate estimation is of paramount importance for many studies such as hydrologic water balance, irrigation system design and management, crop yield simulation and water resources planning and management. Water use efficiency can be improved by proper irrigation scheduling, which is essentially governed by crop evapotranspiration. A common practice for estimating ET from a well-watered agricultural crop is to first estimate reference crop ET, i.e., grass reference ET (ET0) or alfalfa reference ET (ETr), from a standard surface and to then apply an appropriate empirical crop coefficient, which accounts for the difference between the standard and crop ET (Kumar et al., 2002).
Evapotranspiration is not easy to measure. Specific devices and accurate measurements of various physical parameters or the soil water balance in lysimeters are required to determine evapotranspiration. The methods are often expensive, demanding in terms of accuracy of measurements and can only be fully exploited by well-trained research personnel. Although the methods are inappropriate for routine measurements, they remain important for the evaluation of ET estimates obtained by more indirect methods (Allen et al., 1994). Indirect methods based on climatological data are used for ET0 estimation. These methods vary from empirical relationships to complex methods based on physical processes such as the Penman (1948) combination method. The combination approach links evaporation dynamics with the flux of net radiation and aerodynamic transport characteristics of a natural surface based on the observations that latent heat transfer in plant stems is influenced not only by these abiotic factors, Monteith (1965) introduced a surface conductance term that accounted for the response of leaf stomata to its hydrologic environment. This modified form of the Penman equation is widely known as the Penman-Monteith evapotranspiration model.
Allen et al. (1994) recommended using a modified Penman-Monteith equation
for estimating ET0. The modified Penman-Monteith equation, which
was first presented by Allen et al. (1989), has received widespread acceptance
internationally for estimating ET0. This equation is currently recommended
by the United Nations Food and Agriculture Organizations (FAO) and by World
Meteorological Organization (WMO). However many different methods for estimating
ET0 have been developed, most of which are complex and require a
significant number of weather parameters (Doorenbos and Pruitt, 1977). But later
studies also showed lesser differences between estimated and measured ET0
with the Penman-Monteith equation than with others (Choisnel et al.,
1992; Hussein, 1999; Ventura et al., 1999).
Crop ET is computed by multiplying the ET0 with a crop coefficient (kc) to account for differences between the grass and crop ET. Due to variation in crop canopy and climatic conditions, ETc differs with the crop irrigation, planning and decision making on a regional scale is performed on the basis estimated crop ET, which in turn depends on the crop coefficient. The crop coefficient represents crop specific water use and is essential for accurate estimation of irrigation requirement of different crops in the command area (CSSRI, 2000; Kashyap and Panda, 2001).
The aim of this study is to compare the actual ET values with ET values calculated by different prediction methods and to define the method which gives the most accurate results of ET for sub-humid climatic region.
MATERIALS AND METHODS
Experimental site and meteorological data: This study was carried out in the summer of 2001 in a 3 year old and in the summer of 2002 in a 4 year old sweet cherry trees (Prunus avium, variety Z-900) on Mazzard rootstocks in Bayramic-Canakkale which is located in the west part of Turkey. The co-ordinates of research area are as follows: latitude 39° 48 N, longitude 26° 37 E and altitude 70 m above sea level.
The local climate is temperate, summers are hot and dry and winters are mild
and rainy. According to long term meteorological data which were taken from
State Meteorological Station, located 1 km away from the orchard, annual mean
rainfall, temperature and relative humidity are 624.3 mm, 14.0°C and 69%,
respectively (Anonymous, 1992). Although sub-humid climate prevails in the region
according to mean rainfall amount (from 600 to 700 mm of annual precipitation)
(Jensen et al., 1980), rainfall amounts are extremely low in the summer
period. Meteorological data used for ET measurements for June-October 2001 and
May-September 2002 was monitored by Metos early warning system and forecasting
station which is located in research area. Some meteorological data as average
values were given in Table 1 for irrigation season. Daily
meteorological conditions during the irrigation season, May to October, were
also given in Fig. 1 and 2 for 2001 and 2002.
Meteorological data for 2001, 2002 and long-term averages
* 2001-2002 Metos Early Warning System and Forecasting Station
**1970-2000 Bayramic State Meteorological Station Records
Daily meteorological conditions during the irrigation season,
May to October 2001, in Bayramic, Canakkale
Daily meteorological conditions during the irrigation season,
May to October 2002, in Bayramic, Canakkale
Data about daily maximum and minimum temperature, relative humidity, rainfall,
wind speed and solar radiation were used in calculating ET. Other parameters
needed for calculation were measured and estimated from meteorological data.
Field layout and experimental design: Soil texture characterization
was carried out from 12 profiles. Samples were taken with an auger at 0.30 m
intervals and with maximum depth to 0.90 m. The granulometric composition was
determined for each sample (Clay, <0.002 mm; Silt, 0.002-0.05 mm; Sand, 0.05-2
mm), as were the d50 parameter, the mean granulometrics and the mean
d50 for each profile and the mean values for each depth (Liu and
Evet, 1984). No vertical variability in the texture could be observed. The samples
were analyzed for soil texture, field capacity, wilting point, bulk density,
total salt content, pH and productivity level of the soil samples were found
according to methods given by Hansen et al. (1980). Soils in research
area have clayey loam texture. The main soil properties were given in Table
The trial was carried out with three replications in random blocks. However,
replications have been distributed to the random blocks in such a way that following
same range in three blocks not to disturb the existing irrigation system. The
trees were planted in 1998, spaced 6x6 m apart. Each plot contained three plant
rows and 39 trees, each block consists of 156 trees and total number of trees
is 468 on the trial plot. In order to prevent the water in any one plot from
affecting its neighboring plots, the two rows on the outer edges of each plot
were left untouched and only the one middle row was monitored. The alleyway
was kept under grass with an herbicide stripe (3 m broad) along the tree rows.
The experimental plots were fertilized with mineral nitrogen (1.5 kg/tree (NH4)2SO4),
potassium (1.2 kg/tree K2SO4), MAP (1.6 kg da-1)
and magnesium (7 kg da-1) in two experimental years. A routine pesticide
program was maintained.
Plant material studied was sweet cherry trees (Prunus avium, variety
Z-900) on 3-4 years old Mazzard rootstocks. Mazzard rootstock produces a vigorous
tree with very good anchorage and best compatible with sweet cherries. Z-900
grafted on Mazzard rootstock is large, firm, juicy, sweet variety with bordeaux
color and is adaptable to grow in different altitudes and different climates.
Irrigation water amount and irrigation scheduling: Micro-sprinkler irrigation was selected as the irrigation method. The laterals with the micro-sprinklers are laid along the rows of the trees, one line at each row; with one micro-sprinkler per tree. Sprinkler are operated under 1.4 bar pressure head and discharge of each is 35.8 L h-1 and sprinklers wetted diameter is 4.2 m. To determine the applied irrigation water along a lateral, water measurement devices were used for each lateral.
The trees were subjected to four micro-sprinkler irrigation treatments (T-1, T-2, T-3 and T-4). T-1 and T-2 treatments were programmed using two reduction percentages of the US Weather Bureau Class A pan evaporation. The water applied in treatment T-3 was considered sufficient to satisfy fully needs of the crop (100% of ETc) and to allow good rooting and tree growth.
The total amount of irrigation water (TIW) applied in treatment T-2 was calculated from (Allen et al. 2000);
where Kp is the pan coefficient (0.70; Doorenbos and Pruitt 1977),
Kc the crop coefficient (0.85; Allen et al., 2000), Kl
the shade coefficient (0.97; Wermeiren and Jobling, 1986) taking into account
that the estimated mean shade surface provided by the tree canopies was 85%
of the total surface of the orchard, Ea water application efficiency
(0.85; Armoni, 1986; Burt and Styles, 1994), Eu the coefficient of
uniformity of emitters (0.9).
T-3 treatment which is applied water amount in farm was selected as the control.
In this treatment, full of evaporated amount from Class A pan (100 % of Epan)
was applied to the trees. The difference at 25% level between T-2 and T-3 treatments
was considered as deficit amount and T-1 treatment was defined according to
this difference. To determine the impact of excessive water application on cherry
trees, another treatment (T-4) with the same amount of difference was selected
and applied to the trees. Thus, irrigation treatments were as follows:
The amount of irrigation water to be applied during a particular week was calculated
from the weekly evaporation values measured in the Class A pan during the preceding
week. Irrigation water was supplied weekly.
Calculation of crop evapotranspiration (Etcrop): Measurements of soil water content were initiated immediately after the completion of the flowering period with the ratio of 70% and ended with first frost appearance. The soil water content was measured every 7 days from 25 June 2001 to 29 October 2001 and from 27 May 2002 to 29 September 2002 (i.e., during the irrigation season) on 2 trees root zone. Elisea (2002) reported that the young cherry trees irrigated with micro-sprinklers shows that the bulk of the root system is located at 40-50 cm soil depth. Since the trees are young, efficient root depth was taken as 0.90 m. The soil water content was determined by gravimetrical method and soil samples were taken at 0.30 m intervals and with a maximum depth of 0.90 m. For each treatment, samples were taken from the points which are 0.50, 1.0, 2.0 and 3.0 m far from stem under tree crown. As it is reported by Abrisqueta et al. (2001); the three-dimensional aspect of water flow in the soil-plant-atmosphere system means that it is essential to determine the areas and volumes of soil in which water moves or is stored. It is customary to relate the water balance to the plantation spacing (Sharples et al., 1985), down to a depth slightly below that reached by the roots.
The water balance in the soil is estimated by means of the mass conservation equation by James (1988);
where ETc is the evapotranspiration (mm), I is the applied irrigation
water amount (mm), P is the precipitation (mm) and ΔS is soil water content
variation in crop root depth (mm/90 cm), D is drainage below the root zone and
R is the runoff. Since the clayey loam soil characteristics are fully
dominant in the field and lower sprinkling velocity of sprinkler (Isp =
5.81 mm h-1) than soil infiltration (Is = 8.00 mm h-1),
runoff and drainage (deep percolation) were assumed to be negligible.
Calculation of reference crop evapotranspiration (ETo): Reference
evapotranspiration was calculated by four methods for all months where sufficient
data available (Table 3).
Doorenbos and Pruitt (1975, 1977) presented a modified equation for estimating
reference ET for grass.
Different types of evapotranspiration estimated methods used
for the study
The major modifications consisted of a more sensitive wind function than that
used by Penman (1948),an adjusted factor c that is based on local climatic conditions
and the assumption that soil heat flux (G)=0 for daily periods. Thus the form
of FAO-Penman became
where c the adjustment factor and Wf = 1+0.864 x uz.
When the value of c is set to 1.0 it is called FAO-Penman and when the value of c calculated using the following equation:
Studies in the Sahel (Monteith, 1991) and in other arid regions (Jensen et
al., 1980) have shown the Penman-Monteith equation to be the most reliable
of the commonly used reference evapo-transpiration equations in arid and semi-arid
environments. The form of the Penman-Monteith equation for reference evapotranspiration
(Allen et al., 1994) requires values of temperature, humidity, sunshine
and wind speed. Penman-Monteith method (Allen et al., 1994).
where Rn is the net radiation (MJm-2 day-1), G is the
soil heat flux (0 MJ m-2 day-1), ea—ed
is the vapour pressure deficit (kPa), is
the slope of the vapour pressure curve (kPa °C-1), Á is
the psychrometric constant (kPa °C-1), rc is the canopy resistance
(70 sm-1), ra is the aerodynamic resistance (sm-1), Ú
is the atmospheric density (kg m-3), cp is the specific heat
of moist air (1.013 kJ kg-1 °C-1), Ï is the latent
heat of evaporation (MJ kg-1).
Doorenboss and Pruitt (1975, 1977) reported an equation, which is called FAO-Radiation method and is given as:
where c is an adjustment factor which depends of mean humidity and daytime
wind (Fervert et al., 1983), Δ is the slope of vapour pressure curve
(kPa °C-1), γ is the psychrometric constant (kPa °C-1),
Rs is the mean solar radiation (MJ m-2 day-1), λ
is the latent heat of evaporation ( MJ kg-1).
Doorenboss and Pruitt (1977) presented the most fundamental revision of the Blaney-Criddle method, which is given as
where P is the mean daily percentage of annual daytime hours (%), T is the
mean air temperature (°C), a, b are empirical factors (dimensionless), E
is the altitude of the station (m). a and b estimated from equations produced
by (Allen and Pruitt, 1986) and (Fervert et al., 1983), respectively,
using long-term average values of minimum relative humidity, sunshine fraction
and daytime wind speed.
It has been suggested (Smith et al., 1996) that where full climate data are not available, the Penman-Monteith equation can be used with long-term monthly average values of wind speed and estimates of mean daily vapour pressure, assuming dew point temperature equal to mean monthly minimum air temperature. As both wind speed and sunshine duration data are frequently missing for the chosen stations, the performance of the Penman-Monteith equation, using mean monthly values for both sunshine and wind, was tested. In order to maintain independence, the long-term monthly average values were taken from a published source (FAO, 1984) rather than the original data set.
Calculation of crop coefficient Kc: Crop evapotranspiration (ETcrop)
refers to evapotranspiration of a disease-free crop, grown in very large fields,
not short of water and fertiliser (Doorenbos and Pruitt, 1977). Estimation of
ETcrop is essential for computing the soil water balance and irrigation
scheduling. ETcrop is governed by weather and crop condition. Mathematically,
ETcrop can be expressed as
Where, Kc is the crop coefficient which varies for different crops and their growth stages and ETo is the reference crop evapotranspiration. Most of the current demand models are non-spatial models which uses point data of ETo and the Kc values from available literature (Doorenbos and Pruitt, 1977).
Comparison of methods: The results from each evapotranspiration calculation
method were compared to those from the crop evapotranspiration (ETcrop)
for which the latter could be calculated. The Root Mean Square Error (RMSE)
criterion was used to compare the daily values of ETo estimated by
various climatological methods. The RMSE provides a good measure of how closely
two independent data sets match (Ventura et al., 1999). The RMSE values
were calculated as
Where, n is number of observations.
RESULTS AND DISCUSSION
The annual amount of irrigation water applied is shown in Table
4 in which the amounts of irrigation water applied were ranged from 346.0
to 846.7 mm and from 313.3 to 783.3 mm for 2001 and 2002, respectively.
The soil water content was firstly measured on 25 June in 2001 and 27 May for
2002 to determine the crop water requirement. Daily measurements of the evaporation
from the Class A pan were started at 26 June for first experimental year and
28 May for second year. Total evaporation from Class A pan were 689.9 and 626.6
mm for 2001 and 2002, respectively. According to those values, measured evapotranspiration
at applied irrigation levels were realized as 365.0-839.0 and 447.0-866.0 mm
for first and second year, respectively (Table 5). The average
value of both years was found as 535.0 and 406.0 mm for T-2 and T-1 treatments,
Amount of water applied for each irrigation treatments and
rainfall in two years (mm)
*Total precipitation during the observation of
||Monthly and seasonal evapotranspiration (ETc mm)
a June 2001: ETc values in between 25-30
June, b October 2001: ETc values in between 01-29
October, c May 2002: ETc values in between 27-31 May,
d September 2002: ETc values in between 01-29 September
Ranking of ETo estimation methods on the basis
of root mean square error(RMSE) for 2001 and 2002
Monthly ETc has reached a maximum in August, July and September
in 2001 and in July, June and August in 2002, respectively.
The relationships between referens evapotranspiration values (ETo)
and ET values estimated by four different methods were investigated in the study.
In this context, the results of RMSE are presented in Table 6
for 2001 and 2002. The lowest values of RMSE were observed at Penman-Monteith
method (Table 6). In addition, the coefficient of determination
(R2), which expresses the proportion of the total variation in the
values of estimated variable can be explained by a linear relationship with
the values of the measured variable, was also used as a criterion for relative
performance of the estimation method (Fig. 3 and 4).
The values of R2 are listed in Table 6.
In 2001, the coefficient of determination (R2) and RMSE were found as 0.79 and 0.03 mm/day in the Penman-Monteith method for T-2 treatment, respectively. On the contrary, in 2002, coefficient of determination (R2) and RMSE were found as 0.84 and 0.05 mm/day in the Penman-Monteith method for T-3 treatment, respectively. The value of the coefficient of determination (R2) was the highest and the RMSE was the least in this case. The reason for the best performance is that this method uses data of maximum number of weather variables to estimate daily ETo.
Though, Penman-Monteith method is considered as the most rational and elaborate
approach, it requires a large number of climatological data. Often such data
are very scanty forcing the user to choose some other alternative empirical
methods for reliable estimation of ETo. Another reason to find an
alternative method is that the Penman type equations, having calibrated wind
functions, often fail in locations with a different climate than where they
were calibrated (Ventura et al., 1999).
Comparison of measured daily ET0 with that estimated
by different methods: (2001), (a) Penman-Monteith (b) Blaney-Criddle (c)
Fao-Penman (d) Fao-Radiation
Comparison of measured daily ET0 (2002) with that
estimated by different methods: (a) Penman-Monteith (b) Blaney-Criddle (c)
Fao-Penman (d) Fao-Radiation
Comparison of seasonal ET0 estimated by different
methods with respect to measured ET for 2001 and 2002
Comparison of seasonal ETo estimated by different methods with respect
to that measured is given in Fig. 5. Results of all the tests
mentioned above indicated that the Penman-Monteith method were close to measured
ETo, particularly T-3 and T-4 treatments in all years.
From the results of the study it is inferred that Penman-Monteith method is most ideal for estimation of ETo. Results of all the tests mentioned above indicated that the Penman-Monteith method is the best followed by the Blaney-Criddle method though FAO-Penman also gave reasonably good estimation of seasonal ET0 value.
The following conclusions were drawn from the results of the study.
||Comparison measured data revealed that the Penman-Monteith
gives the best estimate of daily reference evapotranspiration in a sub-humid
||The alternative methods to the Penman-Monteith, for estimation of reference
evapotranspiration, are Blaney Criddle, FAO-Penman and FAO-Radiation in