102 M. Smith Agricultural and Forest Meteorology 103 2000 99–108
Fig. 2. Average irrigation water losses.
ration, loss of energy height and deterioration of water quality.
Most of the water loss 40 occurs at farm and field level with a direct effect on crop production due
to inadequate water supplies causing water stress or excessive water and resulting in reduced growth and
leaching of plant nutrients. Considerable scope exists for a more accurate and efficient crop water application
by improved field irrigation methods and better crop water management techniques through the introduction
of irrigation scheduling and water supply control.
To introduce an effective crop water supply system adequate information is required on crop water require-
ments as determined by crop and weather conditions. In Paragraph 4 below an example is given how to evaluate,
plan and manage irrigation supplies and field irrigation practices based on crop and weather conditions for a
given irrigation management system.
3. FAO methodologies on crop water requirements
The FAO Land and Water Development Division has been instrumental in developing guidelines for the pre-
diction of crop water requirements, which have been widely introduced for the design and management of
irrigation systems. The methodology, published first in 1974 as FAO Irrigation and Drainage Paper No.
24 and revised in 1977 Doorenbos and Pruitt, 1977, has become an international standard, extensively used
worldwide.
Advances in research and the more accurate assess- ment of crop water use have revealed weaknesses in
the original methodologies of FAO No. 24 Batche- lor, 1984; Allen et al., 1989; Jensen et al., 1990, and
have made a review necessary. In collaboration with the International Commission for Irrigation and Drainage
ICID and the World Meteorological Organization WMO, a consultation of experts and researchers was
organized in May 1990, in Rome, to review the method- ologies and to advise on the revision and update of pro-
cedures. Based on findings of comparative studies on the performance of various ET
o
estimation methods, the panel of experts recommended the adoption of the
Penman–Monteith method and a revised definition and calculation procedures for estimating reference evapo-
transpiration Smith et al., 1991.
The review and update of the methodologies have been recently completed and are contained in FAO
Irrigation and Drainage Paper No. 56 Crop Evapotran- spiration Allen et al., 1998. The main elements of the
review and revised procedures for crop evapotranspi- ration are presented further.
3.1. Review of reference evapotranspiration methods A range of more or less empirical methods have been
developed over the last 50 years by numerous scien- tists and specialists, worldwide, to estimate evapotran-
spiration of a reference crop from different climatic variables. Relationships were often subject to rigorous
local calibrations and proved to have limited global validity.
The large number of ET
o
estimation methods, with various locally adapted or modified parameters, have
often been confusing and prevented transparency and
M. Smith Agricultural and Forest Meteorology 103 2000 99–108 103
uniformity in calculated parameters. To evaluate the performance of some of the main ET
o
estimation proce- dures, under different climatological conditions, stud-
ies were undertaken under the auspices of the American Society of Civil Engineers Jensen et al., 1990 and the
European Community Choisnel et al., 1992.
The studies demonstrated the superior performance of the Penman–Monteith approach, in both arid and
humid climates, and convincingly confirmed the sound underlying concepts of the method. Based on these
findings, the method was recommended by the FAO Panel of Experts, convened in 1990, to be adopted as
a new standard for reference crop evapotranspiration estimates. The conclusions of the comparative studies
are summarized in Table 1.
3.2. FAO Penman–Monteith equation By introducing the aerodynamic and canopy resis-
tance in the original combination method, a better
Table 1 Performance ET
o
estimation method Jensen et al., 1990 Locations
Humid. Arid
Performance indicator Rank No.
Over-estimate Standard error
Rank No. Over-estimate
Standard error Combination methods
Penman–Monteith 1
+ 4
0.32 1
− 1
0.49 FAO-24 Penman c=1
14 +
29 0.93
6 +
12 0.69
FAO-24 Penman corrected 19
+ 35
1.14 10
+ 18
1.1 FAO–PPP-17 Penman
4 +
16 0.67
5 +
6 0.68
Penman 1963 3
+ 14
0.60 7
− 2
0.70 Penman 1963, VPD 3
6 +
20 0.69
4 +
6 0.67
1972 Kimberley Penman 8
+ 18
0.71 8
+ 6
0.73 1982 Kimberley Penman
7 +
10 0.69
2 +
3 0.54
Businger-van Bavel 16
+ 32
1.03 11
+ 11
1.12 Radiation methods
Priestley Taylor 5
− 3
0.68 19
− 27
1.89 FAO-Radiation
11 +
22 0.79
3 +
6 0.62
Temperature methods Jensen-Haise
12 −
18 0.84
12 −
12 1.13
Hargreaves 10
+ 25
0.79 13
− 9
1.17 Turc
2 +
5 0.56
18 −
26 1.88
SCS Blaney-Crddle 15
+ 17
1.01 15
− 16
1.29 FAO Blaney-Criddle
9 +
16 0.79
9 0.76
Thornwaite 13
− 4
0.86 20
− 37
2.4 Pan evaporation methods
Class A pan 20
+ 14
1.29 17
+ 21
1.54 Christiansen
18 −
10 1.12
16 −
6 1.41
FAO Class A 17
− 5
1.09 14
+ 5
1.25
simulation of wind and turbulence effects and of the stomatal behavior of the crop canopy was achieved
Monteith, 1965. The earlier difficulties in the use of the method, related to the estimation of the resistance
values, have been largely overcome by progress in re- search and reliable estimates of the two parameters for a
range of crops, including the reference crops, grass and alfalfa.
To adopt the Penman–Monteith method globally, valid estimates on bulk surface and aerodynamic re-
sistance were required. Procedures, as elaborated by Allen et al. 1989, were adopted and fixed values for
crop surface resistance and crop height defined for the reference crop. This required an adjustment of the con-
cept and definition of reference evapotranspiration as the rate of evapotranspiration from a hypothetical refer-
ence crop with an assumed crop height 12 cm, a fixed crop surface resistance 70 s m
− 1
and albedo 0.23, closely resembling the evapotranspiration from an ex-
tensive surface of green grass cover of uniform height,
104 M. Smith Agricultural and Forest Meteorology 103 2000 99–108
actively growing, completely shading the ground and with adequate water.
Thus defined, the FAO Penman–Monteith equation can be further derived from standard crop parameters
to the following relationship for daily reference crop evapotranspiration:
ET
o
= 0.4081R
n
− G+γ 900T + 273U
2
e
a
− e
d
1 + γ 1 + 0.34U
2
where ET
o
: reference crop evapotranspiration [mm per day]; R
n
: net radiation at the crop surface [MJ m
− 2
per day]; G: soil heat flux [MJ m
− 2
per day]; T: average air temperature [C]; U
2
: wind speed measured at 2 m height [m s
− 1
]; e
a
− e
d
: vapor pressure deficit [kPa]; 1: slope of the vapor pressure curve [kPa C
− 1
]; γ : psy- chometric constant [kPa C
− 1
]; 900: conversion factor. Full details of the FAO Penman–Monteith method,
including procedures for determining hourly ET
o
val- ues and the various parameters, algorithms, recom-
mended values and units, have been published in the proceedings of the consultation Smith et al., 1991, in
the ICID Bulletin Vol. 43, No. 2 Allen et al., 1994a, b, and the new FAO Irrigation and Drainage Paper No.
56 Allen et al., 1998.
3.3. Use of FAO Penman–Monteith with limited climatic data
The limited availability of the full range of climatic data, in particular, data on sunshine, humidity and wind
data, has often been the main restriction in the use of the combination methods and resulted in the use of empir-
ical methods, which require only temperature, pan or radiation data. This has contributed to the confusing use
of different ET
o
methods and conflicting evapotranspi- ration values. To overcome this constraint and to further
standardize on the use of one single method, additional studies have been undertaken to provide recommenda-
tions on the use of the FAO Penman–Monteith when no humidity, radiation or wind data are available. As
a result, procedures are presented to estimate humid- ity and radiation from maximumminimum tempera-
ture data and to adopt global estimates for wind speed. The availability of world wide climatic databases fur-
ther facilitates the adoption of values from nearby sta- tions. Such procedures have proved to perform better
than any of the alternative empirical formulas and will largely improve transparency of calculated evapotran-
spiration values Smith et al., 1996. 3.4. Crop evapotranspiration
Procedures for estimating crop evapotranspiration have been well established in FAO Irrigation and
Drainage Paper No. 24 Doorenbos and Pruitt, 1977, using a series of recommended crop coefficient values
K
c
to determine ET
crop
from reference evapotranspi- ration ET
o
, as follows: ET
c
= K
c
ET
o
Although the adoption of the Penman–Monteith approach would allow a direct estimate of crop evapo-
transpiration by introducing the appropriate crop pa- rameters, research on crop canopy and aerodynamic
resistance, for different crops, has so far not been suffi- ciently conclusive to allow the development of reliable
global valid parameters. The crop coefficient approach is therefore maintained which integrates all those ef-
fects that distinguish the concerned crop in the differ- ent growth stages from the standard crop reference,
including soil evaporation.
A review of the crop coefficients has resulted in an update of K
c
values to the FAO Penman–Monteith method and procedures to reach better estimates under
various climatic conditions and crop height and ex- panding the range of crops and crop types Allen et al.,
1996. The procedure for adjusting crop-coefficients for non-standard climatic conditions is given below:
K
c mid
= K
c midTab
+ [0.04u
2
− 2
− 0.004RH
min
− 45]
h 3
0.3
with the tabulated standard K
c
values K
cmidTab
ad- justed for deviating wind u2 ms, humidity RH
min
45 and crop height h0.33 m. A dual crop coefficient approach is introduced, which
allows to estimate separately soil evaporation K
e
and basal crop transpiration K
cb
as illustrated in see Fig. 3. More accurate estimates of crop evapotranspiration
can thus be made as appropriate for daily calculations. Procedures for estimating soil evaporation and crop
evaporation under conditions of soil-moisture stress and salinity, double cropping and various crop, soil and
M. Smith Agricultural and Forest Meteorology 103 2000 99–108 105
Fig. 3. Dual crop coefficient curve showing the basal K
cb
thick line, soil evaporation K
e
thin line and the corresponding single crop coefficient curve K
c
= K
cb
+ K
e
curve dashed line.
water management practices are included. Further de- tails on the revised K
c
values and length of growing stages, are included in the new FAO publication Allen
et al., 1998.
4. Computerized crop water use simulations
