CHAPTER 4: EVAPORATION, EVAPOTRANSPIRATION AND SOIL MOISTURE
4.1 EVAPORATION, EVAPOTRANSPIRATION
Evaporation and transpiration are the primary
abstractions of the hydrological cycle. These abstractions are small during a runoff event and can be
neglected. The bulk of evaporation and transpiration takes place during the time between runoff
events, which is usually long. Hence, these abstractions are the most important during this time
interval. The combined effect of evaporation and
transpiration is called evapotranspiration. Over
large land areas in temperate zones, about two thirds of the annual precipitation is evapotranspired and the remaining one third runs off in
streams and through the groundwater to the oceans.
In arid regions, evapotranspiration may be even
more significant, returning up to 90 per cent or
more of the annual precipitation to the atmosphere.
Evaporation also links hydrology to atmospheric
science and, through transpiration, to agricultural
The process by which water is changed from
the liquid or solid state into the gaseous state
through the transfer of heat energy is known as
In the hydrological cycle evaporation is an important process, so much so that on a continental basis,
approximately 70 to 75 per cent of the total annual
precipitation is returned to the atmosphere by
evaporation and transpiration. In hot climates, the
loss of water by evaporation from rivers, canals and
open-water storage equipment is a vital matter as
evaporation takes a significant proportion of all
water supplies. It is significant in the sense that
most of the water withdrawn for beneficial uses
ultimately returns to streams and aquifers and
becomes available for reuse, while the loss of water
due to evaporation is entirely lost from the usable
supply. Even in humid areas, evaporation loss is
significant although the cumulative precipitation
tends to mask it so that it is ordinarily not recognized except during rainless periods.
Storage reservoirs expose wide surfaces to evaporation and thus are a major source of water loss
even though they may lessen natural evaporation
by confining floods in deep storages instead of
spreading over wide flood plains.
The factors controlling evaporation have been
known for a long time, but evaluating them is diffi –
cult because of their interdependent effects.
However, in general, evaporation is affected by
temperature, wind, atmospheric pressure, humidity, water quality, water depth, soil type and nature,
and shape of surface.
Transpiration is defined as a natural plant physiological process whereby water is taken from the soil
moisture storage by roots and passes through the
plant structure and is evaporated from cells in the
leaf called stomata.
The amount of water held in storage by a plant is
less than 1 per cent of that lost by it during the
growing season. From the hydrological standpoint, therefore, plants are like pumps that
remove water from the ground and raise it to the
It is difficult to make precise estimates of the
water transpired because of the many variables
responsible for the process. Available estimates
should be used with due caution taking into
consideration the conditions under which these
estimates were obtained. Adequate relationships between climatic factors and transpiration
are prerequisites if the data derived in one
climatic region are supposed to have general
Transpiration is affected by physiological and
environmental factors. Stomata tend to open and
close in response to environmental conditions
such as light and dark, and heat and cold.
Environmental factors that affect transpiration
are essentially the same as for evaporation, but
can be considered a bit differently. For practical
purposes, vapour pressure gradient, temperature,
solar radiation, wind and available soil moisture
are the most important factors affecting
The term evapotranspiration (ET) is defi ned as the
water vapour produced from the watershed as a
result of the growth of plants in the watershed.
Evapotranspiration and consumptive use include
both the transpiration by vegetation and evaporation from free surfaces, soil, snow, ice and vegetation.
Here it will be important to give the difference
between evapotranspiration and consumptive use.
Consumptive use differs from evapotranspiration
only in that it includes the water used to make plant
tissues (Singh, 1994). In computing evapotranspiration both transpiration and soil evaporation are
included. The actual evapotranspiration can be
determined by the analysis of the concurrent record
of rainfall and runoff from a watershed.
There is an important difference between evapotranspiration and free surface evaporation. Transpiration
is associated with plant growth and hence
evapotranspiration occurs only when the plant is
growing, resulting thereby in diurnal and seasonal
variations. Transpiration thus superimposes these
variations on the normal annual free water-surface
The potential evapotranspiration (PET) is defi ned as
the evapotranspiration that would result when
there is always an adequate water supply available
to a fully vegetated surface.
This term implies an ideal water supply to the
plants. In case water supply to the plant is less than
PET, the deficient would be drawn from the soilmoisture storage until about 50 per cent of the
available supply is utilized. With further moisture
deficiency, the actual evapotranspiration (AET) will
become less than PET until the wilting point is
reached, and when the evapotranspiration stops.
Interception is that portion of the precipitation
that, while falling on the Earth’s surface, may be
stored or collected by vegetal cover and subsequently evaporated. The volume of water thus lost
is called interception loss.
In studies of major storm events and fl oods the
interception loss is generally neglected. However, it
may be a very signifi cant factor in water balance
studies. Precipitation falling on vegetation may be
retained on leaves or blades of grass, fl ow down the
stem of plants and become stem fl ow or fall off the
leaves to become part of the throughfall. The
amount of water intercepted is a function of (a) the
storm character, (b) the species, age and density of
plants and trees and (c) the season of the year.
Usually about 10 to 20 per cent of the precipitation
falling during the growing season is intercepted and
returned to the hydrological cycle through
evaporation. Under very dense forest conditions, it
may be even as high as 25 per cent of the total
precipitation. In temperate regions, evaporation of
water intercepted by the vegetation represents an
important part of the evapotranspiration. There is a
wide variety of techniques used to measure rain
interception (water stored in the canopy), canopyinterception-storage capacity, time of leaf wetness,
throughfall, canopy evapotranspiration, and
interception evaporation (often, but less
appropriately, called interception loss). Reviews of
interception measurement and leaf wetness
methods are given by, for example, Bouten and
others (1991) and Lundberg (1993), whereas
canopy-storage-capacity measurements are
summarized by Klaassen and others (1998).
Micrometeorological evaporation methods are
described by, for example, Garratt (1984) and
4.1.3 Measurement of evaporation
For a general reference on measurement instruments, see the Guide to Meteorological Instruments
and Methods of Observation (WMO-No. 8).
220.127.116.11 Direct methods
Reasonably accurate methods of measurement of
evaporation and evapotranspiration are available
from pans and small bodies of water and soil, but
direct measurement of evaporation or evapotranspiration from large water or land surfaces is not
possible at present. However, several indirect methods have been developed that give acceptable
results. Evaporation pans and lysimeters are used
in networks for this purpose, and are discussed in
this chapter. For existing reservoirs and plots or
small catchments, estimates can be made by waterbudget, energy-budget, and aerodynamic
approaches and other available methods. These
latter techniques are discussed in this chapter only
from the point of view of instruments and observational requirements.
Computation of evaporation
and evapotranspiration from water and land
surfaces by the various indirect methods is also
discussed separately in this chapter. Some of the
direct methods are as follows.
For estimation of evaporation from open water
bodies, evaporation records of pans are generally
used. The pans could be either square or circular
section, mounted entirely above the ground or sunk
in the ground so that the water level is approximately that of the ground. They may be mounted
on anchored floating platforms on lakes or other
Three types of pans deserve special mention: the
United States Class A pan (Figure I.4.1), the
GGI-3000 pan (Figure I.4.2) and the 20-m2 tank of
the Russian Federation. The United States Class A
pan has been recommended by WMO and the
International Association of Hydrological Sciences
as a reference instrument as its performance has
been studied under a range of climatic conditions
within wide limits of latitude and elevation. The
GGI-3000 pan and 20-m2 tank are used in the
Russian Federation and some other countries with
different climatic conditions, as they possess reliable operational qualities and an extremely stable
relationship with the meteorological elements that
infl uence evaporation. WMO sponsored comparative observations (WMO, 1976) of the Class A pan,
the GGI-3000 pan and the 20-m2 tank in several
countries, which eventually led to some operational
recommendations on the suitability of these pans
in diverse climatic and physiographic conditions.
In addition to the pan, a number of other instruments, such as integrating anemographs or
anemometers, non-recording precipitation gauges,
thermometers or thermographs for pan water
temperature, maximum and minimum thermometers or thermographs for air temperature or hygro-thermographs or psychrometers, are also
When installing evaporation pans it is important to
ensure that the site of the pan is reasonably level
and free of obstruction. At sites where normal
climate and soil do not permit the maintenance of
a soil cover, the ground cover should be maintained
as near as possible to the natural cover common in
the area. Obstructions such as trees, buildings,
shrubs or instrument shelters should not be closer
than four times the height of the object above the
pan. Under no circumstance should the pan or
instrument shelter be placed on a concrete slab or
pedestal, or over asphalt or gravel.
The instruments should be located on the evaporation station plot so as to prevent them from casting
shadows over the pan. The minimum size of the
plot should be 15 m x 20 m. The plot should be
fenced to protect the instruments and to prevent
animals from drinking the water. The fence should
be constructed so that it does not affect the wind
structure over the pan. At unoccupied sites, particularly in arid and tropical regions, it is often necessary
to protect the pans from birds and small animals by
using chemical repellants and a wire mesh. To estimate the error introduced by the wire-mesh screen
on the wind fi eld and thermal characteristics of the
pan, readings from the protected pan should be
compared with those of a standard pan at the nearest comparable occupied site.
The water level in the pan must be measured accurately before and after water is added.
This may be done in two ways:
(a) The water level may be determined by means
of a hook gauge consisting of a movable scale
and vernier fitted with a hook enclosed in a
still-water chamber in the pan. An alternative arrangement is to use a float. A calibrated container is used to add or remove water at
each observation so as to maintain the water
level to a pre-specified point;
(b) The water level may be determined by the
(i) A vessel of small diameter fitted with a
valve is placed on top of a benchmark
below the water surface in the pan;
(ii) The valve is opened and the water level in
the vessel is allowed to equalize with the
water level in the pan;
(iii) The valve is closed and the volume of
water in the vessel is determined accurately in a measuring tube.
The height of the water level above the
benchmark is determined from the volume
of water in the vessel and the dimensions of the
Daily evaporation is computed as the difference in
water level in the pan on successive days, corrected
for any precipitation during the period. The
amount of evaporation that has occurred between
two observations of water level in the pan is determined by:
E = P ± Δd (4.1)
where P is the depth of precipitation during the
period between the two measurements, and Δd is
the depth of water added (+) to or removed (–) from
Several types of automatic evaporation pans are
in use. The water level in the pan is kept automatically constant by releasing water into the
pan from a storage tank or by removing water
from the pan in the case of precipitation. The
amount of water added to or removed from the
pan is recorded.
The major difficulty in using a Class A pan for the
direct measurement of evaporation arises
because of the use of coefficients to convert the
measurements from a small tank to large bodies
of open water. Fuzzy logic as suggested by Keskin
and others (2004) can provide an alternative to
the classical evaporation estimation.
Evaporimeters made of polyethylene or colourless
plastic are used in many countries for measuring
evaporation from, or condensation on, snow cover.
Snow evaporimeters should have an area of at least
200 cm2 and a depth of 10 cm.
A sample of snow is cut to fi ll the evaporimeter,
the total weight is measured and the evaporimeter is set fl ush with the snow surface. Care should
be taken that surface characteristics of the sample
in the evaporimeter are similar to those of the
snow cover in which it is placed. At the end of
the measurement period, the evaporimeter is
removed from the snow cover, the outside is
wiped dry and a second measurement of weight is
made. The difference between initial and final
weights is converted to evaporation or condensation in centimetres. Measurements during periods
of snowfall or blowing snow are not valid. During
melt, the evaporimeters should be weighed and
new samples should be cut at more frequent
intervals as the snow cover will be settling, exposing the edge of the evaporimeter and altering air
fl ow over the sample.
18.104.22.168 Indirect methods
Because of problems encountered in making direct
measurements of evaporation from lakes and reservoirs, a number of indirect methods, such as the
water-budget, the energy-budget, the aerodynamic
approach or combination of these, are frequently
used. The meteorological elements incorporated
into these methods are solar and long-wave radiation, air and water-surface temperatures,
atmospheric humidity or vapour pressure, and
wind. Instruments and observational procedures
for measuring these elements are described in the
following subsections. The manner in which observations of the above elements are used in various
indirect methods for estimating evaporation is
described below in this chapter.
Incident total solar (short-wave) radiation should
be measured at a site near the reservoir with a
pyranometer, and the output should be recorded
continuously. Incoming short-wave radiation on
a horizontal surface is measured with a pyranometer. Most modern types of pyranometers are
based on multi-junction thermopiles and are
covered by single or double glass domes that
allow only radiation in the 0.3–3 µm range to
reach the sensitive pyranometer surface
(Figure I.4.3). Some types of pyranometer have
the entire surface blackened with half the thermojunctions attached to it, with the other junctions
located so that they sense the slowly varying
reference temperature of a large, shielded brass
block. Other types have a sensitive surface that
consists of white and black painted surfaces, with
thermojunctions attached to both.
Long-wave radiation is measured indirectly with
fl at-plate radiometers. These instruments are not
selective in response to different wavelengths and
thus measure all wavelengths. The long-wave radiation is computed as the difference between the total
radiation received from sun and sky as observed
with a radiometer; the solar radiation is measured
with a pyranometer at the same site.
One type of long-wave radiometer consists of a
flat 5-cm2 plate mounted horizontally in the
exhaust of a small blower. The plate is a sandwich
with a blackened aluminium upper surface and a
polished aluminium lower surface. A thermopile
measures the vertical temperature gradient across
an insulating sheet that forms the centre layer of
the sandwich. The thermopile voltage is
proportional to the heat fl ow down through the
plate, which in turn is proportional to the energy
received at the blackened surface after deduction
of the black-body radiation. To correct for the
black-body radiation, a separate thermocouple
measures the black-surface temperature. The
function of the blower exhaust is to minimize the
effects of wind on the calibration coefficient of
Another type of instrument, a net pyrradiometer,
measures the difference between total (short-wave
and long-wave) incoming (downward) and outgoing (upward) radiation. The instrument consists
of a horizontally mounted plate with two blackened surfaces. Half the junctions of a thermophile
are attached to the upper surface and the others are
attached to the lower surface, so that the thermopile output is proportional to net radiation in the
0.3–100 µm band. These instruments are divided
into two types: those that are ventilated and those
that are shielded to reduce convective heat transfer
from the sensing element. Instruments should be
mounted at least 1 m above representative vegetation cover.
Air temperature should be measured 2 m above
the water surface near the centre of the reservoir. For small reservoirs, the air temperature
may not be greatly modified in its passage
across the water surface, in which case satisfactory measurements can be made at an upwind
Although observations of air temperature at intervals of one, four or six hours may be satisfactory,
continuous records are desirable, especially in
connection with humidity measurements. Electrical
thermographs, utilizing thermocouple thermometers, are suitable for recording on the
multichannel recording potentiometers used for
In measuring air temperature, thermometers must
be shaded from the sun without restricting natural
ventilation. Special radiation shields have been
designed for thermocouple thermometers.
Measurements of air temperature should be accurate to within ±0.3°C.
Several types of thermometers, such as mercury-inglass or mercury-in-steel (including maximum and
minimum and reversing thermometer), platinumresistance or thermistor elements with electronic
circuit and meter or recorder and thermocouple
thermometers, with voltmeter, with or without
recorder, are used for the measurement of water
Particular applications will determine which thermometer is most suitable. For example, direct
observations are best carried out with a mercury-inglass thermometer, whereas continuous records
may be obtained with resistance or thermocouple
Thermographs, which produce a continuous
record of temperature, usually comprise a mercuryin-steel sensing element immersed in the water,
which is connected to a circular or cylindrical
chart recorder with a Bourdon-tube transducer.
Care should be taken in the installation of thermographs to ensure that measurements taken are
representative of the water temperature (Herschy,
In the case of automatic stations where the measurement, which will usually include other variables,
is recorded on a magnetic tape or transmitted over
direct wire or radio-telemetry systems, the platinum-resistance or thermistor thermometers are
used most frequently. As these have no moving
parts, they are more reliable and offer greater accuracy and sensitivity of measurement. The sensing
element is usually connected to a Wheatstonebridge circuit and an electronic amplifi er to produce
an output signal that is suitable for recording or
In general, the precision required for the measurement of water temperature is ±0.1°C, except
for special purposes where a greater accuracy may
be required. However, in many circumstances
precision of observation of ±0.5°C is adequate
and there are many instances where statistical
temperature data are quoted to the nearest 1°C.
Thus, it is important to specify the operational
requirement so that the most suitable thermometer is selected.
Humidity or vapour pressure of the air
Humidity measurements are made at the same location as air temperature. Psychrometers utilizing
thermocouple thermometers are best suited for
recording purposes. The thermocouple thermometers described in the preceding section on Air
temperature, with an additional thermocouple
thermometer to record wet-bulb temperatures, will
give adequate results. Wet-bulb thermocouples
require a wick and a reservoir that should be so
arranged that the water will arrive at the wet-bulb
temperature. Wet-bulb thermometers must be
shielded from radiation and must, at the same time,
maintain adequate ventilation to obtain a true wetbulb temperature. A shield similar to the one used
for air temperatures will provide adequate ventilation if wind speeds are greater than 0.5 ms–1. In
practice, the shield for the wet-bulb thermometer is
placed just below the air temperature shield.
If measurements of dry- and wet-bulb temperatures
are made to within ±0.3°C, the relative humidity
should be within ±7 per cent for moderate temperatures. This is adequate for determining vapour
Wind speed should be measured near the centre of
the lake or reservoir at a height of 2 m above the
water surface. In practice, an anchored raft is used
to support the instrumentation.
Any type of standard anemometer suitable for
remote indication or recording should be adequate
to determine the average daily wind speed. The
three-cup rotor fan anemometers are most suited
for remote recording. Accuracy of wind measurements by the three-cup or fan anemometers is
usually within ±0.5 m s–1, which is considered
acceptable for evaporation measurements.
If a totalizing anemometer is used, provision must
be made to read the counter at fi xed intervals (preferably daily). If an electrical-contact anemometer is
used, a recorder must be provided. This can be done
by an electrical event marker on the margin of the
4.1.4 Measurement of
Soil evaporimeters and lysimeters
Evapotranspiration can be estimated by the use of
soil evaporimeters and lysimeters, by the waterbudget or heat-budget methods, by the
turbulent-diffusion method, or by various empirical formulae based on meteorological data. Use of
soil evaporimeters and lysimeters allows direct
measurement of evapotranspiration from different
land surfaces and evaporation from the soil between
cultivated plants. These instruments are simple and
accurate if all requirements concerning their installation and observational techniques are fulfilled.
Transpiration of vegetation is estimated as the
difference between measured evapotranspiration
and contemporaneously measured evaporation
from the soil.
Soil evaporimeters and lysimeters are categorized
according to their method of operation:
(a) Weight based, which use mechanical scales to
account for changes in water content;
(b) Hydraulic based, which use the hydrostatic
principle of weighing;
(c) Volumetric based, in which water content
is kept constant and evapotranspiration is
measured by the amount of water added or
There is no single standard instrument for measuring evapotranspiration.
General requirements for the location of evaporation plots are as follows:
(a) The site selected for the plot should be typical of
the surrounding area with respect to irrigation,
soil characteristics (texture, layering, genetical
type), slope and vegetative cover;
(b) The evaporation plot should be located beyond
the zone of influence of individual buildings
and trees. It should be situated at a distance
not less than 100 to 150 m from the boundaries of the field and not more than 3 to 4 km from the meteorological station. Soil monoliths
for inclusion in evaporimeters and lysimeters
should be taken from within a radius of 50 m
of the plot, and the soil and vegetative cover
of the monolith should correspond to those of
4.1.5 Remote-sensing measurements
of evaporation and
Remote-sensing observations combined with ancillary meteorological data have been used in obtaining
indirect estimates of ET over a range of temporal
and spatial scales (Schulz and Engman, 2000).
Recently there has been a lot of progress in the
remote-sensing of parameters, including:
(a) Incoming solar radiation;
(b) Surface albedo;
(c) Vegetative cover;
(d) Surface temperature;
(e) Surface soil moisture.
Remote-sensing of evaporation variables
Measurements of radiation and air temperature are
usually made at the same locations, either at the
centre of the lake or reservoir or at an upwind shore
station. This permits recording several items in
sequence on a single multichannel recorder.
Integrating devices are sometimes used with stripchart recorders.
These devices present a visual
readout of the average value of each item for the
time period for which evaporation is to be computed
(usually 10 days or two weeks).
Remote-sensing of several important parameters
used to estimate evaporation is made by measuring
the electromagnetic radiation in a particular waveband reflected or emitted from the Earth’s surface.
The incoming solar radiation can be estimated from
satellite observations of cloud cover primarily from
geosynchronous orbits using Multispectral Scanner
(MSS) in the visible, near-infrared and thermal
infra-red parts of EMS (Brakke and Kanemasu, 1981;
Tarpley, 1979; Gautier and others, 1980). The
surface albedo may be estimated for clear-sky conditions from measurements covering the entire visible
and near-infra-red waveband (Jackson, 1985; Brest
and Goward, 1987). The surface temperature may
be estimated from MSS measurements at thermal IR
wavelengths of the emitted radiant fl ux (Engman
and Gurney, 1991).
However, there has been little progress in the direct
remote-sensing of the atmospheric parameters that
affect ET, such as:
(a) Near-surface air temperature;
(b) Near-surface water vapour gradients;
(c) Near-surface winds.
Furthermore, remote-sensing has a potentially
important role because of its areal coverage in the
spatial extrapolation process of ET.
Remote-sensing of evapotranspiration variables
Recently, researchers have begun using satellite data
(for example, Bastiaanssen and others, 1998;
Choudhury, 1997; Granger, 1997) to estimate
regional actual evapotranspiration. Remote-sensing
of several important parameters used to estimate ET
is made by measuring the electromagnetic radiation in a particular waveband reflected or emitted
from the Earth’s surface. Estimates of incoming
solar radiation, surface albedo and surface temperature may be done by the same satellite measurements
described in 4.1.3. The soil moisture may be estimated using the measurement of microwave
properties of the soil (microwave emission and
refl ection or backscatter from soil). However, there
are uncertainties in such soil moisture estimates
due to previously mentioned factors such as surface
roughness and vegetative cover.
The most practical remote-sensing approach for the
future will include repetitive observations at the
visible, near and thermal infra-red, and microwave
lengths. Components for determining the sensible
heat fl ux will be measured by the EOS instruments.
The latent heat fl ux cannot be measured directly
but EOS instruments will provide some sampling
capability. Furthermore, the future programme such
as EOS should provide the necessary data for evaluating ET on local, regional and global scales.
4.2 ESTIMATING EVAPORATION FROM
4.2.1 General [HOMS I45]
Evaporation from water surfaces can be determined
by various methods, such as:
(a) Water budget;
(b) Energy budget;
(c) Mass transfer methods;
I.4-8 GUIDE TO HYDROLOGICAL PRACTICES
(d) Combination methods;
(e) Empirical formulae.
Any of the methods described can be employed to
determine evaporation. Usually, instrumentation
for energy-budget and mass-transfer methods is
quite expensive and the cost to maintain observations is substantial.
For these reasons, the water-budget method and use of evaporation pans are more common. The pan method is the least
expensive and will frequently provide good estimates of annual evaporation. Any approach
selected is dependent, however, on the degree of
accuracy required. As the ability to evaluate the
parameters in the water budget and energy budget
improves, so also will be resulting estimates of
4.2.2 Water budget
The method is based on the continuity equation
and can be utilized for the purpose of computing
E = I – O – ΔS (4.2)
where E = evaporation, I = inflow, O = outflow and
ΔS = change in storage.
By adding the suffixes s and g to the various components in equation 4.2 to denote vectors originating
above and below ground surface respectively, the
equation can be expressed as:
Es = P + R1 – R2 – Rg – Ts – F – ΔSs
where Es = reservoir evaporation, P = precipitation,
R1 = surface runoff coming into the reservoir,
R2 = surface runoff going out of the reservoir,
Rg = groundwater inflow, Ts = transpiration loss,
F = infiltration (or seepage) and ΔSs = change in
If the net transfer of seepage (Rg – F) = Os
transpiration term Ts
equals zero, then equation 4.3
can be rewritten:
Es = P + R1 – R2 + Os – ΔSs
All the terms are in volumetric units for a time
period of interest that should be not less than a
week. The water-budget method, although having
the obvious advantage of being simple in theory,
has the disadvantage in that the errors in the measurement of the parameters used in equation 4.4 are
reflected directly in the computed amounts of
evaporation. Therefore, it is not recommended that
the method be applied to time periods of less than
a month if the estimate of evaporation is expected
to be within ±5 per cent of the actual amount.
Probably the most difficult term to evaluate is the
seepage, F. This component can be estimated knowing the hydraulic conductivity of the lake bed and
the hydraulic gradient. Nevertheless, it should be
recognized that the water-budget method of determining evaporation will prove most successful when applied to relatively impervious lakes inwhich the seepage is negligible in comparison with
the amount of evaporation.
To evaluate ΔSs
, an accurate area-capacity curve for
the lake should be available. Even with these data,
the bank storage component can introduce an error
in the water budget. However, if the bank storage
component is neglected, the water budget would
not be useful on an annual cycle.
Although it is theoretically possible to use the
water-budget method for the estimation of evaporation from any free surface, it is usually impractical
to do so because of the effects of errors in measuring various parameters. Evaporation, estimated by
this method, is residual and, therefore, may be
subject to considerable error if it is small relative to
In summary, the method is difficult and inaccurate
under most conditions, particularly for short averaging time periods. Some of the most difficult
parameters to measure are change in storage, seepage, groundwater flow and advected flows.
4.2.3 Energy budget
The energy-budget method illustrates an application of the continuity equation written in terms of
energy. It has been employed to compute the evaporation from oceans and lakes, for example, at
Elephant Butte Reservoir in New Mexico (Gunaji,
1968). The equation accounts for incoming and
outgoing energy balanced by the amount of energy
stored in the system. The accuracy of estimates of
evaporation using the energy budget is highly
dependent on the reliability and preciseness of
measurement data. Under good conditions, average
errors of perhaps 10 per cent for summer periods
and 20 per cent for winter months can be
The energy-budget equation for a lake may be written as (Viessman and others, 1989):
where Q0 = increase in stored energy by the water, Qs
= solar radiation incident at the water surface,
Qr = reflected solar radiation, Qa = incoming longwave radiation from the atmosphere, Qar = refl ected
long-wave radiation, Qbs = long-wave radiation
emitted by the water, Qv = net energy advected (net
energy content of incoming and outgoing water)
into the water body, Qe
= energy used in evaporation, Qh = energy conducted from water mass as sensible heat and Qw = energy advected by evaporated water.
All the terms in equation 4.5 are in watt per square
metre per day (W m–2day). Heating brought about
by chemical changes and biological processes is
neglected, as it is the energy transfer that occurs at
the water–ground interface. The transformation of
kinetic energy into thermal energy is also excluded.
These factors are usually very small, in a quantitative sense, when compared with other terms in the
budget if large reservoirs are considered. As a result,
their omission has little effect on the reliability of
Each of the various terms in the energy-budget
equation is either measured directly or computed
from known relationships. The procedure used in
evaluating each term is described below.
The terms of equation 4.5 that can be measured are
and Qa, and the net radiation balance is:
Rf = Qs – Qsr + Qa – Qar – Qbs (4.6)
All of the above values are expressed in W m–2.
Detailed descriptions of the instruments and measuring techniques concerning the above-mentioned
elements can be found in 4.1.3, 4.1.4 and 4.1.5, or
in the Guide to Meteorological Instruments and Methods
of Observation (WMO-No. 8).
Reflected long-wave radiation (Qar) may be taken as
3 per cent of the long-wave radiation received by
the water surface.
Long-wave radiation emitted by the water (Qbs) is
computed according to the Stefan–Boltzmann law
for black-body radiation, with an emissivity factor
of 0.970 for water. The equation for computing
radiation emitted by the water surface is:
Qbs = 0.97σθ4 (4.7) where Qbs is the radiation emitted by the water
surface in W m–2, σ is the Stefan-Boltzmann constant
(5.67 x 10–8 W m–2 °K–4), and θ is the temperature of
the water surface in °K. For computing purposes,
the average temperature of the water surface, as
recorded near the centre of the reservoir, is determined for each period of study. The temperature is
converted to °K, and the average radiation emitted
by the water surface is computed for the period of
study in W m–2.
The thermal energy of the volume of water in the
reservoir for a given date is computed from a
temperature survey made on that date. These
temperature measurements, which should be accurate to within 0.1°C, are usually made at biweekly
or monthly intervals. The reservoir may be divided
into several layers from the surface to the bottom.
The volume of water for each of the layers is determined from the stage–volume relationship. All
temperature observations made in a particular layer
are averaged to obtain a mean temperature for that
volume of water.
The summation of the products of volume and
temperature (assuming a base temperature of 0°C)
will give the total energy for that particular date.
Density and specific heat are considered as unity for
the range of temperatures that occur in the reservoir. In order to determine the energy utilized in
evaporation, Qe changes in energy storage resulting
from advection of energy in the volumes of water
entering or leaving the reservoir must be evaluated.
Again, a base temperature of 0°C is usually chosen
in computing the amount of energy in these
volumes. Their temperatures are determined by
observation or recordings (4.1.3) depending on the
variation of temperature with the rate of fl ow. If the
temperature of the water changes with the rate of
fl ow, the mean temperature of the volume should
be weighted according to the rate of fl ow. The
temperatures of bank storage and net seepage are
considered as being equal to the mean annual air
temperature. This assumption is admittedly subject
to error, but is not considered serious if the surface
infl ow is a large item in the water budget.
If precipitation is a significant item in the water
budget, then the energy of this volume of water
must be taken into account. The temperature of
rainfall is assumed to be that of the wet bulb at the
time of rainfall. In computing the energy for each
of these volumes, centimetre-gram-second units are
used, and density and specifi c heat are considered
as unity for the range of temperatures that occur in
these volumes. The product of temperature times
volume will give the amount of energy for each
volume in joules (net energy advected, Qv). The
difference between the computed energies of
stored water for the thermal surveys made at the
I.4-10 GUIDE TO HYDROLOGICAL PRACTICES
beginning and end of the period of study determines the change in energy storage (Q0).
During winter months when ice cover is partial
or complete, the energy budget only occasionally
yields adequate results because it is diffi cult to
measure reflected solar radiation, ice surface
temperature and the areal extent of the ice cover.
Daily evaporation estimates based on the energy
budget are not feasible in most cases because
reliable determination of changes in stored energy
for such short periods is impractical. Periods of
one week or longer are more likely to provide
In using the energy-budget approach, it has been
demonstrated that the required accuracy of measurement is not the same for all variables. For example, errors in measurement of incoming longwave radiation as small as 2 per cent can introduce
errors of 3–15 per cent in estimates of monthly
evaporation, while errors of the order of 10 per
cent in measurements of refl ected solar energy
may cause errors of only 1–5 per cent in calculated
monthly evaporation. To permit the determination of evaporation by equation 4.5, it is common
to use the following relation:
where B is known as Bowen’s ratio (Bowen, 1926)
where cp = the specifi c heat of water (cal/g°C) that is
equal to 4186.8 J/kg°C, Te
= the temperature of
evaporated water (°C) ; Tb= the temperature of an
arbitrary datum usually taken as 0°C and L = the
latent heat of vaporization (cal/g) that is equal to
2260 kJ/kg. Introducing these expressions in equation 4.5 and solving for Qe we obtain: Q e = Q s − Q r + Q a − Q ar − Q bs − Q o + Q v
1 + B + c p (Te − Tb ) / L (4.10)
To determine the depth of water evaporated
per unit time, the following expression may be
where E = evaporation (m sec–1) and ρ= the mass
density of evaporated water (kg m–3).
The energy-budget equation thus becomes:
The Bowen ratio can be computed using:
where p = the atmospheric pressure (mb), To = the
water-surface temperature (°C); Ta = the air temperature (°C), eo = the saturation vapour pressure at the
water-surface temperature (mb) and ea = the vapour
pressure of the air (mb).
This expression circumvents the problem of evaluating the sensible heat term, which does not lend
itself to direct measurement.
Remote-sensing of several important parameters
used to estimate evaporation is made by
measuring the electromagnetic radiation in a particular waveband refl ected or emitted from the Earth’s
surface as discussed earlier in 4.1.3.
Applicability of energy-budget approach
The points summarized below should be recognized
fi rst in order to apply the energy-budget approach
for estimating the evaporation from free surfaces:
(a) The fl ow of heat from the bottom of the lake
has not been accounted for. This, however, is
important in the case of shallow lakes;
(b) Bowen’s ratio is assumed to provide a suffi –
ciently accurate estimate of Qh;
(c) The approach neglects the effect due to radiative diffusivity, stability of the air and spray;
(d) The applicability of the approach hinges greatly
on the ability to evaluate the advective energy
4.2.4 Mass-transfer method
The mass-transfer approach, as the name implies, is
based on the determination of the mass of water
vapour transferred from the water surface to the
atmosphere. To better understand this, an insight
into the physics of air movement is first discussed.
When air passes over land or water surfaces, the air
thickness in the lower atmosphere may be divided
into three layers: (a) the laminar layer near the
surface; (b) the turbulent layer; and (c) the outer
layer of frictional influence. The laminar layer, in
which the airflow is laminar, is only of the
order of a millimetre in thickness.
In this layer the temperature, humidity and wind velocity vary
almost linearly with height, and the transfer of
heat, water vapour and momentum are essentially
molecular processes. The overriding turbulent layer
can be several metres in thickness depending on
the level of turbulence. In this layer, temperature,
humidity and wind velocity vary approximately
linearly with the logarithm of height, and the transfer of heat, vapour and momentum through this layer are turbulent processes.
The mass-transfer approach is based on Dalton’s
aerodynamic law giving the relationship between
evaporation and vapour pressure as:
E = k (es – ea) (4.14)
where E = direct evaporation, k = a coefficient and
depending on the wind velocity, atmospheric pressure and other factors, es and ea = saturation vapour pressure corresponding to the water-surface temperature and the vapour pressure of the air, respectively.
Mean daily temperature and relative humidity may
be used in determining mean vapour pressure ea
and mean saturation defi cit (es
Equation 4.14 was originally proposed by Harbeck and Meyers
4.2.5 Combination of aerodynamic
and energy-balance methods
Perhaps the most widely used method for computing lake evaporation from meteorological factors is
based on a combination of aerodynamic and
where Ei is the estimated evaporation from a freewater surface, Δ = es – esz Ts – Tz
is the slope of the saturation vapour-pressure curve at any temperature θa, which is tabulated as γ/Δ versus Tz in Brutsaert (1982,
Figure 10.2), Rn is the net radiation, γ is the constant
in the wet and dry bulb psychrometer equation,
and Ea is the same expressed in equation 4.14.
The psychrometer constant γ for °C is the same
constant of the Bowen ratio, and its value at 1000-mb
pressure is 0.61.
The net radiation Rn (in MJ m–2 day) can be estimated by the following equation: where n/N is the ratio of actual to possible hours of sunshine, S0 is the extraterrestrial radiation (in
MJ m–2 day), ed is the actual vapour pressure of the
air in mm of mercury, σ is the Stefan–Boltzmann
constant, also expressed in equivalent evaporation
in mm day–1, and T is the mean air temperature
(absolute) expressed in degrees Kelvin.
Although it may be necessary to use the above
equation, it would be preferable to use measured
values of solar and long-wave radiation.
A similar approach was used by Kohler and others
(1959) and a graphical presentation of the relationship is shown in Figure I.4.4. The meteorological
observations of solar radiation, air temperature,
dewpoint and wind movement at the anemometer
height of a Class A pan are required for application
of this technique. In the absence of solar-radiation
observations, radiation may be estimated from the
percentage of possible sunshine or cloud-cover data.
Lake evaporation computed for short periods by
this method would be applicable only to very shallow lakes with little or no advection of energy to
the lake. For deep lakes and conditions of signifi –
cant advection due to inflow and outflow, it is
necessary to correct the computed lake evaporation
for net advected energy and change in energy storage. These terms are described under the
energy-budget method in 4.2.3. However, all of the
advected energy and change in energy storage is
not utilized for evaporation. The portion of this
energy used for evaporation can be obtained from a
relationship such as shown in Figure I.4.5.
Observations of water-surface temperature and
wind movement at 4 m above the water surface are
required for application of this relationship.
I.4-12 GUIDE TO HYDROLOGICAL PRACTICES
evaluation is made of the energy-advection and
4.2.6 Extrapolation from panmeasurements [HOMS C46]
The evaporation from pans exposed in or on the
ground is influenced by the characteristics of the
pan. Sunken pans are subject to undetected leaks,
accumulation of debris on the water surface, and
boundary conditions with the soil different from
those of a large lake. Pans exposed above the ground
are subject to heat exchange through the sides and
to other effects that do not occur in lakes. Floating
pans are subject to splash-in and splash-out, and
are costly to install and operate.
Pans have much less heat storage than lakes and
tend to experience a different annual cycle of
evaporation, with pan-evaporation extremes
occurring earlier in the season. Reliable estimates
of annual lake evaporation can be obtained by
multiplying the annual pan evaporation by the
appropriate pan-to-lake coefficient.
will be reliable only if it can be assumed that, on
an annual basis, any energy advected to the lake is
balanced by a change in heat storage. The pantolake coefficient for a particular pan is determined
by comparison with actual lake evaporation, if
available, or more commonly by comparison with
a pan large enough to simulate a lake (sunken pans
4 m or more in diameter). The coeffi cient for a
specific pan is also dependent, to a degree, upon
the climatic regime, that is, different for arid or
humid conditions. For an evaporation pan to serve
as a valid index to lake evaporation, the exposure
of the pan should avoid the environmental effects
of the lake. Such an exposure would be near the
lake, but on the side toward the prevailing wind
direction. An island exposure would not be
One method for determining the climatic variation
of the pan coefficient is by fi eld comparisons with
large pans under the various conditions. This
method is applied in the Commonwealth of
Independent States with the GGI-3000 and 20-m2 tanks. The pan-to-lake coefficients thus derived for the GGI-3000 range between 0.75 and 1.00. For estimates of monthly average evaporation, the coeffi cient for a floating GGI-3000 evaporation pan is
estimated by the following equation:
is the average monthly vapour pressure, in
hPa, estimated from the surface temperature of
water body, e’o
is the average monthly vapour
pressure, in hPa, estimated from surface-water
temperature in the fl oating GGI-3000 pan, e200 is
the average monthly vapour pressure at 200 cm
above the water surface, in hPa, β is a correction
factor for the area of a water body, and γ is a factor
that depends on the distance l along the average
direction of wind from the shore to the pan
The ratio, β/γ, needs to be determined only for water
bodies located in tundra, forest and forest-steppe
zones and when the pan is located at a distance of
up to 500 m from shore. In all other cases, this ratio
is assumed to be equal to 1. For water bodies of
approximately round or square shape, β is determined from the area of the water surface by using
For water bodies of irregular shape (long with
islands and gulfs), the area used is that of an
assumed circle with a diameter equal to an average
distance, l, weighted with the frequency of wind
direction in per cent from the eight points of the
compass. The weighted distance can be computed
by the equation:
Where Ni is a frequency of wind direction from the
eight points, in per cent; γ can be determined from
Another method is the adjustment of the pan evaporation for heat gain or loss through the sides and bottom. An example of this method is the technique in estimating evaporation by using data from
the Class A evaporation pan. In humid seasons and
climates, the pan water temperature is higher than
the air temperature, and the pan coefficient may be
0.80 or higher. In dry seasons and arid areas, the
pan water temperature is less than air temperature,
and the coefficient may be 0.60 or less.
A coefficient of 0.70 is assumed to be applicable when water and
air temperatures are equal. The relationships for
estimating lake evaporation by adjusting Class A
pan evaporation for heat gain or loss are shown in
Figures I.4.7 and I.4.8. Owing to the important variation of wind with height, standard instrument
heights are an essential requirement of the Class A
To obtain short-period estimates of lake evaporation
with the pan method, it is also necessary to evaluate
the net energy advection to the lake and change in
energy storage as described in 4.2.3. It is useful to
have pan evaporation near a lake or reservoir as a
source of alternative data in the absence of other
meteorological data and to help verify estimates
made by the energy-budget and aerodynamic
4.2.7 Empirical formulae
The energy-budget and mass transfer methods,
though theoretically sound, require data which,
for many studies, are not readily available.
Moreover, in many cases even the economics of
acquiring such data through instrumentation of
the lake is also questionable. Thus, one has to
make use of empirical formulae to obtain estimates of evaporation. Many empirical formulae
to obtain estimates of evaporation have been
developed (Mutreja, 1986) either on the basis of relationship
the energy-budget or mass transfer method.
However, most of the equations are base
on the simple aerodynamic equation given as
Where U2 = the wind speed at 2 m above the water
surface, es = saturation vapour pressure at water
surface temperature and ea = vapour pressure of the
air at the specified height.
Marciano and Harbeck’s formulae, United States
(Marciano and Harbeck, 1954):
A few of the more common of these empirical
formulae used for estimating the evaporation from
lake surfaces are given below:
Penman’s formula, United Kingdom – small tank
United States Geological Survey (USGS), United
States and Bureau of Reclamation’s formula (USGS,
where es = saturated vapour pressure at the water
surface temperature (cm Hg–1) and ea = actual vapour
pressure (cm Hg–1)
Unless specified in the above equations the wind
speed (U) is in km x h–1 and vapour pressure is in
cm of mercury. Further, the subscripts attached to
the terms refer to the height in metres at which the
measurements are to be taken. Also, the vapour
pressure term e is frequently taken as the saturated
vapour pressure at the mean air temperature during
the interval of measurement.
The equations require surface temperature of the
body of water, which is very difficult to measure. If
this is substituted by the mean air temperature,
then the effects of advected energy to the lake on
evaporation are not considered. This may introduce
considerable error in the computed amounts of
evaporation, as small errors in temperature induce
large errors in the computations. Furthermore, the
measurements of the wind speed and vapour
pressure should be measured at the height specified
by the equation being used. Usually, it is difficult to
adjust the data collected at different heights because
neither an accurate wind law nor laws defining the
variation in humidity with height are currently
The greatest appeal for the use of these empirical
formulae lies in the fact that they are simple to use
with the standard available meteorological data.
Nevertheless, the limitations of these empirical
formulae must be clearly understood
4.3 EVAPOTRANSPIRATION FROM
DRAINAGE BASINS [HOMS I50]
Evapotranspiration considers evaporation from
natural surfaces whether the water source is in the
soil, in plants, or in a combination of both. With
respect to the cropped area, the consumptive use
denotes the total evaporation from an area plus the
water used by plant tissues, thus having the same
meaning as evapotranspiration.
The determination of evaporation and transpiration as separate
elements for a drainage basin is unreliable. Moreover,
their separate evaluation is not required for most
Evapotranspiration is one of the most popular subjects of research in the field of hydrology and irrigation. Numerous procedures have been
developed to estimate evapotranspiration. These fall in the categories of: (a) water balance methods, such as evapotranspirometers, hydraulic budget on field plots, and soil moisture depletion; (b) energy
balance method; (c) mass-transfer methods, such as
wind speed function, eddy fl ux and use of enclosures;
(d) a combination of energy and mass-transfer
methods, such as the Penman method; (e) prediction
methods, such as the empirical equations and the
indices applied to pan-evaporation data; and (f)
methods for specific crops. These have been
described in the National Handbook of Recommended
Methods for Water Data Acquisition (USGS, 1977).
In the context of evapotranspiration, Thornthwaite
and Holzman (1941) introduced the term “potential
evapotranspiration” to defi ne the evapotranspiration that will occur when the soil contains an
adequate moisture supply at all times, that is, when
moisture is not a limiting factor in evapotranspiration. The prediction methods estimate potential
evapotranspiration. Most other methods apply to
estimation of actual evapotranspiration under the
condition of suffi cient water at all times. The actual
evapotranspiration from potential evapotranspiration is derived using a simple soil moisture function,
f(φ) (Saxton and others, 1986):
λEactual = f(φ)* λE (4.25)
where λΕactual is the actual evapotranspiration and
the soil moisture function is a dimensionless variable estimated by a simple linear model. The soil
moisture function is defined by the following:
f(φ) = M/Field capacity (4.26)
where M is soil volumetric moisture at 20-cm depth
(at rooting zone). Field capacity can be defined as
the percentage of water remaining in a soil two or
three days after it has been saturated and after free
drainage has practically ceased. It has been shown
(Brandes and Wilcox, 2000) that simple linear
models of the evapotranspiration/soil moisture
process are appropriate for hydrological modelling.
4.3.2 Water-budget method
The water-budget approach can be used to estimate
evapotranspiration, ET, when precipitation, P,
stream runoff, Q, deep seepage, Qss, and changes in
storage, ΔS, can be measured or estimated. The
ET = P – Q – Qss ± ΔS (4.27)
The annual evapotranspiration from a basin for a
water year can be estimated as the difference
between precipitation and runoff if it can be established by hydrogeological studies that deep seepage
is relatively insignifi cant. The date chosen for the
beginning and ending of the water year should
coincide with the dry season, when the amount of
water in storage is relatively small and the change
in storage from year to year is negligible.
If evapotranspiration is to be estimated for a shorter
period, such as a week or a month, the amount of
water storage in the ground and in the stream channel must be measured. This is feasible only on small
basins, and application of the water-budget
approach for such short periods is generally limited
to experimental plots or catchments of a few acres.
For average annual evapotranspiration, the change
in storage is usually negligible, and evapotranspiration can be estimated by the difference between
average annual precipitation and average annual
The various terms of the above equation can be
measured by conventional methods. The precipitation measurements can be made by a network of
raingauges. For this purpose non-recording raingauges are adequate. The number of such raingauges
would depend upon the expected variability of
precipitation over the catchment. The streamflow
measurements can be done by continuous measurement (Chapter 5). The change in water storage in the
ground can be measured in two separate components, that is, the saturated and unsaturated
components. For this purpose measurement of water
table elevation in wells and measurement of soil
moisture in the saturated zone are required. The
elevation of the water table can be determined by
measuring the distance from reference point to the
water surface in wells at the end of each time period
for which evapotranspiration is to be computed. The
change in volume of water storage is equal to the
average change in water elevation x the specifi c yield
of the formation x the area of the catchment. Soilmoisture profi les from the saturation level (or to a
point of constant soil moisture in arid regions) to the
ground surface should be measured at the end of
each computation period at a number of points over
the catchment. The gain or loss of soil moisture
during the period can then be computed. The
amount of water that moves from or to the catchment as deep seepage cannot be measured directly. A
hydrogeological study of the hydraulic characteristics of the underlying formations should indicate the
relative magnitude of this flow, which must be
considered when choosing the experimental area.
This item should be small enough so that it can be
neglected in water-budget studies.
4.3.3 Energy-budget method
This method (WMO, 1966) may be applied for the
estimation of evapotranspiration when the difference between radiation balance and the heat flux
into the soil is significant and exceeds the errors of
measurement (4.2). This method is applied for estimation of evapotranspiration for periods of not less
than 10 days. For shorter periods, the estimation of
evapotranspiration by the energy-budget method is
Assuming that the surface energy balance equation
is the primary boundary condition to be satisfied in
computing ET, there are three techniques to solve
the energy-balance equation. The fi rst technique
uses semi-empirical methods, the second employs
analytical methods and the third utilizes numerical
The semi-empirical methods represent an effort to
obtain a manageable model to estimate ET. These
modern operational approaches are derived chiefl y
from Penman’s original formulation, which is a
combination of the diffusion and energy-balance
approaches (Bailey, 1990). The Jackson model
(Jackson and others, 1977) was later evaluated using
empirical and theoretical results (Seguin and Itier,
1983). The energy-balance model is integrated over
a 24-hour period and thus assumes that the soil
heat fl ux is negligible. Furthermore, observations
(Itier and Riou, 1982; Brunel, 1989) suggest that the
daily ratio of sensible heat fl ux to the net radiation
fl ux, Rn, can be approximated by that ratio estimated near midday under clear sky conditions.
With some further approximations the energybalance model can be recast as:
where LE is the latent heat fl ux (evapotranspiration,
is the surface temperature estimated remotely,
say from a satellite-based thermal IR sensor, Ta is
the near-surface air temperature obtained from a
nearby weather station, the subscript i represents
the “instantaneous” observation by the satellite
over the area of interest, and A and B constants
which vary with location (Caselles and Delegido,
1987). In practice, however, A and B vary with a
wide range of both meteorological and surface
factors (Bailey, 1990). This expression and derivatives of it have been tested and shown to produce
reasonable estimates of daily ET (Brunel, 1989; Kerr
and others, 1987; Nieuwenhuis and others, 1985;
Rambal and others, 1985; Thunnissen and
Nieuwenhuis, 1990; Riou and others, 1988).
Although equation 4.28 is characterized by low
demands for data provision and ease of operation,
it is also characterized by limited spatial and temporal areas of application together with poor accuracy
especially in the presence of cloud when using satellite thermal infra-red methods to obtain Ts
According to WMO, Germany is utilizing NOAA
AVHRR data for input into numerical evaporation
models in small-scale agricultural areas. Satellite
data include vegetation, land-surface temperature
gradients, soil moisture, diurnal temperature variations and solar irradiance. Extrapolation of the
model results are to be tested (WMO, 1992a).
4.3.4 Aerodynamic approach
The application of this method (WMO, 1966) for
the estimation of evapotranspiration is diffi cult
because of the lack of reliable methods to determine
the turbulent-exchange coeffi cient (4.2). Thus, it is
seldom used. It is used only for approximate estimation of evaporation.
In some countries, evapotranspiration is estimated
by empirical methods, the Penman method and the
Thornthwaite formula. Penman’s method is used in
conditions of sufficient moisture, and the
Thornthwaite formula (Thornthwaite and Holzman,
1941) is applied for regions with climatic conditions similar to those of the middle Atlantic coast of
the United States on which this formula was
In the Commonwealth of Independent States,
Konstantinov’s method (Konstantinov, 1966) is
applied for the estimation of evaporation based on
observations of temperature and humidity of the
air in a psychrometer shelter at 2 m above the
ground. This method is mainly applicable for the
computation of long-term mean monthly, seasonal
or annual evapotranspiration.
4.3.5 Penman–Monteith method
The combination equation 4.14 represents the
energy budget at the land surface and the transfer
of water vapour and heat between the surface and
the atmosphere. The Penman–Monteith method
(Monteith, 1965) introduces aerodynamic and
surface resistances. The former describes the effect
of surface roughness on heat and mass transfer
and the latter describes the resistance to the fl ow
of water vapour between the evaporating surface
and the air. Surface resistance for water surfaces is
zero. In the case of vegetation, the surface resistance represents biological control of transpiration
and is largely controlled by stomatal resistance.
For drying soil, the surface resistance depends on
soil moisture availability. This method may be
used on an hourly or daily basis. However, its use
is restricted because it requires sub-models for the
The Penman–Monteith model is expressed as:
λE = (ΔΔ + CpρD / raa) / (Δ + γ + γ (rcs / raa)) (4.29)
where raa is the aerodynamic resistance above the
canopy, and r
cs is stomatal resistance of the canopy.
For the Shuttleworth–Wallace model (Shuttleworth
and Wallace, 1985), λE is separated into evaporation from the soil (λEs
) and transpiration from the
), which are derived from the Penman–
Monteith combination equations:
- ρcpD0/rsa)/(Δ + γ(l + rss/rsa)) (4.30)
= (ΔΔ( – As
- ρcpD0/rca)/(Δ + γ(l + rcs/rca)) (4.31)
Where As is available soil energy, D0 is vapour pressure defi cit in the canopy, rsa is the aerodynamic resistance between the substrate and canopy source height, rca is the boundary layer resistance of the
vegetation, and rss is soil resistance.
The aerodynamic resistance above the canopy (raa) and the
aerodynamic resistance between the substrate and
canopy source height (rsa) are functions of leaf area
index, eddy diffusivity decay constant, roughness
length of the vegetation (function of vegetation
height), zero plane displacement (function of vegetation height), a reference height above the canopy
where meteorological measurements are available,
wind speed, von Karman’s constant, and roughness
length of the substrate. D0 is derived from the
Ohm’s law electrical analog for the vapour pressure
and temperature difference between the canopy
and the reference height above the canopy where
fluxes out of the vegetation are measured. D0 is a
function of the measurable vapour pressure defi cit
at the reference height, D:
D0 = D + (ΔΔ – raaλEc
(Δ + γ))/ρcp (4.32)
and D can thus be substituted for D0 into the combination equations. The total evaporation from the
crop, λE, for the Shuttleworth–Wallace model is the
sum of the Penman–Monteith combination equations with D substituted for D0:
λE = Cc
4.3.6 Priestley–Taylor (radiation)
The method of Priestley and Taylor (Priestley and
Taylor, 1972) is based on the argument that, for
large, wet areas, radiation controls of evaporation
must dominate rather than advective controls. If
the atmosphere remains saturated when in contact
with the wet surface, then the latent-heat transfer
(evaporation) may be expressed by:
where Q* is the available net radiation, G is the soilheat flux, and ε equals sλ/cp, with s equal to the
slope of the saturation specifi c humidity curve, λ is
the latent heat of vaporization, and cp is the specifi c
heat of water.
For equilibrium evaporation, it is proposed that:
with α = 1.26, an empirical constant. This expression is used as an estimate of potential evaporation
in the absence of local advection. It also gives good
estimates for evaporation from well-watered but
not wet vegetation in much smaller regions.
4.3.7 Complementary method
The complementary method, first suggested by
Bouchet (1963), is increasingly used in hydrological
applications for large areas because it essentially
uses standard climatic data.
The method considers that potential evaporation is
as much the effect of the actual evaporation as its
cause. Heat and moisture released from the surface
will modify the temperature and humidity of the
air above it. It has been suggested that the increase
in potential evaporation observed when an area
dries out may be used as a measure of the actual
If actual evaporation E is reduced below the potential rate Epo for an extensive wet region, then an
amount of energy Q would be released, so that:
λEpo – λE = Q (4.43)
This energy change will affect temperature, humidity, turbulence and hence evaporation. If the area is
big enough so that the change in energy does not
result in changes in the transfer of energy between
the modifi ed air mass and that beyond, Q should
equal the increase in λEp, the potential evaporation
for the drying region.
λEp – λEpo = Q (4.44)
E + Ep = 2 Epo (4.45)
Most applications of the complementary relationship (Morton, 1982) have been concerned with
fi nding appropriate expressions for Ep and Epo. These
may be estimated with equation 4.15 and the
Priestley–Taylor method given in 4.3.6, respectively.
The approach does not consider advection and
assumes Q to remain constant. Also, the vertical
exchange of energy, that is, with air masses brought
in by large-scale weather systems, is not
4.3.8 Crop coefficient and reference
In 1998, Crop evapotranspiration – Guidelines for
computing crop water requirements (FAO-56 report),
recommended a new standard for reference crop
evapotranspiration using the Blaney–Criddle,
Penman, radiation and pan evaporation methods. The FAO-56 approach (FAO, 1998; Allen
2000) first calculates a reference evapotranspiration (ETo) for grass or an alfalfa reference crop
and then multiplies this by an empirical crop
) to produce an estimate of crop
potential evapotranspiration (ETc
). The ETc calculations used the dual crop coefficient approach
that includes separate calculation of transpiration and evaporation occurring after precipitation
and irrigation events.
The FAO-56 Penman–Monteith method computes
reference evapotranspiration from net radiation
at the crop surface, soil heat fl ux, air temperature,
wind speed and saturation vapour pressure defi –
cit. The crop coeffi cient is determined from a
stress reduction coeffi cient (Ks
), a basal crop coeffi cient (Kcb) and a soil water evaporation coeffi cient (Ke
). The Kcb curve is divided into four growth
stages: initial, development, mid-season and late
season. Field capacity and wilting point estimates
determine soil water supply for evapotranspiration. The downward drainage of the topsoil is included but no upward flow of water from a
saturated water table was considered, possibly
causing some overprediction of water stress
between the known irrigations. Water stress in
the FAO-56 procedure is accounted for by reducing the value of Ks
4.3.9 Large aperture scintillometer
Estimation of actual evapotranspiration using the
energy-balance method requires knowledge of
the sensible heat fl ux. According to the Monin–
Obukhov similarity theory, the sensible heat flux,
H, is related to the structure parameter of temperature, CTA large aperture scintillometer is an instrument to collect path-average values of CT 2
(de Bruin and others, 1995). The scintillometer
directs a light source between a transmitter and
receiver and the receiver records and analyses
fluctuations in the turbulent intensity of the
refractive index of the air. These fluctuations are
due to changes in temperature and humidity
caused by heat and moisture eddies along the
path of the light. Additional data on temperature,
pressure and humidity are necessary to compute
the characteristic parameter of the refractive
index. This can then be converted to sensible
heat fl ux.
An important feature of the scintillometer technique is that although the measurement is along the path of the light beam, because of the
effects of wind, this is actually an estimate of H
over an area. The method therefore forms an
intermediate level between the field scale measurements and the large area remote-sensing
4.4 EVAPORATION REDUCTION
4.4.1 From free surfaces
Evaporation losses from a fully exposed water
surface are essentially a function of the velocity and
saturation defi cit of the air blowing over the water
surface, and the water temperature. Evaporation
losses are held to a minimum by:
(a) Exposing the least possible water-surface area.
This in turn means that streams and reservoirs
should be kept deep instead of wide;
(b) Covering the water surface;
(c) Controlling aquatic growth;
(d) Creating afforestation around reservoirs that
would act as windbreakers. However, this
method has been found to be useful under
limited conditions for small ponds;
(e) Storing water underground instead of creating
a surface reservoir. To accomplish this there are
physical and legal problems in preserving the
water so stored from adverse withdrawal;
(f) Making increased use of underground water;
(g) Integrated operation of reservoirs;
(h) Treatment with chemical water evaporation
The first seven methods mentioned above are
direct and easily understandable methods.
However, the last method needs some explanation. This method comprises dropping a fl uid on
the surface of the water so as to form a monomolecular fi lm. The problem with the fi lm, however,
is that it becomes damaged by wind and dust, and
is too rigid to enable repair of the film thus
damaged. Chemicals such as hexadecanol and
octadecanol, of course, can be used for the purpose
Studies by the Bureau of Reclamation indicate that
evaporation may be suppressed by as much as
64 per cent with hexadecanol films in 1.22-m diameter pans under controlled conditions. Actual
reduction on large bodies of water would, of course,
be significantly less than this because of problems
of maintaining the films against wind and wave
action. Evaporation reduction to the extent of 22 to
35 per cent has been observed on small lakes of
roughly 100 ha in size with reductions of 9 to
14 per cent reported on larger lakes (La Mer, 1963).
In Australia, evaporation reduction to the extent of
30 to 50 per cent has been observed on medium
lakes of roughly 100 ha in size. Although the use of
the monomolecular film is still in the research stage,
its relative case means that some measure of evaporation control can be obtained through this
4.4.2 From soil surface
There are various methods of controlling evaporation losses from soil (Chow, 1964).
(a) Dust mulch: This is an age-old practice in cultivation of soil to keep it loose on the surface. It is
based on the theory that loosening the surface
will permit rapid drying and reduce contact
between soil particles. Rapid drying will develop
dry soil to act as a blanket to suppress evaporation. Reducing points of contact between soil
particles will lessen capillary rise.
It has been found that soil cultivation by tillage may be necessary only to kill weeds and
keep the soil in a receptive condition to absorb
water and deep tillage is futile as a means of
overcoming drought or increasing yield. Experiments have also shown that mulching not only
decreased the amount of water in the soil, but
also caused loss of more moisture than in the
bare, undisturbed soils. In tank and field trials it
has also been found that mulching by thorough
cultivation at weekly intervals failed to save
soil moisture, but the surface shallow layer, by
drying quickly, acted as a deterrent to further
loss of moisture.
Since these early investigations, the results of
many others have been published. Many agricultural experiment stations have studied this
problem, resulting in conclusions similar to
those mentioned. Various experiments have
also indicated that the soil mulch can reduce
moisture loss only when the water table,
perched or permanent, is within the capillary
rise of the surface;
(b) Paper mulch: Covering the soil with paper to
reduce evaporation was widely used in the late
1920s, but is now rarely done as it has been
found that the effect of paper mulch is confi ned
to limited surface of soil, which again is due to
condensation of water beneath the paper;
(c) Chemical alteration: Experiments in the early
1950s indicated that chemical alteration of the
soil moisture characteristics may decrease evaporation. The addition of polyelectrolytes to soils
decreases the rate of evaporation and increases
the water available to plants;
(d) Pebble mulch: In China this method has been
used for partial control of evaporation in some
4.5 SOIL MOISTURE MEASUREMENT
Below the surface of the Earth there exists a huge
reservoir of freshwater. This subsurface water can
be divided into soil moisture, vadose water, shallow groundwater and deep groundwater. The
zones of soil moisture and vadose water are
together known as the zone of aeration. The
amount of water held as soil moisture at any time
is an insignifi cant amount by comparison with
the Earth’s total available water, but it is crucial
to plant life and food production and thus vital
Soil moisture can be defi ned as the water held in
the soil by molecular attraction. The forces acting
to retain water in the soil are adhesive and
cohesive forces. These forces act against the force
of gravity and against evaporation and
transpiration. Thus, the amount of moisture in
the soil at any given time is determined by the
strength and duration of the forces acting on the
moisture, and the amount of moisture initially
Natural sources of soil water such as rainfall and
snow melt are normally greatly reduced during
drought. Slope shape, gradient and soil surface
roughness will affect soil water content since
surface or subsurface run-on from adjacent
upslope sites can add to the soil moisture, while
surface runoff can remove water from a site.
Evaporation, evapotranspiration and deep percolation beyond rooting depth are other factors that
deplete soil moisture.
Hence, soil water content must be defined in
specific quantitative terms to accurately indicate
the amount of water stored in the soil at any
given time. At saturation, after heavy rainfall or
snow melt, some water is free to percolate down
through the soil profile. This excess water is
referred to as gravitational water and can percolate below the rooting depth of some plants. Here
it is important to define some terms in relation to
soil moisture. Field capacity is defined as the
amount of water remaining in the soil after percolation has occurred. Wilting point is defined as
the soil water content at which the potential of
plant roots to absorb water is balanced by the
water potential of the soil. The amount of water
between field capacity and wilting point is generally considered plant available water content
although plants can also extract gravitational
water while it is available.
The moisture content of the soil is a key component in making irrigation scheduling decisions.
The root zone serves as a reservoir for soil moisture. During the rainy season the moisture content
is high, but at harvest time the soil is commonly
depleted of moisture. Thus the measurement of
soil moisture is an important factor in preventing
overirrigation resulting in wastage of water and
leaching of fertilizers or under-irrigation, that
result in water deficit.
Soil moisture is measured in two distinctly different methods: quantitatively and qualitatively,
which is an indication of how tightly the water is
held by the soil particles.
4.5.2 Quantitative methods
22.214.171.124 Gravimetric method (Oven dry and
The gravimetric method is one of the direct methods
of measuring soil moisture. It involves collecting a
soil sample (usually 60 cm3), weighing the sample
before and after drying it, and calculating its moisture content. The soil sample is considered to be dry
when its weight remains constant at a temperature
of 105°C. Many different types of sampling equipment, as well as special drying ovens and balances,
have been developed and used for this method.
The gravimetric method is the most accurate
method of measuring moisture content in the soil
and serves as the standard for calibrating the equipment used in all other methods. However, it cannot
be used to obtain a continuous record of soil moisture at any one location because of the necessity of
removing the samples from the ground for laboratory work.
The procedure for collecting a sample for the
gravimetric method depends on whether the soilmoisture determination is to be based on the dry
mass of the sample or on its volume for dry-mass
determination, but not for volumetric determination. The sample can be disturbed for dry-mass
determination, but not for volumetric determination. Soil sampling is fraught with difficulties if
the soil is very dry or very wet or if it contains
stones or other material that preclude easy cutting
by the sampling equipment.
The technique and equipment used for sample
collection should be such that the samples do not
lose or gain moisture or otherwise become altered
or contaminated during sampling and transportation. When sampling through a wet layer into a
dry layer, care must be taken to keep the sampling
equipment as dry as possible and to prevent water
from running down the hole into the drier material. If there is free water in the soil, the measured
moisture content probably will be less than the
correct value because some water will drip off as
the sample is removed from the ground, or some
may be squeezed out by compaction during
When dry, hard, fine-textured sediments are
encountered it is difficult to drive the core barrels
or to rotate the augers. When dry, coarse-textured
sediments are sampled, the sample may slide out
at the end of the core barrel or auger as it is withdrawn. Stony soils are very difficult to sample,
especially volumetrically, because of the likelihood of hitting a stone with the cutting edges of
the equipment and because representative
samples must be large. Soils that contain a considerable amount of roots and other organic matter
also present difficulty.
The amount of soil taken for the gravimetric
moisture determination of a gravel soil needs to
be substantially more than for non-gravel soils
and depends proportionally on the size and
content of the gravel. Moisture is determined as a
percentage by mass (weight). If multiplied by
bulk density, moisture as a percentage of volume
In soil-moisture sampling, it is essential that all
sampling operations, as well as the transfer of
samples to cans, and the weighing of the moist
samples be done as rapidly as possible to minimize moisture losses. Many difficulties in the use
of sampling equipment may be avoided if the
equipment is kept clean and free of moisture and
Description of samplers
Auger samplers (Figure I.4.9)
The simplest equipment for soil-moisture sampling
is the hand auger. Hand augers, with shaft extensions of aluminium pipe, have been used in
sampling to depths as much as 17 m. One of the
most useful types of hand augers consists of a cylinder 76 mm in diameter and 230 mm long, with a
1.4-m extension pipe on the top and two curved,
cutting teeth on the bottom. Because the barrel is a
solid cylinder, the sample is not as likely to become
contaminated from the side of the test hole. Thus,
a good, representative, but disturbed, sample is
obtained by using this equipment. For ease in
sampling at depths greater than 1.5 m, 0.9 m extensions of 19-mm aluminium pipe are added, as needed (Figure I.4.10).
To obtain a sample by the hand-auger method, the
auger has to penetrate usually about 80 mm of the
material in order to fi ll the cylinder barrel. The auger
is then raised to the surface, and the barrel is struck
with a rubber hammer to jar the sample loose.
A soil-sampling tube, core barrel or drive sampler
offers an advantage in soil-moisture sampling as
volumetric samples can be obtained for calculating
moisture content by volume. Core samplers provide
uncontaminated samples if the equipment is kept
clean. Oil should never be used on the samplers,
and they should be kept free of dirt, rust and moisture. A two-person crew is normally recommended
for deep sampling, and depths of 20 m may be
sampled (Figure I.4.11). A volume of soil core of at
least 100 cm3 is recommended.
The open-drive sampler consists of a core barrel of
50 mm inside diameter and 100 mm long, with
extension tubes of 25 mm in diameter and 1.5 m
long for sampling at depth. Brass cylinder liners,
50-mm in length, are used to retain the undisturbed
core samples. The samples are removed from the
core barrel by pushing a plunger. A light drill rod or
15-mm pipe may be used as extensions.
A simple and economical sampler for obtaining
volumetric core samples from shallow depths
consists of a thin-walled brass tube 50 mm in diameter and 150 mm long mounted on the end of a 90-cm T-handle of 19-mm pipe. After samplers are removed from the hole, they are pushed out of the
core barrel by the central plunger. Since the inside
diameter of the core barrel is known, volumetric
samples may be obtained easily by cutting off predetermined lengths of the core as it is removed from
First, the wet soil samples are weighed individually
in their transport containers. The containers are
then opened and placed in a drying oven that is
capable of maintaining a temperature of 105°C
±0.5. For samples that contain peat or significant amounts of gypsum, the oven temperature should
be 50°C ±0.5, which will then require a longer time
for the sample to reach a dry state.
After drying, the samples are reweighed in their
containers. The difference in the wet and dry weights
for a sample is the measure of its original water
content. Other drying processes that are faster than
the standard oven may be used, for example, alcohol
roasting, infra-red lamps and microwave ovens.
If the samples contain gravel and stones, the above
procedure can be modified if the weights or volumes
of the gravel and/or stones can be determined
The advantages and disadvantages of the method
are given below.
Advantages: This technique is relatively inexpensive, simple and highly accurate.
Disadvantages: This technique is time-consuming,
labour-intensive and difficult in rocky soils.
126.96.36.199 Neutron scatter method [HOMS C58]
The neutron method indicates the amount of water
per unit volume of soil. The soil volume measured
by this method is bulb-shaped and has a radius of
1 to 4 m, according to the moisture content and the
activity of the source.
This method is based on the principle of measuring
the slowing of neutrons emitted into the soil from
a fast-neutron source (Greacen, 1981). The energy
loss is much greater in neutron collisions with
atoms of low atomic weight and is proportional to
the number of such atoms present in the soil. The
effect of such collisions is to change a fast neutron
to a slow neutron. Hydrogen, which is the principal
element of low atomic weight found in the soil, is
largely contained in the molecules of the water in
the soil. The number of slow neutrons detected by
a counter tube after emission of fast neutrons from
a radioactive source tube is electronically indicated
on a scale.
Instruments A typical set of equipment consists of a portable
battery-powered or spring-wound timer that has a
time-accounting range of 0.5 to 5 minutes and
weighs approximately 16 kg, and a moisture probe
containing a 100-mCi fast-neutron source of
americium-241 and finely ground beryllium (halflife, 458 years). The probe has a length of about
400 mm, a diameter of about 40 mm and a weight
of 20 kg when complete with a lead and paraffin
shield that is 150 mm in diameter and 100 mm
long (Figure I.4.12). These probes have been used
with up to 60 m of cable.
The source and detector are lowered into the soil through a hole cased with aluminium tubing, and readings can be taken at any depth except close to the surface. The inside diameter of the tube should
be only slightly larger than the diameter of the
probe. The tube should be installed by augering the
soil inside the tube, if possible, to ensure close
contact between the outside surface of the tube and
Similar gauges have been developed to make measurements in the surface layers of the soil. In this case, the equipment is placed on the ground surface and gives the moisture content of a hemispherical
volume of 15- to 40-cm radius. Access tubes
The installation of access tubes must be performed
carefully to prevent soil compaction and to ensure
soil contact around the outside of the tubes, that is,
no voids in the soil should be created outside the
tubes during their installation. Access tubes may be
(a) By inserting the tubes into prepared holes of the
same or slightly smaller diameter (the holes can
be prepared by using either a hand-powered or
(b) By driving the tubes into the soil with a hammer
and then removing the soil from inside the
tubes with an auger.
The bottom ends of the tubes should be sealed to
prevent infiltration of groundwater. The top ends of
the tubes should be sealed with a cap or a stopper
when not in use.
The probe should be calibrated by gravimetric
sampling (188.8.131.52) of the type of soil that is to be
tested and in the size and type of casing into which
the probe is to be lowered. Sufficient samples
should be taken around the test hole to define the
soil moisture profile. It is difficult to obtain a good
calibration in heterogeneous soil or when soil
moisture is changing rapidly with depth. An
approximate calibration can also be carried out in
the laboratory by using a container fi lled with soil
material. The type and size of casing and the
method of installation of the access tube have a
considerable effect on the readings, and new calibration curves should be obtained for each type of
Measurements and accuracy
The access tubes must be kept free of excess moisture or erroneous readings will result.
After lowering the probe to the proper depth in the
access tube, the number of counts over a known
time period is determined. The average count is
converted to soil moisture content by using the
calibration curve. The accuracy of a determination
depends primarily on:
(a) The validity of the calibration curve;
(b) The number of counts per determination.
Because of the randomness of the emission and
the impact of neutrons, random count errors can
occur. Timing errors may be kept to a minimum by
using a standard-count timing cycle of two
Salt concentrations in the range ordinarily found in
soil moisture do not materially affect data obtained
by the neutron method, but at salt concentrations
at the level of seawater, the effect is appreciable.
There is some evidence of a temperature effect
Readings close to the surface are affected by the
position of the probe with respect to the air-soil
interface. Proximity of the interface causes lower
counts than would be indicated for the same moisture content at a greater depth.
When the error sources are minimized, the accuracy of an individual determination can reach 0.5
to 1 per cent. For repeated determinations over
time, such as might be performed in a water-balance
study, the changes in water content of soil can be
even more accurate because of the elimination of
The advantages and disadvantages of the method
and the availability of instruments for its use are
summarized below (Prichard, 2003):
Advantages: The neutron probe allows a rapid, accurate, repeatable measurement of soil moisture
content to be made at several depths and locations.
Disadvantages: The use of radioactive material
requiring a licensed and extensively trained operator, the high equipment cost and extensive
calibration required for each site.
Readily available instruments: Neutron probes are
184.108.40.206 Dielectric methods [HOMS C60]
The dielectric constant methods seek to measure
the capacity of a non-conductor (soil) to transmit
high-frequency electromagnetic waves or pulses.
The resultant values are related through calibration
to soil moisture content.
The basis for use of these instruments is that dry soil
has dielectric values of about 2 to 5 and that of water
is 80 when measured between 30 MHz and 1 GHz.
Two approaches have been developed for measuring the dielectric constant of the soil water media
and estimating the soil volumetric water content:
(a) Time domain reflectrometry (TDR);
(b) Frequency domain refl ectrometry (FDR).
Neither TDR nor FDR use a radioactive source,
thereby reducing the cost of licensing, training and
monitoring when compared with the use of the
Time domain reflectrometry
The TDR device propagates a high-frequency transverse electromagnetic wave along a cable attached
to a parallel conducting probe inserted into the soil.
The signal is refl ected from one probe to the other,
then back to the meter, which measures the time
between sending the pulse and receiving the
refl ected wave. By knowing the cable length and
waveguide length, the propagation velocity can be
computed. The faster the propagation velocity, the
lower the dielectric constant and thus lower soil
Waveguides are usually a pair of stainless steel rods,
which are inserted into the soil a few centimetres
apart. The measurement is the average volumetric
water content along the length of the waveguide if
so calibrated. Waveguides are installed from the
surface to a maximum depth of usually 45–60 cm.
Pairs of rods can be permanently installed to provide
water content at different depths. If deeper measurements are needed, a pit is usually dug after which
the waveguides are inserted into the undisturbed
pit wall. The soil disruption can change water movement and water extraction patterns, resulting in
TDR units are relatively expensive. However, once
properly calibrated and installed, the TDR technique is highly accurate. Since surface measurements
can be made easily and in multiple sites, it works
well for shallow rooted crops.
Frequency domain refl ectrometry
This approach uses radio frequency waves to measure soil capacitance. The soil acts as the dielectric
completing a capacitance circuit, which is part of a
feedback loop of a high-frequency transistor oscillator. The frequency varies between instrument
manufacturers but is generally about 150 MHz. The
soil capacitance is related to the dielectric constant
by the geometry of the electric fi eld established
around the electrodes. The dielectric constant is
then related to the volumetric water content as
discussed in the TDR method. Two distinct types of
instruments use the FDR techniques – an access
tube method and a hand-held push probe.
Access tube type
An access tube of PVC material similar to one being
used in the neutron probe and the electrodes is
lowered into the access well and measurements are
taken at various depths. It is necessary to ensure a
very close fi t between the walls of the access tube
and the soil to ensure reliable values as air gaps
affect the travel of the signal in the soil. Calibration
to soil volumetric water content is required (especially in clayey soils and those with high bulk densities) to ensure accurate values. If properly calibrated and installed, the accuracy of the probe can
Many of the advantages of the neutron probe are
available with this system, including rapid measurements at the same locations and depths over time.
Another variant of this technology is the use of a
permanent installation, which reads multiple
depths. These are used in conjunction with electronics to make frequent readings and transmit
results to a central data-collection device.
The other type of capacitance device is a hand-push
probe, which allows rapid, easy, near-surface
readings. These probes provide a qualitative
measurement of soil water content on a scale from
1 to 100 with high readings indicating higher soil
moisture content. Probe use in drier soils and those
containing stones or hard pans is diffi cult. Deeper
measurements are possible using a soil auger to gain
access to deeper parts of the root zone. The probe is
best used in shallow-rooted crops.
Advantages: The advantages of the TDR and FDR
equipment is that they are relatively accurate
(±1–2 per cent); can provide direct readouts of volumetric, available plant soil moisture percentages or
continuous readings if used with a data logger; do
not require calibration; and are relatively unaffected
by salts in the soil. TDR is more accurate and less
affected by salts while FDR can detect “bound”
water in fi ne particle soils, which is still available to
plants. Thus, the TDR instrument would be preferable for extensive acreage of salt-affected soils.
However, if dealing with primarily fi ne-textured,
non-saline soils, the FDR instrument would be preferable. In general, these instruments are accurate,
reasonably priced, easy to use and very suitable for
Disadvantages: Owing to the cost of the instruments,
these methods are more expensive than others.
Readings can be affected if good contact is not made
with the soil, and prongs can be damaged in hard or
rocky soils. TDR has complex electronics and is the
most expensive, whereas FDR is more susceptible to
soil salinity errors. Data logger readings are in the
form of graphs requiring interpretation.
220.127.116.11 Gamma-ray attenuation
The intensity of a gamma ray that passes through a
soil section undergoes an exponential decrease that
principally depends on the apparent density of the
soil, the water contained in the soil and the coeffi –
cients of attenuation of the soil and of the water,
which are constants. The method consists of concurrently lowering a gamma-ray source (generally
Caesium 137) and a gamma-ray detector (scintillator-photomultiplier) down a pair of parallel access
tubes that have been installed in the soil. At each
measurement level, the signal can be translated
into the apparent wet density of the soil or, if the
apparent dry bulk density of the soil is known, the
signal can be converted into a measure of the volumetric soil-moisture content.
The measuring equipment permits tracking of the
evolution of wet density profi les and of the volumetric soil-moisture at several tens of centimetres
of depth below the soil surface if the dry density
does not vary with time.
The method has the advantage of having a high
spatial resolution (it measures over a slice of soil
20 to 50 mm in thickness with the access tubes
separated by about 3 m). However, the measurements are not specific to water alone. The apparent
variations in dry density can confound the measurements of soil moisture.
Some complex equipment has two energy sources
with different intensities of gamma rays, which
permit the joint study of the variations in both
apparent density and soil moisture. Such equipment is used primarily in laboratories and not under
4.5.3 Qualitative methods
18.104.22.168 Tensiometric method [HOMS C62]
The components of a tensiometer include the
porous cup, the connecting tube and/or the body
tube and the pressure sensor. The porous cup is
made of a porous, rigid material, usually ceramic.
The pores of the cup wall are small enough to
prevent the passage of air. A semi-rigid connecting
tube and/or a rigid body tube are used to connect
the tensiometer cup to the pressure sensor. The
system is filled with water and the water in the
point or cup comes into equilibrium with the moisture in the surrounding soil. Water flows out of the
point as the soil dries and creates greater tension, or
flows back into the point as the soil becomes wetter
thereby decreasing the tension. These changes in
pressure or tension are indicated on the measuring
device. Multiple tensiometres located at several
depths permit the computation of a soil-moisture
Tensiometers provide data on soil-water potential
(pressure components). If a tensiometer is used for
moisture determinations, a calibration curve is
needed. The calibration curve may be a part of the
soil-moisture retention curve, but it is recommended
that fi eld data from the gravimetric method (22.214.171.124)
and tensiometer readings be used for the calibration. Even so, the moisture data are only approximate,
because of the hysteresis between the wetting and
drying branches of the soil-moisture retention curve.
The range of use is restricted to 0 to 0.8 bars (0 to 8
m of negative hydraulic head). Therefore, the
method is suitable only for wet regions.
The pressure measuring device is usually a Bourdontube vacuum gauge or a mercury manometer. The
tensiometer may also be attached to an electrical
pressure transducer to maintain a continuous record
of tension changes. Because the system is under a
partial vacuum during unsaturated soil conditions,
it is necessary that all parts or joints be impermeable to air. For field use, Bourdon vacuum gauges are
more convenient than mercury manometers, but
they have a lower accuracy. Electrical pressure transducers are both convenient and precise.
The tensiometer response time is much faster with
pressure transducers that have small volume
displacements than with other pressure sensors.
The disadvantage of the cost can be offset by using
only one electrical pressure transducer connected
to several tensiometers via a scanning device.
Another solution consists of using a measuring
apparatus that briefl y samples the pressure in the
tensiometer by means of a needle. This needle
perforates a special bulb on the tensiometer tube
only during the moment of the measurement. A
single needle apparatus can be used to sample
numerous tensiometers placed in the soil. However,
unlike the system described above, this type of
tensiometer cannot be used to record changes of
Tensiometers should be filled with de-aerated water.
Then it would be possible to remove air trapped
inside the system by using a vacuum pump.
Tensiometers are generally inserted vertically into
the soil in pre-augered holes of the same diameter
as the porous cup. The centre of the porous cup is
located at the depth where pressure measurement is
required. Tensiometers are affected by temperature
fluctuations that induce thermal expansion or
contraction of the different parts of the system and
that infl uence the pressure readings. In the fi eld,
protection from solar radiation is recommended for
tensiometers that are above ground to minimize
this infl uence. Similarly, tensiometers used in the
winter should be protected against frost damage to
the water tube and the pressure sensor. Tensiometers
need to be purged periodically to remove
accumulated air from the system.
A tensiometer reading indicates the pressure in the
porous cup minus the pressure difference caused by
the water column between pressure sensor and
porous cup. Therefore, the pressure potential of the
soil water at the depth of the cup is the pressure
sensor reading plus that of this water column. If the
pressure is expressed in terms of suction, that is,
atmospheric pressure minus gauge pressure, then
the pressure potential of the soil equals the sensor
reading minus the pressure difference caused by the
water column in the tube. Corrected pressure potential of the soil can be generated directly with pressure transducer systems.
It is difficult to state the precision of a tensiometer
measurement of soil-water pressure potential. The
accuracy of a measurement is influenced by temperature, the accuracy of the pressure sensor and the
quantity of air accumulated within the system.
Moreover, the response time of tensiometers can
cause erroneous measurements if the soil-water
potential is changing quite rapidly in time. In this
case, equilibrium between the soil water and the
tensiometer water cannot be obtained. Recent
studies have shown that semi-permeable plastic
points provide much faster response than ceramic
points (Klute, 1986).
The tensiometer is probably the easiest to install
and the most rapidly read of all soil-moisture measuring equipment. However, tensiometers are not
suitable for installation at depths greater than 3 m.
At normal atmospheric pressures, the method is
limited to a range of pressure potential down to
about –85 kPa. Tensiometers require frequent servicing to obtain reliable measurements under field
Advantages: Tensiometers are not affected by the
amount of salts dissolved in the soil water. They
measure soil water tension with a reasonable accuracy in the wet range.
Disadvantages: Tensiometers only operate between
saturation and about –85 kPa. Thus they are not
suited for measurement in dry soils.
126.96.36.199 Porous blocks/electrical resistance
blocks [HOMS C60]
Porous blocks are made of gypsum, glass/gypsum
matrix, ceramic, nylon and fibreglass. They are
buried at the depth of measurement desired. Over
time, the blocks come to equilibrium with the moisture content in the surrounding soil. Therefore, the
subsequent measurement is related to soil water
In the case of electrical resistance blocks, two electrodes are buried inside the block with a cable
extending to the surface. The electrical resistance is
measured between the two electrodes using a meter
attached to the cable. Higher resistance readings
mean lower block water content and higher soil
Porous blocks require the same careful installation
as tensiometers and good soil contact is important.
Maintenance requirement is small and is much
less than for tensiometers. Gypsum blocks are
proven to break down in alkaline soils and will
eventually dissolve, necessitating an abandonment or replacement. Soils high in soluble salts
may cause erroneous readings, as salts infl uence
soil conductivity and resistance. Gypsum blocks
are best suited for fi ne-textured soils, as they are
not generally sensitive below 1 000 hPa For most
sandy soils, this would be outside the level of available water.
A newer type of gypsum block consists of a fi ne
granular matrix with gypsum compressed into a
block containing electrodes. The outside surface of
the matrix is incised in a synthetic membrane and
is placed in a perforated PVC or stainless steel
protective cover. The construction materials
enhance water movement to and from the block,
making it more responsive to soil water tensions in
the 300–2 000 hPa range. This makes them more
adaptable to a wider range of soil textures.
Thermal dissipation blocks: These are made of a
porous, ceramic material. Embedded inside a porous
block is a small heater and temperature sensor
attached by cable to a surface meter. A measurement is made by applying voltage to an internal
heater and measuring the rate at which heat is
conducted away from the heater (heat dissipation).
The rate of heat dissipation is related to moisture
Thermal dissipation sensors are sensitive to soil
water across a wide range of soil water contents;
however, to yield water content they must be individually calibrated. These blocks are considerably
more expensive than electrical resistance blocks.
Advantages: The method is quick, repeatable and
Disadvantages: The blocks do not work well in
coarse-textured, high shrink-swell, or saline soils.
Accuracy is rather poor unless blocks are individually calibrated for the soil being monitored using a
pressure plate extractor or gravimetric method.
Blocks should be replaced every one to three years.
Major consideration is that the sensitivity of the
blocks is poor in dry soil conditions. The blocks
need to be soaked in water for several hours before
they are installed in the fi eld.
4.5.4 Remote-sensing [HOMS D]
The remote-sensing technique is the most recent
tool being used to estimate soil moisture properties
at or near the surface. This information may be used
to infer soil moisture profiles down to several
metres. Remote-sensing of soil moisture can be
accomplished using visible, infra-red (near and
thermal), microwave and gamma data (Engman
and Gurney, 1991; Schultz and Engman, 2000).
However, the most promising techniques are based
on the passive and active microwave data. The visible and near-infra-red techniques, which are based
on the measurement of reflected solar radiation, are
not particularly viable because there are too many
noise elements that confuse the interpretation of
the data. The thermal infra-red techniques are based
on the relationship between the diurnal temperature cycle and soil moisture, which depend upon
soil type and is largely limited to bare soil conditions. A main problem associated with thermal
infra-red techniques is cloud interference.
Microwave techniques for measuring soil moisture
include both passive and active microwave
approaches; each has distinct advantages.
Microwave techniques are based on a large contrast
between dielectric properties of liquid water and
dry soil. The variation of natural terrestrial gamma
radiation can be used to measure soil moisture
because gamma radiation is strongly attenuated by
water. It appears that operational remote-sensing of
soil moisture will involve more than one sensor.
Furthermore, both active microwave and thermal
infra-red applications need much additional
research before they can be used to extract soil
The reflection from bare soil, in the visible and
near-infra-red parts of the electromagnetic spectrum, can only be used under limited conditions to
estimate soil moisture. The accuracy of this method
is poor and absolute values of soil moisture cannot
be obtained. More spectral bands and a much higher
geometrical resolution in the (VIS/NIR) infra-red
visible/near range are needed for soil moisture and
agricultural purposes, than that available from
Landsat, SPOT and the NOAA satellites. Soil moisture has been estimated by using precipitation
indices; operational applications have been developed by FAO using geostationary imagery over
intertropical regions (WMO, 1993). With the advent
of the International Geosphere–Biosphere
Programme (IGBP) the need for high-resolution
data is increasing.
Thermal infra-red techniques have been successfully used to measure the few surface centimetres of
soil moisture. A limitation to the thermal approach
is that it cannot effectively be applied to surfaces
with vegetation cover.
Attempts have been made to evaluate the soil moisture through observation of the Apparent Thermal
Inertia using both AVHRR data from Landsat and
SPOT and geostationary images; applications have
been more of pilot projects rather than operational
Microwave techniques have shown a lot of potential for measuring soil moisture but still need
varying amounts of research to make them operational. In order to progress to operational soil
moisture monitoring by remote-sensing techniques,
multi frequency and multipolarization satellite data
will be required; such data are needed to quantify
different surfaces and thus reduce the amount of
ground truth required.
Only in the microwave region is there a direct physical relationship between soil moisture and the
refl ection or emission of radiation. A unique advantage of using the microwave region is that at long
wavelengths the soil moisture measurements can
be made through clouds. It has also been illustrated
that the synergistic use of optical and microwave
data in agrometeorological applications is advantageous. The passive microwave region has been
exploited the most so far. At present, microwave
radiometers capable of measuring soil moisture are
available only on aircraft. These are being used in
both research and a few operational applications.
Soil moisture information at a depth of several
metres can be obtained from short pulse radar
(wavelengths of 5–10 cm) techniques. In the Russian
Federation, this aircraft-based method is used for
soil moisture measurements in forested areas and
for detecting zones of saturation down to a depth of
5–10 m. The use of gamma radiation is potentially
the most accurate of the remote-sensing methods
developed for soil moisture measurement. The
attenuation of gamma radiation can be used to
determine changes in soil moisture in the top
20–30 cm of the ground. This technique requires
that some fi eld measurements of soil moisture be
made during the measurement fl ight, because it
does not give the absolute values of soil moisture.
References and further reading
Allen, R., 2000: Using the FAO-56 dual crop coeffi –
cient method over an irrigated region as part of an
evapotranspiration intercomparison study. Journal of
Hydrology, Volume 229, Nos. 1–2, pp. 27–41.
Bailey, J.O., 1990: The potential value of remotely sensed
data on the assessment of evapotranspiration and
evaporation. Remote Sensing Reviews, Volume 4,
No. 2, pp. 349–377.
Bastiaanssen, W.G.M., M. Menenti, R.A. Feddes and
A.A.M. Holtslag, 1998: A remote sensing surface
energy balance algorithm for land (SEBAL): Part 1
Formulation. Journal of Hydrology, Volumes 212–213,
Bouchet, R.J., 1963: Evapotranspiration réelle et potentielle:
signifi cation climatique. Proceedings of the International
Association of Hydrological Sciences General Assembly,
Committee for Evaporation, 19–31 August 1963,
Berkeley, International Association of Hydrological
Sciences Publication No. 62, pp. 134–142.
Bouten, W., P.J.F. Swart and E. de Water, 1991:
Microwave transmission: a new tool in forest hydrological research. Journal of Hydrology, Volume 124,
Bowen, I.S., 1926: The ratio of heat losses by conduction
and by evaporation from any water surface. Physics
Review, Volume 27, pp. 779–787.
Brakke, T.W. and E.T. Kanemasu, 1981: Insolation
estimation from satellite measurements of refl ected
radiation. Remote Sensing of Environment, Volume 11,
Brandes, D. and B.P. Wilcox, 2000: Evapotranspiration
and soil moisture dynamics on a semiarid
ponderosa pine hillslope. Journal of the American
Water Resources Association, Volume 36, No. 5,
Brest, C.L. and S.N. Goward, 1987: Deriving surface
albedo measurements from narrow band satellite
data. International Journal of Remote Sensing,
Volume 8, pp. 351–367.
Brunel, J.P., 1989: Estimation of sensible heat
fl ux from measurements of surface radiative
temperature and air temperature at two meters:
application to determine actual evaporation
rate. Journal of Agricultural and Forest Meteorology,
Volume 46, pp. 179–191.
Brutsaert, W., 1982: Evaporation into the Atmosphere.
D. Reidel Publishing, Dordrecht.
Caselles, V. and J. Delegido, 1987: A simple model to
estimate the daily value of the regional maximum
evapotranspiration from satellite temperature and
albedo images. International Journal of Remote Sensing,
Volume 8, No. 8, pp. 1151–1162.
Choudhury, B.J., 1997: Estimating land surface evaporation using multispectral satellite and ancillary
data. In: Applications of Remote Sensing in Hydrology:
Proceedings of Symposium No. 17, G.W. Kite,
A. Pietroniro and T. Pultz (eds), National Hydrology
Research Institute, Saskatoon.
Chow, V.T. (eds), 1964: Handbook of Applied Hydrology.
McGraw-Hill, New York.
de Bruin, H.A.R., B.J.J.M. van der Hurk and W. Kohsiek,
1995: The scintillation method tested over a dry
vineyard area. Boundary-layer Meteorology,
Volume 76, pp. 25–40.
Engman, E.T. and R.J. Gurney, 1991: Remote Sensing in
Hydrology. Chapman and Hall, London.
Food and Agriculture Organization of the United
Nations (FAO), 1998: Crop evapotranspiration –
Guidelines for computing crop water requirements – FAO
irrigation and drainage paper 56 (R.G. Allen, L.S.
Pereira, D. Raes and M. Smith). Rome, FAO.
Garratt, J.R., 1984: The measurement of evaporation
by meteorological methods. Agricultural Water
Management, Volume 8, pp. 99–117.
Gautier, C., G. Diak and S. Masse, 1980: A simple physical model to estimate incident solar radiation at the
surface from GOES satellite data. Journal of Applied
Meteorology, Volume 19, pp.1005–1012.
Granger, R.J., 1997: Comparison of surface and satellite derived estimates of evapotranspiration using
a feedback algorithm. In: Applications of Remote
Sensing in Hydrology: Proceedings of Symposium No. 17,
G.W. Kite, A. Pietroniro and T. Pultz (eds), National
Hydrology Research Institute, Saskatoon, pp. 71–81.
Greacen, E.L. (ed.), 1981: Soil Water Assessment by
the Neutron Method. CSIRO, Special Publication,
Gunaji, N.N., 1965: Evaporation Reduction Investigations:
Elephant Butte Reservoir. New Mexico Engineering
Experimental Station, Las Cruces, New Mexico State
University, Technical Report No. 25.
Gunaji, N.N., 1968: Evaporation Investigations
at Elephant Butte Reservoirs in New Mexico.
International Association of Scientifi c Hydrology,
Volume 18, pp. 308–325.
Harbeck, G.E. and J.S. Meyers, 1970: Present day evaporation measurement techniques. Proceedings of the
American Society of Civil Engineers, Journal of the
Hydraulics Division, Volume 96, No. HY7,
Herschy, R.W., 1971: River Water Temperature. Water
Resources Board, TN5.
Irmak, S, A. Irmak, J. Jones, T. Howell, J. Jacobs,
R. Allen and G. Hoogenboom, 2003: Predicting daily
net radiation using minimum climatological data.
Journal of Irrigation and Drainage Engineering,
Volume 129, No. 4, pp. 256–269 (http://unh.edu/
Itier, B. and C.H. Riou, 1982: Une nouvelle méthode
de détermination de l’évapotranspiration réelle par
thermographie infrarouge. Journal de Recherches
Atmosphériques, Volume 16, pp. 113–125.
Jackson, R.D., 1985: Estimating evapotranspiration at
local and regional scales. Institute of Electrical and
Electronics Engineers Transactions on Geoscience and
Remote Sensing, Volume GE-73, pp. 1086–1096.
Jackson, R.D., R.J. Reginato and S.B. Idso, 1977: Wheat
canopy temperature: a practical tool for evaluating water requirements. Water Resources Research,
Volume 13, pp. 651–656.
Kerr, Y.H., E. Assad, J.P. Freleaud, J.P. Lagourde and
B. Seguin, 1987: Estimation of evapotranspiration
in the Sahelian zone by the use of METEOSAT and
NOAA AVHRR data. Advances in Space Research,
Volume 7, No. 11, pp. 161–164.
Keskin, E.M., Ö. Terzi and D. Taylan, 2004: Fuzzy logic
model approaches to daily pan evaporation estimation in western Turkey. Hydrological Sciences Journal,
Volume 49, No. 6, pp. 1001–1010.
Klaassen, W., F. Boseveld and E. de Water, 1998: Water
storage and evaporation as constituents of rainfall
interception, Journal of Hydrology, Volumes 212–213,
Klute, A. (ed.), 1986: Method of Soil Analysis. Part 1:
Physical and Mineralogical Methods, Second
edition, American Society of Agronomy and
Soil Science Society of America, Madison,
Kohler, M.A., T.J. Nordenson and D.R. Baker, 1959:
Evaporation Maps for the United States. United States
Weather Bureau, Technical Paper 37.
Konstantinov, A.R., 1966: Evaporation in Nature.
Translated for the United States Department of
Commerce and National Science Foundation by the
Israel Program for Scientifi c Translations, Jerusalem,
Available from the United States Department of
Commerce, National Technical Information Service,
Springfi eld, Virginia.
Kuzmin, P.O., 1957: Hydrological investigation of land
waters. Bulletin of the International Association of
Scientifi c Hydrology, Volume 3, pp. 468–478.
La Mer, V.K., 1963: The case for evaporation suppression.
Chemical Engineering Journal, Volume 70, p. 213.
Lundberg, A., 1993: Evaporation of intercepted snow:
review of existing and new measurement methods.
Journal of Hydrology, Volume 151, pp. 267–290.
Marciano, T.T. and G.E. Harbeck Jr, 1954: Mass transfer
studies. In: Water Loss Investigations, Lake Hefner
Studies, Technical Report, USGS Professional
Monteith, J.L., 1965: Evaporation and environment. In:
The State and Movement of Water in Living Organisms,
Symposium of the Society for Experimental Biology,
Volume 19, pp. 205–234.
Morton, F.I., 1982: Operational estimates of areal evapotranspiration and their signifi cance to the science
and practice of hydrology. Journal of Hydrology,
Volume 66, pp. 1–76.
Mutreja, K.N., 1986: Applied Hydrology. Tata McGrawHill, New Delhi.
Nieuwenhuis, G.J., E.H. Smidt and H.A.M. Thunnissen,
1985: Estimation of regional evapotranspiration
of arable crops from thermal infrared images,
International Journal of Remote Sensing, Volume 6,
Penman, H.C., 1948: Natural evapotranspiration from
open water, bare soil and grass. Proceedings of the
Royal Society of London, A193, pp. 120–145.
Prichard, Terry L., 2003: Soil Moisture Measurement
Technology. University of California, Davis (http://
ucce.ucdavis.edu/fi les/fi lelibrary/40/975.pdf).
Priestley, C.M.B. and R.J. Taylor, 1972: On the assessment of surface heat fl ux and evaporation using
large-scale parameters. Monthly Weather Review,
Volume 100, No. 2, pp. 81–92.
Rambal, S., B. Lacaze, H. Mazurek and G. Debussche,
1985: Comparison of hydrologically simulated and
remotely sensed actual evapotranspiration from
some Mediterranean vegetation formulations.
International Journal of Remote Sensing, Volume 6,
No. 8, pp. 1475–1481.
Riou, C., B. Itier and B. Seguin, 1988: The infl uence of
surface roughness on the simplifi ed relationship
between daily evaporation and surface temperature.
International Journal of Remote Sensing, Volume 9,
No. 9, pp. 1529–1533.
Saxton, K.E., W.J. Rawls, J.S. Romberger and
R.I. Papendick, 1986: Estimating generalized soilwater characteristics from texture. Soil Science Society
of America Journal, Volume 50, No. 4, pp. 1031–1036.
Schultz, G.A. and E.T. Engman, 2000: Remote Sensing in
Hydrology and Water Management. Spinger-Verlag,
Seguin, B. and B. Itier, 1983: Using midday surface
temperature to estimate daily evaporation from
satellite thermal IR data. International Journal of
Remote Sensing, Volume 4, No. 2, pp. 371–383.
Sharma, L.M., 1985: Estimating evapotranspiration.
In: Advances in Irrigation, Volume 3, D. Hillel (ed.),
Academic Press, New York, pp. 213–282.
Shuttleworth, W.J. and J.S. Wallace, 1985: Evaporation
from sparse crops: an energy combination theory.
Quarterly Journal of the Royal Meteorological Society,
Volume 111, pp. 839–855.
Singh, V.P., 1994: Elementary Hydrology. Prentice Hall of
India, New Delhi.
Tarpley, J.D., 1979: Estimating incident solar radiation
at the surface from geostationary satellite data.
Journal of Applied Meteorology, Volume 18,
Thornthwaite, C.W. and B. Holzman, 1941: Evaporation
and transpiration. Yearbook of Agriculture, 1941,
United States Department of Agriculture,
Washington DC, pp. 545–855.
Thunnissen, H.A.M. and G.J.A. Nieuwenhuis, 1990: A
simplifi ed method to estimate regional 24-hr
evapotranspiration from thermal infrared data.
Remote Sensing of Environment, Volume 31,
United States Department of Agriculture, Natural
Resources Conservation Service, 1997: Soils. National
Engineering Handbook 210-VI-NEH, Part 652:
Irrigation Guide, Chapter 2.
United States Geological Survey, 1977: National
Handbook of Recommended Methods for Water Data
Acquisition. Reston, Virginia.
Viessman, W. Jr, G.L. Lewis, J.W. Knapp, 1989:
Introduction to Hydrology. Harper and Row Publishers,
World Meteorological Organization, 1966: Measurement
and Estimation of Evaporation and Evapotranspiration.
Technical Note No. 83, WMO-No. 201, Geneva,
World Meteorological Organization, 1976: The
CIMO International Evaporimeter Comparisons.
WMO-No. 449, Geneva.
World Meteorological Organization, 1983: Guide to
Meteorological Instruments and Methods of Observation.
Sixth edition, Part I, Chapter 6, WMO-No. 8,
World Meteorological Organization, 1992a: Application
of Satellite Technology. Annual Progress Report,
SAT-10, WMO/TD-No. 569.
World Meteorological Organization, 1992b: Remote
Sensing for Hydrology: Progress and Prospects
(R. Kuittinen). Operational Hydrological Report
No. 36, WMO-No. 773, Geneva.
World Meteorological Organization, 1993: Executive
Council Panel of Experts on Satellites: Final Report.
9–10 March, Geneva