-
J Grinùley
Estimation and mapping of evaporation
J. Grindley Meteorological Offce Bracknell, U. K.
SUMMARY: Most evaporation data in the United Kingdom are
obtained using the Penman formula, which permits calculations of
potential evaporation to be made. By making assumptions about‘ the
way in which actual evaporation falls below the potential as soil
moisture becomes limiting it is possible to calculate and map soil
moisture deficit. Careful assessment of land use permits an
assessment of actual evaporation and soil nioisturc deficit over
any specific catchment area for use in water balance studies. A
provisional map of potential evaporation has been prepared for
England and Wales and a start has been made on the preparation of a
map of actual evapora- iion using a 10 km network of grid
intersections with an appropriate spectrum of land Lise for each tn
tersection.
ESTIMA TI O N ET ETABLISSEMENT DE CAR TES R ~ S U M B : La
plupart des données relatives à l’évaporation dans le Royaume Uni
sont obtenues en utilisant la formule de Penman qui permet de faire
des calculs dc l’évaporation potentielle. En faisant des hypothèses
sur la manière dont l’évaporation actuelle est inférieure a
l’évaporation potentielle quand l’humidité du sol apporte certaines
limitations, il est possible de calculer et de cartographier le
déficit en humidité du sol. Une détermination soigncuse de
l’utilisation du sol permet unc évaluatjon dc l’évaporation
actuelle et du déficit d’humidité du sol sur l’étendue d’un bassin
spécifique, pour leur utilisation dans des études de bilans d’eau.
Une carte de l’évapo- ration potentielle a été préparée pour
l’Angleterre et le pays de Galles et on a commencé la pré- paration
d’une carte de l’évaporation actuelle en utilisant un réseau avec
des mailles de 10 km. avec un spectre approprié d’utilisations du
sol pour chaque intersection.
DE L’E VA POR A TIO N
ESTíMACíÓN Y ELABORACJÓN DE MAPAS RELATIVOS A LA EVAPORACIÓN
RESUMEN: La mayor parte de los datos relativos a la evaporación
se obtienen, en el Reino Unido, mediante la utilización de la
fórmula dc Penman, que permite calcular la evaporaciun potencial.
Estableciendo hipótesis sobre la forma en que la evaporación actual
desciendc por debajo del potencial a medida que la liuniedad se
limita, resulta posible calcular y elaborar del déficit de la
humedad del suelo. Una evaluación cuidadosa del suelo permite
calcular ia evaporación real y definir CI déficit de la humedad del
suelo para una cuenca de captación determinada, lo que sirve para
llevar a cabo los estudios del balance hidrológico. Se ha prcparado
un mapa provisional de la evaporación potencial para Inglaterra y
el país de Gales y se ha iniciado la preparación de otro para la
evaporación real, utilizando una red de 10 kilómetros dc
intersecciones en forma de cuadrícula, con una imagen adecuada de
Ia utilización de la tierra para cada una de las interseccion.
es.
200
-
Estimation anil mapping of evaporation
1. JNTRODUCTION
Although a number of evaporation tanks and a few lysimeters are
maintained in the United Kingdom, the main source of information
concerning evaporation is based on estimates using the well-known
Penman formula. This formula has received wide acclaim since its
first publication [i] as one of the most soundly based methods of
calculating evaporation using readily available meteorological
data. Basically the Penman formula provides estimates of potential
evapotranspiration, the amount of water which would be transpired
by a vegetation cover when water is at all times freely available
to the root system of the vegetation. By making certain assumptions
it is possible to estimate the actual amount of evaporation which
occurs when soil moisture limitation does impose a restriction on
potential evaporation and, as an important corollory to this, to
calculate and map soil moisture deficit.
2. THE CALCULATION OF POTENTIAL EVAPOTRANSPIRATION
The merit of Penman's method for calculating potential
evapotranspiration lies in the combination of two of the classical
approaches to the estimation of evaporation, the energy budget and
the aerodynamic. By combining these two approaches, the require-
ment for the measurement of the temperature of lhe evaporating
surface, a measurement which is often difficult and rarely carried
out on a routine basis, is eliminated. Although the formula is
soundly based physically some empiricism is inherent in the
derivation of incoming and outgoing radiation and particularly
in the aerodynamic terni. Much of the empiricism in the radiation
terms can be eliminated if measurements of global or net radiation
are available. The version of the formula used in the
hydrometeorological branch of the United
Kingdom Meteorological Office is, with one important amendment,
that published by Penman in 1963 [2]. The calculations are carried
out by computer and the formula is programmed to calculate
evapotranspiration from a vegetated surface, the albedo of which is
taken to be 0.25. This albedo is generally considered
representative of grass, most agricultural crops in most phases of
their development and deciduous woodland in leaf (but not
conifers). N o account is taken in the calculations of zero plane
displace- ment or roughness length. The formula used is:
AH + YE, A+Y
E =
where d is the slope of the saturation vapour pressure curve at
air temperature, y is the hygrometric constant (taken as 0.49 in
the program, where temperature is expressed in "C and vapour
pressure in nim Hg).
H = 0.75 R, (0.18 + 0.55;) - 0.95 Ta4 (0.10 + 0.90 ij (0.56 -
0.092JG). Here: R, is the anionnt of short wave radiation reaching
the outside of the earth's atmo-
sphere expressed in mrn water equivalent; n/N is the ratio of
observed hours of sunshine to possible number of hours; aTa4 is the
theoretical black body radiation at mean air temperature T,
(expressed in
degrees absolute) and e, is actual vapour pressure at mean air
temperature Ta.
20 1
-
J. Griiidlry
The coefñcieiit 0.95 is the important amendilient rcferred to
and is intended to allow for vegetation not radiating as a perfcct
black body [3]. Ea is given by the equalion Eo = 0.35 (e,-e,,) (I +
ü/iOO) where e, is the saturation
vapour pressure at the mean air temperature, ed is the actual
vapour pressure and U is the run of wind in miles per day.
Beginning with I970 data, estiinates of potential
evapotranspiration have been published
in the United Kingdom for about 80 stations which have
meteorological elements measur- ed four times a day averaged to
give mean values of each element for the month. Values are also
calculated but not published for stations which measure the
required meteorolo- gical elements less frequently than four times
a day. The question of the extent to which time and frequency of
measurement affect calculated values of evapotranspiration is a
subject for investigation at the moment.
3. THE MAPPING OF POTENTIAL EVAPOTRANSPIRATION
The number of stations for which estimates of long-period
average annual potential evapotranspiration are available is quite
inadequate to permit a valid detailed map of the distribution of
average annual potential evapotranspiration over the United
Kingdom. This is particularly true for upland areas. The number of
stations for which evapotrans- piration data are available is about
150 of which 75 per cent lie below 100 metres. There are only five
stations situated at a height greater than 500 metres, the highest
being at 847 metres. Nevertheless, for water balance studies and
assessment of yield from catchment areas
it is necessary to make soine assessment of areal average
evapotranspiration. To this end an altempt has been made to
construct a map showing the distribution of average annual
evapotranspiration over Britain, which, whatever its limitations,
will permit a quantitative estimate of evaporation over an area.
Such a map has been prepared, so far for England and Wales only, on
a scale of 1 :625 000. The averages are mainly based on periods of
not less than seven years within the period 1954-1966. The basis of
the construction of the map was to apply multiple regression
analysis
of evapotranspiration on altitude and any other factor which
appeared significant, commonly grid northing of the United Kingdom
National Grid. Separate analyses were carried out for River
Authority areas (major river basins
such as the Thames, Great Ouse) and estimates of potential
evapotranspiration were made, using appropriate regression
equations, for each 10 km intersection grid. At boundaries between
different areas agreement between estimates obtained for a common
grid intersection by different regression equations was generally
good, although in hilly areas discrepancies of up to 15 per cent
were occasionally discerned. For about 4- of the country, a simple
height relationship was used to estimate evapotranspiration for
cach grid intersection. The relationship, obtained from reference4,
allowed for a variation of .35 inch evapotranspiration per LOO feet
in altitude, the quantity to be added or sub- tracted to the value
for a mean county altitude depending on the difference between the
mean county altitude and that of the grid intersection. The map (a
simplified version of which is shown at figure i) must be
considered as
tcntative at the moment, particularly for altitudes above 200
metres where little factual data exist and extrapolation is wide.
It is susceptible to considerable modification and it is hoped
improvement in accuracy as data from. more stations, partic~ilarly
at greatcr altitudes, become available. Neverthelcss, it is hoped
that for heights below 200 metres thc accuracy is within 10 per
cent.
202
-
Esrimaiion und mopping of euuporuiion
O F y 460
FIGURE I. Auernge unnuul porentiul eunporniion in rnilliineters
for u surfuce with nlbeùu 0.25. Isopletlis ur 350, 400, 460, 500,
530, und 560 m m interuuls
4. THE CALCULATION OF SOIL MOISTURE DEFICIT
Soil moisture deficits are considered to have been set up when
evapotranspiration exceeds precipitation and vegetation has to draw
on reserves of moisture in the soil to satisfy transpiration
requirements. Such deficits may occur in winter but sustained
deficits do not usually arise until late spring. In drier areas of
Britain, they commonly persist until autumn or early winter and
occasionally, in exceptional cases, throughout the following
winter. Often, the sustained period of soil moisture deficit starts
in mid-month in spring and
allowance needs to be made for this by estimating
evapotranspiration for the dry period.
203
-
J. Grindley
Adjustment is made for the rapidly increasing rates of avcrage
evapotranspiration which occur in spring months. A typical exaniple
of the building up of soil moisture deficits would be:
Month Rainfall Evapotranspiration Accumulated soil milliinetrei
moisture deficiii
April 15th-30tli Nil 30 30 May 50 79 59 June 75 96 80
In assessing soil moisture deficits one needs to take account of
the fact that vegetation has increasing difficulty in extracting
moisture from the soil (because of increasing soil moislure
tension) as accumulated potential evapotranspiration becomes
greater than accumulated rainfall. The question of the manner and
extent to which actual soil moisture deficit falls below
the potential is extremely controversial and has been discussed
extensively in the liter ature. Some models postulate a divergence
between potential and actual evaporation (and hence soil moisture
deficit) at a rate proportional to the remaining sojl moisture as
soon as a deficit is set up; other models propose a linear
divergence, sometimes with threshold values at which the rate of
divergence changes sharply and others suppose evaporation to
continue at near potential rate until near wilting when actual
evaporation drops to nil. Baier’ has a useful discussion of various
models. The model used in the hydrological branch of the United
Kingdom Meteorological
Office is that proposed by Penman6 where the concept of root
constant is introduced. The root constant defines a specified
amount of soil moisture (expressed in mm equivalent depth) which
can be extracted from the soil without difficulty by a given
vegetation on a given soil. A further 25 mm of moisture can be
extracted with increasing difficulty and extraction thereafter
becomes minimal. Typical figures for a root constant of 75 mm
are:
Potential 100 125 150 175 250mm Actual 99 109 113 115 121mm
In the model used in the United Kingdom Meteorological Office
for the preparation of soil moisture deficit maps it is assumed
that each station is representative of a typical cross-section of
catchment area and a system of variable storage basin accounting is
adopted. It is assumed that 50 per cent of the area is covered by
short-rooted vegetation (grass etc.) which can draw up to 75 nun of
moisture from the soil before actual evapo- transpiration (and
hence soil moisture deficit) starts falling below the potential, 30
per cent of the area is covered by long-rooted vegetation (trees,
etc.) which can draw freely on 200 mm of soil moisture, and 20 per
cent or the area is riparian where the permanent ground water is so
near the surface that moisture is always freely available to
rooting systems and evapotranspiration is never restricted. This,
essentially, is the model adapted by Penman in his study of the
water balance of the Stour Catchment [7]. Adopting the procedure
for divergence between actual and potential evaporation
outlined in reference [6], the modified procedure for
estiinating soil moisture deficits over a catchment area
becomes:
~ 2 0 0 C,, Catchment area
C, Month Raiiifall (R) Evapotrans- R-E pirution (E)
millimïlres
June 75 96 -21 80 80 80 64 July 13 I o0 - 87 167 167 I14 107
August 2 90 - 88 255 237 I22 132
20 4
-
E.stimution und mapping of evuporrition
____ 3 4 5 6
ESTIMATED SOIL MOISTURE DEFICIT
0900 GMT 29..Qç.r~har.I969 .....
Areas with no sail moisture deficit are shaded. Remaining areas
are bounded
FIGURE 2.
Here C, indicates the accumulated potential soil moisture
deficit; C,,, the accumulated deficit over the zone with
long-rooted vegetation and C,, the accumulated deficit over the
zone with short rooted vegetation. The “areal” soil moisture
deficit represents the average of the deficit over the C,, zone (50
per cent of the total area), the deficit over the Czo0 zone (30 per
cent of the area) and zero deficit over the riparian zone (?O per
cent of the area). W h e n rainfall exceeds evapotranspiration the
soil moisture deficits are reduced by the
difference between the two amounts. For a number of years, end
of month values were obtained by using inontlily totals of rainfall
and evapotranspiration. It becaine apparent
205
-
J. Grindley
that such a procedure may seriously underestimate the
accumulated potential soil niois- ture deficit if, for example, lhe
bulk of a month's rainfall should fall in the last few days of the
month following a very dry period in the earlier part of the month.
Estimates are now made on a daily basis therefore. Additional
complications arise with this system of daily accounting,
particularly when the actual rate of evapotranspiration has fallen
below the potential and alternating wet and dry spells occur. Such
complications are dealt with in reference [8] where the estimation
of soil moisture deficits is discussed more fully. It should be
noted that excess rainfall over evapotranspiration will make an
almost immediate contributioii to run-off over the riparian zone
(C,,) and that the zone with
0900 GMT .12.Nnvernber..1969 ....-
Areas with no soil moisture deficit are shaded. Remaining areas
are bounded by -.a, ~.55,'.2.,IB.4.jnch.!.ffl~ ._.__.._ _..__
1
9
7
I
7
FIGURE 3.
206
-
Estitmiion and niupping of evuporotion
short-rooted vegetation (C75) in general will have deficits madc
good (and hence be contributing to run-off) before zone C,,, . The
United Kingdom Meteorological Office has been preparing regular
estimates of
soil moisture deficits in map form and as tabular data ror River
Authority areas for a number of years. This information,
accompanied by a verbal description relating soil moisture
variations to changes in weather patterns, has been distributed,
usually twice a month, to interested autorities since September
1962. A sequence of maps is shown in figures 3-4. The maps are
intended as an aid to authorities responsible for flood warning and
to engineers interested in the delay which might be expected before
appreciable contribution to surface and ground water reserves
occurs.
FIGURE 4.
207
-
J. Grindley
5. THE DERJ.VATION OF ACTUAL EVAPORATION FROM SOIL MOISTURE
DEFICIT
Actual evaporation can be derived quite simply for a month or
any other specified period by subtracting accumulated soil moisture
deficit at the beginning of the period from that at the end and
adding to the difïerence any rainfall which has occurred during the
period. Thus, for the example in the previous section, actual
evaporation would be:
Month R PE R-E CP czoo c7 5 Area
mm mm mm SMD AE SMD AE SMD AE SMD AE
June 75 96 -21 80 96 80 96 80 96 64 96 July 13 100 -87 167 100
167 100 114 47 107 74 August 2 90 -88 255 90 237 72 122 10 132
45
Here R signifies rainfall; PE potential evaporation; SMD soil
moisture deficit and A E actual evaporation. Actual evaporation
over the short-rooted zone (C,,) is only 47 mm in July compared
with a potential of 100 mm, and in August 10 mm compared with a
potential of 90 mm. The areal actual evaporation is obtained by
multiplying the value over the riparian zone, Cp, (always the
potential rate) by 0.2, the value over the C,,, zone by 0.3, the
value over the C,, zone by 0.5 and summing the products.
6. SOIL M O I S T U R E DEFICIT AND ACTUAL EVAPORATION OVER
SPECIFIC CATCH MENT AREAS
From the outset it was realised that the generalised catchment
model with three storage capacities, used in the preparation of the
soil moisture deficit maps, could not be applied to specific
catchment areas where widely different land use might apply.
Accordingly a more detailed model has been developed which takes
into account a
much wider spectrum of vegetation type and associated root
constants. Advice was obtained from Lhe agricultural branch of the
United Kingdom Meteorological Office on themaximum amount of water
likely to be extracted before wilting became permanent. These
amounts for a wide variety of crops are shown in table 1. These
maximum deficits have been equated with root constants as
follows:
Maximum soil moisture deticit Root constant mm m m
250 200 150 125 1 O0 50
200 I40 97 75 56 13
208
-
Estimation und mapping of evuporution
TABLE 1. Land Utilization Survey
Crops Maximum Actual SMD millimetres
Wheat Barley Oats Mixed Corn Rye Potatoes, first early varieties
Potatoes, maincrop and second earlies Beans and Peas Turnips,
Swedes and Fodder Beet Mangolds Rape (or Cole) Kale Cabbage, Savoys
and Kohl Rabi Mustard Other Crops Sugar Beet (for Sugar) Hops
Orchards, grown commercially Orchards, not grown commercially Small
fruit, grown commercially Vegetables; Hardy Nursery Stock; Flowers;
Crops under Glass Small Fruit and Vegetahles, not grown
commercially Bare Fallow Lucerne Clover, Sainfoin Temporary Grasses
Permanent Grass Rough grazing Permanent woodland
200 200 200 200 200 1 O0 150 1 O0 150 150 125 125 125 125 125
150 200 225 200 150 I O0 I O0 25 150 1 O0 1 O0 125 50 250 but 125
on poor land
The amount by which actual evaporation falls below the potential
for each root constant has been expressed in tabular form following
the model in reference [6]. The root constants refer strictly to
vegetation type and make no reference to soil type
except in so far as the soil type is reflected in the vegetation
which it carries. Where soils are known to be poor allowance can be
made by reducing the root constant. Certain seasonal restrictions
are imposed for different vegetation cover and other
forms of land use are also taken into account as follows:
1. Bare ground, fallow. Evaporation is assumed to take place at
the potential rate until 25 mm of moisture has been extracted when
evaporation ceases until more rain has fallen.
2. Riparian area. Evapotranspiration always takes place at the
potential rate.
3. Open water. Evaporation is considered to be 20 per cent
higher than the potential evapotranspiration from a vegetated
surface.
4. Urban areas. These are taken to be the areas coloured grey on
the United Kingdom I :63 360 Ordnance Survey maps. It is assumed
that 25 per cent of the area so obtained is water-proofed
(pavements, roofs, etc.), the remaining 75 per cent consisting of
gardens, parks, open ground etc. Evaporation from the water-proofed
zone is consid- ered to occur only on days of rainfall and is taken
to be the rainfall or the potential evaporation whichever is
smaller.
209
-
J. Gritidley
5. After harvest (convential data for United Kingdom, 1st
August) it is assumed that evaporation from cereal areas takes
placc as froni fallow ground.
6. For certain crops when ground cover is incomplete in the
spring it is assumed that evaporation takes place at 4- the
potential rate until cover is complete (conventional data Ist May).
The proviso is made that up to 75 nim may evaporate at the
potential rate .:as with bare ground).
7. Rough grazing. An upper limit of 50 mm extractable water is
imposed.
8. A major weakness of the model is that no proviso is made for
direct surface run-off. For much of lowland Britain the assumption
of no surface run-off may give rise to no serious, error, except in
intense thundery rain, but for mountainous Britain the assumption
is clearly not tenable. For such areas an attempt has been made to
allow for direct surface run-off by measuring the areal extent of
precipitous (gradient greater than 1 in 3), vegetationless slopes;
the information is obtained from Ordnance Survey maps. For such
slopes it is assumed that all rainfall runs off and no evaporation
occurs.
Typical examples of distribution of land use for (u) an area
containing a good-deal of urbanisation in northwest England and (b)
a predominantly farming area in East Anglia are:
Land use type Root constant Per ccnt of total area mm (4 íb)
Cereais, etc. Rootcrops, etc. Permanent grass, etc. Temporary
grass, etc. Rough grazing Woodland Fallow Urban (waterproofed area)
Riparian O p e n water
140 97 75 56 13 200
14.7 3.3 39.4 16.7 10.7 1.7 0.4 7.2 4. I 1.8
60.8 10.7 10.2 9.8 1 .o 2.6 0.9 0. I 3.9 0
100.0 100.0
In carrying out water balance studies of catchment areas, the
qproach has been to use cclumpeá” estimates of rainfall aiid
evaporation over the catchment area. Areal rainfall has beeil
estimated on a daily or mcnthly basis by using a network of
stations covering the ai-ca aad preparing arcal gcneraJ values by
(o) ariíhmetic mean; (E) Thiessen weighting: (e) e?cpressing
rainfall at each station as a percentage of station long-period
coverage, meaning percentage values an¿ applying percentage mean to
thc areal 1.sng-period average to obtaiii a general quantitative
value for the area; (dj cartographicülly. Method (e) has been the
one most generally used in the U iiited Kingdom Meteorological
OfTice. Estimates of areal general potential evaporation ûre
prepared by íìrst of all assessing
a long-period avcrage cartographically, using the distribution
of i:;ûpleths of potential evaporation from the niap discussed in
Section 3 and illustrated in figurc I. Areal potential evaporation
for a given month is obtaincd by a method analogous to method íc)
for obtaining areal gcncral rainfall i.e. a network of evaporation
stations is used lo repre- sent the catchment area, potential
evaporation at each station is expressed as a percentagc
210
-
Estiinution unci mapping of evuporniioii
of station long-period average, percentage values are ineaned
and the mean value applied to the areal average to obtain a general
quantitative value. Potential evaporation is not calculated for
periods shorter than a month. Where estimates of daily evaporation
are required long period estimates of daily evaporation for an
annual average of 500 mm obtained from empirical curve are
increased in the ratio of calculated monthly potential ’
evaporation for the month to long period monthly evaporation.
Estimates for any particular day may be in error using this method
but errors are not likely to be accumu- lative and in any case must
cancel out by end of month. With the data for rainfall and
potential evaporation so obtained estimates of general actual
evaporation over an area are obtained using the methods set out
above with appropriate land use apportionments.
O 1 miles IO 20
SCALE I:b25,000 With National Grid
I I I I I I I I I
O
- Rivers -.- Boundary of River Authority
FIGURE 5. Mersey und Weauer River Authority. Average unnual
potential eucrporulion (mt r i )
7. MAPPING ACTUAL EVAPORATION
The calculation of evaporation using “lumped” data masks a
considerable areal variation in rainfall and evapotranspiration,
the variation moreover acting in an opposite sense with rainfall in
general increasing with altitude and evapotranspiration decreasing.
A more desirable procedure would be to carry out calculations on a
point basis using a
21 1
-
J. Grindley
O 2 0 SCALE 1:625.000 W t h National Grid I miles
io
I I I I I I I I I I 40 I
- Rivers -.- Boundarg of River Authority FIGURE 6. Mersey und
Weuuer River Authority. Averuge unnuul ucfucrl euaporation
(nini)
close grid network, each point having its appropriate land use
value or spectrum of values. Isopleths of actual evaporation would
then be drawn and the point values integrated (by planimeter) to
obtain an areal value of actual evaporation. Such a proceûure has
been carried out for one River Authority area and a map of
actual evaporation constructed for points at the intersection of
a 10 km grid network. Average annual rainfall at each point was
obtained from survey maps and average annual potential evaporation
from the larger scale version of the map illustrated at figure 1. A
spectrum of land use was made for an area of 4 km2 surrounding each
major (10 km) grid intersection. The distri bution of derived
potential and actual evaporation arc shown in figures 5 and 6
respectively. The areal estimate of annual evaporation over a
ten-year period using lumped data for the catchment was 488mm and
that using integrated point values was 480 mm. Average annual
potential evaporation was 531 mm. It is hoped ultimately to extend
the map of long-period actual evaporation over the
whole country. When such a map is available average annual
actual evaporation at each 10 km grid intersection can be
substracted from long period rainfall to provide an esti- mate of
annual run-off at the point. A map of the derived distribution of
annual run-off can then be prepared.
21 2
-
The prediction of actual euuporation it2 semi-arid arem
REFER ENCES
1. PENMAN, H.L. (1948): Natural evaporation from open water,
bare soil and grass, Proc. R.
2. PENMAN, H.L. (1963): Woburn irrigation 1951-1959: (I)
Purpose, design and weather, J. Agric.
3. BUDYKO, M. I. (1956): The heat balance of the earth’s
surface. Leningrad, Gidrotneteoizduf. 4. Ministry of Agriculture,
Fisheries and Foods, (I 967): Technicol Bulletin No. 16, Potential
transpiration, London, Her Majesty’s Stationery Office.
5. BAIER, W. (1967): Relationships between soil moisture actual
and potential evapotranspiration, Proc. of Hydrology Syitiposiirm
No. 6 heldat Utliuersity of Snskatcheiuun on 15 und 16 Noueniber
1967.
6. PENMAN, H.L. (1949): The dependence of transpiration on
weather and soil conditions, Jnl. Soil Sci. Oxford, 1, pp.
74-89.
7. PENMAN, H.L. (1950): The water balance of the Stour catchment
area, Jnl. Innstn. Waf. Etigrs., London, 4, pp. 457-469.
8. GRINDIXY, J. (1967): The estimation of soil moisture
deficits, Mel. Mug. London, vol. 96, pp. 97-108.
Soc., London, A, 193, pp. 120-145.
Sci. Cambridge, 58, pp. 343-348.
The prediction of actual evaporation in semi-arid areas
J. V. Sutcliffe, Institute of Hydrology, Wallingford and C. H.
Swan, Sir Alexander Gibb and Partners, London
SUMMARY: A study of the world water balaiice based on rainfall
and potential evaporation alone will not lead to a prediction of
actual evaporation in semi-arid areas. O n the other hand, it would
be an immense undertaking to measure the total runoff from a region
to deduce empirically actual evaporation by comparison with
rainfall. It is, therefore, essential also to consider the role of
the soil moisture regime in dividing rainfall between actual
evaporation and runoff. The dangers of the simple comparison of
rainfall and potential evaporation are illustrated by
comparing the two sidcs of the Alborz range in Iran. The
southern slopes receive moderate winter precipitation but because
of this seasonal distribution the actual evaporation is low
compared with he high annual potential evaporation; the runoff is
relatively high. The northern slopes facing the Caspian Sea receive
summer rainfall as well but the actual evaporation is high and the
runoff low. The clue to this apparently anomalous situation is the
forest on the Caspian slopes which transpires at the potential rate
where the perennial rainfall allows it to survive. A study which
took account of the limited soil moisture storage and the seasonal
precipitation pattern would predict this result at least in
qualitative terms. While global and regional rainfall maps,
together with maps of potcntial evaporation, are an
essential first step towards a water balance, any realistic
study must take account of the existing information on regional
runoff, soil depths, and vegetation.
213