$
;411@'Q .t*
9 .
&@.*:k
r:,
,lep*.

2%
COMMISSIE VOOR HYDROLOGISCH ONDERZOEK TNO MMITTEE FOR HYDROLOGICAL RESEARCH TNO
Verslagen en Mededelingen No. 28 F'roceedings and Informations No. 28
y:,
&
"F,@
% ? ,;
IRATION IN RELATION ' HYDROLOGY
R
EVAPORATION IN RELATION TO HYDROLOGY
COPYRIGHT @ BY CENTRAL ORGANIZATION FOR APPLIED SCIENTIFIC RESEARCH IN THE NETHERLANDS TNO, 1981
1981 .COMMISSIE VOOR HYDROLOGISCH ONDERZOEK TNO COMMITTEE FOR HYDROLOGICAL RESEARCH TNO Verslagen en Mededelingen No. ONDERZ. COMM. HYDROL. MEDED. 1981) VERSL.THE HAGUE . 28 Proceedings and Informations No. TNO 28 . 28 EVAPORATION IN RELATION TO HYDROLOGY PROCEEDINGS OF TECHNICAL MEETING 38 (February.
.
1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4. . . . . . . . . . . . . Introduction . . . . . . . . . . . . . Physicalmathematical models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2. . . . . . . . . . . . . . . . . . . 4 . . . Soil physical properties . . . . . Boundaries of the unsaturated zone . . . . . . . . . . . . . . . . . Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5. . . . . 5 . . . .R. . . . . . Combination methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Users of evaporation data . . . . . . . . What can be achieved with additional weather observations? . . . . . . .A. . . . . . VAN BAKEL . . . . . Introduction . . . . . . . . . . . . . . . . . . . . . . . The routine evaporation data published by the KNMI . . . 2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1. . . . . . . . . . . . . Problem formulation . . . . . . . . . . . . . Summary . . . . . .J. . . . .1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reference crop evapotranspiration and crop water requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. . . . . . Comparison between ET and ETo . . . . 3. . . 5 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. .2. . . . . . . . . . . . . . . . . . 2. . . . . . . . . . . . . . . .1. . . . . . . . . . . . . . . . . . . . . . . . General theory . . . . . . . . . . . . 3. . Alternative methods for the determination of ETo . . . . . . . . . . . . . . 4. . . . . . . . Actual evapotranspiration . . . . . Introduction . . . 3. The energy balance of the earth's surface .3. . . . . . 3. . . . . . . . . . . . . . . . . . . 4 . . . . . . . . . . . . . . . . . .1. . . . . . . . The determination of (reference crop) evapotranspiration from routine weather data. . . . 7 . . . . . . . . 3 . . . . . . . . . . . . . . . . . . KEIJMAN . . . . . . . . . . . . . . Unsaturated zone and evapotranspiration. . . . . References . . . . . . . . . . Theoretical background of some methods for the determination of evaporation. . . . . . . . . References . . . . . . Steadystate capillary rise . . . . . . . . . . . . . . . . . . . . . . 4 . . 6. . . . . . . . . . . . . . . . . . . . . . . . . . . Transport equations . . . . . . . . . . . . . . . . . . . . . . . . . . . H. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Parametric models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Central theme of the meeting . . . . . . . Nonstationary flow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2. . . . . . . . . . . . . . . . . . . . . Water uptake by p h t roots . . . . . . . . . . . . . . . J. . . . . 5. . . . . . . Long term records of E and rainfall . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. 2 . . . . . . . . . . . . . Evaporation under practical conditions. . . . . . . . . . . . . . . . .2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3. . . . . . . . . . . . . . . . . . . . . . . . . . . .Q. . The routine weather data .3. . . . . . . . . . . . .CONTENTS 1. . . . . . Summary . . . . . . . . . . . . . Comparison of the different methods . Atmospheric system . . . . . . . Concluding remarks . . . . . . . . . . . . . . DEBRUIN . . . . . . . . . . . . . . . . . . . . . . .T. Unsaturated zone . . . . . . . . . . . . . . . . . . . . . . . Empirical methods . . . . . . . . . . . . . . . . . . . . . . . . 1. . . . . . . .WARTENA . . . . .2. . . . . . 5 . . . . . . . . . . . . . . . . . . . 4 . . . . . . . . . . . . . . . . 4. . . . . . . . . . . . . . . . . . . . . . . . 3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. . . . . . 3 . . . . 3 . . . . . . . . . . . . . . . . . . . . . . 3 . . . . . . . . . The advectionaridity approach . . . . Introduction . . . . . 4 . . . . . . .4. . . . . . . . . . . . . . . . . . . . . . . . Acknowledgement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. . . . . . . . . . . P. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . Conclusions and recommendations . . . . . 2 . . 2 . . . . . . . 5. . . . . . . . . . . . . . . Integated approach to evaluate effects of changes in soil water conditions on evapotranspiration . . . . . . . . . . . 3. . L . . . . . . . . . . . . . . . . . . . . . . . . . . . Saturated groundwater system . . . . . . . 2 . . . . . . . . . . . . 2. .
. . . . . . . . . . . . . . . . . . . . . . . . . . . . 4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Results and discussion . . References . . . . . . . . . . . References . . . . . . . . . . . . . .. . . . . . . . . . . . . . Summary . . . . . . Evapotranspiration: advance and future outlook. . . . . . . . 3. . . . . . . . .7. . . . . . . . . 4. . . . . . . . . . . . . . . . . . . 3. . . . .. . . . . . . .6. . . . .. . . . . . . 3. . . . . . . VELDS Summary . . . . . . . . . . . . . . . . . . . Estimation of actual evapotranspiration by the advectionaridity method . . . . . . . 3 Calculation of potential and actual evapotranspiration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. . Actual evapotranspiration . . . . . . . .N. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Water balance . . 2. . . . . Potential evapotranspiration . . . . . .1. . .2. . . . . . . . . . Potential evapotranspiration by the methods of Penman and Thom & Oliver . . . .6. . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Problems with evaporation research today . . . . . . . J. . . Lakeevaporation . . . . . . . . . . . . . . . . . . . . . . . . . 3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Rainfall excess: Period March t o August . . . . . . . 3. . . .. . . . 4. . . . . . 4. . . . . . . . . . 3. . . . . . . . . . .. . ... . . . 1. . Inst~um_entation . . . . . . . . . . . . . . . . .. . . . . . . 4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Conclusions and recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1. . . .5. . . . . . . . . . . .. . . . . . . . . . . . .1. . . . . . . . . . . . . . . . . . 5 Future outlook . . . . . . . . . . . Advection . . . . . . . . . . . . . . . . 4 . . 6 . . . . . . . . . . . . . . . . .3. . . . . .. . . . . . . . .4.. . . . . . . . . . . . . . . . . 6 . . C. . . . 2. . 3 Development during the last 20 years . . . . . State of the art of evaporation 19581960 . . . . . . . . . . . . STRICKER . . . . . . . . . . . . . . . . . 4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . Development in theory and formula's . .2.1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5. . . . . . . . . . . . . . . . . . . . . . . . 3. . . . . . .A. . . . . . . . . . . . 3. Introduction . . Interception . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . Interception . . . . . . . . . . . . . . . . Methods of estimating evapotranspiration from meteorological data and their applicability in hydrology. . . . . 4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . .4 Use of the methods of Penman and Thom & Oliver or of Priestly & Taylor under moisturelimiting conditions. . . . . . . . . . . . . . . . . . . . . . . . .M. . . . . . . . . . Introduction ... . . . . . . . . . . . . . . . . . . Acknowledgement . . . . . . . . . . . Evapotranspiration of vegetation . . . . . . . . . . . . . . .5. . . . Soil water evaporation . . . . . . . . . .2. . .. . . .7. . . . . . . . . . . . . . . Experimental site and available data . . . . . . . . . . . . . . . . . . . . . . . . . . Potential evapotranspiration by the method of Priestly & Taylor . .3.
N.R.M. Department of Hydraulics and Catchment Hydrology. P. J. KEIJMAN Drs. STRICKER Dr.A. Wageningen Royal Netherlands Meteorological Institute. DE BRUIN Ir.T. De Bilt Royal Netherlands Meteorological Institute. Wageningen Royal Netherlands Meteorological Institute. VELDS Agricultural University.A. ir. dr.AUTHORS Prof. C. Wageningen Agricultural University. H. Department of Physics and Meteorology. De Bilt . VAN BAKEL Ir. J.J. L. De B i t Institute for Land and Water Management Research. WARTENA Drs.Q.
.
That has evidently influenced the very character of this research. WARTENA 1. and nowadays data can in situ be processed successfully. do not always emphasize identical aspects. the very questions from users are rather divergent and. there were no basic data enabling the application of models. the user of meteorologic knowledge still had to exercise some patience. hydrology has clearly done more than using meteorological knowledge thus offered or otherwise available. over the last few decades. In 1958. understandably and rightly SO. It is an indication that. That meeting spent a considerable part of the time in describing the physical processes underlying evaporation. More robust sensors have been developed. PROBLEM FORMULATION Today's emphasis will be on the empirical cum practical approach. Hydrology has distinctly stimulated the development of evaporation research. It is gratifying to note that a special day can be devoted to hydrological aspects of evaporation science. Generally speaking. They are principally interested in the . many requisite types of measuring were too complicated. Knowledge has deepened. therefore. over the last twenty years or so. it covered three days. and measuring turned out to be more difficult. and the pertinent instruments were not suitable for long use in a catchment area.9 EVAPORATION UNDER PRACTICAL CONLlITIONS L. the stimulus has worked well. In fact. Meteorologists themselves use evaporation data. under trying weather conditions and without supervision. INTRODUCTION In meteorology some topics enjoy particular interest from other disciplines. Attention has meanwhile been paid to all these points. Today's papers will focus on hydrological application. 2. and results have been forthcoming. The previous technical meeting on evaporation was held in 1958. In other words. USERS OF EVAPORATION DATA Incidentally. The interest of hydrologists for the evaporation of water is a case in point. 3. The expectable came true: The process of evaporation was found to be more complex than first thought. the physical models had hardly ever been tested in terms of hydrological problems.
and that is a good thing for a1 its users. hopes will have to be set on renewed joint thinking. This interalia means that the requirements of accuracy for evaporation shall be the same as those for sensible heat. is today's central theme. please. methods for calculation and measurement that are useful in present practice. The timescale such a farmer will use in his thinking tends to cover a few days. and by Stricker and Van Bakel. The culprit is often found to be a narrowness of physical base. a hydrologist cannot keep waiting indefinitely. Therefore: a little more patience. have evolved. perhaps a few weeks. 4. The lengthscale is usually a few thousand kilometers. He want to know evaporation values for his soil plots. or sometimes days. He really needs evaporation figures. He is moreover particularly interested in the differences. Research workers trying to solve rather fundamental problems cannot deal with all subproblems simultaneously. the smaller the hazard of unpleasant surprise. CENTRAL THEME OF THE MEETING Meteorologists thus interested in hydrology. and that advective transports shall not be ignored. there is a much higher or a much lower crop. Application research is by definition much needed. it is to be considered by De Bmin. On the other hand. That.changes that will be caused in terms of atmospheric behaviour due to the vapour input. In the event. The more these empirical models contain some basic physical elements. higher up and lower down. That is the price eventually to be paid for accepting an empirical approach. he will tend to gamble. all sorts of advective influences play a role. of a plant or crop externally moistened by dew or rain. He is generally interested in advection as relevant to distances of some meters at the borders of soil parcels. Therefore. relying on sprinkling for his crop. But he does want to know whether evaporation is highly affected by environmental factors. However. if necessary. two hydrologists oriented in terms of meteorology. using a fair amount of physical sense in the process. if for instance. a phytopathologist usually handles timescales of one to a few hours. Neither is he interested in sensible heat transmission. there are ways that lead more quickly to usable results. a meteorologist with hydrologic affinities that is known in this group of scientists. as the three central presentations will show. one should be more wary in this respect now that . I mean the empirical techniques. and hydrologists comparably familiar with meteorology. Just the same. I briefly mentioned the physical backgrounds of empirical formulae. and is less interested in boundary effects or divergency of evaporation. along the route I just sketched. Essential shortcomings of empirical models are liable to show up after years of application. The timescale involved may run from a few hours to a few days. Actually. Fundamental research on evaporation is continued. This. may well be interested in the pF of its rootzone. An "all weather" farmer. I suggest. The wishes of the hydrologist are different again.
In other words. also.twenty years ago. led to specialisation. There is a danger at present that. that one then tries to amend the empirical formulae without making any essential correction. the connection with meteorologists engaged in evaporation research is missed. And. all those are some of the fields now under investigation through evaporation research that corrects or diverts its boundaries. coherent turbulence structures. At the time. self preserving flows. The complexity of inherent problems. most of the evaporation research activities were performed to benefit hydrology and quite a number of hydrologists were actively engaged. Mr. then experienced. which I specified earlier. roughness transitions. He will restrict himself to the fundamental science used in the operational methods featuring later in today's programme. Before embarking on the items of practical approach. Keyman will deal with the physical backgrounds of evaporation. . Internal turbulent boundary layers. with only little change of attaining limited success. in the event of shortcomings shown up.
INTRODUCTION The purpose of this lecture is to give a survey of the equations which will be used in the subsequent lectures. The theoretical foundations of the methods for the determination of evaporation are twofold: on the one hand one has balance equations and on the other hand transport equations.0. Evaporation from forests requires a special approach owing t o the great aerodynamic roughness of a forest and the large fraction of intercepted precipitation. The method has been extended by a. Then one can use simplified versions of which those of PriestleyTaylor and Makkink are wellknown examples. 2 . In many cases the necessary input data for this method are lacking. . 1 .Q. the sensible heat flux. the evapotranspiration. The combination method was introduced by Penman. Subsequently the equation for the energy balance of the earth's surface. KEIJMAN SUMMARY The theoretical foundations of the methods for the determination of evaporation are balance equations and transport equations.THEORETICAL BACKGROUND OF SOME METHODS FOR THE DETERMINATION OF EVAPORATION J. With regard to balance equations we will confine ourselves to the energy balance of the earth's surface. Fluxes towards the surface are given a positive value. The water balance will not be considered. The transport equations are based on the similarity theory of MoninObukhov. the heat flux into the soil. The concept of a diffusion resistance is introduced in the transport equations. THE ENERGY BALANCE OF THE EARTH'S SURFACE The energy budget of the earth's surface can be expressed by where Q* H ET h AET G is the net radiation. Of the balance equations only the energy balance of the earth's surface is treated. Rijtema and ThomOliver. the latent heat flux. the latent heat of vaporisation of water. transport equations and combination methods will be treated. A powerful1 method for the determination of evaporation is the combination of the energy balance and the transport equations. Monteith.
As the reflection coefficient r is in good approximation a constant during t the day. The quantity K' is called global radiation.The net radiation. The erratic course of the global radiation K\S due to variations in cloudiness. This radiation has a wave length between 4. The short wave radiation is emitted by the sun. The arrow .15 to 4. U the StefanBoltzmannconstant (5. Average values are given in table 1.0 and 50 pm.74 pm. The figure shows that the variation of the long wave components is very small. thus where r is the reflection coefficient of the surface. K shows the same course as K'. 1.400. Kt is the fraction of K' which is reflected at the earth's surface. emitted by the atmosphere. . The comt ponent L has a maximum in the middle of the day because surface temperature is then maximal. Many natural surfaces can be consisdered black surfaces.K4) and To the temperature of the surface (K). The variation of L' and Lt is so small because they are mainly a function of the absolute temperature. The C component L decreases slightly as a consequence of decreasing cloudiness. also called radiation balance. The downward long wave radiation. is made up of the following components: Here K means short wave and L long wave electromagnetic radiation. Then the net short wave radiation can be written as The long wave radiation emitted by the earth's surface can be expressed by where e is the emissivity of the surface.0 pm. About 45% of it consists of visible light with a wave length between 0.1denotes incoming and the arrow f outgoing radiation. The difference of these two radiation flues is called the net long wave radiation: An example of the daily course of the components of net radiation is given in fig. The long wave radiation is emitted by the earth's surface and certain components of the atmosphere as cloud particles and the gases water vapour and carbon dioxide.67 X 10S Wm* . t L'. which implies that E = 1. It is confined to the wavelength interval from 0. is usually smaller than L .
. .......400 .. 3 5 7 L .... Daytime is period with K$ > 0............. .......... .... Q* is the resulting net radiation L$) Table I Daytime and daily averages of the components of the net radiation above a grass cover measurred 14 Aug........... It has a sensor consisting of a small blackened plate equipped with a thermopile in such a .. radiation. .........._... ............. daytime averages daily averages The net radiation can be measured directly with a net radiometer or net pyrradiometer...........................................Kt)and long wave (Lf... I 9 11 13 15 17 19 21 +GMT Fig.. I Example of the diurnal variation of incoming and outgoing short wave (K(..... 1980 at Cabauw.
k the Von Khrmhn constant. TRANSPORT EQUATIONS The transport of sensible heat and water vapour can be calculated from the vertical gradients of wind speed. The soil heat flux G is usually measured with a heat flux plate similar to that used to measure net radiation. So the output of the thermopile is proportional to the heat flux in the soil. stability functions. L. zoh and zov the roughness lengths for momentum.way that one set of junctions is in contact with the upper face while the other set is attached to the lower face. e. it can be estimated from standard weather data. $h and $. T(z) and e(z) are wind speed (msl). T. The factor of proportionality has to be determined experimentally. K + L . 3.. If no direct measurements of Q* are available. The plate is placed horizontally in the soil near the surface. 1979. So Q* is proportional to the thermopile output. First the transport equations for the quxes of sensible and latent heat will be reviewed. and the outgoing. zom. temperature (K) and vapour pressure (mbar) respectively. the Obukhov length (m).3 to 100 pm. De Bruin and Kohsiek.g. These data are sunshine duration or cloud cover. temperature and water vapour. The quantities U. To overcome this effect the plate is either forcefully ventilated at a constant rate or protected by hemispheric domes of polyethylene. one can arrive at the evaporation in a number of ways which will be treated in the subsequent sections. This material is virtually transparent to radiation with wave lengths in the band of 0. Gm. When the plate is aligned parallel to the surface. These gradients are related to the corresponding fluxes by the socalled fluxprofile relations which are based on the simiiarity theory of MoninObukhov (1954): where u(z). 1978). are defined by: . radiation fluxes. sensible heat and water vapour transport respectively (m). the tempera& ture difference between the two faces is proportional to difference of the incoming. temperature and humidity of the air at screen height (see e. The heat flux through the plate is approximately equal to the flux in the soil at the same depth. and L.. K t f t L'. With Q* and G measured as described above. This factor of proportionality depends however on the wind speed over the plate. Oke.
1975). the specific heat of the air.This assumption gives comparatively small errors in combination formulas (Thom.h = z = zom. In practice for a wide range of crops the value of d is approximately given by d = 0. See also Wieringa (1980) for a revaluation of the results of Businger et al. For a vegetated surface it can be estimated with z.K') ' (K) (m. 1970. (8) and (10) the flux of sensible heat can be determined if the temperature at the heights zl and zz and the wind speed at the height 2. 1972). Similarily to (1 1) the flux of latent heat is given by . = 0. Generally one assumes JIv = $h. The stability functions $. (1971). Then one has By iteration H and u can be determined from (1 1) and (12). 1973. 1974). are measured.~) (0. the zero plane displacement d has to be introduced and zd must be taken as height of measurement instead of z. simply assumes z. If the surface is covered by vegetation.7 h (Monteith. Paulson. and approximately a constant. About zoh and zov much less is known and in many cases one . the absolute temperature. 1971. 1973. T g y is the surface stress.~. the density of the air.66 mbar K' at sea level) About the roughness length zom much information is available..where T p c. Dyer. the acceleration due to gravity. Brutsaert.. This wl also be il done in this lecture. With (7). 2) (kg.m 3) (Jkg .13 h where h is the vegetation height (Monteith. and $h have been determined by experiment (Businger et al. 1975). Brutsaert.
is given by With Q* and G measured as described before and p from (14). In fig. Q* is the net radiation. 2 an example is given of the energy balance of a grass cover. 3 refers to a situation with a shortage of soil water. regardless of stability. The soil heat flux G is very small compared with Q*. Fig. We see that E T is much greater than H. Now H is during a part of L L 5 6 7 8 9 10 1 12 l 11 15 16 17 18 19 20 21 22 23 21 1 1 3 2 B 3 L GMT F g 2 Example of the diurnal variation of the 4 terms of the energy balance of a grass cover well i. It is a reliable method and often used as a reference. supplied with water. There was no shortage of water in the soil. 1977). . one has for the flux of latent heat This method for the determination of XET is usually called the energy balancelBowen ratio method. It is often neglected or estimated as a small fraction of Q*.From (1 1) and (13) it follows that the Bowen ratio p. AET and H the fluxes of latent and sensible heat and G the soil heat flux (from De Bruin and Kohsiek.
the surface temperature (K) and the vapour pressure at the surface. One can write the equations for H and AET as H = pc. The concept of transport resistance is often introduced in combination formulas. TO eo are the resistance to the transport of heat and water vapour in the air layer between the surface and the height of measure(s. Comparing (16) and (1 7) with (8) and (9) and assuming zoh = z. 2.100 I I I I I I I I I I I I I I I .. The evaporation reaches early in the morning a certain value which cannot be exceeded because of the restricted water supply to the roots. rah To where rah and r.. 3 The same as fig.m') ment z. the day greater than AET. (mbar) = zo.gives . but the soil drying out. . ( 4 5 6 7 8 9 10 11 12 13 14 15 16 17 l8 19 20 21 22 23 24 1 2 4 c GMT 3 Fig. I I I I I I I .
1965. Combination equations as (23) can only be used if the vapour pressure in or at the surface has the saturation value of the . 1965): This method is based on the work of Penman (1948). The vapour pressure at the surface.. in m.h = zov = z. from (17) and (21) gives 4. Then it follows from (18) and ( ) assuming $m = $ h = @ The errors introduced by the assumption z. then (19) can be rewritten as According to Thom (1972) B' = 6. " ~ (U.. introduces the quantity B' (Owen and Thomson. But one may assume that in the stomatal cavities of the plant leaves the vapour pressure has the saturation value at the surface temperature To: e.sl) for many types of vegetated surfaces. COMBINATION METHODS By combining the energy balance equation (1) with the transport equations (16) and (22) one can derive an expression for the flux of latent heat in which the surface temperature does not occur (Monteith. is practically always unknown. are not well known at this is for a rough surface probably larger than zOh. Rijtema. e.27 u . We can now introduce a canopy resistance r.. defined by Eliminating e. The magnitude of B' is rather uncertain at this moment. 1979). If one moment. The quantity z. (To). 3 ms' @e Bruin and Kohsiek. 1963) defined by In (zom/zoh) = k B'. This gives an increase in rh in many cases of 3040% compared with (19).The influence of (in)stability on the transports is small for wind speeds greater than 7.
is called the aerodynamic evaporation or the drying power of the air. It is also possible to refer the concept of potential evaporation to a specified surface e.Q* G. r. (23) can also be applied if soil moisture limits the evaporation. 1977). ET.g. Eq. In that case r. leaf water pressure (Van Bakel. 1981). evaporation from the soil surface has to be taken into account (Van Bakel. Eq. If the vegetation is not completely closed and radiation penetrates to the soil surface. If a vegetation completely covers the ground and soil moisture is not a limiting factor. Natural . f (4 is defined by ET. = 0 and introducing a wind function f(u) defined by Then we have ET=S S + 'l Q* + E T . (23) shows clearly the factors on which evaporation in nature depends: on the one hand on the available energy. and rh depend on the vegetation. m small roughness length namely z = 1. and on the other on the dryness of the air. See for this problem De Bruin (1 98 1).(T2)e2).g. S + 'l ' X where s is the slope of the saturation vapour pressure curve at air temperature and AET.surface temperature. on the water status of the vegetation e. The conventional Penman equation can be derived from (23) by setting G = r. The first term at the right hand side of (24) is called the radiation term and the second term the aerodynamic term.a.~ (Thom and Oliver. 1981). e. then we speak of the potential evaporation of that vegetation. a grass cover of a certain height. and the turbulent exchange which is expressed by rh. is it clear that potential evaporation may differ from one vegetation to another. depends i. Penman used the wind function With this choice of f(u) the transport resistance is Comparing (26) at higher wind speeds with (19) shows that Penman used in fact a very .37 X 1 0 .{e. (T(z)} e(z). = . As the quantities Q*G.
see Stricker (198 1) and De Bruin (1 981). On the other hand the ratio of the aerodynamic term to the radiation term increases by the factor m. For rural areas in Southern England Thom and Oliver estimated m Z 2% and rc n =rh  1. (24a) can be simplified.4. Therefore Thom and Oliver proposed to introduce into (25) a factor defined by is the roughness length of the surface under consideration. Substitution of this expression in (24b) leads to  . which decreases when the contribution of interception to total evaporation increases. In winter Q* is very small but the aerodynamic term of (24a) cannot be neglected. Thom and I Oliver (1977) deduced the relation rc = rCd(I) where r. Instead of (25) where. 0. During the time the intercepted water evaporates. since then n >> 1 and (24a) becomes in good approximation The application of (24b) is complicated by the fact that forests intercept a fairly large proportion of the precipitation. z and (26) we then have These simple expressions $0 take into account the influence of stability at low wind speeds.d is the resistance of the dry E surface and I the interception. Instead of (24) we then have It turns out that the evaporation calculated with (24a) is often nearly equal to the evaporation calculated with (24). For a application of (24a) with a slightly different value of m. This means that we have to use in (24b) an effective value of r.surfaces have usually a much larger roughness length. In case of a very rough surface as a forest. we have r. The effects of the introduction of m and n are often approximately selfcancelling. Therefore the approximate equality of (24) and (24a) does not apply to all circumstances.
1976.Evaporation from extremely rough surfaces depends primarily on saturation deficit and interception. eq. . 1977). They considered first the case that the atmosphere is saturated with water vapour in the lowest layer. Davies and Men. Then they added an experimental constant oc to account for the fact that the atmosphere is usually not saturated.H (27) Soer (1980) used this approach and calculated H with (16) taking into account the influence of stability on rh. This value has been confirmed by a number of other investigators. 1977. Moreover. (12) and (13). As net radiation Q* is highly correlated with global radiation K'. 1973.1960). this is a point measurement in contrast to the measurement of surface temperature by IRLS which surveys an area. (25) can be approximated by Values of cl and cz are given by De B ~ i (1981). The surface temperature was remotely sensed by infrared line scanning (IRLS). De Bruin (1981) and Stricker (1981). 1975. An important simplification of (24) with a wide range of applicability has been introduced by Priestley and Taylor (1972). Another possibility is to measure air temperature at two or more heights and use eqs.26. Therefore it is often feasible to determine the evaporation from AET = Q * . A further simplification of (24) has been proposed by Makkink (1957. This method has been used by Stricker and Brutsaert (1978) for a small rural catchment. Then from (14) we have 0 = y/s and from (1) AET = s/(s + y) (Q*G). The latter method avoids the difficulties about the quantity z. Furthermore in saturated or nearly saturated circumstances H is small compared to XET. even in the case of large areas well supplied with water. Mukarnmal and Neumann. For applications of (25) in The Netherlands see De Bruin and Keijman (1979). Stewart and Rouse. AET = oc S (Q* G) S+Y They found a= 1. n From an experimental point of view it is easier to determine H from a transport equation than AE.h but a high instrumental accuracy is required for measuring vertical temperature differences in the air.G . (Ferguson and Den Hartog..
Pudoc. DE (1981) .H.R. potential and actual evaporation from cropped surfaces in southern Ontario. Comm. 36. Eng. 417448.A review of fluxprofile relationships.R. 7710. Legg and Monteith.5. P. BRUIN.L. In: D. of Wat. BoundaryLayer Meteor.L. H. MAKKINK.A. it still has serious limitations. De Bit. CONCLUDING REMARKS Although the theory presented in the preceding sections is based on sound physical principles.The PriestleyTayler evaporation model applied to a large. H.Comments on surface roughness parameters and the height of dense vegetation. 793.F. . WYNGAARD. H.A. and J. W. DAVIES. Ontario. App.A. There exist of course models for these processes but they require detailed information about the vegetation which is seldom available (see Goudriaan. J. Sc.A. This layer is represented in the theory by a few parameters such as roughness length. Inst. (1977) .Testing the Penman formula by means of lysimeters.Equilibrium. J.Fluxprofile relationships in the atmospheric surface layer.). John Wiley and Sons London. Met.J. 3. pp. 11: 277288. J. J.Toepassingen van de Penmanformule. 18: 898903. BRUIN. In reality the exchange processes in a vegetation layer are too complicated to be accounted for by a few parameters.Crop micrometeorology: a simulation study.Unsaturated zone and evapotranspiration. 1975).R. The Hague.Meteorological studies of evaporation at Perch Lake. 1977. LEGG. Met. BRUTSAERT. VAN (1981) . J.R. Appl. 12: 649657. REFERENCES BAKEL. B.A. DE and W. KNMI. Met. KEYMAN (1979) .J. Comm. BUSINGER.P. Y. One of these limitations is the fact that the theory is one dimensional. ALLAN (1973) . So different values of the evaporation can be expected on these fields. De Bilt. IZUMI and E. G. J. and G. The surrounding fields may have different properties of soil and vegetation. J. Proceedings and Informations 28: 3858. shallow lake in The Netherlands. Soc.Heat and mass transfer within plant canopies. MONTEITH (1975) . DE and W. KNMI. (1975) . Atmos. : FERGUSON. DYER. pp. KOHSIEK (1979) .T. Another difficulty is presented by a vegetation layer.R. P.A.J. Research. de Vries and N. DEN HARTOG (1975) .. Notwithstanding these limitations and others satisfactory results can be obtained with the theory presented. The Hague. BRUIN.De verdamping uit gras op kleigrond in de zomer van 1976 vergeleken met de Penmanformule. KOHSIEK (1977) . Hydrol. J.F.Q. AECL Chalk River Nuclear Laboratories. (1974) . 7 363372.). DE and J. TNO. H. Proceedings and Informations 28: 2537. Heat and mass transfer in the biosphere.D.A. H. GOUDRIAAN.J. Hydrol.R. Barry (Ed. Wageningen.The determination of (reference crop) evapotranspiration from routine weather data. This means in practice that the necessary measurements have to be taken on a reasonably homogeneous field of sufficient dimensions. A. Afgan (Eds. (Evaporation from grass on clay soil in the summer of 1976 compared with the Penman formula) Scientific Report W. as will be shown by the subsequent authors. (1957) .C. Japan 53: 9697. BRADLEY (1979) . BRUIN. (Applications of the Penman formula) Scientific Report W. and C. Research TNO. Hydrological studies on a small basin on the Canadian shield . 28: 190201.Evaporation studies. zero plane displacement and surface diffusion resistance.
Methuen.A.B. J.S.R. A. 16: 649650. Biol. and AM. and H. Met.Application of the PriestleyTaylor evaporation model to assess the influence of soil moisture on the evaporation from a large weighing lysimeter and Class A pan. Quart.. Proceedings and Informations. NEUMANN (1977) . Res. (1980) . Appl. Inst. Pudoc. (1970) .A simple method for determining the evaporation from shallow lakes and ponds.H. 39: 139157. C. WIERINGA. Soc. and W. Acad.Momentum. Met. R. Proc. MONIN. (Evaporation from vegetations in relation to Penman's formula) Comm. IS: 321334. . J. Soc. and W.. J. Res.R. P. Met. Comm. MONTEITH. OBUKHOV (1954) . Rep. G.J.L. PRIESTLEY. J.MAKKINK. J. 193: 120145. BRUTSAERT (1978) . 18: 411430. London. bare soil and gxass. ROUSE (1976) . Research TNO.NM. J. (1978) Boundary layer climates. and H.26 for potential evaporation in high latitudes. PAULSON. J. 98: 139157.H.Principles of environmental physics. OLIVER (1977) . Soc.R. THOM. R. Weath. TAYLOR (1972) .B.S. J.Estimation of regional evapotranspiration and soil moisture conditions using remotely sensed crop surface temperatures. Exp. J. H. Rev. 12: 623628. Proc.Basic laws of turbulent mixing in the atmospheric surface layer.B. Wageningen. 107. E. mass and heat exchange of vegetation. 9: 857861.R. The Hague. Res.R. Proceedings and Informations 4 : 90115. PENMAN.N. T. Met.Roy. BoundaryLayer Meteorol. OKE. Soc. (1980) . Roy. Agric. Edward Arnold. STEWART.The mathematical representation of wind speed and temperature profiles in the unstable atmospheric surface layer. A.Roy. 96: 6790. pp. (1972) . A. (Month. BoundaryLayer Meteorol. Hydrol.J..Actual evapotranspiration over a summer period in the Hupsel Catchment.R. 659. 19: 205234.On the assessment of surface heat flux and evaporation using largescale parameters. Fluid Mechs.M. and W. (1960) .Evaporation and environment.Natural evapotranspiration from open water.De verdamping uit vegetaties in verband met de formule van Penmar.F.. (1948) .L.On Penman's equation for estimating regional evaporation. London.S. Sci U. The Hague. Wat. Appl.An analysis of actual evapotranspiration. STEWART. RIJTEMA.Heat transfer across rough surfaces. THOM.Substantiation of the Priestley and Taylor parameter a = 1.J.Methods of estimating evapotranspiration from meteorological data and their applicability in hydrology. ROUSE (1977) . Tmdy Geofii. Hydrot.. C. 12: 243256. Research TNO. THOMSON (1963) .A revaluation of the Kansas mast influence on measurements of stress and cup anemometer overspeeding. J.R. of Hydr. STRICKER.S.L. 1. G. P. SOER. 24 (151): 163187. No.) 100: 8192. (1981) . MUKAMMAL. Quart. (1965) . A.E.S. Symp. OWEN. and R. (1965) .R. Remote Sensing of Environment 9: 2745. 28: 5975. MONTEITH. STRICKER. (1973) . and W.
the hydrologist in the field needs evaporation data and mostly he has only routine weather data at his disposal and some rough information on soil and canopy. However.A. at first sight it seems to be impossible to evaluate evaporation from routine weather data only. evapotranspiration depends on many other factors such as type and stage of development of vegetation. So. In the first part the use of simple routine weather data for the determination of crop water requirements is considered. ET. which often yields satisfactory results. PriestleyTaylor and Makkink are considered and a comparison is made..THE DETERMINATION OF (REFERENCE CROP) EVAPOTRANSPIRATION FROM ROUTINE WEATHER DATA H. (= the potential evapotranspiration of a short grass cover). 2. In this paper such approaches will be treated with respect to the evapo(transpi)ration problem. DE BRUIN SUMMARY The first part of this paper deals with the determination of reference crop evapotranspiration. In the second part of this paper ET. besides on weather elements. probably. THE ROUTINE WEATHER DATA Almost every country in the world is a member of the World Meteorological Organiza . of Penman. This quantity depends on weather elements only. A first analyses leads to the conclusion that. soil type and soil moisture. ThornOliver. will be discussed. INTRODUCTION This paper deals with the problem of the determination of evaporative water losses from vegetative surfaces. Under such circumstances a special type of researchers comes foreward who simply ignore the theoretical difficulties and construct straightforward ad hoc models. groundwater table etc.R. can be applied in the future on a routine base. as an old proverb from China says: "One can not applaud with one hand". In the second part ETo is compared with the actual evapotranspiration ET. Finally. The formulas for ET. will be compared with the actual evapotranspiration of grass as measured at Cabauw in The Netherlands. 1. or. some measuring techniques for ET are discussed which. Several methods for the evaluation of ET. Finally the question will be touched if it is possible to get more information on actual evaporation when a small number of additional observations is made on routine weather stations. For this the socalled reference crop evapotranspiration (ETo) is introduced.
wind speed at 10 m. but.tion of the United Nations.). this quantity can be estimated from sunshine duration (or cloud cover). REFERENCE CROP EVAPOTRANSPIRATION AND CROP WATER REQUIREMENTS An important example of how standard weather data can be used is the determination of crop water requirements. among which The Netherlands. Often. the concept of "potential evapotranspiration" is introduced to account for the influence of weather on the transpiration rate of crops.The reference crop evapotranspiration (ET.temperature at screen height. green grass cover of uniform height. but there is no unanimity about the crop that is considered. For evaporation studies we need net radiation (Keijman. Unfortunately there is confusion about the interpretation of this term. . To avoid this confusion. also on crop factors. the terminology of the FAOresearchers Doorenbos and Pruitt (1977) will be followed. besides on weather. completely shading the ground and not short of water. This organization demands of its members. which are defined as the depth of water needed to meet the water loss through evapotranspiration of a diseasefree crop. have also established stations at which duration of sunshine and global radiation are observed. In the Technical Regulations (WMO) rules are given for the meteorological quantities to be measured. methods of observation. are sometimes called "potential evapotranspiration" e. . These authors introduce the following quantities: The crop water requirements (ETcmp). . the term potential evapotranspiration will not be used. in this paper. such as crop height. As noted before in literature both ETcmp and ET. for this. Fortunately.humidity at screen height.. 3. One group of authors refers to the actual vegetation of interest.g. reflectivity and shading percentage of the ground. 1981) which is not measured directly at a weather station.cloud amount. In this way the hydrologist has the following elements relevant for evaporation at his disposal: . When available use van be made of global radiation data for these estimates. Van . which is defined as the rate of evapotranspiration from an extensive surface of 8 to 15 cm tall. temperature and humidity. growing in large fields under nonrestricting soil conditions including soil water and fertility and achieving full production potential under the given growing environment. . It is obvious that ETaoP depends. minimum accuracy and time of observation. among other things. Bake1 (1981) who takes ETpot = ET. . t o establish in their territories a network of synoptic and climatological stations. actively growing. whereas others consider potential evapotransipiration the maximum transpiration rate from a hypothetical reference crop (which mostly is a short grass cover). Some countries. Mostly potential evapotranspiration is defined as the maximum transpiration under the given weather conditions.
(mbar) the actual vapour pressure at screen height.The introduction of ET. @. Keijman. the soil heat flux.mw2) the slope of the saturation water vapour pressure . . notably ET. and is expressed in kg. at screen height. through its greater roughness length. is a crop factor. This is important to note.. (mbar. 1981). (1) where k. The windfunction f(u) refers to a hypothetical water surface. (Jkg the saturation vapour pressure at air temperature T. 1) The reference crop evapotranspiration (ET.m' . (mbar. aerodynamic properties which leads to a higher wind function (see e. T = k.S' This is the windfunction proposed by Penman and used by Kramer (1957). are related as follows: E. 2) ET. Strictly speaking Doorenbos and Pruitt adopted a somewhat different form of f(u).) ea is the net radiation of the grass surface.. (mbar) ' U2 G ET.g. it appears that the following assumptions can be made which are rather realistic in a wide range of climatological conditions.~') the psychometric constant. 3) The best results are obtained when ET.) is a function of weather elements only and is independent of plant and soil. Mostly the soil heat flux can be ignored. ET.~' ) the latent heat of vaporization of water. Fortunately. is only useful when they can be applied in practice. and ET. Note: the wind speed at 2 m.temperature curve at air temperature.ea) s+7 S where Q* Y h es(T. and ET. = $Q* G)/h + Y f(4 (es(Ta) . is evaluated with the wellknown Penman equation in a somewhat modified form. since a grass cover has.
is related to ET. the evaporation totals are given for the first and the second 1. (2). (2). the KNMI. E. is based on the work of Penman (1948. the calculations are carried out with eq. by the simple expression (Wesseling. with two differences (De Bruin. expecially. in my opinion. thus determined. It is noted that there is an alternative approach based on the work of Rijtema (1965) and Monteith (1965) for the determination of ETmop. It appears that for the period AprilSeptember E. the time intervals are not exactly 10 days.s month.days of a month and of the remaining period. when one askes for verifications that will convince physicists. (4) The E. from a practical point of view. the reader is referred to Doorenbos and Pruitt (1977). it has the disadvantage that weather and plantsoil factors are not separated explicitly as is the case in the approach of Doorenbos and Pruitt (1977). Since 1956 the Royal Netherlands Meteorological Institute (KNMI) publishes in its monthly weather review the socalled "evaporation from a free water surface" often denoted as "open water evaporation". This method is e. Buishand and Velds. the latter has a variable length between 8 and 11 days. 1956). Until 111971 monthly totals were presented of 12 stations. since we are dealing with a practical problem. 3. The routine evaporation data published by the KNMI. falls outside the scope of this paper. However. It will be clear that the treatment of the crop factor k. (Strictly speaking.g. a) The actual net radiation Q* is replaced by the net radiation of a hypothetical water surface. The advantage of this technique is that it has a certain physical basis. For this.8 E. the experimental evidence that can be found in literature supports the assumptions made sufficiently in order to adopt them as very useful working hypotheses for practical applications. = 0. data are published by the KNMI in its monthly weather review with a retardation of about 1. Thus.. Doorenbos and Pruitt (1977) introduced a multiplication factor in the right hand side of eq. For many purposes this is too late. b) Instead of 24hourly averages of temperature and humidity the means of these quantities over daytime are used for the computations. 1977): ET. while after that date the evaporation amounts of 10days periods are published. for the common conditions in The Netherlands this factor can be taken equal to 1. 1979.) The determination of E. The evidence given by Doorenbos and Pruitt (1977) is certainly too meagre to be convincing for a physicist.1. ..Strictly speaking. mentioned by Van Bake1 (1981) and explained by Keijman (1981). Experimental confirmation of the above assumptions isvery difficult to give. however. 1980). Therefore. however. however.
values of 5 stations of the previous day are given in the morning radio weather bulletin.2. and rainfall Kramer (1956) publishes monthly values of E.started in 1980 with a daily "fast" evaporation data provision. 1980). 1980)  0. since at the time of computation no bright sunshine observations are available @e Bruin and Lablans. Every day between 1 April till 1 November the E.8 E. l Isolines for cumulative frequencies of potential precipitation surplus (R computed from 1 April. The evaluation method used for these E. It is very likely that these data provision will be continued in the fiiture. 3.) of De Bit. values differs slightly from eq. (2). Long term records of E. . (after Buishand and Velds. of 12 Netherlands stations for 1931 Fig.
g. 3. Moreover De Bruin (1979) presents cumulative frequency distributions of E. since mostly Q* > G. On the other hand the number of stations is diminished to 6.ET.G) S+Y S Where G is the soil heat flux. G can be ignored in eq. A second simplification of eq. which possibly can serve as an attractive alternative for eq. (= R . 1. what can be called "potential rainfallsurplus". further for grass r . It is seen e.. 1979a). These are of practical importance (e. Alternative methods for the determination of ET. It is noted that this expression applies also to a water > surface (see e.) for the growingseason.1951. This data set has been extended by De Bruin (1979) to 19111975. De Bruin and Keijman.g. They arrived at  XET. while he added the corresponding rainfall totals (R).2 mm day l . say 0.g. (6) refers to daily mean values. An example is shown here in fig. which refers to (R . (2) is given by Priestley and Taylor (1972). (6). for the assessment of the crop water requirements) for hydrologists and agricultural researchers. is less or equal to about 300 mm. The six stations for which this type of information is available are Den Helder De Bilt Winterswijk Oudenbosch Gemert en "Avereest/Den Hulst/Wijster/Witteveen/Dedemsvaart".). and r is the reflectivity.ET. R and R 0. In literature several methods can be found for the determination of ET. Since eq.ET. It is noted that cl 0. (2). Finally it is worthwhile to mention the expression of Thom and Oliver (1977) which is . = (lr)S s+Y cl K C + cz 1 (W . = 1. K is the incoming shortwave radiation (which is obsewed directly on a routine base at several stations in The Netherlands).3.8 E. A graphical representation of a set of these frequency distributions for De Bilt is given by Buishand and Velds (1980). that over the period April to August inclusive there is a probability of 1 percent that R .9 and c2 is small. which can be written as: XET.25.m' ) where cl and c2 are constants.2: 0. since it appeared that the quality of the basic weather observations (if present) was too poor of the other stations for 19111950.26 (Q* . The most simple expression is given by Makkink (1960).
while it needs only the global radiation and air temperature as input.G ) + m77. determined with eq.(T. = 0. the model of Makkink appears to be a very attractive alternative. while n describes the effect that in the case of grass the water vapour transport goes mainly through the plant's stomata resulting in an extra resistance which is not present in the case of open water. the concept ET.g. are evaluated from the 10day totals as published by the KNMI with a method proposed by De Bruin and Kohsiek (1977): where the index i refers to the ith day (i = 1. c) In 1977 (a "normal" year) there is no significant difference between the models of Penman. 5 4 ~ ~(e. since the bright sunshine was not observed at this station. 2. (2). Keijman. The quantity m accounts for the greater roughness of a grass cover with respect to open water.m") where m and n are constants. It reads AET. .4. Comparison of the different methods In table 1 the different methods to determine ET. while for the evaluation of ET.G ) are used. With respect to table 1 the following remarks can be made: a) The interrelation between the different models for ET.) . Table 1 refers to daily values and is based on a set of micrometeorological data collected at Cabauw during the summer period of 1976 and 1977. are compared. 3. . of Cabauw.) of a 10day period. Therefore it was decided to compare E.4 (1 + 0 . In the calculations of ET. This table contains some statistical features such as mean values. ThomOliver and PriestleyTaylor. = s(Q* . is high.) ) S + ?(l t n ) . It was not possible to determine E. since it has about the same skill.8 E.e. The daily values of E. with the equations of Penman.. of De Bilt (ca 25 km north western of Cabauw) with ET. (~. (see e. (2)] gives the best results. with Makkink's formula the measured global radiation has been applied. and the index "10" to the entire period of 10 days. The correlation coefficients for daily values (!) are all greater than 0.. 1981).a variant of the PenmanMonteithRijtema combination equation.. The advantage of this approach is that it has a better theoretical bases than eq. and Thom and Oliver the measured values of (Q* .9. yields a somer what ( 8%) greater value. Priestley and Taylor. correlation coefficients and the least square estimate of the regression constant a from the regression equation 9 = ax.. (2).. for Cabauw. b) When it is believed that Penman's formula [eq.
= regression coefficient a from 9 = ax.94 . = Makkiik [eq. (7). 1977 (119 daysin MaySeptember) X Y X Y Corr. Comparison of different methods for the determination of ETo (mm day .5 b. which reduces the application of the Priestley and Taylor concept. 1979) and that then.06 1. MAK THOL = ThomOliver [eq.97 . under very dry circumstances. E. . Coef. = 01. m = 1.. among other tings. (lr)cl = 0. (2) which depends. Brutsaert and Stricker.6 3.G) measured] . these effects counteract. we must realize that under very dry conditions the common formula's for the estimation of net radiation are not verified.5 3.2 2.3 2. Data of Cabauw. There is evidence that under very dry conditions the latter overestimate ET. 1976 (55 days in JuneAugust) X Y X Y  Corr. the reader should use the different ET. e) There is a great resemblance between the results of eq. 15% can be obtained very easily.5 3.99 . . (Q* . (6).2 2.5 3.8 E. (2) in the denominator the stomata1 resistance is ignored. (S).2 2.5 3.G) measured]. PNM = Penman [eq.9. So.91 . In e.65.3 1.00 . This is a typical example of how two errors can cancel each other.93 .97 .3 2. Anyhow.G) measured] . They introduce a correction factor in eq. PNM PNM PNM PNM 2.l ) . PRTA = PriestleyTaylor [eq. (2).3 2. (2) yields such good results.q. a too low wind function is taken. Possibly. except E. estimates with caution when these are applied for crop water requirements calculations in very dry years.06 0. Coef.8 E. the method proposed by Doorenbos and Pruitt (1977) offers an alternative solution in this case. = open water evaporation of De Bilt (see text).1 3. on radiation and relative humidity.92 d) In 1976 (a dry year) the PriestleyTaylor model differs from the other ones.02 1. say. PNM PNM PNM PNM 3. but at the same moment.00 PRTA MAK THOL 0. Apparently. (Q* .2 (roughness length = l cm). . n = 1.12 1. C. Slope a.Table I. (see e.96 1.g. the Priestley and Taylor concept must be prefered. However. (Q* . it is rather fortuitous that eq.93 Slope 1. (2) and those of the Thorn and Oliver equation. in the numerator. In such years an overestimation of.5 3.91 Slope PRTA MAK THOL 0.5 2.
1 2. as evaluated with the models of Penman [eq. Table 2.82 . Coef.^:!^^ ET ET ET 3. and b) with the model for ET of Brutsaert and Stricker (BRSTR). it is to be expected that the actual evapotranspiration (ET). regardless the weather circumstances.89 . 1977).68 . ACTUAL EVAPOTRANSPIRATION As noted before.4 2. . apart on weather. will depend.5 3. it is worthwhile to examine what is the "weight" of the weather influence.~ ET 2. further they are restricted to grass. . under Netherlands conditions. daily values.1 It is important to remark that the above results apply to the "summer" period MaySeptember. Coef.93 PNM ET E T ~ ET..4 b. whereas the PriestleyTaylor concept give an overestimation of. 1977 (119 days in MaySeptember) X Y X Y Corr. as determined with the models of Penman and PriestleyTaylor respectively.77 .B9 .g. In the next section a comparison will be made between ET and ET.3 2. Nevertheless. O p ~ . then Penman's equation (2) overestimates ET about 4076.1. e.7 2. using Bowen's ratio (De Bruin and Kohsiek. in the extremely dry year 1976 ET is much less than ET. 25%.4.1 2. 1976 (55 days in JuneAugust) X Y X Y Corr. This will be done for an extreme dry year (1976) and a more or less normal year (1977). say. 4. on soil and plant factors.72 Slope . (2)] and PriestleyTaylor [eq. (6)] respectively.75 . It will be clear that this "weight" is a function of plant and soil conditions. .86 ET. Data of Cabauw. ET has been determined with the energybalance method. From table 2 it is seen that in the "normal" year 1977 the two ETo estimates are only slightly greater than ET. ETBRSTR ET.2 2. a.4 2.87 Slope . so it is unlikely that ET can be determined from weather elements only. As to be expected. Comparison between ETand ETo In table 2 a comparison is given between the actual evapotranspiration ET of a grassland and ET. when there is no water in the soil the evapotranspiration will be zero. Comparison between the actual evapotranspiration ET (mmday ") and a) ET.
Brutseart and Stricker (1979) constructed such a model and they tested it in The Netherlands. Therefore in winter the expression of Thom and Oliver must be preferred.g. is independent of the soil conditions. Recently Stricker and Brutsaert (1978) used the socalled profdeenergybudget method for the determination of ET from grass. (6) and (2) respectively. at the moment. no simple rules are available to decide whether it is dry or not. if there is an interaction between ET. However. 4. (9) must be used. and the soil conditions. the problem is that quantities such as ET. Penman's equation must be applied. the exact interaction between ET. which they called themselves "heuristic" reads: where ET. This makes that then the second term of Penman's formula plays a dominant part. Anyhow. and soil condition is unknown. 5.PRTA and ET.2. however.g. while otherwise e. it can be imagined that use of this interaction can be made for the determination of ET. Under very dry circumstances this assumption is questionable. in normal years the random scatter is greater than that of e. the models based on this interaction must be seen primary as empirical. which implies that discussions on the behaviour of such quantities under rare conditions will be hypothetical themselves and are therefore rather vague and difficult to understand. a. The modified aerodynamic approach. This model.In wintertime net radiation will be small. are hypothetical. the Penman model. while also in 1977 the skill is rather good (see table 2). and a better description of the surface roughness is needed. This supports the conclusion that in very dry periods eq. The advectionaridity approach In chapter 2 the assumption is made that ET. The choice of these methods is rather arbitrary.'^" are evaluated with eqs. The implications of this for very rough wet surface is discussed by Keijman (1981). This method is discussed in this report by Keijman. Now. WHAT CAN BE ACHIEVED WITH ADDITIONAL WEATHER OBSERVATIONS? In this brief chapter we will mention some methods for the determination of actual evapotranspiration for which only a limited number of weather observations has to be done. At the moment. The only quantity that has to be measured apart from the standard . For the very dry year 1976 the results of this approach are satisfactory. while furthermore the canopy is often wet. Because.
Businger 50 100 A. The surface temperature can be measured with a remote sensing (IR) thermometer. is the vertical temperature gradient. The theoretical consequences are discussed by Keijman (1981).g.weather observations. 2 Measured values of ET plotted against the computed ones using the temperature fluctuation method. see e. The temperature fluctuation method Finally I will treat a method which is based on the fact that the sensible heat flux H is an empirical function of the standard deviation of the temperature. The remote sensing technique. It is certainly possible. b. The only difference is that the lowest level at which the temperature is measured is at the ground. c. hourly values) .E T measured ( ~ r n . This method is similar t o the previous one. that in the future thisinstrument can be used on a routine base.~ ) Fig. De Bruin and Keijman (1979b) tested this method. They concluded that it has a good skill.
VAN (1981) . DE (1979) .R.An advectionaridity approach to estimate actual regional evapotranspiration. Res. PRIESTLEY. Staatsdrukkerij. Neth. (1980) .Q. and VELDS. H. ET can be obtained from the energybalance equation.T. (in Dutch) KNMI Meded. Research TNO. Met.L. H. Res. (Evaporation form grass on clay soil in the summer of 1976 compared with the Penman formula) Scientific Report W. (1956) . J. TNO. Soc.R. MAKKINK. of Agric. DE and LABLANS. Proceedings and Information.The PriestleyTaylor evaporation model applied to a large. T. PENMAN. KNMI Verslagen 357. H. Irrigation and Drainage Paper No. DOORENBOS.N. In: Workshop on Micrometeorology (ed.On the assessment of the surface heat flux and evaporation using largescale parameters. '~Gravenhage. H.Evaporation: an introductory survey. where ET as measured with the energybalance method.Guidelines for predicting crop water requirements. Exp.Neerslag en Verdamping. Klimaat van Nederland 1. (1979b) . but. 19.A. Climate of The Netherlands 1) BUSINGER. (1980) . PENMAN. Research TNO.B. C. ten behoeve van snelle voorlichting omtrent de verdamping. 4: 929.A. A (very. Proceedings and Informations 28: 3858. The Hague.A. W. (Evaporation from vegetations in relation to the formula of Penman) Comm. (1981) .R.Neerslag.205234. Hydrol. KEIJMAN. De Bilt. Biol.(1973). 24.Natural evapotranspiration from openwater. 193: 120145. The Hague. 7 9 4 . Proc.M. A.A. After having determined H.L.O.en Uitgeverijbedrijf. 30.De verdamping uit gras op kleigrond in de zomer van 1976 vergeleken met de Penmanformule. DE and KEIJMAN. Hydrol.Unsaturated zone and evapotranspiration. H. 100: 8192. Met. DE and KEIJMAN. 18: 898903. (1973) . 2. At the moment the present author examines this approach. W. W. Sci. BRUIN. J. and PRUITT. 4: 901 15. C. en Verh.H. Comm.Een test van een nieuwe berekeningswijze van de openwater verdamping volgens Penman. bare soil and grass. J. (1977) .Turbulent transfer in the atmospheric surface layer. (1972) . G. DE and KOHSIEK.J. REFERENCES BAKEL. Rev. Proceedings and Informations 28: 1223. De Bilt. J.Evaporation from a uniform and a nonuniform surface determined with a simplified aerodynamic method. every second. J. Scientific Report W. in the near future it will be very easy to do with a microprocessor. using Bowen's ratio. 70. Research. shallow lake in The Netherlands. BRUTSAERT. J. This fast sampling is a problem at the moment. BRUIN. BRUIN.Evaporation and environment. Wat. BRUIN. (1965) .Q. KNMI. MONTEITH. say. (in Dutch with English Summary) KNMI. Comm.Theoretical background of some methods for the determination of evaporation. BRUIN. HAUGEN. 15: 443450. Symp. C.R. (1979) .R. (1979a) .Q. (Precipitation and Evaporation. Soc. The Hague. and R.A. H. Roy.A. J. preliminary result is shown in fig. and STRICKER.A.J.R.) Am. Weath.De verdamping uit vegetaties in verband met de formule van Penman. KNMI.L. Trans. H can be measured with a fastresponding thermometer which is sampled.N. openwaterverdamping en potentieel neerslagoverschot in Nederland. but it is far too early to state that it is a reliable method applicable at a routine weather station.A. KNMI. 7710. (1960) . H. Am. KRAMER. (1957) . Rome.A. BUISHAND. TAYLOR. J.F. F. Union 60: 582.A. Month. Soc. J. . of Appl.O. It is seen that in this example the approach works splendid. Proc. is compared with ET determined with the temperaturefluctuation method. Hydrol. De Bilt.Berekening van de gemiddelde grootte van de verdamping van verschillende delen van Nederland volgens de methode van Penman. D. (1977) . Geoph. (1948) . Frequentieverdelingen in het groeiseizoen. W.R. P.
0.An analysis of actual evaporatranspiration. P E . (1977) . 48. No. 107.O.Technical Regulations. Rep. Res. W. Wageningen. (1978) . (1977) . W Y .N.RIJTEMA.O. . H. (in Dutch) H. STRICKER.Actual evapotranspiration over a summer period in the "Hupsel Catchment". Agric. J. Soc.M. W. of Hydr. Publ. Pudoc. THOM. Quart.S. pp. (1965) . A. J.M.10: 185186. and OLIVER. Geneva. 103: 345357.On Penman's equation for estimating regional evaporation. J. WESSELING. J. Met. .De Landbouw in de droge zomer van 1976. 39: 139157.R. 659. and BRUTSAERT.Roy.
These models concern the entire system of surface water . The hydrological part of this evaluation can be made with the aid of. For the domestic and industrial water supply the main objective is a continuous delivery of sufficient water of good quality.38 UNSATURATED ZONE AND EVAPOTRANSPIRATION P3.one needs a weighing of the conflicting interests. Proper water management of any region with competing interests implies an evaluation of the various measures that are available. Agriculture can be characterized by an intensive use of the soil.soil water . An overview of the various categories of methods available t o evaluate effects of changes in soil water conditions on evapotranspiration is presented. In order to find the optimal distribution of water among the users .e. The upper zone of the soil. An important issue is the mathematical description of water uptake by plant roots. Firstly the most relevant theory concerning this zone is given. also from surface/groundwater resources. INTRODUCTION In The Netherlands the demand for water by agriculture. Finally conclusions and recommendations are made. The latter is obtained from surface water and/or groundwater. Only the most common methods to quantify this uptake are treated. This zone is very important for the biological.T. 1. The relationship between these three systems are discussed in short. mathematicalhydrological models. among others. The problems encountered in practice mostly relate to the translation of model results to field situations. industry and domestic water supply and also for nature is steadily increasing.groundwater . a need for good drainage conditions during winter and an additional water supply in summer. Often this implies conservation of the existing hydrological situation. VAN BAKEL SUMMARY This paper mainly deals with the hydrological processes in the unsaturated zone.plant system. the unsaturated zone. The various aspects with respect t o this problem are discussed. The main objective for managing nature is the creation of such environmental conditions that the diverse and mutual relationships are guaranteed. environment and pollution . This calls for proper water management schemes.taking into account the effects on safety. .evapotranspiration. The saturated groundw~tersystem and the atmospheric system constitute the boundaries of the unsaturated zone. chemical and physical processes occurring in the soil . constitutes the medium between the atmosphere and the saturated groundwater system. i.
When dealing only with the matric head. h (cm). part of the pores are filled with air.Hydrological measures taken will affect the processes in the unsaturated zone. cm and potential is then denoted as 'head'. arising from the gravitational force.dl) ~can be written as: where K is the hydraulic conductivity (cm. and the soil water content. i.cm 3). General theory Water in soil moves from points where it has a high energy status t o points where it has a lower one. In the following the soilphysical aspects of the unsaturated zone will be treated with respect to groundwater movement and evapotranspiration. 2. SO: To describe the flow of water in soil systems.d . h. For each soil there exists a relation between the pressure head. B (cm3 . however. arising from local interacting forces between soil and water and gravitational head. total (hydraulic) head. Under conditions of atmopheric pressure and nonswelling soils the matric head can be denoted as pressure head.. can be expressed as: where the vertical coordinate z is considered positive in upward direction. UNSATURATED ZONE 2. Therefore K is not a constant but depends on the soil moisture content B or because [B = f(h)] on the pressure head Substitution of eq. the volumetric flux q ( ~ m ~ c . it is customary t o use Darcy's law. In hydrology potential is usually expressed as energy per unit weight of soil water.e.1. With unsaturated flow. The energy status of water is the water potential which consists of several components. (3) yields: . with the dimension of length. For mone dimensional vertical flow. z. (1) into eq.l ) For saturated (groundwater) flow the total soil pore space is available for water flow and the hydraulic conductivity is constant. H. Potentials are defined relative to the reference status of water at atmospheric pressure and elevation datum zero.
In order to get a complete mathematical description we apply the continuity principle (Law of Conservation of Matter): Substitution of eq. see Mualem and Dagan (1975) and Royer and Vachaud (1975). 2. can be represented as a "pFcurve". according to pF = ''log(h). (7) yields the onedimensional equation for water flow in heterogeneous soils 2. the derivative of 0 with respect to h can be introduced. 1). (5) in eq. The pFcurves are usually determined by removing water from an initially wet soil sample (desorption). In practice hysteresis curves are not determined at the laboratory and are hardly applied in models. The hysteresis problem is greatest in sandy soils at low suction.l ) Writing do dh and substitution of eq. (8) in to eq. The hydraulic conductivity curve can be represented in different ways.2. which is known as the differential soil water capacity C C = . . Whether or not considering hysteresis depends mainly on the type of soil and if the soil is exposed to frequent wetting and drying. This phenomenon is referred to as hysteresis. If one adds water to an initially drier sample (adsorption). 8 = f(h). Soil physical properties The soil water characteristic. (6) yields: To avoid the problem of the two dependent variables 0 and h.(cm . Examples of such expressions are given in fig. With the aid of numerical models it is possible to account for effects of hysteresis. For more information about hysteresis. the moisture content in the soil will be different at corresponding tensions (see fig.
where KO = saturated hydraulic conductivity. hl = some pressure limit. fig. = effective conductivity attained after rewetting.a and n are constants. = the slope of the curve and h. and according to Bloemen (1979). ha = pressure head at airentry point. where K. KW is attained. n. fig. . = pressure head at which the hydraulic conductivity is negligibly small. h = pressure head a t which . 2B.h (cm) F g 2 Generalized hydraulic conductivity functions according to Rijtema (1965). a. i. 2A.
Writing eq. (10) in a finite difference from yields: The functional relationships C(h) and K(h) now must be 'averaged' over time and space. and vertical coordinate. similar remarks as for the 0 (h)relation can be made. the maximum time step to be allowed during the computation is relatively small. This upward flow is usually termed capillary rise. Steadystate capillaiy rise In The Netherlands the water use of crops is effected by the upward flow from the relatively shallow groundwater table. Integration of eq. In heavy clay soils swelling and shrinking during the year affects the K(h)relationship. For special types of K(h)functions q. Because of the high nonlinearity in the C(h) and K(h) functions. pressure head. is always possible. both for homogeneous and heterogeneaus (layered) profiles. For a unique solution of h with respect to time and space initial and boundary conditions must be applied. Solution of this equation for nonhomegeneous soil profiles is only possible with numerical and analog models. Nonstationary flow In case of a nonstationary flow. As initial condition either the pressure head as a function of depth z must be given h(z. Therefore eq. (5) in steadystate conditions yields the relationship between flux.3. analytical solution of the integration is possible. however. In fig. 2. (5) can be written in a finite difference notation as: c.4. 2B an average scanning curve between drying and wetting was taken to account for this effect. (10) is valid. Bouma and De Laat (1980) have shown that reduced K(h)curves should be used for a column as compared with the curve obtained for aped by standard laboratory methods. Solution by numerical integration. t = 0) = h. Vauclin et al. z. (z) or the moisture content as a function of the depth z (1 1) . and applied to each layer separately.As far as hysteresis in the K(h) curve is concerned. 2. 6h/6 t # 0 and eq. (1979) give a good description of the consequences of this averaging process.
Functional relationship (Cauchy condition): the flux is a function of the dependent variable h at z = ZB Use of this type is only possible in iterative computations. In the next chapter the systems at the boundaries will be treated. at the top and the bottom of the unsaturated zone can be specified in three different ways: .Dirichlet condition: the pressure head is specified as a function of time h(z = z. Through these boundaries relations can be established with the atmosphere and the saturated zone. Boundary conditions at z = zB i. changes in the water table depth were calculated vvithout considering the actual respons of the unsaturated zone and a mostly unique relationship between mean groundwater table depth during the growing season and evapotranspiration/crop yield was taken (see fig.e. In general at the top and at the bottom of the unsaturated zone different types of boundary conditions can be used at the same time. Even during the process of computation the type of boundary conditions may be changed. l . . 3. The connection with the saturated zone is usually established by the flux . as far as they affect the unsaturated zone and vice versa. .. For the unsaturated zone the boundaries are constituted by the soil surface and the phreatic surface. Saturated groundwater system As "lower" boundary of the unsaturated zone the phreatic surface is acting.Neumann condition: the flux is specified as function of time (13) . it is possible to consider both the saturated and unsaturated zone as one continuum (see Neumann et al.groundwater table depth relationships which van be derived from some drainage formula. with the use of numerical models.must be applied. BOUNDARIES OF THE UNSATURATED ZONE 3 . Recently. In the more traditional approach of evaluating water management measures.t) = hB(t) . 3). 1974).
again. Here. GD. 1979). In this way these fluxes constitute a Neumann type of boundary condition. the kind.2. Often the combined water loss from the crop . GD. the potential rate of infdtration is controlled by atmospheric (or other) external conditions. and Tpot are the potential evapotranspiration flux. i. stage and condition of the crop and the soil water status. respectively. Similarly if the potential rate of infiltration (e. the rain or irrigation intensity) exceeds the absorption capacity of the soil. per unit area of soil. The water uptake by roots.soil surface is considered as one term. Atmospheric system The atmosphere influences the unsaturated zone in two ways: at the soil surface through infdtrationlevaporation and at the crop surface through transpiration.. zero.see Van den Berg. = highest groundwater table depth where reduction is appr.g.. i. the transpiration is dependent on the conditions of the atmosphere.e. The soil can lose water to the atmosphere by evaporation or gain water by infiltration. during the growing season on crop yield reduction. 3. . part of the water will be lost by surface runoff. the actual flux across the soil surface is limited by the ability of the porous medium to transmit water from below. whereas the actual flux depends on the soil physical properties and on the soil water conditions. evapotranspiration. The maximum possible flux through the soil and crop surface can be defined as where ETpot. soil evaporation flux and transpiration flux. at it's maximum (after Grootentraast. E. i.mean groundwater table depth 0 a GDrnox I  max F g 3 Effect of groundwater table depth... GDo = lowest groundwater table depth where reduction is appr.e. While the maximum possible (potential) rate of evaporation from a given soil depends only on atmospheric conditions.
the potential or actual transpiration flux must be equal to the water uptake rate of the plant roots. Neglecting storage in the crop. S The potential evapotranspiration flux.S l ) where. Keijman. rl .Potential evapotranspiration flux can be derived from a combination of the energy balance equation and the vapour transport equation (Rijtema.. Epot = . is the aerodynamic diffusion resistance and r. the internal canopy resistance built up of a stomatal resistance dependent on light intensity. The evapotranspiration flux of a wet crop surface ETwet. r. Also the way in which nonoptimal soil moisture conditions influence the transpiration rate will be discussed. According to eq. according to Rijtema (1965). 1981). 1965.) S+ Y ET. Stricker (1981) compares the results of the calculations defined with the three expressions with water balance measurements. rc = rl + rsc (sm l) (19) Alternative expressions for estimating maximum possible evapotranspiration fluxes are the Priestley and Taylor (1972) equation and the Thom and Oliver (1977) equation.^ (kg m ' .391 (kgm ' Q* (S + 7) S . A is latent heat of vaporization (Jkgl).K l) and ET. Monteith. The potential evaporation of a cropped soil can be computed by neglecting the aerodynamic term and taking into account only that fraction of Q* which reaches the soil surface (Ritchie. is the aerodynamic evapotranspiration flux (kgm ' . 1973). (16). 1965.and a resistance dependent on the fraction of soil cover.me2). . ETpot. l). the potential transpiration flux now can be calculated as the potential evapotranspiration flux minus potential soil evaporation flux (minus evaporation flux of intercepted precipitation)./r. He also discusses the practical applicability of the various methods.S l) where I is the leaf area index. This index generally can be related to soil cover. can be written as: where s is the slope of the saturation vapour pressure curve (mbar. y is psychrometer constant (mbar.Kl). can now be defined as: ETp~t = s + y ( l + r. Q* is net radiation flux @.0.. In the next chapter this water uptake will be described mathematically. r.
Laand Rphnt are the flow resistaiices (d) in soil and plant respectively. Transpiration is expressed here as a volume flux.. 1980).hoot . (23) is the determination of this root effectiveness function. (a) Analogous to Ohm's law the steady state flow of liquid water through the soil . boil. i.hoot . one generally prefers to use an equivalent form of eq.e. From eq.. As it is still not possible to measure h. When the transpiration demand of the atmosphere on the plant system is too high or when the soil is rather dry. One of the major difficulties of eq. cm3 . Therefore Rplant does not include stomatal resistance.d is equivalent with a volume flux of 1 mm. and hsaf are the pressure heads (cm) in the soil at the root surface and in the leaves respectively.d l).d ' ) ' ' where h. it appeared that the function b could be changed quite drastically without seriously affecting the transpiration rate. considered as liquidphase resistances.4. WATER UPTAKE BY PLANT ROOTS To describe water uptake by plant roots several approaches are available. and RPM influence hleaf in such a way that transpiration is reduced by closure of the stomata.heaf Rson Rphnt (cm .cm' . From calculations of Van Bake1 (1979) with an unsaturated flow evapotranspiration model (De Laat.hkaf (cm . (22): ' T= hsoil.#. can be written as: where b is an empirical root effectiveness function (cm) and K(h) is the hydraulic conductivity of the root zone.d' ) L a + Rpbt (22) Equation (22) can be applied to either the root system as a whole or to discrete horizontal layers. We restrict ourselves to the most common methods. h .root system can be described as T = hwil. (22) it can be derived that .d . is the rooting depth (cm). Feddes and Rijtema (1972) derived from a number of crops that where z. (Note that a mass flux of 1 kg HzOm .. The soil resistance.
(25) and (26) that ET depends on rh. By definition the integral of the sink term over the rooting depth. (18) as (Rijtema. 27a). (22a): Some empirical rhfunctions for different types of crops are listed by Van Bake1 (1979). Feddes et al. z equals the actual transpiration rate . (1978) considers the sink terms as a function of the soil water pressure head. water uptake by the roots is represented by a volumetric sink term. in principle ET can be solved iteratively. h.l ) ZI In nonoptimal soil water conditions water uptake is reduced according to . Because rh depends in its turn on (E)T (eq. Again. which is added to the continuity equation (6): " where S represents the volume of water taken up by the roots per unit bulk volume of soil in unit time (cm3 . For optimal soil moisture conditions.cm 3 d ). For crops with full soil cover T ET. T = Tpot and S(h) = S . (b) In the second approach. according to eq. It appears from eqs. 1965): with rcl = rc + rh and the diffusion resistance rh is depending on leaf water pressure head. = POf (d .Actual evapotranspiration can now be described similar to eq. one faces the problem of determining the magnitude of the sink term. So (26) or. .
. Incorporation of the sink term into eq. h4 = wilting point. Some authors (e. 4. fig. depending on the evaporative demand of the atmosphere.e. pF = 4. (after Feddes et al. Yang and De Jong. = S(h)/S. 4). Feddes et al.16..2).. With such methods he may evaluate effects of water management upon evapotranspiration. 1978) An example of the shape of a(h) is shown in fig.) and soil water pressure head..000 cm (i. 1978) apply a fixed range. (lob) is essentially the same as for eq. crop production. at which deficient aeration conditions exist. hl = anaerobiosis point... h3 = highest value of h where S = S. For the wilting point (h4) one usually takes a hvalue of . (10a). duration of anaerobiotic situation. 1972) found a varying range.g. water quality. Little is known about the anaerobasis point. In the following an overview of the various categories of methods that are nowadays in use or developed in The Netherlands will be presented. different approaches can be used. In any case. composition of vegetation. different values of h2 and hg are used.. it will depend on type of crop. INTEGRATED APPROACH TO EVALUATE EFFECTS OF CHANGES IN SOIL WATER CONDITIONS ON EVAPOTRANSPIRATION To optimize different schemes of water management it is important to have methods available that can help the water manager in making his decisions. . (h. temperature and type of soil. 5. They range from methods that are based on semiempirical relationships to fysicalmathematical models that describe the system in a very detailed way. h.g. For the hrange (h2 + h3) where transpiration is potential. h2 = lowest value of h where S = S. while others (e.(a(h) i. Restricting ourselves to the effects of changes in soil water conditions on evapotranspiration...F g 4 Relation between sink term variable a. etc. (10a) yields: The numerical solution of eq.
it is evident that the curves may undergo a horizontal shift to the right in wet years and a shift to the left in dry years. Use of it is to be advised against. Paramem'c models This type of models is still empirical but make use of some properties of the unsaturated zone. Yield depressions at high water tables can be ascribed to lack of aeration of the soil. Then the groundwater table depth GDo is .This point corresponds with a theoretical groundwater table depth G . Since both conditions may occur dependent on climatological circumstances.yield depression ( O h ) Il I heavy rlverandseo clay lmmy humusnch a s n d 1 m mean depth of water table during the growing period (cm below the surface) F g 5 The effect of the mean depth of the groundwater table during the growing season on fimal i. but the disadvantage that they are not applicable to other areas than for which they were developed. (after Visser. 1958) 5. Visser (1958) gives the yield depression for seven soil classes with various mean depths of the water table during the growing seasons (see fig. 5).1. . where the capillary rise approximately is zero. crop yield for seven groups of soils. depressions at deeper water tables are due to shortage of water. however. This gives the maximum yield D depression D.2. 1979) determined the yield depression of grassland on sandy soils due to an artificial drawdown of the groundwater table according to the principle presented in fig. These curves aTe reflecting the behaviour of the unsaturated zone and its effect through evapotranspiration on crop production. Empirical methods This category of methods has the advantage that they are simple t o apply. It should be remarked that attempts have been made to adopt the curves to changed circumstances. For a number of years the relative evapotranspiration/yield is computed for conditions where no water table is present. Hence the unsaturated zone is used here as a 'blackbox'. Grootentraast (see Van de Berg.. 5. 3.
is water storage of soil profile at t = t and t = o respectively. For a unit area the water balance of the unsaturated zone over a certain period can be written as: STt = ST. One of the most important parameters is the water storage ST of the unsaturated zone. sprinkling. water supplied from outside the system (precipitation. A reservoir model that takes into account the influence of the groundwater table has been developed by a Working Group of ICW (1979). seepage)..determined at which a reduction in evapotranspiration/yield never will occur and the points GD. EvapotranspirationET depends on ST as follows: . t Q . ET actual evapotranspiration. and GD. ST. With this method a number of properties as the soil moisture retention curve and the hydraulic conductivity implicitly are taken into account.ET (cm) where Q ST. (34) applies to (vegetated) soil with deep groundwater tables (no capillary rise). and ET = ETpot it follows thst Eq.. Thornthwaite and Mather (1955) calculated the actual evapotranspiration ET according to under the assumption that and taking at t = to : ST = ST. are connected by a smooth curve.
+ STt ST = 2 Because STt is not know beforehand.2 < pF < 4. From eq. i. This type of relationship can also be obtained from experiments of Denmead and Shaw (1962) and Van Bavel(1967).2 pF = 4. If this value differs significantly from the initial STtvalue. (5a) yields the relationship between q. a. Numerical integration of eq. the computing process is repeated. Steadystate models These models can be applied when calculating percolation profdes or steady state capillary rise (evaporation) from the groundwater table towards the root zone or the soil surface. Federer. the procedure of calculation starts with a first estimation of STt.3. Otherwise the next timestep is taken. In the above mentioned two methods the reduction factor is independent of the magnitude of (potential) evapotranspiration. Physicalmathematical models This group of models is based on fysical processes and mathematical relationships as described in the chapters 24. (39) ET is computed. For examples and computer programs see AlivertiPiuri and Wesseling (1979) and Bloemen (1980).g. 6. Therefore some authors (e. for OGpF93. From a computational point of view there is no problem in handling nonhomogeneous soil profiles. One can imagine that at low potential evapotranspiration rates reduction in evapotranspirationoccurs at lower values of available soil water than at high potential evapotranspiration rates. z and h.where c= l for for c=O The value of ST is derived as ST. (38) yielding a new value for STt .e. STt = ST.2 3. An illustration of the effect of a disturbing coarse sand layer in a fme sandy soil on the capillary flux from the groundwater table towards the root zone is presented in fig. and the ratio of available soil water (or availability factor) and potential (evapo)transpiration.. 1979) found a relation between the reduction factor. This value is substituted in eq. 5.2 .
The pressure head at the lower boundary of the root zone is in both cases . 6 Relation between groundwater table depth. precipitation (infdtration).1000 cm. i. Changes in water storage of the soil profile are found from the difference in two subsequent steadystate situations. q) through the pFcuwes. The pressure head profiles h(z. no vertical gradient over this zone exists. In the root zone all water is taken to be transported through the roots. change in available soil water. He distinguished two zones in the soil profde: the root zone and the subsoil.ST(ql).e. Pseudo steadystate models A nonstationary process of water flow can be approximated by a subsequent series of steadystate flow situations. ST(q2) ST(q1). It is assumed that the evapotranspiration rate is at its potential value when the tension . and capillary rise. The model computes actual evapotranspiration over a certain period from the water balance of the root zone: ET=P+Q+AST (cm) where ET P Q AST is evapotranspiration. A ST = ST (qz) . The subsoil is considered as a homogeneous single system.1 = f~ne sandy sot1  2 = flne sandy so11wlth a coarse layer at 5060cm below the soil surface  50 100 150 I I 200 250 (cm below the so11 surface) Fig. GD. (after Mooy. in a homogeneous (1) and a heterogeneous profile (2). capillary rise from the subsoil. Rijtema (1971) used a pseudosteady state approach to compute relative evapotranspiration during the growing season. personal communications) b. q . q) profilescan beintegrated over a certain soil depth yielding water storage ST(q).The B(z. q) can be converted into water content profiles 8(z.
. When this limit is exceeded. (30) and (3 l). c.". Ehat.layered soil profiles. Rijtema (1971) assumed a futed value of hl of . The traditional disadvantage of a large amount of computing time inherent to this approach becomes less and less significant as recent trends in computer technology are directed towards faster computers.!. for two... (27a) and (25) to calculate ET.16. Feddes et al. De Laat (1980) developed a computer program for this model and extended it for periods with rainfall surplus and for heterogeneous subsoils.E':!~. (after Feddes et al. a reduction in evapotranspiration occurs. balance. 1978) . Theoretical cumulative flow (cm) 30r June l July I August I September I time (days) October Fig. Transient models The nonstationary process of water flow can be approximated by a numerical solution of eq.000 cm. (lob) with the sink term described according to eq. transpiration p.::' E ! and measured cumulative evapotranspiration.. De Laat can apply subsequently the eqs.soil evaporation. (18).in the root zone is smaller than a certain limit h. (lob). (22a). 7 Computed cumulative evapotranspiration. (1978) developed a computer program based on eq. In fact this approach is the most simple one as it needs less restricting assumptions.
i. Important properties are: . From experience it appears that most problems are encountered with regard to the following items: . where flow takes place under cropped field conditions in a fivelayered anisotropic soil. secondly the cumulative transpiration Tcomp as computed with the model by integration of the sink term over depth. (1974) developed a twodimensional finite element model for transient flow in saturatedunsaturated soils.soil use as far as it influences evapotranspiration through its roughness.e. Van der Ploeg et al. . (1975). from the computed terms of the thirdly the cumulative soil evaporation Ez:Pderived water balance. . collection of data depends on the type of model chosen. Secondly it depends on the availability of the above mentioned data. In fig.) as obtained from the lysimeter. Reversely.results predicted by the model were compared with a field experiment in which red cabbage was grown on a heayy clay soil in the presence of a water table. CONCLUSIONS AND RECOMMENDATIONS In the previous section a number of methods and models have been discussed. Nimah and Hanks (1973a/b) and Childs and Hanks (1975). In the first place the choice of method or model depends on the objective of the study (communication between policy maker and model user). Water balance studies were performed with a specially designed nonweighable lysimeter. In practice there are difficulties in choosing the right method or model and translating the onedimensional model results obtained at one point to larger areas. 6. (1978). The above mentioned model applies to onedimensional vertical flow.influence of very wet conditions upon plant growth and transpiration. slope. The latter take also into account the presence of salts. rooting depth.estimation of rooting depths. etc.topographical data such as altitude. The complex reality has to be schematized into a proper way.soil physical parameters as soil water characteristic. etc. also as function of time. From sensitivity analyses @e Laat. .b. which is very poorly understood. one has to distinguish the soilplantatmosphere system in a limited number of characteristic situations. Neuman et al. . 1980) it appeared that a good estimation of this depth is very important for the calculated evapotranspiration.. b. 7 curves of cumulative flow are given: first the measured cumulative evapotranspiration (ETwat. the boundaries of which include two ditches on the side and a pumped aquifer at the bottom. water availability for the plant and storage coefficient. Other examples of such models are given by Van Keulen (1975). capillary conductivity and derived properties such as infiltration capacity. An example of this approach can be found in Feddes et al.
.000 Fig. is seldomly taken into account. rs.representativity of meteorological data.soil loosening and compaction have distinct effects on soil physical properties (Boone et al./ 2000 0 10. as for example analyzed by a Working Party (1968). 8. .spatial variability of soil physical properties. Of all the models discussed in section 5 none has the possibility to incorporate this phenomenon. . Nielsen et al. From experiments (Thomas et al..6000 I 4000 X X 2 stresses .. .in simulation models topograhical variation of the soil surface within the area of investigation. 1976) .000 20. . Bouma (1977) stated that much testing remains to be done to extrapolate data measured in a particular soil to unmeasured soils with identical classification elsewhere. hl. due to differences in soil water stress history. .f. This problem is discussed elsewhere in this paper.fast and simple determination of the soil physical properties. (after Thomas et al. and leaf water pressure head. Especially the attempts of Ernst (1978) to establish drainage formulas for areas with composite water course systems Should be emphasized. however.000 30.reaction and adaptation of an active growing plant system on changes in external circumstances..000 40.field determination of the drainage respectively infdtration resistance of open water courses. 1976) it can be deduced that the history of water stress can be very important for the subsequent reaction of crops on water Shortages. (1973) found from their investigation in California that even seemingly uniform soil areas manifest large variations in hydraulic conductivity values. This is clearly demonstrated in fig.translation of the soil map into a map of soil physical characteristics. 8 Differences in relationship between stomata1 diffusion resistance. 1978).. . they are taken as invariant with time. based on soil texture and organic matter content of the soil. Mostly. as developed by Bloemen (1979). . A promising approach is for example the estimation of K(h)relationships.
. SEBASTIAN (1978) . however. Maybe one could take advantage of the experience obtained in this respect in the USA (Bachmat et al. BACHMAT. Nota ICW 1156. This is a drastic simplification of reality. would be very useful. 1980) Some knowledge of the physicalmathematical background of methods and models will always remain necessary for the field hydrologist. Feddes for his critical remarks made after reading the first draft of this paper. while on the other hand the soil physical models are bounded by the atmospheric conditions. A kind of a clearing house for models. Micrometeorological approaches are in fact bounded by the soil water pressure head in the root zone. HOLTZ and S.Calculation of height of capillary rise in non homogeneous soil profiles. model user and model builder. 1978). . D. one may conclude that many problems remain to be solved. A way out of this dilemma would be t o let model specialists compute a great number of selected flow cases and let them derive from it some generalized relationships.choice of the method or the model that is best suitable to apply to a certain situation. Looking over the present state of the scientific knowledge dealing with the unsaturated zone in relation to evapotranspiration.accessibility of data. .G. including new developments. drain depths and weather conditions as carried out by Feddes and Van Wijk (1977) in the framework of the working group on Reevaluation of Land Reclamation Projects (see also Working Group H.E. pp.A.. ACKNOWLEDGEMENT The author is greatly indebted to dr. .formulation of the problem in hydrological terms. B. 7.. 46. Especially data on crops. A&.coupling of the meteorological system and the soil plant system.. R. soil physical and hydrological properties for larger areas that can be applied on a routine basis are scarce and difficult to obtain. As an example of such an approach one could refer to the evaluation of water management on crop production for a number of soils. WESSELING (1979) . Oklahoma. 177. ANDREWS. The presentday models need much data which should be computer accessible. This will help to select criteria for practical field situations. .L.Utilization of numerical. pp. E. that can present an overview of models that are available and that contains directions for the user. groundwater models for water resources management. Y. From the field hydrologist it is not to be expected that he keeps himself fully informed about all the methods and models available.P. This paper may serve as a small contribution to that end. There seems to be large discrepancies between policy maker. REFERENCES ALIVERTIPIURI. and J.
NIELSEN.Spatial variability of fieldmeasured soilwater properties. BEESE. and RH.An integrated modelapproach to the effect of water management on crop yield.Berekening capillaire doorlatendheid uit granulaiue samenstelling en humusgehalte. Agron.The water balance of . DE SMET (1978) . R. 26(4): 405420.Simulation of field water use and crop yield. Biol.Berekening van de capillaire opstijging in niethomogene bodemprofielen. pp.L. Deelrapport voor de Werkgroep Landbouwkundige aspecten van de Commissie Invloed grondwateronhekking op de omgeving. 34. 126. LAAT. Repr. R.Finite element analysis of two dimensional flow in soils considering water uptake by plant roots. 11 (3): 452460.Sensitivity analysis of models for evapohanspiration. STREBEL and M. pp. NEUMAN and E. pp. Wageningen. plant and atmospheric interrelations. HANKS (1973) . and P. Agric. (in press). Waterstaat Gelderland. FEDDES. FEDDES and E.A. and R.Evaporation and environment.. of Hydr. Agr. Agr. Misc. BOUMA.Finite element simulation of flow in saturatedunsaturated soils considering water uptake by plants. BRESLER (1975) . SP. DE (1980) .R. Res. (1980) .Estimation of moisture supply capacity of some swelling clay soils in The Netherlands. VAN WIJK (1976) . 189. RENGER (1978) .J.H. MONTEITH. P.R. KEULEN. TNO. BLOEMEN. HANKS (1975) . and A. J. ICW 252.Model for unsaturated flow above a shallow water table. FEDERER. Res. 117131.Inventarisatie van de methoden ter bepaling van landbouwschade ten gevolge van grondwaterwinning. VAN DER.A case study of soil compaction on potato growth in a loamy sand soil. 176. P. Pudoc.A soilplantatmosphere model for transpiration and availability of soil water. (1978) . 39: 231237. Field applications. Landbouwk. FEDDES. RIJTEMA (1972) .. Hydrol. 17: 3359.M.W.Model of salinity effects on crop growth. Misc.T.M.A. Proc. Soc. Deel I: Berekening van opbrengstdepressies. J. J. BERG. VAN (1979) .J. (1977) .N. Res. Arnhem. 39: 617622.R. R. P.E. Wageningen. Physical measurements and rooting patterns.Availability of soil water to plants as affected by soil moisture content and meteorological conditions. Soc. J. Symp. Tijdschr. Water Res. of Hydr. pp. ERNST.F. CHILDS. Wageningen. of Agr. NIMAH. Y. of Hydr. Hilgardia 42(7): 215259. pp. 0. pp.T. Simulation Monograph. DE LAAT (1981) . Am.P. Soil Survey Inst. O.Drainage of undulating sandy soils with high groundwater tables. Comm. FEDDES. M.Q.M. BLOEMEN. PLOEG. Rep. L. 11.A dependent domain model of capillary hysteresis. Research. 54: 385390. Exp. ZARADNY (1978) . F./pt 92(4): 183189.BAKEL. Techn. C. 104. KEIJMAN. Proc. Hydraulic Engineering Laboratory. R.J. Proc. ERH (1973) . DAGAN (1975) . pp.M. Proc. Am. Wageningen. Pudoc.. R.J. Israel.Soil survey and the study of water in unsaturated soil. (1981) Theoretical background of some methods for the determination of evaporation. Soil Sci. 4: 165176. VAN DEN (1979) .G. and G. J. 895. J. R.Model for estimating soil water. D. GermanDutch symposium: Models for water management 11.. 37: 522532. 15(3): 555562.W. J. BOUMA and L. S.Simulation of water use and herbage growth in arid regions. SHAW (1962) . S. Wat. VAN (1967) . Pudoc. Landbouwk. Repr. J.H. Bull. BRESLER (1974) . MUALEM. BIGGAR and K..L.Changes in canopy resistances to water loss from alfalfa induced by soil water depletion. J. Am. 39: 150. J. Soil Sci.Water withdrawal by plant roots. KOWALIK and H. J.W. ICW 245. 19: 205234. Soil Sci.A. 1.. Proc. and R.W. BAVEL. Res. Sci. The Hague: Proceedings and Informations 28: 1224.J.T. CP. F.(1979) . G. (1965) . Prov. Soil survey paper 13. Water Management I (1): 320.J. BOUMA.A. ICW 103. and P.A. Met.A. VAN (1975) . J. FEDDES. Neth. Simulation Monograph. (1979) . Tijdschriftlpt 91 (12): 329335. NEUMAN. H. BOONE. DENMEAD. 107.
68: 706708.a sugar beet crop: a model and some experimental evidence.P. NotaICW 1115. 86. Agron. Soc. 141: 313328.TNO.On Penman's equation for estimating regional evaporation. Centestor. ICW 56. A. and R. WORKING GROUP of ICW (1979) .Field determination of hysteresisin soilwater characteristics. THOMAS. WORKING PARTY of ICW (1968) .C. PRIESTLEY.Resolution numerique d'une equation de diffusion nonlineaire. THOM. 28: 5975. Month. HAVERKAMP and G.J. J. The Hague. 85. Wat. No. and J. USA. Publ.C. OLIVER (1977) .R. Met. THORNTHWAITE. C. Res. S.Kwantitatieve waterbehoefte van de hoogheemraadschappen Deltland en Schieland. Presses Universitaire de Grenoble. VISSER. J. 39(2): 221223. 8(5): 12041213.W. VAUCLIN.Methods of estimating evapotranspiration from meteorological data and their applicabiiity in hydrology. Rep. '~Gravenhage.L. C. W. (1972) . inClimatology VIII. Pudoc. 107. Amer.A model for predicting evaporation from a row crop with incomplete cover..De landbouwwaterhuishouding van Nederland. pp. M. Techn. WORKING GROUP H.Roy. Comm. Res.W.On the assessment of surface heat flux and evaporation using largescale parameters. pp.pp.Determination of the optimum combination of water management systems in areas with microrelief. Weather Rev.M. pp. (1981) . (1958) . Onderzoek Landbouwwaterhuishouding Nederland . pp. 103: 345357. Agron. pp.H. and G. Planzenern. l. Agric. MATHER (1955) . Comm. 659. 2. Wageningen. Research TNO. Res..Stomata1 response to leaf water potential as affected by preconditioned water stress in the field.S. 188.E. and E. RITCHIE.M. K.Effect of aerial environment and soil water potentialon the transpiration and energy status of water in wheat plants.R. J.Methode voor de evaluatie van landinrichtingsprojecten. Quart. . 100: 8192.An analysis of actual evapotranspiration. and Bodenk. STRICKER. P E . (1965) .J. 103. Hydrol. J. DE JONG (1972) . VACHAUD (1975) . TAYLOR (1972) . Soil Sci. Proc. pp.B. J. VACHAUD (1979) .T. 159. ROYER. and H. Bull. J.N. Soc. R. (1980) . Proceedings and Informations. J. 64: 571578. YANG. BROWN and W.The water balance. RIJTEMA. JORDAN (1976) .R. 183.
). including the waterbalance. daily values of ET have been calculated by different methods. Meteorologists and hydrologists have come to accept that more testing is needed of existing equipment and approaches and more development of new tools for estimating actual and potential ET. because of the spatial variation in vegetation and soil moisture regime over a catchment. for which many experimental and representative basins are well equipped. defmed by (N0. and some conclusions are drawn concerning their usefulness and applicability. careful measurements of ET need more equipment and daily management than rainfall or runoff measurements. preliminary comment is given on interception of grass and the socalled rainfalldeficit. The Final Report . Greenwood (1979). INTRODUCTION In the water balance.N. for example ascertained: "that hydrologists have always found it hard to come to grips with transpiration". One may wonder why. From March 1976 to December 1978. Besides. The data are from e!.METHODS OF ESTIMATING EVAF'OTRANSPIRATION FROM METEOROLOGICAL DATA AND THEIR APPLICABILITY IN HYDROLOGY J. Therefore it is of interest to develop and test methods for estimating ET. such as The Netherlands.8 E. measurements of runoff produce spatially integrated values. are compared. For example. The situation now looks more promising. the process of ET seems less attractive and less urgent to be studied. A third reason may be the complexity of the process of evapotranspiration. But for actual evapotranspiration the representativity is always open to question. In my opinion this fact may be considered as a shortcoming in the program of the International Hydrological Decade (IHD): no timeseries of actual ET from several experimental basins are available. Nevertheless the number of studies on rainfall and runoff far exceed the number on the process of evapotranspiration in hydrological literature.M. Or may be. STRICKER SUMMARY Next to rainfall evapotranspiration (ET) is the most important term in the waterbalance of catchment areas in temperate regions. 1. pointmeasurements and were collected in the experimental basin "Hupselse Be? The results of different methods. Perhaps it is because of the background of many hydrologists that makes them unfamiliar with the subject. Rainfall measurements normally represent an estimate for an area. even to determine potential ET. evapotranspiration (ET) is normally of the same order of magnitude as rainfall or runoff. but are poorly equipped with meteorological instruments. Furthermore.
their validity and restriction. and no deep seepage or infiltration interferes in the water balance. 2. 1975) stated. 1972.50 km2 and may be considered representative of sandy regions in The Netherlands. The catchment (fig. that ET functions in existing catchment models may be inadequate and need improvement by future research. EXPERIMENTAL SITE AND AVAILABLE DATA For a detailed description of the Hupselse Beek Experimental Catchment see the biennial reports of the Hupselse Beek Study Group (1971. Welr 0 500 lOOOm Fig. recommended more emphasis on such problems. 1) is situated near Groenlo and Eibergen. It is supposed to be watertight in the subsoil.W O . The previous papers indicate that progress has been made in the last years and will certainly continue. Some conclusions are drawn about spatial representation of ET by different methods. 1 Experimental catchment "Hupselse Beek" . This paper enlarges on some methods discussed earlier by De Bruin (1981) and which I have tested against data from the Hupselse Beek Experimental Catchment. It covers 6. 1976). A report on Operational Hydrology W O . 1977) of the Technical Conference on the Assessment of Areal Evapotranspiration. Only basic information is given here. held in Budapest. 1974. It discussed the results of an intercomparison of catchment models and concluded that blackbox and conceptual models have a weakness in the way of modelling rainfall to effective rainfall. A complicating factor is the local presence of 7 0 P 50 100 R ~ v u l e t =Road t .
50 m in that year. : Additional data from Lambrecht cupanemometer at "Hupselse Beek". In 1978.10 and 1. Land use is predominantly agricultural: 70% pasture. Station Soil moisture Period complete complete * 950 days*l complete until*' 770709 + 1000 days*3 917 days ~ o m p l e t e * ~ at 2 or 3 levels. and 6% forest. temperature from the sensor at 3. which causes a high local resistance t o vertical transport of water. : No definite judgement exists about the accuracy of the data and about their representation of the catchment. Table l Available data for calculations over the period March 1976 to December 1978 (1036 days or 34 months) Data Rainfall Runoff Net radiation Rel.00m became unreliable because of insufficient shielding against radiation. 20% arable with maize as the main crop. Hair hygrometer Thiess anemometer Semiconductors 60 rnin 60 rnin * 60 rnin 15 rnin bimonthly Semiconductors nearly ~omplete*~ at six sites *' *' *3 *= *6 : Additional data from Wageningen.tilt layers in the upper 1. Table 1 summarizes the available records. Station "Winterswijk".O.S. Careful comparison between RH from Hupsel and Winterswijk over the period March 1976 to the first half of 1977 provided a onetoone relationship between the two sets of data for daily mean values above 60% RH. A full discussion on the accuracy of the meteorological data was given by Stricker and Brutsaert (1978). 1981). Station "Winterswijk" (see text). RH had to be corrected upwards by 1 or 2% and for 5 days by 3% or more. However below 60% RH. So. Rainfall was also collected from other types of rain gauges (Warmerdam. . Over the period March 1976 to December 1978 daily rainfall. humidity Windspeed Air temperature (three levels) Soil temperature Groundwater at Assink Met. Station of the Department of Physics and Meteorology. application of the profde method was restricted to temperature at 0.I. : Additonal data from KNMI Met. : Additonal data from KNMI Met. Met. t 80 days at one level available 1000 days*= complete Interval 15 rnin 1 rnin 5 60 rnin 60 rnin Instruments Recover at groundlevel HLflume C.R. 10 items from the Winterswijk data needed some correction: for 5 days. runoff.S0 m. Data on relative humidity (RH) became erroneous from June 1977 and data from Winterswijk were taken. : Additional data by interpolation. and actual and potential evapotranspiration were collected or calculated from catchment data.
3. CALCULATION OF POTENTIAL AND ACTUAL EVAPOTRANSPIRATION 3.1. Potential evapotranspiration Several formulations of potential evapotranspiration have been applied to mean daily data over different periods of the year. The following have been used, with the period of the year given in brackets: XETPEN= (Q*  G) + z ( 3 . 7 S+Y
S+Y
+ 4.0 C)
(all days)
=a
s (Q*  G) (April, May, June, July, August, Sept.) S+Y
(3)
List of symbols: X : latent heat of vaporization (J/kg) :potential evapotranspiration above grass under measured conditions ET^^^ (kg/m2 .S) of net radiation (mbar/K) S : change of saturated vapour pressure of air per O K Y : psychrometric constant: 0.66 at 295OK (mbar/K) Q* : net radiation or net radiant flux density ON/m2 G : soil heat flux or heat flux density of ground @/mZ) U : mean daily wind velocity (m/$ (mbar) : mean daily water vapour deficit in the air at 2 m height Ae : potential evapotranspiration of a dry canopy by the method of Thom and Oliver (1977) (kg/mz .S) n : ratio of canopy resistance (r,) to aerodynamic resistance (r,). Here 1.2 has been adopted for grass. With eq. 2 for a wet canopy n becomes zero. r, 2 65 s/m m : ratio of the aerodynamic resistance as expressed by the Penman equation to the aerodynamic resistance over a grass surface. Here 1.9 has been adopted. (1) ET^^+^ : potential evapotranspiration formula, as proposed by Priestley and (kg/m2 .S) Taylor (1 972) a : empirical factor which has to be calibrated in areas of different climates. Here a = 1.28. (1)
"
Note:
86 400 X kg/m2 S = 1 kg/m2 .d For water, 1 kg/m2 .d A 1 mm/d 
3.2. Actual evapotranspiration From temperature measurements at heights of 0.10, 1.50 and 3.00 m, sensible heat flow into the air can be calculated by the aerodynamic profile method. This method was already applied to Hupsel data for the summer period of 1976 (Stricker and Brutsaert, 1978). From the daily sensible heat flow, the daily rate of evapotranspiration can be derived by means of the energy budget at ground level by the equation:
AETAC, Q*, G and H are the actual rate of vapour loss, net radiation, soil heat flux and sensible heat flux respectively, all expressed in W/m2. Spil heat flux was calculated by the calorimetric method (Kiiball and Jackson, 1975). For 917 days, daily AETAC could be calculated, but values from the winter halfyear were of limited value, because of the relatively low accuracy in the estimate of H. For 119 days, values were missing. For the month June, July and August, actual evapotranspiration was also estimated by the advectionaridity method, as outlined by Brutsaert and Stricker (1979). The basic idea stems from Bouchet (1963). Two empirical formulations are suggested here: XETAD1= 2* X E T ' ~ ' ~ and hETAD2= 2 * X E T ' ~ ' ~
 X E T ~ ~ ' O(June,  XETPEN (June,
July, August)
(5)
July, August)
(6)
in which XETAD' and A E T ~ are supposed to be estimates of daily actual evapotrans~' piration (ET), expressed in W/mZ . No further comment will be given here on the different methods, but the reader is referred to the preceding papers (Keijman, 1981; De Bruin, 1981) or to the papers cited.
4.
RESULTS AND DISCUSSION
4.1 Water balance
.
Al relevant components of the water balance were measured independently at Hupsel. l Thus ET can be estimated as a remainder by the simple equation
Table 2 Monthly totals of daily net radiation, rainfall, runoff and evapotranspiration, estimated by different methods and expressed in mm/month. Note: The symbol E in tables and fig. is identical to ET in the text.

Time
EAC
Rn
not av.: 32.6 63.0 73.6 64.1 56.9 34.4 18.7 3.7 2.0
( 3.7)* ( 23.6) ( 28.2) ( 9.4) ( 28.7) ( 1.3) ( 0 ) ( 0 ) ( 0 ) ( 94.9)
Precipitation 36 (estim)
Runoff
March '76 April May June July August September October November December
Subtotals
385
January '77 February March April May June July August September October November December
Subtotals
9) ( 4.1 88.3 27.2 15.1 70.6 75.47.8  m 64.8 75.4 20.4 48.12.5 64.0 49.5 83.7 18.2 20.3 * Values between brackets have been taken from either ETHcO or EAD2 .5 65.9 62.6 19.5 76.2 10.4 36.9 78.8 11.8) (1.6 18.1 75.5 54.9 .3 38.3 436.5 3.2 59.8 16.5 2.2 35.3 47.2    Subtotals 408.1 11.3 2.1 .9 91.1) ( 5.0) ( 3.3 26.8 5.6 63.4 25.4  January '78 February March April May June July August September October November December (1.1 9.2 51.7) V .  462.3 57.6 61.6) ( 20 ) 226.9 62.3 31.4 67.0 101.7 12.1 438.7 9.4 70.9 48.5 9.2) ( 0 ) ( 0 ) ( 1.1 53.4 65.2 72.5 55.3.7 55.6 13.4 88.3 ( 0 ) ( 0.7 18.8 17.4 110.5 17.2 17.4 39.1 8.8 15.4 32.5 10.9) ( 1.5 56.3 61.2 16.4 3.3    62.8 673.Table 2 (second part) Time Rn Precipiration Runoff EAC EPEN ETH+O EPR+T EAD~ EAD~ .4) ( 0.1 7.2 57.7 13.3 57.8 7.4 10.1 39.
(4). which represents the average of observations from 18 groundwaterwells. id. Table 2 gives montly sums of ET. Table 3 shows the results and explains the combinations selected. No moisture deficits would occur in all other months and several reliable estimates of actual (catchment) evapotranspiration could be made by taking monthly values from different columns of table 2. Several effects may be responsible for these relatively small deficits: .1977 and 1978. 0 97 82 15 24 20 40 **) E A D ~ : id. temperature and relative humidity would not vary Table 3 Summary of the water balance: March 1976 .representativity of the meteorological site. ~m+O. Mean *) E A D ~ Water balance 1858 EAC id. Equation 7 then yields 1338 mm for ET^. P and D were 1858 mm and 480 mm respectively. EPEN.where P D AB :precipitation (mm) : discharge (mm) : increase in moisture storage over the considered period (mm) : actual evapotranspiration from the water balance (mm) ET^ Note : This ET has a different dimension from that in energy balance. July and August 1976. eq. ranging from 20 to 97 over nearly 3 years. Storage increased by 40 mm.errors in data recording.d . From 760301 to 781231 (34 months). July and August 1976. i id. If the catchment suffered from drought only in June. . id. accompanied by a rise of the watertable of about 40 cm. Particularly errors in net radiation would influence the water balance result.E*D~') id. id. ***) EAD1 and EAD2 : applied in June. The meteorological data are point measurements. calculated from daily rates with equations 16.~Am**) id.December 1978 (ET = evapotranspiration) Method for ET P(mm) D(mm) ET (mm) Increase of storage (mm) Deficit (mm) Deficit as % of P 480 id. 1338 1241 1256 1323 1314 1358 40 id.~~~~*** EPEN. The various water balances gave deficits. : only in June. id. id.EADI***) id. Because of nonpotential ET during the dry summer of 1976. Daily mean values of wind. except for ET^'. id. ) id. July and August of 1976. one cannot simply take the sum of each column as a reliable and meaningful estimate of actual ET of the catchment. . potential ET could not be reached. ~m+o.
.................. PL0 r m P m L" L" m L" F g 2 Graphical representation of accumulated precipitation (P) and accumulated evapotranspiration i.... (resp... ...... EAC D .[mm1 18501800 1750  1700 1650160015501500145014001350 130012501200115011001050 1000 950 900 850 800750 700650600550500 P ETH*Oor EAD2 + D . and E ~ or E ~ ~ plus~~ + discharge (D) against time........... Period: March 1976 to De~ ) cember 1978....
. 1977. O continuously exceeds A E T ' ~ ~ and annual values confirm that conclusion. 4.2. Monthly. X E T ~ ~ + O during winter (OctoberMarch). MC. Under moist soil conditions local differences in the rate of vapour loss will not be significant.. For Hupsel it amounts to about 5 mm. Naughton et al. The equations are applied outside the boundary conditions for which they were formulated. Priestley and Taylor.3. 1973. 1972) indicate that it estimates reasonably well potential ET of vegetation with nonlimiting soil moisture and during the period of the year with relatively high net radiation. Particularly during June. after careful analysis. used to calculate AET*'. Davies and Allen. It is therefore important to examine in which period of the year the Priestley and Taylor equation was applicable in Hupsel. especially during summer. July and August actual ET dropped below its potential level and results by the methods of Penman and of Thom & Oliver must be carefully interpreted.28) to for 10d internals was calculated for 1976 (except June. But under moisture limiting circumstances it becomes relevant whether the profile data are collected at a representative site. . Potential evapotranspiration by the methods of Penman and Thom & Oliver De Bruin (1981) showed that for the year 1977 and 1978 daily actual evapotranspiration. 4. soils became dry in the catchment.some seepage may occur in westerly direction. 1977 and 1978 (figure 3). July and August). They found a redistribution of ET of about 27 mm in favour of the winter period. 1979. if one applies the equation of Thom and Oliver instead of that of Penman. reflecting average conditions in the catchment. A E T ' ~ ~and X E T ~ ~ +are in good agreement. So measuring sites have to be selected carefully and with some feeling. Thom and Oliver (1977) obtained a similar result for a catchment study in England and concluded. But the temperature profdes. But results from the literature (Rouse et al. During the summer of 1976. Nevertheless either method is appropriate to estimate potential evapotranspiration. that it is caused by the different structure of the ventilation term (second term) in the equations. 1977. .inaccuracy of the ET methods applied. Potential evapoiranspiration b y the method of Priestley & Taylor Basic to the use of an empirical relationship to estimate evapotranspiration is the need for local or regional calibration and defmition of the conditions under which it can be applied. Seasonal influence is evident.significantly inside the catchment and would be fairly representative.. Priestley & Taylor (1972) defined their formula as an estimator of evaporation from saturated surfaces. . would represent the more local conditions. The average ratio of ET^^+^ (a = 1. Figure 2 shows a graphical representation of accumulated P en (D + XET) against time for two methods to calculate evapotranspiration. Mukarnmal et al.
A value of 1. the ratio is between 0. This suggests that a value of a equal to 1. Periods: April. June.April May June July A ug Sept Per~odsof 1 days 0 Fig. G ~ For a is 1. as can be seen from the equations of Priestley & Taylor and Thom: the first term of hETTH+O.34 in eq. May and Sept. 2 and 3.28 seems somewhat to low and with 1.9 and 1. July and August 1976. 4. over / E ~ ~ + ~ 1976. over 1977 and 1978. 3. 3.l.34 for a ET^^+^ would be in better agreement with results from comparable methods.1977 and 1978. July and August 1977. 1978 (potential conditions!) are comparable for eqs. From the second half of May through August. is about 50% of G X E T ' ~ ' ~ and so X has about twice as much influence on A E T ' ~ +as on XETTH+O. 3 and actual evapotranspiration over the indicated period.38 would be necessary in eq.4. there were noticeable differences between X E T ' ~ ~ and XETTH+O on one hand and X E T ' ~ ' ~ on the other hand. 3 Average ratio of E ~ ~ + ~for periods of 10 days. In spring. This fact was (partly?) caused of nonoptimal moisture conditions in the soils in the region. in which X appears. the sums of daily evapotranspiration over June. And to get comparable results of ETPEN and ET'^'^ over the same period 1. Use of the methods of Penman and Thom h Oliver or of Priestley & Taylor under moisturelimitingconditions For June. During periods of ' .34 results from comparison of eq. July and Aug. This is fairly high for grassland. soil heat flux influences the ratio.
Period: June. 4 and 5 give daily ET^^' and ET^^^ against ET^'. Fig.moisture deficit. the conversion of available energy. estimates of ET^^^ and A E T ~ ~ + will change too and will decrease. the second terms are in ET^^^ and A E T ~ ~ + O determined not only by climatic or largescale atmospheric conditions. 3 is a component of eqs. say 10 m.5. .1978. 4 EAC against EAD1 for days with no or few precipitation (P < 0. 4. Thus. however for the period June. in my view provides a more realistic estimate of regional potential ET. resulting in an increase in ET. 1 and 2. the applicability of eqs. layer of air.5 mmlday). 5 and 6 i the s same as for eq. 5 and 6. July and Aug. July and August. This change in the lower atmosphere enhances the ventilation term in eqs. Eq. into latent and sensible heat shifts in favour of the last component. 1 and 2 estimate potential evapotranspiration under actual conditions of the lower. eqs. Estimation of actual evapotranspiration by the advectionariditymethod Because eq. In fact. That means for instance that if the micro climate is affected by irrigation at a regional O scale. but also by regional moisture conditions. 3: Figs. 3 of Priestley and Taylor is considerably less sensitive to regional conditions of soil moisture and therefore. so that at least the lower layer of the atmosphere becomes warmer and drier than with potential evapotranspiration. of 1976. Q*.1977.
. Aug. of 1976. 1977.1978. Fig. 6 EAC against EAD2 as sums of 3 consecutive days. July. Period: June. 5 EAC against EAD2 for wet and dry days.1977. Aug. Total of 69 points. of 1976. July. 1978.Fig. Period: June.
ETH+Ofor wet days (P > 05. but needs more theoretical and practical evidence before it may become operational in future. .4) of eqs. May. The method. 6 shows ET^^^ (preferred over ET^^') against ET*' for sums of three consecutive days. 5 also includes values for wet days. the 1:l line was drawn. Interception Returning to the water balance. no evidence was found that interception played a marked role in evapotranspiration for pasture. Fig.. it covers the period of possible moisture deficit under Dutch conditions. 3 on the other hand.1978. Several other quantitative formulations for the Bouchet hypothesis are possible. In all figures. mm). proved encouraging. In fact. which were inexplicable left out of fig. but climatological calibration is always indispensable. Sept.5 mm). of which one was dry (1976).. June. 2 and 3 under sufficiently moist conditions and on the different behaviour with limiting moisture conditions (section 4. Aug. 1 . July. Although the method can only be used for a relatively short period of the year. 4. 1 and 2 on the one hand and eq. Philips (1979) estimated the increase in evapotranspiration at 2 . applied to summerperiods of three years. ~ i g 7 E ~ ' against . 1977.3 mm a month (with a canopy storagecapacity of 0. 4.6. Period: Apr. the method was quantitatively based on the identical behaviour of eqs.Fig.
1960). Linear regression may yield a small deviation from the 1:1 line in fig.5 mm/month as an average value. dry and partially wet conditions. 8 against for wet days. However research in the last twenty years on interception of coniferous and hardwood forests has proved that interception is an important component of water loss from forests (e. Whether a forest lose more or less water than other vegetation is not easy to answer and depends highly on the amount.1975..2) and ET^^^ are drawn against ET^' for days with rainfall exceeding 0.l d o y ] E [rnrn Fig. 7 in favour of ET^'. Gash. He applied Ruttertype interception models (e.1 mm/d on average or about l mmlmonth (1. The estimates of (with n = 1. 1979).g. 7 and 8. intensity and distribution of rainfall. that interception would not markedly influence ET above grassland @cMillan and Burgy. on the type of forest and on the area covered.5 mm/d. 1979) to hourly data of Hupsel over a winter period. 2 for wet. Period: as in figure 7. 1972. The few. making full use of eq. the 1:l lines would not fit the data. Rutter. More evidence of the minor role of interception for pasture is given by figs. more sophisticated studies with lysimeters indicated. but by no more than 0.1977. rather a different value to compensate several deficits in table 3. If we take 2. If interception plays any role.g. interception over 34 months would increase ET by 85 mm. Pearce and Rowe.10 wet days). .
years 1976. .) Prec. (Hup.1978 ETHcO Eo(Wiit.ETHcO) ( ' .0 7 E. ) .) (P .EPEN)and (P .32 270 + 70 + 17 2 * Rain gauge: 2 dml at 40 un above ground level.0. Hupsel .1977.)  1976 1977 1978 187 430 359 187 429 366 461 346 348 487 341 345 653 514 488 274 + 83 + 18 300 88 21 336 + 19 . Year Prec.Table 4 Comparison of (P . Eo) (P .* (Wit.0 7 Eo) l08 .E ~ ~ of+ ~ )against (P .8 E and (P .) of Wintermijk. Period: March to September.EPEN) (P .
from about the second half of May until September for Hupsel. Rainfall excess: Period March to August In a report of CHOTNO (Van der Heide. Thus.. 1977. Rainfall excesses between Hupsel and Wmterswijk were compared for sixmonth period of 1976.8 X E.. eqs.) overestimated potential ET of grass. openwater evaporation. The calculation (0.0.8 X E. a factor of 0. for which this method is more sensitive than the other methods. with a factor 0.8 X E.XG) t 3.7.0.7 seems a better estimate of potential ET for Winterswijk. amounted to 653 mm in 1976 against an average of 544 mm. Rainfall excess in Hupsel was defined by (PET'~~) and PET~~") and rainfall excess at Winterswijk by (P . with factor 0.8 X E.) (mm). The results are given in table 4. especially during winter. CONCLUSIONS AND RECOMMENDATIONS . are given for several meteorological stations in The Netherlands. and by (P .7 X Eo) gave similar results to those from Hupsel. potential rainfall excess was calculated over a sixmonth period from 1 March. All methods were based on meteorological observations and not linked to soil moisture storage.. C and G Kramer. . 1 to 3 produce reliable estimates of real potential evapotranspiration. whereas (P . gave too high an estimate of potential ET for grass and a better estimate seems to be E.8 X E.4. 3 can only be used within a limited period. Potential ET of grass is often expressed by 0. The profileenergy budget method showed the highest deficit. but in 1976 (P .0. 12 and 13). =. For Witerswijk E. perhaps because of inaccuracies in the method itself or in available profile data.) decreased to 344 mm. for the year 1976 and the average E. 'T. Openwater evaporation.S sty y myer. E.8.).1977 and 1978. .7. The average for Witerswijk is 61 mm.7 X Eo). figs.0. 5.7t4..If there is no regional moisture deficit. whose wider application still needs to be tested.Ae sty X are mean values over daylight hours (see X is normally neglected and R. Rainfall excess was defmed by the equation: (P . as usual. 13 of the same report.The small deficits in the different water balances of table 3 indicate that the methods of calculating actual and potential evapotranspiration are satisfactory. or because of local variation in actual evapotranspiration. E.0. However eq. E. In fig. 1957).0ii.
The advectionaridity method then seems to provide a fair estimate of actual ET. Comm. The method is attractive. Research TNO.C. W. Verstraate and the Sections of Information Processing and of Data Systems Management. The advantage is that less data are needed. (1963) . signification climatique. Wat. In my view it makes the catchment very attractive for testing of operational hydrological models. Daily values (and for rainfall and discharge at much smaller intervals) of a period of about three years are available now and the period will be extended. 62. J.J. BRUTSAERT.A. J. Observations of 80 wells are also available and landuse is surveyed annually. 1 to 3 cannot be used to estimate actual ET.. As shown for pasture. Proceedings and Informations 28: 2537.28 in the Priestley & Taylorexpression may be somewhat too low. Arnhem.A. (1979) . Kole.M. such as numerical groundwater models and conceptual rainfallrunoff models.In hydrology.In the Hupselse Beek Experimental Catchment. Mr. However a value of a = 1. of App. because it needs routinely measured meteorological data.W. E. and he thanks Mrs. 12: 649657. J.J. without whose help the results would not have been obtained. REFERENCES BOUCHET. I. 15: 443450. Climatological calibration is indispensable. Met. Rigg (Pudoc) for revising the text. . But it could also be attractive for multidisciplinary research on water management and the environment.The determination of (reference crop) evapotranspiration from routine weather data. a Penmantype equation may correctly be replaced by the Priestley & Taylor equation in summer. STRICKER.N. He is also grateful to Mr.. Hydrol.. as earlier indicated. pp.vd. DAVIES. F..R. wide use is made of Penmantype equations and reduction functions based on moisture storage to estimate actual evapotranspiration. BRUIN. van Ernst for making the drawings and Mr. Res.Evapotranspiration re6lle et potentielle. H. eqs. for obtaining and processing the data.D. 6 . Molen and Miss M. (1973) . in particular Mr.Equilibrium. The Hague. ACKNOWLEDGEMENT The author would like to thank the Ministry of Public Works (RWS). But a more important reason to recommend replacement is that during drought the Priestley & Taylor formula produces a more realistic estimate of regional potential evapotranspiration than a Penmantype equation. and J. .An advectionaridityapproach to estimate actual evapotranspiration. DE (1981) . andC.potential and actual evaporation from cropped surfaces in Southern Ontario. 134142. a l l components of the water balance were determined independently.Proceedings I.H. Res. .A.If actual evapotranspiration drops below potential during drought. R.S. Kessel for typing the manuscript. ALLEN. Heijnekamp .
De invloed van de wind op regenwaarnemingen.M. PRIESTLEY. .14: 1620.F. (n Dutch) In: Rapporten en Nota's No. Perth. Derivation of the model from observations in a plantation of Corsican pine.Influence of vegetation on the hydrological cycle. Hydrology and Water Resources Symposium. Hydrol. J. for Water Management Gelderland. E. MUKAMMAL.J.H. 1 4 .Actual evapotranspiration over a summer period in the "HupselCatchment". verschenen over de droogte van 1976.M. KERSHAW. OLIVER. J. Research TNO.C..Roy. Geneva.J.The phenomenon of transpiration and its importance in the management of the Hydrologic Cycle. (1972) . A.Roy. ROBINS. The Hague.H.F. of Appl.J. JACKSON. ROBINS. New Zealand. (1978) .'S Gravenhage. 70.H. RUTTER. CLOTHIER. pp. Ecol.K. C. Sensitivity of the model to stand parameters and meteorological variables.On the assessment of surface heat flux and evaporation using largescale parameters. HEIJDE. B.A.De droogte van 1976. VAN DER (1977) ..J. and P. A.D.J. (1977) . MORTON. and A. (Ed.I. WORLD METEOROLOGICAL ORGANIZATION (1975) . Research TNO. A. Generalisation of the model and comparison with observations in some coniferous and hardwood stands. 14: 567588.id. Met. A. K. GREENWOOD. J. J.S. BURGY. and H. J. RUTTER. J. ROUSE. and W.Forest management effects on interception. Budapest. (Evaporation Symposium) McMILLAN. TNO. en Verh.Berekening van de gemiddelde grootte van de verdamping voor verschillende delen van Nederland volgens de methode van Penman. A. Ackroyd). P. Proceedings and Informations 4: 901 15.Evaporation from land surface. (1957) . Met.Operational Hydrology Report No. of Appl. D.Evaporation in high latitudes. MILLS.M.R. J.G.een vergelijkend regenmeteronderzoek. Soc. PHILIPS.L. Res.An analytical model of rainfall interception by forests. of Hydr.P.Final Report of the Technicai conference on Assessment of h e a l Evaporation. Quart.I.D. Proceedings and Informations.N. and J.H. P. Staatsdrukkerij en Uitgeverijbedrijf. 12: 367380.E. Ecol. pp. K.C. (1960) . Weather Rev. (1975) Soil heat flux determination: a nullalignment method.M. Boundary Layer Meteor. and R.On Penman's equation for estimating regional evaporation. 100: 8192. KEIJMAN. Agric. Water Res. P. J. KRAMER. 2.Q.Evapotranspiration from vegetations in relation to Penman's formula. 12: 243256.1974. Biennial reports no. (1960) . Hydrol. 133137. 105: 4355.J. Quart. Comm. The Hague. 65: 23892394. 103: 345357.id.J. and L. 1. Een samenvatting en overzicht van de literatuur.Theoretical background of some methods for the determination of evaporation. (1977) . (1979) . MORTON. B. H. Met.1973.A.J. MORTON. Murray and P. and R. Med. 3: pp. RUTI'ER. 28: 1224. STUDY GROUP "HUPSELSE BEEK" (1971.M.Hydrological Research in the catchment of the Hupselse Beek. KIMBALL. of Hydr. E. NEUMANN. Amhem. i 113. (1979) . evaporation and water yield. (1979) . J. ( Dutch) Report of m the Study Comm. (1979) . C. (1979) . Physical Hydrology. Comm. 39: 139157.Application of the Priestley & Taylor evaporation model to assess the influence of soil moisture on the evaporation from a large weighing lysimeter and class A pan. (1981) .GASH. W. (1971/1972) . 9: 367384. 3.A. BRUTSAERT. Met. THOM. 18: 7387... PEARCE. WARMERDAM.G. Comm. McNAUGHTON.A predictive model of rainfall interception in forests. 1. 97119.1976) . and H. and R. (1977) . Month. (in Dutch) K. Agr. Research. TAYLOR. and A. O. of Geoph. (1975) . MAKKINK. Soc. Hydrol. 7: Intercomparison of conceptual models used in operational hydrological forecasting. and P. 15: 19.N.K. KERR. RE. G.B.Interception loss from grass.R. ROWE. 13: 909914. W. Res.. J. (in Dutch with English summary). A. The Hague. WOLRD METEOROLOGICAL ORGANIZATION (1977) . (1977) .N. (1981) . STRICKER.
1960). STATE O F THE ART O F EVAPORATION 19581960 Many articles in the proceedings of the Evaporation symposium (CHOTNO. polderboards. The calculation of the evaporation from a free water surface according to Penman was simplified by using tables. hydrologists. (CHOTNO. graphs and nomograms.A. 1. We will start by looking to the state of evaporation research in 19581960 and try to figure out the developments in different subjects of evaporation during the last 20 years. 1960) concentrated on the computation method by Penman. TNO (CHOTNO. VELDS SUMMARY In this paper a review is given of the developments in evapotranspiration research in The Netherlands since the last Evaporation Symposium of the Committee for Hydrological Research. At last we will have a future outlook on the need of evapotranspiration data and on the way the research is expected to go. soil and vegetation. horticulturists. The problems in evapotranspiration research that still exist today are discussed and a future outlook is given on the need of evapotranspiration data and the way the research is expected to go. For all these users it is interesting to see 'if. the formula of ThornthwaiteHolzman has been used for . TNO. environmentalists. The emphasis has shifted from empirical methods (such as Penman's equation) t o more physically justified models and later on t o physiologically controlled processes (equations in which a canopy resistance is taken into account). INTRODUCTION The principal objective of evapotranspiration research is to find methods for calculating the loss of water under varying conditions of climate. 1960) some 20 years ago. Various groups need evapotranspiration data in their work. such as fluxes of heat and water vapour shall be taken into account. Besides the Penmanequation. Besides these the predictive numerical models for mediumrange weather forecasts (210 days) require that the effects of the planetary boundarylayer. for example agriculturists. drinkingwatersuppliers. 2. the evapotranspiration research has evolved since the last Evaporation symposium of the Committee for Hydrological Research. Doing so we will arrive at the state of art at this moment with the problems yet unsolved. and if so how.78 EVAPOTRANSPIRATION: ADVANCE AND FUTURE OUTLOOK C.
further experimental methods were the water balance. The use of lysimeters was still in full swing. He concluded that it is possible to calculate Ep according to Penman. 3. The aerodynamical method or profile method was developed further by De Bruin and Kohsiek (1979). such as the study of the stomatal resistance. the energy balance and the aerodynamical method (vertical transport of water vapour). The eddycorrelation technique was developed only very recently at that time. More detailed studies were wanted at that time on vegetation factors for the relation between the actual evapotranspiration and Ep or E. 3. Development in theory and formula 's Measurements of the energy budget at ground level were continued in the Rottegatspolder until 1969. The oldest combination formula. Research has been done on the uncertainty in the evaporation of a free water surface computed according to Penman's method and on the influence of a lake's depth on the validity of the Penman formula and the assumption that the soil heat flux G = 0.an atmosphere in neutral equilibrium. The Rottegatspolder acted as an experimental catchment area for evapotranspiration research. Results of the energy balance were compared with those obtained by the water balance and the aerodynamical method.1. As already stated by Keijman (1981) energy balance equations can be combined with transport equations leading to socalled combination formula's. and measured global radiation and Rijtema compared calculated values of the potential evaporation Ep after Penman. In 1975 extensive evaporation research started at the KNMIground in Cabauw with the energybalance method and eddycorrelation measurements. Makkink and Turc from meteorological data with the same accuracy as is obtainable with lysimeters and evaporation pans. The concept for potential evaporation . Makkink. Turc and Haude with measured evapotranspiration values from lysimeters and evaporation pans. The last years signaled a shift in evapotranspiration research from physically controlled processes to evapotranspiration as a physiologically controlled process. DEVELOPMENT DURING THE LAST 20 YEARS Briefly it can be said that the emphasis in evapotranspiration changed from the early simple empirical methods to methods which more closely represent the physical and biological processes.. Makkink reported about the influence of advective heat on the actual evapotranspiration from any vegetation. the model of Penman has been discussed in this symposium by De Bruin (1981) and Keijman (1981). Makkink pointed out the lineair relation between openwater evaporation E. Furthermore there was a need for cumulative frequency distributions of precipitation surplus in the growing season.
On account of this Rijtema derived an equation in which the factors determining the actual evapotranspiration are taken into account. the evapotranspiration has under many conditions no direct relation with the evaporation from a free water surface.Taylor formula was tested by De Bruin and Keijman (1979) on Lake Flevo.Monteith estimate. Instrumentation The search for a simple economical.Ep = f. It is there fore necessary to take all the factors Boverning the real evaporation into account". So the canopy resistance term. who combined the energybalance method with the profile method. found already in Monteith's model was expanded by Rijtema and later by Feddes to include terms for stomatal resistance.the rate of supply of water to the evaporating surface.the transport of water vapour from the air layers close to the evaporating surface to higher layers. Rijtema (1965) stated: "due to the development of the crop and t o a possible lack of water. E. 1978). and resistance dependent on the fraction of soil cover. under various conditions of atmospheric stability (Brutsaert and Chan. this requires additional research into the functional form of the similarity functions for sensible heat and bulk water vapour transfer. Van Wijk and De Wilde (1962) concluded that it was impossible to use futed pan coefficients for vegetation under different climatic and exposure conditions. reliable evaporation pan continued in the first years of the period. who found a diurnal and seasonal variation in the constant a (mean value 1. 3. . Rijtema's thesis gives an analysis of the most important factors determining the actual evapotranspiration. These are: . The usefullness of the Priestley .type model a substantial improvement upon the Penman .the amount of energy available for the vapourization of water. Recently the planetary boundarylayer theory has been applicated to evaporation models.the aperture of the stomata in connection with the diffusion of water vapour through them.26). Various other combinationformula's have been set forth such as Stricker . . Ziemer (1979) called the Rijtema .Brutsaert (1978). Each of these factors can act as alimiting factor for the evapotranspiration. is an approximation method with a strongly empirical character. The advectionaridity method of Brutsaert and Stricker (1979) has been applied to the Hupselse Beek catchment. . . The WMO Technical Conference p n Assessment of Areal Evaporation in Budapest (1977) discouraged the further development of evaporation pans. resistance dependent on the availability of soil moisture and on liquid flow in the plant. due to the empirical constant in the windterm and f.2.
Lake evaporation In 1967 evaporationmeasurements have been made at Lake Flevo where water balance. the aerodynamical profile equations. As the surface temperature of the leaves depends on the amount of water available to evaporation. later Hupselse Beek area. Furthermore it is possible to calculate actual evapotranspiration rates from the remotely sensed surface temperatures with the aid of the energybudget equation. Furthermore he showed that the evaporation of a lake can be estimated rather well from simple meteorological data of a station alee. Keijman and De Bruin (1979). Remote sensing measurements have been made from low laying platforms (at Cabauw. incorporating soilplantwater relations (Soer. The lysimeter research has contributed highly in the development of relations which can be applied in hydrological research.3. without determining T. . De Jong and Keijman (1971) treated the evaporation of an estuary where a periodical tide run has to be taken into account. improvements have been made in measuring the different components of evapotranspiration models such as neutron moistureprobes for soil moisture. from aeroplanes and from satellites (NOAATIROSN. Keijman (1974) published a method to calculate E from isothermal lakes.and Penman models have been compared. Great strides have been made in the development of remote sensing techniques for measuring reflectance and surface temperature. has to be known. The profile method. pointed to large discrepancies that can occur between the surface temperature measured by Infra Red Line Scanningtechniques and the theoretical temperature. 1980). the Bowen ratio and the water balance have been compared to assess the representativity of point measurements with routinely determined data for areal evapotranspiration. Tellusproject centered around EXPLORERAbearing sensors for Heat Capacity Mapping Missipn). for example at Castricum the interception on trees has been studied. remote sensing can be used to determine the water stress of vegetation. Verhagen (1977) published an article on nonisothermal lakes with a sudden temperature jump and Sweers (1976) connected the wind function in Penman's formula to the area of the lake. \ 3.. types of soil and climates. energy balance. however. aerodynamical. Besides all this. with the purpose of studying the effects of reclamation works and the testing of instruments. For the evaporation from lakes the watersurface temperature T. Barnes and Heijmann).The research with lysimeters yielded a large number of data for setting up water balances for various vegetations. fastresponse sensors for wind. In 1969 research started in the Sleen catchment on mechanisms determining the relation between the components of the water balance and research on the relation between evapotranspiration and soil moisture availability. Experiments in catchment areas have been done in TielerwaardWest and in the Leerinkbeek area. temperature and water vapour and of course the enormous developments in registration and data collection by the advancement of the computer.
has to be considered. In the last twenty years understanding of the factors on which stomatal resistance depends has increased.6. According to Rijtema the aerodynamical resistance in the PenmanMonteith's equation can be related to the inverse of a roughness function. 3. Evapotranspiration o f vegetation In this case the soilplantatmosphere system. Van Bavel and Hillel (1976) stated that the transition from the energylimiting phase to the soillimiting phase is not due to changes in albedo. Soil water evaporation The complexity of the evaporation of water from the soil has been found from laboratory and lysimeter experiments. (1976) proposed a model that takes into account a root extraction term. In the Hupselse Beek catchment it has been found that interception did not influence evapotranspiration in a noticeable way for grassland (Stricker. . Precipitation and irrigation have direct and indirect effects on soil temperature and hence on evaporation. As the soil dries energy availability becomes less important and the rate of soil water conduction becomes more important.3 A.S. on grassland (Cabauw) and wheat (Flevoland). Rijtema and Ryhiner (1968) found that interception gives a big increase in evapotranspiration of coniferous forests. soil moisture content and the depth of the roots. but to the hydraulic properties of the soil and to a reduction of the relative humidity at the surface to values less than one. Feddes et al. depending on potential evapotranspiration. 3. They proposed an extension of the Penman equation by incorporating terms related to the hydraulic and thermal properties of the soil profile. of which no one part operates independently of the other. some species close their stomata. describing the terms of the energy balance of a vegetable surface. This function is the product of vegetation height and a dimensionless function of wind speed. Stomatal resistance has been measured by Stigter (1975) who used a closed diffusion porometer as the most suitable device for field measurements on separate leaves and by Kohsiek (1979) who used a perspex chamber in open or closed condition with circulating air to measure the bulk stomatal resistance of a grass surface. Soer (1977) tested the TERGRAmodel. 1981). The research concentrated on 4 lysimeters at Castricum of which one has a foliagetree vegetation and one a coniferous vegetation. When evaporative demand exceeds the ability of the roots to supply the necessary water. The role of stomata in the regulation of transpiration has been widely studied. Interception Relatively little research has been done on the evaporation of forests and interception in The Netherlands.
In many wildland areas. Ziemer (1979) put it this way: "We have advanced a great deal in understanding the physical and biological controls on evapotranspiration. Another point which is still under investigation is the relation between actual and potential evaporation. There are still problems in incorporating evapotranspiration into integrated models for watermanagement: It is impossible to verify the evapotranspiration calculated with an integrated groundwater current model with the measured evapotranspiration for a whole catchment area and for short periods (10 days). With this phenomenon we arrive at one of the problems in evapotranspiration research which still exists today. This has two causes: . one of the problems still existing is the influence of regional and local advection upon calculated evapotranspiration. The ability to apply models to field situations is less succesful. we are unable to measure adequately even areal precipitation let alone the rather detailed meteorological data required to calculate evapotranspiration with the PenmanMonteith equations. Advection The influence of advection of sensible heat from a relatively dry area to a more moist area is a problem which has long plagued attempts to evaluate evapotranspiration. The uncertainty in the evaporation of a free water surface and the evapotranspiration of a wet vegetation calculated with the Penmanequation is about 20 percent. changes in species composition.3. Our ability to calculate accurately evapotranspiration within that cover condiiion between bare soil and full cover is still weak. The energy used for evapotranspiration can exceed the energy supplied by net radiation by a factor of two. . particularly as to areal water loss from scattered vegetation of different species and sizes". as can be seen from tables l and 2.7. Under advective conditions the Bowenratio underestimates evapotranspiration. particularly in forested areas and in other areas where data are lacking. We are still unable to predict the effect of timber cutting. For the lysimeters at Castricum negative advection with moist sea breezes can reduce the evapotranspiration.the methods which are suited for estimating areal evapotranspiration (remote sensing) are not reliable enough.the methods measuring evapotranspiration accurately to this only for point measurements. wildfire. which have been taken from De Bruin and Kohsiek (1979). for what period and how reliable? What is the value of such a balance for the evapotranspiration research? . Is it possible with the collected data to form a water balance. or other cultural activities on watershed water balances. The same problem arose from the Hupselse Beek project. 4. PROBLEMS WITH EVAPORATION RESARCH TODAY As stated in the last sentence of the preceding paragraph. Application on a bigger scale (more measuring points) is too costly.
Error analysis for the evapotranspiration of a wet vegetation. It is to be expected that in the future remote sensing techniques will be more reliable than nowadays. A drawback of using satellite data is. however.Table I . So the diurnal variation in evapotranspiration can only be estimated by mathematical models that are verified with satellite data. Error analysisof the Penmanequation applied to a free water surface (10day or monthly totals). that polar orbiting satellites pass only once a day. FUTURE OUTLOOK The data mostly wanted are reliable areal evapotranspiration data. Moreover remote . It is still questionable whether remote sensing techniques from satellites will give an exact answer to this. Error source neglecting the soil heat flux G wrong estimation of net radiation Q* wrong wind function random errors from input data and reading errors Systematid error +20% spring: autumn: 20% year: 0% summer: *l@&*) autumn: +20% +20% year: year: 5% Random error 10% summer: 10% winter: 20%  *) If Q* less than 100 Wml then +10% if Q* greater than 140 Wm * then 1 0% Table 2 . Error source neglecting the soil heat flux G wrong estimation of net radiation Q* wrong wind function vapour pressure deficit ese Total Systematid error summer: year: 3% 0% Random error 5% 10% 10% summer: 10% autumn: 10% spring: 10% ? 1013% +? 2025% 5.
sensing nowadays gives only relative surface temperatures, so that on the other hand satellite data have to be calibrated with additional measurements on the spot. It can be expected that in the future the research on catchment areas will go on, to improve our knowledge of evapotranspiration as a function of atmospherical and soil physical circumstances. The WMO should promote the conduct of a number of largescale intensivelyinstrumented experiments, such as the present GARP landphase experiment in France, in which large heterogeneous areas are broken down into smaller separate homogeneous subareas, in each of which evaporation is measured by suitably accurate techniques. Some improvements can be expected in measuring meteorological data for evapotranspiration calculation, in datahandling and data archiving; humidity measurements by means of ultra violet Lymancutechniques; wind measurements with sonic or thrust anemometers and temperature measurements with fast responding thermistors, thermocouples or Ptwire in behalf of the eddycorrelation method. Movable masts for eddycorrelation measurements with microprocessors will help to extend point measurements to areal means. At last the temperature fluctuation method that has been developed recently can be mentioned (De Bruin, 1981). The accuracy of the radiosonde data in the lowest 2 km is expected to improve which can be a stimulus for the regional evapotranspiration model of Brutsaert and Mawdsley (1976) which uses rawinsonde data. Some research will be done on the reliability and comparability of lysimeters. For the routinely determination of evaporation the WMO Technical Conference on Assessment of Areal Evaporation (1977) recommended informally net radiation networks using Funktype instruments. Whereever possible these should be supplemented by hourly measurements of solar radiation, air temperature, humidity and wind. At key stations additional direct measurements of vertical profiles of air temperature and humidity should be taken for energy balance or Bowenratio models. It is possible that in the future more measurements will be made for calculating evaporation according to the Penman or profile method. Maybe that more measurements il will be made of the vertical profile of soil moisture and tension and more work wl be devoted to interception on arable crops. Due to the computer, integrated mathematicalhydrological models are expected to be developed which can take into account the influence of human activities and changes in land use on the evapotranspiration. Furthermore the computer has the advantage that lots of data can be handled and computations can be done on a realtime basis. This means that the user may have the disposal of evaporation data at short notice, in every kind of output he wants (total values, mean values, frequency distributions). In the future the evaporation data might be combined with precipitation data so that realtime groundwater levels can be calculated, dependent on the type of soil. Although evapotranspiration is essentially a physical process it is not to be expected
that in the future models will be developed that can take into account all atmospheric, crop and soil properties influencing the process. As the complexity of the models increases the data requirements to drive the equations often make the model useless for routine applications. This means that also in the future the calculation of evapotranspiration will be a compromise between empiri&d and totally physically justified models.
REFERENCES BAVEL, C.H.M. VAN and D.I. HILLEL (1976)  Calculating potential and actual evaporation from a bare soil surface by simulation of concurrent flow water and heat. Agt. Met. 17: 453476. BRUIN, H.A.R. DE (1981)  The determination of (reference crop) evapotranspiration from routine weather data. Comm. Hydrol. Research, TNO, The Hague, Proceedings and Informations 28: 2537. BRUIN, H.A.R. DE and J.Q. KEIJMAN (1979)  The PriestleyTaylor evaporation model applied to a large, shallow lake in The Netherlands. J. Appl. Met. 18: 898903. BRUIN, H.A.R. DE and W. KOHSIEK (1979)  Applications of the Penman formula. (in Dutch) Scientific Report W.R. 793; KNMI, De Bilt, pp. 36. BRUTSAERT, W. and J.A. MAWDSLEY (1976)  The applicability of planetary boundary layer theory to calculate regional evapotranspiration. Wat. Res. Res. 12: 852858. BRUTSAERT, W. and F.K. CHAN (1978)  Similarity functions D for water vapour in the unstable atmospheric boundary layer. BoundaryLayer Meteor. 14: 441456. BRUTSAERT, W. and J.N.M. STRICKER (1979)  An advectionaridity approach to estimate actual regional evapotranspiration. Wat. Res. Rep IS: 443450. COMMITTEE FOR HYDROLOGICAL RESEARCH TNO (1960)  Evaporation symposium and Report on the lysimeters in The Netherlands (11). (in Dutch with summaries in English) Proceedings and Informations 4: pp. 263. The Hague. FEDDES, R.A., P. KOWALIK, K. KOLINSKAMALINKA and H. ZARADNY (1976)  Simulation of field water uptake by plants using a soil water dependent root extraction function. J. of Hydr. 31: 1326. JONG, H. DE and J.Q. KEIJMAN (1971)  Changes in water temperature in the Haringvliet as a consequence of the closing by dikes. (in Dutch) Scientific Report W.R. 714; KNMI, De Biit, pp. 22. KEIJMAN, J.Q. (1974)  The estimation of the energy balance of a lake from simple weather data. Boundary Layer Meteor. 7: 399407. KEIJMAN, J.Q. (1981)  Theoretical background of some methods for the determination of evaporation. Comm. Hydrol. Research, TNO, The Hague, Proceedings and Informations 28: 1224. KEIJMAN, J.Q. and H.A.R. DE BRUIN (1979)  A comparison of measured and calculated temperatures of a grasscovered surface (abstract only). EOS Transactions, American Geophysical Union 60: 583. KOHSIEK, W. (1979)  A stomata1 resistance meter. (in Dutch) Scientific Report W.R. 7910; KNMI, De Bilt, 55 pp. RIJTEMA, P.E. (1965)  An analysis of actual evapotranspiration. Agricultural Research Reports No. 659, Pudoc, Wageningen. RIJTEMA, P.E. and A.H. RYHINER (1968)  Lysimeter investigations in The Netherlands (111). ( m Dutch) In: Precipitation data (11) Proceeding of Technical Meeting 22 and Report on the lysimeters in the Netherlands (111). Proceedings and Informations Comm. Hydrol. Research TNO, The Hague 14: 86149. SOER, G.J.R. (1977)  The Tergra model, a mathematical model for the simulation of the daily behaviour of crop surface temperature and actual evapotranspiration. NIWARS publ. 46, Delft. SOER, G.J.R. (1980)  Estimation of regional evapotranspiration and soil moisture conditions using remotely sensed crop surface temperatures. Remote Sensing of Environment 9: 2745.

STIGTER, C.J. (1975)  Watervapour diffusion porometry for leaf epidermal resistance measurements in the field. In: D.A. de Vries and N.H. Afgan; Heat and mass transfer in the biosphere. I Transfer processes in plant environment. John Wiley & Sons, London; pp. 403412. STRICKER, J.N.M. (1981)  Methods for determinating evapotranspiration from meteorological data and the applicability in hydrology. Comm. Hydrol. Research, TNO, The Hague: Proceedings and Informations 28: 5977. STRICKER, J.N.M. and W. BRUTSAERT (1978)  Actual evapotranspirBtion over a summer period in the "Hupsel catchment". J. of Hydr. 39: 139157. SWEERS, H.E. (1976)  A nomogram to estimate the heatexchange coefficient of the airwater interface as a function of wind speed and temperature: a critical survey of some literature. J. of Hydr. 30: 375401. VERHAGEN, J.H.G. (1977)  Een model voor thermische stratificatie bij variabele weersomstandigheden. (n Dutch) H,0,10: 207210. i WLTK, W.R. VAN and J. DE WILDE (1962)  Microclimate. Arid Zone Research 18: 831 13. ZIEMER, R.R. (1979)  Evaporation and transpiration. Reviews of Geophysics and Space Physics 17:
11751186.
.
Proceedings and Infomzations No. Proceedings of Technical Meetings 1314 (with summaries in English). No. 13. 1960.89 COMMITTEE FOR HYDROLOGICAL RESEARCH T. The Water Supply of Sandy Soils 12. No. Proceedings of Technical Meetings 15 (with summaries in English). 1958. Observations of Groundwater Levels 4. 4. 1959. .O.N. 7. Water in unsaturated Soil Proceeding of Technical Meeting 15 (with summaries in English and French). The Study of Precipitation Data 8. The Water Supply for Crops I1 6. 6. Quality Requirements for Surface Waters Evaporation Synposium and Report on the Lysimeters in the Netherlands I1 (with summaries in English). The Problem of the increasing Salinity of Ground and Surface Water in the Netherlands Proceedings of Technical Meeting 710 and Report on the evaporation research in the Rottegatspolder 19471952 (with summaries in English). Groundwater Levels and Groundwater Movement in the Sandy Areas of the Netherlands 14. 5. Investigations by Drain gauges in the Netherlands 5. 1961. 1955. 11. Measurements and improvement Works in Basin of Brooks 10. Model Research on Groundwater Flows 9. The Water Supply for Crops I 3. Geoelectrical Research Proceedings of Technical Meetings 1112 (with summaries in English) and Report on the Lysimeters in the Netherlands (in English). The Regime of the Rhine. 2. No. 1952. 1. the Ysselmeer and Zeeland Lake No. 1. Investigations into the Water Balance of the Rottegatspolder 2. No. 3.
1963. 1970. 1969. 16. Precipitation Data (11) and Report on the Lysimeters in the Netherlands (111) (both with summaries in English) Proceedings and Informations no. No. Hydrological Investigations for Masterplan for the future Watersupply in the Netherlands Proceeding of Technical Meeting 25 (in English). No. Recent Trends in Hydrograph Synthesis Proceeding of Technical Meeting 22. 11. compiled by the Hydrologisch Colloquium ( English). No. Flow in the unsaturated Zone (Will not be published). No. Proceeding of Technical Meeting 24. 1963. m Proceeding of Technical Meeting 19 (with summaries in French and German). 1968. 1968. 1964. 1975. 1964. 1966. 1962. 17. The Laws of Groundwater flow and their Application in Practice Proceeding of Technical Meeting 18 (with summaries in English). 15. Proceeding of Technical Meeting 29 (with summaries in English). Water Balance Studies Proceeding of Technical Meeting 21 (in English). Soil . Proceeding of Technical Meeting 16 (with summaries in English). 12. Automatic Processing of Hydrological Data No. 8. 15 (in English).Water . 1966. 9. No. No. Water Nuisance Steady Flow of Ground Water towards Wells. 7. 14. Geohydrological Cartography Proceeding of Technical Meeting 20 (in English). 13.Plant Proceeding of Technical Meeting 23. The dry Year 1959 Proceeding of Technical Meeting 17 (with summaries in English). . No. No. Seepage will not be published). 1973.No. 10. No.
Precipitation and Measurements of Precipitation Proceeding of Technical Meeting 34 (in English). 24. Urbanization and Water Management Proceeding of Technical Meeting 35 (in English). 26. 1979. 1980. 23. 19. 1974. No. No. 1978. No. 1976. 27. 20. Salt distribution in Estuaries Proceeding of Technical Meeting 31 (in English). Research on possible Changes in the Distribution of saline Seepage in The Netherlands Proceeding of Technical Meeting 37 (in English). 1976. 1981. Water Resources Management on a Regional Scale Proceeding of Technical Meeting 38 (in English). Groundwater Pollution Proceeding of Technical Meeting 32 (with summaries in English). Water Quality of Dutch Rivers with respect to Water Management (Will not be published). 22. No. 1974. 25. 18. . Systems Approach to the Management of Water Resources Proceeding of Technical Meeting 33 (in English). Proceeding of Technical Meeting 26 ( English). N0:21. 1977. No. The Relation between Water Quantity and Water Quality in Studies of Surface Waters Proceeding of Technical Meeting 36 (in English). Proceeding of Technical Meeting 30 (in English). m Hydraulic Research for Water Management Proceeding of Technical Meeting 27 (in English). 28. No. Evaporation in Relation to Hydrology No. 1976.No. 1981. The Hydrological Investigation Programme in Salland (The Netherlands) Proceeding of Technical Meeting 28. No. 1973. No.