318
A.M.J. Meijerink
textures and land cover types. The evaporation loss can not be more than the
recharge in the intake area.
Quantitative approaches for the recharge. The degree of sophistication of the
quantitative methods for the spatial recharge used so far, is not necessarily
synonymous with accuracy. Important sources of inaccuracy are the spatial
variation, horizontally and vertically, of the hydraulic parameters of the
unsaturated zone and the values of the root parameters. In addition there is the
difficulty of determining the potential evapotranspiration, because the data from
evaporation pans seldom agree well with those calculated, for example by the
Penman-Montheith mode1.
By calibration of groundwater models. In hydrologic practice the spatial recharge
(i.e. the flux across the groundwater table) is often estimated as a result from
calibration of a groundwater mode1. It is useful to compare the pattern of the flux,
after calibration of the simulated heads with the observed ones and with the
relative recharge values based on image interpretations. Discrepancies are likely to
occur and the problem then arises whether the differences are due to wrong
estimates of the hydraulic model parameters and the lateral boundary conditions,
in particular when specified heads or fluxes are used, or to misinterpretation of the
imagery. Given that the calibration may be non-unique, there is much to say for
retaining the geographic pattern of the recharge.
One-dimensional unsaturated flow models. These models simulate vertical fluxes
of water, with given rainfall and potential evapotranspiration, asa function of soil
textures and uptake of water by roots of vegetation. In principle, the models can be
used to transform vegetation classifications in recharge quantities. However,
information for root functions exist mainly for crops and not for natural vegetation.
Interpolation of data from point observations for the regionalization usually does
not describe well the spatial variations, as is evident from the example of Fig. 14.7
(after Vekerdy, 1996).
The figure shows a small part of an alluvial region in Hungary. The drill holes
are marked on the photo. The lighter tones correspond to sandy textures, the dark
ones to heavier textured soil, but these can only be seen on bare fields.
It would be a step forward if by remote sensing more information on hydraulic
properties of surface and subsurface conditions could be obtained. The airborne
gamma ray spectrometry has shown application to surface soil mapping (Reeves,
1992), and also microwave sensors could be helpful in the regionalization which
still relies on much field data.
Energy balances. Considering the difficulties, the recommendation of Schultz
(1988) to base hydrologic simulations on remotely sensed input could well apply
to the evaluation of recharge. The spatial recharge could be estimated from the soil
water balance (eq.l), whereby evaporative losses are determined by energy
A.M.J. Meijerink
textures and land cover types. The evaporation loss can not be more than the
recharge in the intake area.
Quantitative approaches for the recharge. The degree of sophistication of the
quantitative methods for the spatial recharge used so far, is not necessarily
synonymous with accuracy. Important sources of inaccuracy are the spatial
variation, horizontally and vertically, of the hydraulic parameters of the
unsaturated zone and the values of the root parameters. In addition there is the
difficulty of determining the potential evapotranspiration, because the data from
evaporation pans seldom agree well with those calculated, for example by the
Penman-Montheith mode1.
By calibration of groundwater models. In hydrologic practice the spatial recharge
(i.e. the flux across the groundwater table) is often estimated as a result from
calibration of a groundwater mode1. It is useful to compare the pattern of the flux,
after calibration of the simulated heads with the observed ones and with the
relative recharge values based on image interpretations. Discrepancies are likely to
occur and the problem then arises whether the differences are due to wrong
estimates of the hydraulic model parameters and the lateral boundary conditions,
in particular when specified heads or fluxes are used, or to misinterpretation of the
imagery. Given that the calibration may be non-unique, there is much to say for
retaining the geographic pattern of the recharge.
One-dimensional unsaturated flow models. These models simulate vertical fluxes
of water, with given rainfall and potential evapotranspiration, asa function of soil
textures and uptake of water by roots of vegetation. In principle, the models can be
used to transform vegetation classifications in recharge quantities. However,
information for root functions exist mainly for crops and not for natural vegetation.
Interpolation of data from point observations for the regionalization usually does
not describe well the spatial variations, as is evident from the example of Fig. 14.7
(after Vekerdy, 1996).
The figure shows a small part of an alluvial region in Hungary. The drill holes
are marked on the photo. The lighter tones correspond to sandy textures, the dark
ones to heavier textured soil, but these can only be seen on bare fields.
It would be a step forward if by remote sensing more information on hydraulic
properties of surface and subsurface conditions could be obtained. The airborne
gamma ray spectrometry has shown application to surface soil mapping (Reeves,
1992), and also microwave sensors could be helpful in the regionalization which
still relies on much field data.
Energy balances. Considering the difficulties, the recommendation of Schultz
(1988) to base hydrologic simulations on remotely sensed input could well apply
to the evaluation of recharge. The spatial recharge could be estimated from the soil
water balance (eq.l), whereby evaporative losses are determined by energy