316
A.MJ. Meijerink
detenninations are available it may be worthwhile to try to relate them to the
mapping result. Mixtures of young and old water can be expected in such areas,
only old water in the impermeable formations. However, the interpretation of the
isotopes has to consider flow systems.
The spatial recharge, based on qualitative reasoning using an existing geological
map and a spatial rainfall map, resulting from merging the sparse gauge data with
vegetation patterns derived by image interpretation and by physiographic
interpretation of imagery, has been described by Karanga et al.(1990).
Vegetation and recharge. Since vegetation can be recognized and mapped
rapidly with reasonable accuracy by remote sensing, it may be useful to discuss
briefly the effect of vegetation on the recharge evaluation. Theoretically, there are
arguments pro- and contra to the hypothesis that vegetation increases the recharge.
The arguments in favour are related to the higher infiltration rates under vegetation
because of a favourable micro-climate near the soil surface, lower soil
evaporation, increased organic matter contents and no sealing of the surface. In
addition, decayed roots provide preferred pathways for the percolation flow. The
arguments against it are the transpiration rates whereby infiltrated water may be
fully used as well as the possible presence of deep roots taking up water from the
phreatic level. Rosenzweig (1972) concluded, that for a limestone area (672 mm
annual rainfall) dense thickets of natural forest evapotranspiration rates consumed
all the precipitation, while areas with annual grass needed 280 mm and in this case
the remainder was chiefly recharge. Finch (1990) reasoned, for an area in
Botswana, that if dense vegetation is noted on NDVl's, there must be, at least
temporarily soil moisture and there is a probability that a part of that reaches the
groundwater surface. The effect of distribution of rainfall intensities and duration
in time confound the situation.
Much depends also on the permeability of the upper zone. The Australian data
for sandy environments (see Lerner et al.,1990) suggest that vegetation increases
recharge, probably in the order of grass, broadleaf and pine, as long as the rainfall
is not too low, say < 200 mm.
There is little comparative data for non-sandy environments. It is interesting to
note that the same order is mentioned for the transpiration losses from the many
paired experimental catchments, reviewed by Bosch and Hewlett (1982), although
chiefly pertaining to non-sandy catchments without differentiation in direct runoff
and baseflow. The higher water yields under grass and shrub as a result of lower
transpiration may favour recharge, other factors being equal.
The transpiration (4) estimates appear in the simple equations for recharge (r)
proposed by Issar et al.(1985) for areas like the coastal dunes ofIsrael, in the form
of;
r = c (p - 4) , where the empirical coefficient (c) is related to direct
evaporation and rainfall characteristics (0.4 in their study), and p is the mean
annual precipitation and 4 the transpiration.
Semi-quantitative approaches. The "overburden" units, obtained by
physiographic or geomorphological interpretation of imagery and land cover
(vegetation) categories, obtained by supervised classification of multispectral
A.MJ. Meijerink
detenninations are available it may be worthwhile to try to relate them to the
mapping result. Mixtures of young and old water can be expected in such areas,
only old water in the impermeable formations. However, the interpretation of the
isotopes has to consider flow systems.
The spatial recharge, based on qualitative reasoning using an existing geological
map and a spatial rainfall map, resulting from merging the sparse gauge data with
vegetation patterns derived by image interpretation and by physiographic
interpretation of imagery, has been described by Karanga et al.(1990).
Vegetation and recharge. Since vegetation can be recognized and mapped
rapidly with reasonable accuracy by remote sensing, it may be useful to discuss
briefly the effect of vegetation on the recharge evaluation. Theoretically, there are
arguments pro- and contra to the hypothesis that vegetation increases the recharge.
The arguments in favour are related to the higher infiltration rates under vegetation
because of a favourable micro-climate near the soil surface, lower soil
evaporation, increased organic matter contents and no sealing of the surface. In
addition, decayed roots provide preferred pathways for the percolation flow. The
arguments against it are the transpiration rates whereby infiltrated water may be
fully used as well as the possible presence of deep roots taking up water from the
phreatic level. Rosenzweig (1972) concluded, that for a limestone area (672 mm
annual rainfall) dense thickets of natural forest evapotranspiration rates consumed
all the precipitation, while areas with annual grass needed 280 mm and in this case
the remainder was chiefly recharge. Finch (1990) reasoned, for an area in
Botswana, that if dense vegetation is noted on NDVl's, there must be, at least
temporarily soil moisture and there is a probability that a part of that reaches the
groundwater surface. The effect of distribution of rainfall intensities and duration
in time confound the situation.
Much depends also on the permeability of the upper zone. The Australian data
for sandy environments (see Lerner et al.,1990) suggest that vegetation increases
recharge, probably in the order of grass, broadleaf and pine, as long as the rainfall
is not too low, say < 200 mm.
There is little comparative data for non-sandy environments. It is interesting to
note that the same order is mentioned for the transpiration losses from the many
paired experimental catchments, reviewed by Bosch and Hewlett (1982), although
chiefly pertaining to non-sandy catchments without differentiation in direct runoff
and baseflow. The higher water yields under grass and shrub as a result of lower
transpiration may favour recharge, other factors being equal.
The transpiration (4) estimates appear in the simple equations for recharge (r)
proposed by Issar et al.(1985) for areas like the coastal dunes ofIsrael, in the form
of;
r = c (p - 4) , where the empirical coefficient (c) is related to direct
evaporation and rainfall characteristics (0.4 in their study), and p is the mean
annual precipitation and 4 the transpiration.
Semi-quantitative approaches. The "overburden" units, obtained by
physiographic or geomorphological interpretation of imagery and land cover
(vegetation) categories, obtained by supervised classification of multispectral