Geoscience Reference
In-Depth Information
raingauges become more important but they are more expensive to operate and much fewer in number.
Thus it may still be necessary to use daily gauges to get an estimate of the total volume of rainfall over
a catchment, using the nearest recording raingauge to give an approximate idea of the distribution of
rainfall in time, the
storm profile
.
Even on small catchments, daily rainfalls may be adequate for obtaining acceptable predictions of
runoff volumes (rather than hydrograph peaks), especially when volumes over longer times, such as
monthly steps, are required. This is implied by the successful simulation of
soil moisture deficits
shown
in Figure 2.9, since discharge will complete the water balance for these sites. Hydrograph prediction,
however, is particularly difficult when a storm spans two daily measurements, since the fixed daily
measurement period (often 9 a.m. to 9 a.m.) is hydrologically arbitrary.
Raingauge-measured volumes may be subject to error. In particular, they depend on the design of the
raingauge in relation to wind conditions at the site and rainfall intensities. The best design is thought to
be a raingauge with the orifice set at ground level and surrounded by an anti-splash grid but this is not
always practical, particularly in environments with frequent snow. A variety of designs of wind shield
have been used in different countries to try to mitigate this wind effect. The wind effect can be large;
estimates of reductions of up to 20% have been reported at windy sites for gauges only 30 cm above
the ground compared to ground-level gauges (Rodda and Smith, 1986). High rainfall intensities can also
cause problems for some types of recording raingauge, such as the tipping bucket; if the tips start to occur
too rapidly the buckets start to bounce, so that high intensities may require a specific calibration.
Rainfall volumes and intensities can vary rapidly in space and time, particularly in convective rainfall
events (see Figure 3.1). Thus, as well as an interpolation of rainfall volumes in time to produce the storm
profile, it may also be necessary to interpolate in space since raingauge measurements represent only
Figure 3.1
Variations in rainfall in space and time for the storm of 27 June 1995 over the Rapidan catchment,
Virginia (after Smith et al., 1996, with kind permission of the American Geophysical Union).
