Geoscience Reference
In-Depth Information
great attention. Not least this is because area-average precipitation is a major influence
on human settlement and development. Understanding the spatial organization of
precipitation is also crucial to hydrologists and civil engineers concerned with the
design of infrastructure. On the one hand, long-term area-average precipitation rates
determine available surface water resources while, on the other hand, short-term
area-average rates are needed when designing flood protection systems.
Mapping precipitation
Mapping precipitation is a straightforward and visually convenient way to illustrate
the organization of precipitation in space. However, precipitation is a meteorological
variable that is difficult to map, especially from gauge data which is the most
common source of the information used, because the point samples available may
be unrepresentative. Because precipitation is so spatially variable, there may be
extreme values which are missed, especially if the gauge network used is sparse
and uneven. There is also likelihood that a network of gauges will mis-sample local
systematic variations in precipitation associated with topography. In principle,
measurements made with radar would not be subject to the same shortcomings
but, as discussed in Chapter 12, the errors associated with radar measurement are
so large that such data only become reliable when re-normalized to an underlying
gauge network. Hence, the problems associated with gauge sampling remain.
Consequently, gauge data remain the basis of precipitation mapping, and for many
parts of the world are likely to remain so for some considerable time.
To use gauges for mapping it is necessary to assume precipitation is a spatially
continuous variable with no dramatic discontinuities, and to assume that it is
adequately sampled by the gauge network used. Correlation tests between gauge
measurements might provide some level of reassurance that a comparatively
smooth field with point to point correlation is being adequately sampled. In
principle the mapping process is then simple. It involves drawing
isohyets
, i.e.,
lines of constant precipitation, between the point precipitation data available.
Figure 14.1 provides a simple example of the process used.
The accuracy with which isohyets can be drawn will depend on the density of
the gauge network because spurious high or low values can easily dominate the
resulting map if the gauge network is sparse. The assumptions used when drawing
smoothed isohyets also influences accuracy and, if isohyets are drawn manually,
may be subjective. The presence of topography in the area mapped can also affect
accuracy because its influence on precipitation rate is rarely properly sampled and
will therefore be poorly acknowledged in isohyetal maps. However, with knowl-
edge of the underlying topography, a skilled operator or computer program can
make some recognition of the tendency toward higher precipitation at higher
locations when specifying smooth contours of precipitation. In areas with very
strong relief the influence on isohyets of making such allowances for height might
even match the influence of the available gauge data.
