Geoscience Reference
In-Depth Information
0.9
0.8
0.7
0.6
Coastal
All
Parallel to coast
Normal to coast
Regional
Network
0.5
0.4
0.3
Figure 12.5 Distance Decay
curve for the correlation
coefficient, r , between gauges
in a network installed in a
rectangular array near the
Tanzanian coast with distance
D from the coast. (From
Sumner, 1983, published with
permission.)
0.2
0.1
0.0 0
10
20
30 40
Separation of gauges (km)
50
60
70
80
90
100
basis for deciding gauge separation. For example, Fig. 12.5 shows the distance
decay curve for correlation coefficients of daily rainfall measured using a gauge
network deployed on a rectangular grid near the coast of Tanzania. Average curves
are drawn for correlation coefficient as a function of distance for gauges lying
perpendicular and parallel to the coast and for all the gauges in the network. Note
the different rates of decay in correlation coefficient. If some value of correlation
coefficient is selected as being acceptable for specifying the separation of gauges,
the required average separation of gauges corresponding to this value can be
deduced. In the case shown in Fig. 12.5, the separation is different for gauges
arranged parallel and perpendicular to the coast.
Inter-gauge correlations have been used to make recommendations on
idealĀ  gauge densities. The (albeit rather crude) guidelines from the World
Meteorological Organization are given in Table 12.2. A second approach used
to define the recommended minimum gauge densities is to consider standard
errors for different network areas and densities. Table 12.3 shows an example
of this approach used to define the number of randomly positioned gauges
needed to give an 'adequate' measure of monthly average precipitation for
different areas.
 
Search WWH ::




Custom Search