Geoscience Reference
In-Depth Information
This contrast in rainfall levels can also be applied to the monthly averages. The
statistical distribution of monthly rainfall is not uniform and cannot be represented
by Gaussian theory. Rainfall for a particular month and for a particular year can
only be modeled by a random non-Gaussian function, which is both anisotropic and
non-stationary. The example of Algeria focuses on the difficulties that are
associated with the automatic mapping of the precipitation fields.
7.2.3.1.
Identification of the random variable and the normalization of data
Research has shown that irrespective of the season and of the location, the
statistical distributions of monthly rainfall produce a positive skew. However, for
the rainy months (see Figure 7.6) the distributions were seen as being normal square
roots.
Figure 7.6.
Examples of total monthly rainfall distribution for the rainy months
Figure 7.7.
The relationship between the square roots of monthly rainfall averages
and the standard deviation associated with the square roots of these averages
The distribution of monthly rainfall is, therefore, expressed as square roots and
is made up of the following parameters:
