Geoscience Reference
In-Depth Information
blow in this area (Figure 8.9). By studying the map in closer detail, it is possible to
see the two methods that were used to recreate the different variables:
- for example, the rounded marks that can be seen in Figure 8.9 are caused by
spatial autocorrelation. These marks highlight the influence that the wind has on
some of the measurement sites;
- in areas where the wind seems to be more constant, the slight risk differences
are caused by the differences that exists in the area's relief.
Figure 8.11. Calculation of the meteorological risk index known as Carrega I85/90,
during the night of July 5, 2005. Different successive layers were used, with each
layer representing a different meteorological variable. The index is
generated indirectly. Source: [CAR 07] (see color section)
8.3.3.2. The direct recreation of the nocturnal meteorological risk, recorded at
03:00 UTC
Once the index for each measuring site has been calculated, it is very easy to
recreate the nocturnal meteorological risk and to plot these values on a map. The
determination coefficient of the process of environmental regression used reaches a
value of 0.76 (or in other words 24% of variance, which is not “explained” by the
model) and leads to the creation of Figure 8.12. The effects of bumps and hollows,
which were caused by the interpolation method that was applied to the wind in order
to recreate the index indirectly, are not present here. The only influences that can be
observed here are the effects of relief and the proximity to the sea.
Search WWH ::




Custom Search