Geoscience Reference
In-Depth Information
terms of influencing the variations of temperature and precipitation. In some cases,
however, it is difficult or it may not be possible to obtain the NDVI, and this is
especially true if the area to be interpolated is large. It is possible to overcome this
problem if there is an image that shows land cover and if this image provides
information on land cover. This is made possible by CLC, where the information
provided is of a qualitative nature and cannot be easily integrated into the regression
models. By using a CLC it is possible to obtain an index that provides information
about the potential amount of vegetation that exists in each pixel of an image. The
procedure for carrying out this task is relatively easy. In order to produce such a
piece of quantitative information from CLC, and to measure how good this
information is at estimating data, we will use the example of monthly temperature
averages calculated for the period 1971-2000 by the 1,530 temperature stations in
mainland France.
2.4.2.1.
The location of climatological stations and CLC
The 1,530 climatological stations are projected onto a CLC image, which has
been processed by a 250 m resolution raster GIS. The land cover type of each pixel
with a climatological station is archived in a table that is made up of 1,530 rows
(climatological stations) and 44 columns (one CLC post per land cover type). In the
event where a climatological station and its corresponding CIC post cross paths, this
case is given a value of one. All other 43 columns are given a value of zero. The
sum of all of the rows is equal to one (each climatological station is characterized by
only one land cover post). The sum of all of the columns provides information on
the frequency of the land cover types for the 1,530 pixels where a temperature
station is installed. This disjunctive table can be represented as a graph (see Figure
2.11).
Figure 2.11.
Percentage of climatological stations located
in each of the CLC's 44 land cover posts
The climatological stations that record temperature are found on only four
different types of CLC land cover with a frequency of up to 10%:
- type 2 (discontinuous urban areas): 16% of all climatological stations;
- type 12 (arable ground, outside irrigated areas): 18% of all climatological
stations;
