Geoscience Reference
In-Depth Information
2.4.4.3.
Estimating the total monthly rainfall for 1986
Altitude
The correlation coefficients of altitude/rainfall amount for January, April, and
November 1986 (Figure 2.24) vary slightly when different DEM resolutions are
used (r values range from 0.69 to 0.72). The coefficients are quite similar to the
coefficients that describe the 1971-2000 normal (Figure 2.21). November is the
month that has the lowest coefficient (approximately 0.6), whilst the maximum
coefficient was recorded for January 1986 (0.78). This variation pattern from one
resolution to another remains: only the value of the coefficient experiences large
differences from one month to the next, which means that the independent variable
(causal variable) of altitude is not the same for the entire year. It would be
interesting to carry out further investigations into this climatological variable.
Slope and topographic roughness
Slope (Figure 2.25) and topographic roughness (Figure 2.26), which are
calculated by large-scale DEM (2,500 or 5,000 m), better explains the total monthly
rainfall for April (full line in Figure 2.24) than when the same two variables are
produced by lower resolution DEMs of 50 or 100 m. The difference between the
two monthly extremes (January and February) is quite high, as far as the coefficients
for both slope and roughness are concerned. Topographic roughness, which is
calculated by intermediary resolutions (500 or 100 m), better explains the total
rainfall for December (dashed line in Figure 2.26) than when the results are
produced by low or high-level resolutions.
Figure 2.24.
Correlation coefficients for altitude/rainfall amount (April 1986); the altitudes
are generated by seven DEMs, which have a resolution that varies from 50 to 5,000 m
Figure 2.25.
Correlation coefficients for slope/rainfall amount (April 1986)
