Geography Reference
In-Depth Information
h e chapter introduces some key methods and concepts which will hopefully aid
understanding of other approaches not considered here.
Case studies
8.7
h is section comprises two case studies that make use of the same data set, which is
provided on the topic website. h e data represent elevations in Switzerland and pre-
cipitation amounts for 8 May 1986; the number of observations is 467. h e data are
described by Dubois (2003) and the precipitation measurements are shown in Figure 8.12.
h e i rst case study makes use of Moran's
I
autocorrelation coei cient to explore spa-
tial variation in precipitation amount. h e second study uses GWR to explore spatial
variation in the relationship between elevation and precipitation amount.
8.7.1
Spatial autocorrelation analysis
Moran's
I
was computed using geographical weights with the Gaussian weighting
function dei ned in Equation 8.1. For a 10-km bandwidth,
I
was 0.722, indicating
strong positive spatial autocorrelation in precipitation amounts. h e data were further
Precipitation (mm)
0-100
101-200
201-300
301-585
N
0
100 km
50
Figure 8.12
Precipitation measurements for 8 May 1986 in Switzerland.
Search WWH ::
Custom Search