Results - Response of Flood Events to Land Use and Climate Change - page 138

Table 6.6 The high Moran's I in the first distance

Method Moran's I value Result

IDW 0.041 Perfect correlation

Topo to Raster 0.042 Perfect correlation

Spline 0.041 Perfect correlation

Circular 0.071 Perfect correlation

Exponential 0.089 Perfect correlation

Gaussian -0.019 Perfect dispersion

Stable -0.008 Perfect dispersion

Spherical 0.013 Perfect correlation

Based on the obtained results, the negative (positive) values indicate negative
(positive) spatial autocorrelation. Values range from -1 (indicating perfect dis-
persion) to +1 (perfect correlation). A zero value indicates a random spatial pat-
tern. The high Moran's I in Gaussian and Stable showing perfect dispersion in the
first distance. On the other hand, the Moran's I value for IDW, Topo to Raster,
Spline, Circular, Exponential, and Spherical represents perfect correlation in the
first distance (Table 6.6 ).
As partly conclusion from this part, it could be concluded that Geostatistic
techniques show the strongest results compared to the deterministic. In between,
the Spherical and Gaussian showed better agreement with the observed data and
represents the smoothest and more accurate DEM. In terms of spatial analysis, The
Standard residual diagnosis graphs show that the best results in terms of normal
distribution belong to Gaussian method, because all scattered values spread close
to zero (Skewness is 0.071) between the ranges of -0.5 and 0.5. The residuals
dataset in Gaussian method obeys a normal distribution. The results of Moran's I
investigation showed Gaussian and Stable showing perfect dispersion in the first
distance. On the other hand, the Moran's I value for IDW, Topo to Raster, Spline,
Circular, Exponential, and Spherical represents perfect correlation in the first
distance. In between, the lowest Moran's I value of 0.013 belongs to Spherical.

6.2.3 Hydrological Analysis

To compare the accuracy of the interpolated DEMs, the stream network for each
surface was derived using ArcGIS. The modeled stream network was compared
with the observed stream network (''true'' data). Figure 6.9 shows the results of
stream network delineation using ArcHYDRO tools for Kriging-the Spherical
model as a sample. In this part, the quality of all the DEM was tested hydrolog-
ically by creating drainage networks. Figure 6.9 presents the comparison between
the modeled and the observed stream network for (a) original DEM; (b) IDW; (c)
Topo to Raster (ANUDEM); (d) Spline; (e) Kriging, Circular; (f) Kriging;
Exponential; (g) Kriging, Gaussian; (h) Kriging, Stable and (i) Kriging, Spherical.