نوع مقاله : پژوهشی
نویسنده
Department of Geography, Payame Noor University, PO BOX 19395-3697 Tehran, Iran
چکیده
کلیدواژهها
موضوعات
عنوان مقاله [English]
نویسنده [English]
Topographic factor (LS) in the Revised Universal Soil Loss Equation (RUSLE) is highly sensitive to Digital Elevation Model (DEM) resolution. Previous studies have identified optimal DEM cell sizes for different contour intervals, but the directional variation of spatial dependency—termed topographic anisotropy—has remained unexplored. This study investigates how anisotropy affects LS factor estimation and whether anisotropy indices can serve as additional criteria for DEM resolution selection. Four DEMs (30 m, 50 m, 100 m, and 300 m) derived from 20 m contour interval topographic maps of the Kuala Yong area in Peninsular Malaysia were systematically analyzed. Directional semivariograms were computed at four orientations (0°, 45°, 90°, 135°) with a lag size of 300 m, and key geostatistical parameters—nugget (C₀), partial sill (C₁), total sill (C), anisotropy ratio (AR), precision (ε), mean error (ME), and root mean square error (RMSE)—were extracted. Results showed that anisotropy is strongest at 30 m (AR=0.433, ε=0.624) and progressively declines to near-isotropic conditions at 300 m (AR=0.842, ε=0.188). The 50 m DEM achieved optimal balance: highest spatial dependency (C₀/C=0.966), maximum total sill (0.1566), favorable prediction accuracy (ME=0.0332, RMSE=0.2435), and preserved anisotropy (AR=0.500, ε=0.508). The 30 m resolution exhibited "false anisotropy" due to overfitting. The proposed geostatistical framework integrates isotropic and anisotropic criteria, enhancing LS factor reliability and supporting sustainable watershed management in erosion-prone landscapes.
کلیدواژهها [English]
)