Using the attraction model in remote sensing to evaluation of topographic wetness index (TWI)

The objective of this study is evaluation of topographic wetness index (TWI) using attraction model in northern of Fars province. The first attraction model was used to enhance the spatial resolution of digital elevation model (DEM). In this research to estimate of value of sub-pixels neighboring pixels, touching and quadrant neighboring models were used. After manufacturing output images for sub pixels in the 2, 3 and 4 scales with different neighborhoods, the best scale with most appropriate type of neighborhood was determined using ground control points then the values of RMSE was calculated for them. The total number of ground control points extracted from the mapping maintenance was 2118 points. The results showed that between scales with different neighborhoods, 3 scale and quadrant neighboring model have the most accuracy by the lowest value of RMSE for DEM 90 meter. Then using produced DEM from 3 scale and quadrant neighboring model, topographic wetness index for the study area was determined. The results showed that topographic wetness index (TWI) in the study area is variable between -4.45 to 6.06. Central zones of the study area have the highest values ​​of wetness. Compare of wetness index produced from attraction model (with more spatial resolution) with DEM 90 meter (with lower spatial resolution) showed that with using attraction model, more details of the amount of moisture in the study area, is visible.