An Automated Approach Using Topographic Position Index (TPI) for Landform Mapping (Case Study: Gede Watershed, Malang Regency, East Java, Indonesia)

Abstract

Mostly, the researchers have been carried out landform using survey method and manual delineation. Integrating remote sensing data and GIS technique can automate identifying of landform using Topographic Position Index data as one of remote sensing data. The aim of paper is to identify and categorize landform element using Topographic Position Index in the Gede Watershed derived by DEMNAS with 30 m resolution and output scale of map is 1:25.000. The method was carried by algorithm of slope and TPI automation with Geographic Information System technique. To obtain the appropriate TPI with the real condition of landform, we used radius circle 5m, 10m, 25m, 50m and 100m neighborhood size cell for TPI. The combination of slope and TPI shows the best result which are appropriate with the real condition of landform is radius circles 5m neighborhood size cells. Based on the result shows five heterogeneous of landscape i.e peak (0,6%), upper slope (10%), middle slope (9,49%), lower slope (39,36%) and channel bad (40,49%). To achieve the better result, it needs intensive field survey and using high resolution imagery as accuracy test.