Abstract

Decision makers need cost-effective methods for accurately depicting land surface characteristics as basic tools for locating construction project. ASTER Global Digital Elevation Model V002 with 30 m resolution was used to extract information on terrain surface and drainage network at the micro-catchment level. ASTER data was compared with 39 reference points from integrated Global Positioning System (GPS) and topographic maps. However, there was a gap between ASTER dataset and reference points; thus, a readjustment process of ASTER dataset was required. The results indicate that Global Positioning System (GPS) and topographic maps data sets have good connection with a range of difference around ± 32.7 m. Pearson correlation coefficient of ASTER Pixel values in the connection with GPS and topographic maps data sets indicates a strong positive correlation 0.8, 0.827 respectively. Therefore, a Multiple Linear Regression (MLR) model was used to readjust ASTER data based on topographic maps, and ground points. The ‘best’ fit of MLR model for ASTER was chosen and used to interpolate a multiscale temporal and spatial distribution. The research applied two interpolation techniques: the Inverse Distance Weighting (IDW) and Kriging to better understand the spatial distribution. The results show that there is a slight difference between ASTER data and the other two types of elevation models.

Keywords: Multiple linear regressions; Digital elevation model; Topographic map; ASTER