By Jojene R. Santillan, Meriam M. Santillan
The ASTER GDEM Version 2 (v2) which was released in 2011 is considered to be the highest resolution and readily available global digital elevation model. Having a spatial resolution of 30-m, the ASTER GDEM v2 contains significant improvements of Version 1 in terms of spatial coverage, refined horizontal resolution, water masking, and inclusion of new ASTER data to supplement the voids and artifacts. Despite of these improvements, data anomalies such as abrupt rise (“bumps”) and fall (“pits”) in elevation values in the ASTER GDEM v2 still remains. In this paper, we present a simplified and semi-automatic approach of detecting and correcting data anomalies in the ASTER GDEM v2 of Tago River Basin in Surigao del Sur, Mindanao, Philippines. The anomaly detection procedure consisted of calibrating the elevation values in the ASTER GDEM v2 against a resampled SRTM DEM version 4.1, creating a difference image between the calibrated ASTER GDEM and SRTM DEM, applying a low pass filter to the difference image, employing the K-means clustering algorithm to classify the difference image pixels into various classes, and then labeling them as bumps, pits or neither (i.e., not anomaly) using a known set of pixels pre-selected prior to K-means classification. The accuracy of the detection was at 99.47% based on an independent set of randomly selected validation pixels. The detected anomalies were then masked out from ASTER GDEM v2, and the elevation values of these pixels were replaced by elevation values extracted from the resampled SRTM DEM. The corrected DEM was used for hydrological applications such as computations of flow direction and flow accumulation grids, sub-basin delineations and stream network definition. These DEM derivatives are important inputs in hydrological model development for the river basin.
KEYWORDS: ASTER GDEM v2, SRTM DEM, anomaly detection, DEM correction
Publication Details/Metadata:Paper Title: DETECTION AND CORRECTION OF ASTER GDEM V2 DATA ANOMALIES THROUGH DEM DIFFERENCING Authors: Jojene R. Santillan, Meriam M. Santillan Publication Date: 2014 Conference Name/Journal/Book Title: 35th Asian Conference on Remote Sensing 2014 (ACRS 2014) Volume: 2 Pages: 1261-1266 Publisher: Asian Association on Remote Sensing Link to Full Text: View/Download