Evaluation of global digital elevation model for flood risk management in Perlis

• Main Authors: Pa’suya, Muhammad Faiz, Md. Din, Ami Hassan, Mat Amin, Zulkarnaini, Mohd. Omar, Kamaludin, Omar, Amir Hamzah, Rusli@Ruslik, Noradila
• Format: Book Section
• Published: Springer Nature Singapore Pte Ltd 2018
• Subjects: G Geography (General)
• Online Access: http://eprints.utm.my/id/eprint/96574/; http://dx.doi.org/10.1007/978-981-10-8016-6_70

In flood modelling process, Digital Elevation Models (DEMs) is a valuable tool in topographic parameterization of hydrological models. The release of the free-of-charge satellite based DEMs such as SRTM and ASTER prompted the accurate flood modelling process especially to propose flood mitigation in the Perlis region. In this research, the accuracy of SRTM DEM of spatial resolution 1 arc-sec and 3 arc-sec, as well as ASTER DEM are evaluated. The reference levels produced from GNSS observation and Earth Gravitational Model 1996 (EGM96), as well as local mean sea level are used to analyse the vertical accuracy of each GDEMs in Perlis, Malaysia. The total of 38 Benchmark (BM) and Standard Benchmark (SBM) around the Perlis region were observed by GNSS using static method and processed using TOPCON Tool software. A comparison with the local mean sea level height indicated that SRTM 1″ is the much greater absolute vertical accuracy with an RMSE of ±3.752 m and continued by SRTM 3″ and ASTER GDEMs where the obtained accuracy was ±4.100 and ±5.647 m, respectively. Also, by using orthometric height form the GNSS and EGM96 as reference elevation, the obtained accuracy was ±3.220, ±3.597, and ±5.832 m for SRTM 1″, SRTM 3″ and ASTER, respectively. Statistical results have also shown that SRTM 1″ has a good correlation with Hmsl and HGNSS where both correlations values are 0.9925, while the SRTM 3″ and ASTER show the correlation of 0.9873 and 0.9375.