<title>Evaluating SRTM and ASTER DEM accuracy for the broader area of Sparti, Greece</title>

Abstract

One of the major projects of the Institute of Geology & Mineral Exploration (IGME) is called "Urban Geology". In the frame of that project there is need for a high accuracy DEM covering the whole country. The DEM should be used for the orthorectification of high resolution images and other applications such as slope map creation, environmental planning et.c. ASTER and SRTM are two possible sources for DEM covering the whole country. According to the specifications the ASTER vertical accuracy of DEM is about 20m with 95% confidence while the horizontal geolocation accuracy appears to be better than 50 m. More recent studies have shown that the use of GCP's resulted in a plannimetric accuracy of 15 m and in a near pixel size vertical accuracy. The Shuttle Radar Topography Mission (SRTM), used an Interferometric Synthetic Aperture Radar (IFSAR) instrument to produce a near-global digital elevation map of the earth's land surface with 16 m absolute vertical height accuracy at 30 meter postings. An SRTM 3-arc-second product (90m resolution) is available for the entire world. In this paper we examine the accuracy of SRTM and ASTER DEMs in comparison to the accuracy of the 1/5.000 topographic maps. The area of study is the broader area of Sparti, Greece. After a first control for random or systematic errors a statistical analysis was done. A DEM derived from digitized contours of the 1:5.000 topographic maps was created and compared with ASTER and SRTM derived DEMs. Fifty-five points of known elevation have been used to estimate the accuracy of these three DEMs. Slope and aspect maps were created and compared. The elevation difference between the three DEMs was calculated. 2D RMSE, correlation and the percentile value were also computed. The three DEMs were used for the orthorectification of very high resolution data and the final orthophotos were compared.