SASGI Proceedings 2013 – Stream 1 Comparison and validation of ASTER-GDEM and SRTM elevation models over parts of Kaduna State, Nigeria by Olalekan Adekunle Isioye and Israel Castro Yang, Ahmadu Bello University, Nigeria, Department of Geomatics, Faculty of Environmental Design Samaru, Zaria, Kaduna state Abstract Digital elevation model (DEM) represents a very important geospatial data type in the analysis and modelling of different hydrological and ecological phenomenon which are required in preserving our immediate environment. DEMs are typically used to represent terrain relief and are particularly relevant for many applications such as soil erosion volume calculations, flood estimate, quantification of earth materials to be moved for channels, roads, dams, embankment etc. This study investigates the quality (in terms of elevation accuracy and relative altitudinal differences) of two publicly available elevation model datasets over parts of Northern Nigeria: (i) the 3 arc second Shuttle Radar Topography Mission (SRTM) ver4.1 from CGIAR-CSI; and (ii) the 1 arc second Advanced Space borne Thermal Emission and Reflection Radiometer Global − DEM (ASTER-GDEM) ver1 from NASA/METI. The main features of these datasets are reported from a geodetic point of view. This study represents a follow up to Isioye and Obarafo, 2010 and Isioye et al., 2012, because it is believed that the significant advances in terms of resolution and coverage made by ASTER GDEM justify the need for a new evaluation of elevation data over the study area. The quality of the two elevation models was assessed in two ways: (i) a comprehensive model-to-model comparison was carried out over the study area, providing insight into random and systematic errors among the elevation data, (ii) external validation was carried out based on GPS control points (GCPs). Two test sites identified as “mountainous” and “flat” terrain in Kajuru/Kaduna and Zaria respectively were used within the study area, results of the correlation test among the two elevation data sets and the 55 GCPs show that ASTER is slightly better correlated in the mountainous terrain than SRTM while SRTM showed a significantly stronger correlation in the flat terrain than ASTER-GDEM on the 58 GCPs used for the site. The overall absolute average vertical errors of ASTER-GDEM are 18,93 ±2,85 m, 16,36 ±2,14 m, while the STRM have 12,52 ±3,25 m, 3,17 ±1,17 m for the Kajuru/Kaduna and Zaria sites respectively. Keywords ASTER-GDEM, SRTM, GPS, digital elevation model (DEM), slope and aspect Introduction and background Digital elevation models (DEM) provide basic information on heights of the Earth’s surface and features upon it. The specific terms digital terrain model (DTM) and digital surface model (DSM) are often used to specify the surface objects described by an elevation model. A DTM usually refers the physical surface of the Earth, i.e., it gives elevations of the bare ground (terrain). On the other hand, a DSM describes the upper surface of the landscape. It includes the heights of vegetation, buildings and other surface features, and only gives elevations of the terrain in areas where there is little or no ground cover. DEMs have become an important data source for a range of applications in Earth and environmental sciences. Examples of applications for elevation data are numerous, such as gravity field modeling, hydrological studies, topographic cartography, ortho-rectification of aerial imagery, flood simulation and many more. Generally, DEM datasets can be obtained from range of techniques, such as ground survey e.g., [1], airborne photogrammetric imagery e.g., [2], airborne laser scanning (lidar) e.g., [3], radar altimetry e.g., [4] and interferometric synthetic aperture radar (InSAR) e.g., [5]. Quite often, DEMs are constructed from data sourced from several of these methods and are thus of variable quality e.g., [4, 6, and 7]. Since a number of applications may rely solely on SRTM and/or ASTER DEMs, it is important to assess the quality of these data, i.e., how well does the DEM approximate the shape of the Earth’s surface? Quality of elevation data is commonly expressed in terms of vertical accuracy. It can be determined using comparison data that should be based on accurate and independent methods, such as (terrestrial) topographic surveys, airborne laser scanning or photogrammetric techniques, allowing truly external and independent validation but in the course of this research, the (terrestrial) topographic surveys are being carried out. Another issue affecting the quality of space-based DEMs is the presence of systematic error patterns. For example, this can include artificial structures that are systematically too high or low and therefore not representative of the terrain’s surface. Heights of forest regions or buildings, which are often included in space-collected DEM data (i.e., a DSM), represent an error source for applications exclusively interested in elevations of the terrain (i.e., a DTM). This study is aimed at comparing and validating (in terms of elevation accuracy) SRTM and ASTER Global Digital Elevation Model (GDEM) with GPS Points. The quality of the models is then assessed in two ways. A comprehensive