River cross-section extraction from the ASTER global DEM for flood modeling T.Z. Gichamo a, 1 , I. Popescu a, * , A. Jonoski a,1 , D. Solomatine a, b, 1 a UNESCO-IHE Institute for Water Education, P.O. Box 3015, 2601 DA, Delft, The Netherlands b Water Resources Section, Delft University of Technology, Delft, The Netherlands article info Article history: Received 18 January 2011 Received in revised form 5 December 2011 Accepted 9 December 2011 Available online 24 December 2011 Keywords: ASTER GDEM River cross-sections HEC-RAS/GeoRAS Optimization Tisza River abstract An approach to generate river cross-sections from the Advanced Spaceborne Thermal Emission and Reflection Radiometer Global Digital Elevation Model (ASTER GDEM) is discussed. The low resolution and the inadequate vertical accuracy of such global data present difficulties in differentiating features of hydraulic importance, which necessitate pre-processing of the DEMs before they are used. A vertical bias correction carried out by comparison of elevation points with a high accuracy terrain model produces a considerable improvement to the cross-sections obtained. In a situation where there are some flow/stage measurements at either end of the river reach, an optimization routine combined with a conceptual flow routing method can provide an additional tool to identify the parameters of an equivalent river section. The extracted cross-sections were used in a 1D river modeling tool HEC-RAS/GeoRAS to simulate flooding on a part of the Tisza River, Hungary. Model results are encouraging and show good potential for using the suggested method in the areas of topographic data scarcity. Ó 2011 Elsevier Ltd. All rights reserved. 1. Introduction A successful river flood model requires a sufficient representa- tion of the river channel and floodplain geometries, with an accu- rate description of the model parameters, to make it possible to predict the flow magnitude and water levels along the reach accurately. Software tools have been (and are being) developed and updated to extract spatial features that are useful for hydraulic models, from topographical data sources, both in GIS (Merwade et al., 2008; Tesfa et al., 2011) and non-GIS environments (Schwanghart and Kuhn, 2010). However, obtaining detailed topographical data for every river basin under study is a difficult task as the process involves an expensive and time consuming survey campaign (ground or airborne) and a painstaking post- processing of the survey data (Liu, 2008; Mandlburger et al., 2009; Merwade et al., 2008). A number of studies have been carried out that try to deal with topographical data scarcity in river flood modeling. Most of them rely on the integration of GIS with digital elevation models (DEM) obtained from remote sensing satellites or other globally available data sets (Asante et al., 2008; Herath et al., 2003; Merwade et al., 2005; Sanders, 2007); while others try to use data assimilation techniques to identify a (synthetic) cross-section that is hydraulically equivalent to the real river geometry (Honnorat et al., 2006, 2009; Roux and Dartus, 2008). In addition to topographical data for flood propagation modeling, satellite images provide the locations and types of structures in the area of interest. This is useful for spatially distributed flood damage quantification, particularly in urban areas (Qi and Altinakar, 2011). This study presents two approaches for the extraction of river cross-sections from a freely available, satellite-based DEM. The first method involves reading the dimensions of triangular cross- sections from the DEM and applying vertical bias correction to improve the constructed cross-sections. In the second method, an optimization routine applied to conceptual flow routing equations is used to identify the equivalent channel geometry parameters from the observed flow and water levels for a given flood event. With the obtained river cross-section data, flood simulation and analysis was carried out on a part of the Tisza River Basin, Hungary, using the HEC-RAS/GeoRAS modeling software package. The following main tasks were carried out:  Extracting river cross-sections from a satellite DEM, and comparing the outputs from two hydraulic models, one using surveyed data and the other with the satellite DEM;  Correcting the vertical bias of these cross-sections;  Identifying the equivalent (synthetic) cross-section using parameters determined from the DEM and the optimization routine; * Corresponding author. Tel.: þ31 15 2151895; fax: þ31 15 3122921. E-mail addresses: tseganeh@gmail.com (T.Z. Gichamo), i.popescu@unesco-ihe. org (I. Popescu), a.jonoski@unesco-ihe.org (A. Jonoski), d.solomatine@unesco-ihe. org (D. Solomatine). 1 Fax: þ31 15 3122921. Contents lists available at SciVerse ScienceDirect Environmental Modelling & Software journal homepage: www.elsevier.com/locate/envsoft 1364-8152/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.envsoft.2011.12.003 Environmental Modelling & Software 31 (2012) 37e46