The implications of data selection for regional erosion and sediment yield modelling

✍ Scribed by Joris de Vente; Jean Poesen; Gerard Govers; Carolina Boix-Fayos

• Publisher: John Wiley and Sons
• Year: 2009
• Tongue: English
• Weight: 731 KB
• Volume: 34
• Category: Article
• ISSN: 0360-1269
• DOI: 10.1002/esp.1884

✦ Synopsis

Abstract

Regional environmental models often require detailed data on topography, land cover, soil, and climate. Remote sensing derived data form an increasingly important source of information for these models. Yet, it is often not easy to decide what the most feasible source of information is and how different input data affect model outcomes. This paper compares the quality and performance of remote sensing derived data for regional soil erosion and sediment yield modelling with the WATEM‐SEDEM model in south‐east Spain. An ASTER‐derived digital elevation model (DEM) was compared with the DEM obtained from the Shuttle Radar Topography Mission (SRTM), and land cover information from the CORINE database (CLC2000) was compared with classified ASTER satellite images. The SRTM DEM provided more accurate estimates of slope gradient and upslope drainage area than the ASTER DEM. The classified ASTER images provided a high accuracy (90%) land cover map, and due to its higher resolution, it showed a more fragmented landscape than the CORINE land cover data. Notwithstanding the differences in quality and level of detail, CORINE and ASTER land cover data in combination with the SRTM DEM or ASTER DEM allowed accurate predictions of sediment yield at the catchment scale. Although the absolute values of erosion and sediment deposition were different, the qualitative spatial pattern of the major sources and sinks of sediments was comparable, irrespective of the DEM and land cover data used. However, due to its lower accuracy, the quantitative spatial pattern of predictions with the ASTER DEM will be worse than with the SRTM DEM. Therefore, the SRTM DEM in combination with ASTER‐derived land cover data presumably provide most accurate spatially distributed estimates of soil erosion and sediment yield. Nevertheless, model calibration is required for each data set and resolution and validation of the spatial pattern of predictions is urgently needed. Copyright © 2009 John Wiley & Sons, Ltd.