INTEGRATION OF LAND COVER DATA INTO THE OPEN SOURCE MODEL SWAT

摘要

Conversion of land is known to alter hydrological processesrnsuch as the exchange of energy and water. Thereforernit is important to integrate land cover data in hydrologicalrnor eco-hydrological model approaches. To integraternremotely sensed input parameters into the ecohydrologicalrnmodel SWAT, we developed a grid-basedrninterface (SWATgrid), which enables the computing ofrnthe water balance elements for every model grid cellrnwithout losing its geographical reference. We testedrnSWATgrid in a sub-catchment of the Stoer River, thernBuenzener Au. For this study area, we applied conventionalrnSWAT, which is based on the HRU concept, andrnSWATgrid to compute the water balance between 2003rnand 2010. Results of Landsat land use classificationrnserved as input data. To compute daily discharge at thernoutlet of the Buenzener Au, SWATgrid showed slightlyrnbetter accuracy measures (Nash Sutcliffe = 0.60; r2 =rn0.62) than conventional SWAT (Nash Sutcliffe = 0.57;rnr2 = 0.60). Moreover, the grid-cell interface enabled arnspatially distributed modelling based on the Landsatrnland cover. The grid cells now interact with each otherrnand exchange matter and energy enabling a realisticallyrncomputation of water balance elements and their distributionrnpatterns. SWATgrid overcomes the difficulties ofrnthe conventional SWAT; the user is able both to integraternremote sensing data and to obtain spatially distributedrnmodel output.