Aplicação do MDE de alta resolução espacial em uma abordagem distribuída do modelo NRCS-CN-HUT

Abstract

The NRCS-CN-HUT model is one of the most widely used technique for estimating runoff hydrographs. Despite of this, this model does not explicitly account for the spatially distributed nature of watershed properties. An alternative to this limitation is the development of a distributed approach based on this model. In this approach, the watershed can be divided into cells with distinct hydrologic parameter value, related to land cover or topography, for example, and inside each of these cells can be applied the equations from the NRCS model. The runoff hydrograph in the watershed outlet is obtained by the sum of the flow rate from all contributing cells. To access topography data this distributed method use the Digital Elevation Model (DEM) that represents the terrain surface by raster image where each pixel has the information about the surface elevation. A lot of studies affirms that this pixel size has a significant impact on the hydrologic simulation results. Because of this, in this study was evaluated the performance of the distributed approach of NRCS-CN-HUT by using a high resolution DEM (1 m), obtained from LiDAR technology. Therefore, the runoff hydrograph was simulated to 9 real rainfall events occurred in the Saint Hilaire Watershed, located in Porto Alegre city, south of Brazil. This watershed drains about 6 km². The simulated runoff hydrographs were compared to the observed flow rate data. So, were identified some hydrograph characteristics: significantly model sensibility to considerations of initial abstraction of rainfall; accurately prediction of peak flow; underpredicted time to peak; considerable errors in total simulated volume. Despite of the fact that higher resolution DEMs provide more accurate terrain representation, the hydrologic simulation with this so detailed data required considerable demands of data storage space and computation ability.