Avaliação da influência das resoluções espaciais de MDE na estimativa do escoamento superficial de uma bacia hidrográfica usando o modelo SWAT+

Abstract

The effect that different spatial resolutions of digital elevation models (DEM) have on the simulation of surface flowing off are still poorly known, especially in a semiarid environment. This study aims to analyze the influence that different spatial resolutions of MDE (30 to 1500 m) and sources of obtainment (SRTM, ALOS, ASTER) perform on the outflow forecasts in the hydrological model Soil and Water Assessment Tool + (SWAT +), for the hydrographic basin of the Taperoá river, in the state of Paraíba. SWAT + is a completely revised version of the SWAT model and has been applied to estimate hydrosedimentological processes in several basins, where the topography of the terrain can have a great influence on the simulated variables of the model, especially in the semi-arid region of Northeast Brazil that is characterized by the occurrence of extreme rain and outflow events. The results show that the spatial resolutions of the MDE influence the topographic representation and the extraction of the drainage network, where the higher the resolution, the more outlined are its contours, the greater the number of sub-basins and hydrological response units. As for the sources of MDE, all presented similar behavior in the discretization of the hydrographic basin. It is possible to observe that the differences that the MDE resolutions cause in the discretization of the basin influence the simulation of the flow. The model presented an answer for each rain event that generated flow and the resolutions mainly influence peak outflows. For simulations using the highest resolutions, the model tends to overestimate the outflow and, for the lower resolutions, the peak outflow tends to be underestimated. The values of R² are considered good and very good, varying from 0.89 to 0.75, as the resolution decreases. NS values were unsatisfactory for the highest resolutions (30 to 500m), varying from 0.31 to 0.45, which showed an overestimation of the peak flow. For the others (800 and 1500 m), it was considered satisfactory, ranging from 0.5 to 0.52, which showed a better adjustment of the hydrogram. As for the SRTM, ALOS and ASTER sources, the simulations showed similar behavior in the flow estimation, with similar hydrograms and approximate statistical indexes, with a very good R² of 0.89 and an unsatisfactory NS from 0.26 to 0.31.