PARAMETER SCALE ANALYSIS IN RUNOFF MODELLING IN SEMI-ARID REGIONS OF BRAZIL WITH THE MODEL NAVMO

摘要

In contrast to small scales, large scales are heterogeneous and affect the parameters of relevant hydrologic processes. As a consequence, the parameter values vary as the basin area increase. Therefore, the modelling of a process (e.g., runoff) depends on a model capability of taking into account the heterogeneities that are common at regional scales. Once the relationships between the model parameters and basin area are established, the model can be used to simulate effectively. In this study, the effects of scale on the parameters of the distributed model NAVMO were investigated. The model parameters were calibrated by comparing observed and simulated runoffs at various sub-basins of two large semi-arid areas in the Northeast region of Brazil (SeNeB), the Paraiba (PBRB) and Piranhas (PIRB) river basins. For the PBRB the model was first applied to 9 plots (100 m~2), 4 micro-basins (5000-10000 m~2) and 3 sub-basins (10-140 km~2), and then to watersheds from 600 km~2 - 13000 km~2. For the PIRB, the model was applied to subbasins varying from 475 km~2 to 15000 km~2. The results showed that the simulated daily hydrographs, annual means and peak discharges approached reasonably well the observed values for all basins of both regions. Also, the model parameters changed as land use and rainfall changed. In addition, the parameters scale analysis showed that the SCS curve number, the Manning-Strickler roughness coefficient and direct runoff coefficient decreased as the catchment area increased, while the evapotranspiration parameter had the opposite effect. Based on regression analyses, relationships for determining the effective parameters values in terms of basin area were established for both regions.