Desempenho de modelos hidrológicos em uma bacia hidrográfica de cabeceira

• Autor(a) principal: Melo, Pâmela Aparecida
• Data de Publicação: 2019
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Repositório Institucional da UFLA
• Texto Completo: http://repositorio.ufla.br/jspui/handle/1/38941
• topic: Distributed high spatial resolution model (DHSVM); Height above the nearest drainage (HAND); Modelo Hidrológico Distribuído do Instituto Nacional de Pesquisas Espaciais (MHD-INPE); Geomorphons; Geomorfologia; Serra da Mantiqueira; Rio Grande; Gomorphology; Engenharia Agrícola

The Serra da Mantiqueira region has great importance in Grande River’s water production, the largest formator of the Paraná River, both of great importance for public water supply and power generation in Brazil. To study future impacts on these basins, hydrological models, when representative of natural conditions, play a key role in watershed’s management. In this sense, the objective of this work is to analyze the performance of hydrologic models applied in a headwater basin in the Serra da Mantiqueira region, with complex terrain. For this, high (DHSVM) and low spatial resolution (MHD-INPE) models were used. In addition, soil class maps based on geomorphology (Geomorphons) and height difference between each pixel and drainage hydrography (HAND) were used. The MHD-INPE model was concentrated in a single cell of hydrologic response unit and compared to the distributed DHSVM model’s performance. Despite spatial simplification, the low-resolution model was better representative of some of the hydrological processes in the basin. Additionally, the performance of the distributed high spatial resolution model (DHSVM), coupled to a map based on geomorphology as soil representation, was analyzed. Thus, a better performance of the model was observed when simulating the flow in the control section, with higher precision than other studies that used both pedological survey and moisture zones (HAND) maps in the same basin. Therefore, due to the need for spatialized physical details in the basin, the DHSVM model better fits as the input information is detailed, requiring extensive experimental survey. However, the MHDINPE model has ease of implementation and calibration, which is ideal for use it coupled to atmospheric models, for example.