loading page

How can Distributed Hydrological Models Inform Decision Making? Multi-Site Calibration of SWAT for a Large Brazilian River Basin
  • +2
  • Ana Luisa Sales Pereira Almeida,
  • Hersilia Santos,
  • Sónia Maria Carvalho Ribeiro,
  • Robert Mason Hughes,
  • Diego Rodrigues Macedo
Ana Luisa Sales Pereira Almeida
Universidade Federal de Minas Gerais
Author Profile
Hersilia Santos
University of Iowa,Centro Federal de Educação Tecnólogica de Minas Gerais
Author Profile
Sónia Maria Carvalho Ribeiro
Universidade Federal de Minas Gerais
Author Profile
Robert Mason Hughes
Oregon State University
Author Profile
Diego Rodrigues Macedo
Universidade Federal de Minas Gerais

Corresponding Author:[email protected]

Author Profile

Abstract

Although there are numerous modeling tools available for managing water resources, they tend to focus on relatively small watersheds (<25,000 km2) and evidence is scarce regarding their applications to large watershed management and planning. The use of hydrological modeling tools in decision making is particularly challenging in large tropical countries such as Brazil. Here we tailored and customized SWAT (Soil and Water Assessment Tool) calibration, validation and sensitivity analysis for large Brazilian watersheds (>25,000 km2). Our results show that customized SWAT calibrations successfully simulated flow behaviors across the Rio das Velhas basin, which contains heterogeneous landforms and land uses. The multisite calibration method was adopted because of the large basin area and the availability of flow monitoring stations. As a result of the multi-site calibration, specific regionalized parameters were obtained for each group of sub-basins. Our results showed a good adjustment of the model. NS (Nash & Sutcliffe coefficient) values were 0.73 - 0.97 (calibration) and 0.51 - 0.98 (validation). PBIAS (% bias) was 11.3 to 19.4 (calibration) and -18.6 to 24.6 (validation) and R² values were >0.6 in all sub-basins. We conclude that hydrological models coupled with GIS facilitates simulating complex hydrological processes and can improve decision making by Brazilian water resource managers.