Streamflow predictions in ungauged basins by using the Grunsky\'s Method

• Ano de defesa: 2023
• Autor(a) principal: Marchezepe, Bruno Ken
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Biblioteca Digitais de Teses e Dissertações da USP
• Palavras-chave em Português: aridity index; bacias tropicais; escala temporal; evaporação potencial; grupo hidrológico; hydrological group; índice de aridez; mean temperature; potential evapotranspiration; regionalização; regionalization; temperatura média; temporal scale; tropical basins
• Link de acesso: https://www.teses.usp.br/teses/disponiveis/18/18138/tde-02082023-104722/

Predicting runoff in ungauged basins is a significant challenge in Hydrology, and various advanced techniques have been developed to address this issue. However, returning some simplicity to the predictions might be necessary for practical uses. This master dissertation evaluates a generalized of Grunsky\'s approach and proposes its improvement. This approach was originally developed in the early 1900s for basins in San Francisco Peninsula, USA. First, we focus on the application of the Grunsky generalized method to 716 Brazilian catchments on interannual and monthly scales. The rainfall-runoff relation coefficient (α) is determined, and the method\'s performance is evaluated locally, catchment by catchment. Additionally, we regionalized and analyzed the catchments into hydrological groups. Then, an improved version of the previously tested method is proposed, incorporating a rainfall-runoff relation coefficient (α) based on mean annual temperature. Through multiple linear regression, the values are determined using catchment attributes such as the aridity index, annual average temperature, and potential evapotranspiration. The generalized method presented a median percentage bias and Kling-Gupta Efficiency of -1% and 0.73, respectively, with favorable performances observed in certain groups. On a montly scale, more than 83% of the total studied basins had at least one month with KGE greater than 0.50. The performance of the improved method indicates the suitability of this approach for predicting runoff in Brazilian basins, with KGE = 0.899, R² = 0.82, and RMSE = 27.4% on the interannual scale. Here, we emphasize the practicality and reliability of the Grunsky approach as a simple and easy-to-use equation for predicting runoff. The findings suggest that this approach can serve as a viable alternative to more complex methods, especially in ungauged or poorly gauged basins. By incorporating both theoretical and empirical elements, this study contribute to the ongoing efforts to develop accessible and effective methods for runoff prediction, furthering our understanding of hydrological processes in Brazilian catchments.