Concepção do modelo hidrológico LASH em MATLAB®: aprimoramento da propagação em rios e da calibração

Abstract

In view of the current situation of water resources management, studies on watersheds are essential for supporting projects and, when applicable, watershed committees and master plans. In this sense, hydrological models are strongly recommended and, among them, the Lavras Simulation of Hydrology (LASH) has been successfully used in Brazilian watersheds. The main contribution of this study was the implementation of the LASH model in MATrix LABoratory language (MATLAB®), enhancement of its river routing and calibration modules, and its evaluation taking the Camaquã river watershed (CRW) as a case study, which has a watershed committee and a watershed master plan. Chapter 1 addressed technical and computational issues, such as the development of the LASH in MATLAB®, implementation of the river routing module according to the Kinematic Wave method and the insertion of a multi-objective optimization algorithm, known as A Multi-Algorithm Genetically Adaptive Multiobjective (AMALGAM). On the other hand, chapter 2 focused on the evaluation of LASH coded in MATLAB® taking CRW as a case study. Furthermore, two calibration schemes were considered in this study: i) scheme 1: all the LASH parameters were calibrated following a lumped scheme and; ii) scheme 2: two parameters – initial abstraction coefficient (λ) and response time of the surface reservoir (CS) - were calibrated by subwatershed. The results of chapter 1 evidenced some benefits of coding the model in MATLAB® language, mainly with regard to the: i) river routing module based on the Kinematic Wave method; ii) ease of coupling the AMALGAM multi-objective calibration algorithm with LASH; iii) computational strategies applied to optimize the model's processing time; and iv) ease of editing, updating and enhancing the script, which are considered the the major strength in the implementation of LASH in MATLAB®. The results of chapter 2 were satisfactory and allowed to infer that the: i) formulation of LASH in MATLAB® improved the river routing, adding a hydraulic structure to the model, and enabled the implementation and appraisal of scheme 2; ii) scheme 2 made it possible some discussions about both the heterogeneity of regions present in CRW and the impact of spatialization of λ and CS on hydrological modeling at CRW, such that this scheme can be considered as a powerful tool for the water resources management; iii) implementation of AMALGAM along with LASH allowed comparisons between it and the Shuffled Complex Evolution (SCE-UA), such that the former algorithm presented a slightly superior performance for the objective functions used; iv) use of AMALGAM has the potential to reduce the level of uncertainties and equifinality related to the LASH calibration parameters; and v) formulation of LASH in MATLAB® brought excellent benefits in terms of computational performance.