Evaluación del método de múltiples hipótesis para la conceptualización de procesos hidrológicos en una cuenca de páramo andino con un modelo de lluvia-escorrentía

Abstract

Knowledge of precipitation-runoff processes of mountain basins is crucial in terms of water resources to assure the protection of ecosystem services. In this research, a multiple hypotheses modeling approach is presented, with the objective of identifying a structure and set of parameters that most closely approximate the behavior of an Andean Páramo basin. For this, six structures with different complexity were conceptualized to simulate the hydrological response of the system. Mathematical functions were used to define the transfer of rain water to discharge passing through “compartments” within the basin. Model parameters were calibrated with the NSGA-II genetic algorithm. Based on the results, it was demonstrated that the most complex structure VI (ten parameters), is the best representation of the system with a high performance in validation (NSE=0.86 and logNSE= 0.78). Structure V with six parameters was also able to capture the hydrological processes successfully (NSE=0.83 y logNSE= 0.83). The main contribution to water yield for structures (VI and V) during the rainiest season comes from the lateral water flow, provided by the organic and mineral reservoirs. Although the addition of a fractured rock reservoir generated improvements in the goodness-of-fit parameters of the model, these improvements were not major. Finally, through this study, it can be assured that to model Páramo hydrology, it is necessary to include fast and slow response reservoirs.