Selección de eventos de lluvia para la calibración de un modelo lluvia-escorrentía en SWMM

Abstract

Hydrological modeling is an indispensable tool for optimal water resource management, whose performance and assertiveness depend on the model construction process, where the importance of the calibration and validation phases is highlighted. This research seeks to highlight the importance of carrying out a selection stage of the rainfall events to be implemented in the rainfall-runoff model SWMM with the purpose of guiding modelers in the development of this stage and in the construction of their hydrological model, based on the compilation and analysis of the modeling criteria taken by several researchers about this process. For the development of the research, a compilation of studies was carried out where rainfall series with varied duration and intensity characteristics were implemented, as well as the selection of a determined number of rainfall events for the calibration process. Among the main results obtained in this research, it is found that in a hydrological modeling, more events are used to calibrate than to validate, and these events are characterized by having high intensities. As for their duration, discrete events of short duration are used for calibration, while for the validation process, precipitation time series are implemented, with which the behavior of the model is observed in the face of intense and extreme downpours, as well as in different periods of drought or scarce rainfall. It is also important to verify the status of precipitation information and its measurement. It is recommended to complete missing data by implementing regression analysis and correlation techniques. Likewise, it is recommended to implement series that do not have significant gaps, preferably with gaps of less than 10% of the series. The measurement of rainfall events and the hydrological response of the basin have a significant influence on the quality of the model and its accuracy in the simulation, which is reflected in increased calibration time and number of iterations. This is still a problem that is notorious in some regions where there is a lack of instrumentation in watersheds and urban basins to improve water resource management. Although sensitivity analyses are fundamental for an optimal calibration process, everything starts from the response generated by the model with a rainfall event, making this variable of great importance in hydrological modeling and model efficiency.