Effects of Spatial Patterns of Precipitation on the Hydrologic Response. An Evaluation with Observed Data in a Tropical River Basin for Hydropower Purposes

Characterization of response of a basin based on the spatial patterns of precipitation is relevant for the advancement of the understanding of natural processes, as well as for applied research. Knowledge of effects of spatial patterns of precipitation on the hydrologic response helps to understand how the spatial variability of precipitation propagates to the basin outlet and allows this knowledge to be involved in the development of improvement strategies for rainfall runoff models that currently exist. The Sogamoso river basin, located at northeast of Colombia, contributes to a reservoir with hydropower purposes. The company owner of the reservoir and hydropower plant, can take advantage from knowing how this variability is reflected in the basin, to determine actions to improve modelling of extreme events, which is a key factor in rainy periods where the probability of flow over the spillway is high.

In this paper we assess the influence of the spatial patterns of precipitation in a tropical basin, with simultaneous data of precipitation and discharge at the basin outlet. Indexes of spatial pattern of precipitation developed for higher latitudes in non-tropical regions were validate in this study. Also, we related these indexes with the hydrologic response. We implemented a hydrological model distributed at event scale, calibrated and spatially and temporally validated with real events. Patterns of spatial distribution of precipitation in each of the sub-basins were identified trough of indexes of ‘Spatial Moments’ of precipitation. We generated synthetic storms following this pattern of spatial distribution and we evaluated response of the basin using the hydrological model with these storms as rainfall input. Results allowed to identify the storm types, according to their spatial distribution, for which the response of the basin is more critical in terms of peak discharge time and duration. Patterns of precipitation distribution identified in basin, show a preponderance of unimodal rainfall located downstream of the basin`s centroid, which is related to the most critical responses. This result, allows us to conclude that in the lower area of the basin, the density of the precipitation station network must be increased, in such a way that these events can be identified and involved in the hydrological models implemented by the company, in order to improve the prediction of critical floods that could flows over the spillway.