UNCERTAINTY PRECIPITATION ASSESSMENT IN A HYDROLOGICAL MODEL AT THE COMBEIMA RIVER BASIN, Colombia

Abstract:

Prediction and simulation of hydroclimatological events such as rain, have become an Absolute necessity in the management processes of watershed systems, particularly as it relates to the assessment of water resources and risk management. Precipitation is considered as a trigger to natural phenomena such as landslides, avalanches and floods, which occur as a result of the nonlinear interaction of hydrological dynamics. For the study and analysis of precipitation there are technological tools such as hydrological modeling, characterized by the transformation of input variables, such as precipitation and evapotranspiration rate. Therefore, precipitation is one of the most important variables because the quality and distribution of water resources depends upon it, thus a better understanding of the uncertainties associated with it is required.

The precipitation in the tropics has a high variability at all spatial scales, from the microscale to the synoptic scale, as happens in the time scale (Poveda and Mejia 2004; Zawadzki, 1973). This space-time variability has implications for the modeling and simulation of storms, and in extreme flows. This fact, coupled with the hydrological models are calibrated setting, usually simulated flow against the flow observed using the recorded rainfall, which generates uncertainties.

The main goal of this work was evaluate the uncertainty associated with the precipitation variable performing multiple simulations of synthetic events both in space and time, using the distributed hydrological model TETIS (Velez et al, 2002; Frances et al, 2007). A case study at the Watershed Andean high of Combeima River, in the city of Ibague (Colombia), was used to assess the uncertainty associated with the daily scale simulations.