Development of models for hydrological forecast from meteorological radar data in mountain basins

Two of the greatest challenges in mountain areas are the estimation of flows in non -monitored basins and the prognosis of sudden floods. In both cases the main obstacles are the measurement of rain and numerical modeling for the estimation of the flows. In March 2015, a weather radar was implemented in the boxes, with which rain maps are obtained. Thus, uncertainty is reduced to modeling. Due to the complexity of applying distributed models because they require long time series for calibration, an interesting alternative is little explored is the use of models based on other less demanding techniques. Therefore, the objective of this study is to develop models based on artificial intelligence for hydrological forecast from radar data. To achieve this, it will start from analyzing extreme events and knowing the abilities of the models to simulate them, for different progress (4-24 hours) and different income data conditions. Finally, to increase the time in advance of the prognosis, a quantitative precipitation forecast model that would allow to obtain a longer time forecast will be developed. Moreover, the development of the models, it is expected to achieve greater knowledge of the hydrometeorological processes that generate flows and to determine a models development strategy that can be replicated in other mountain basins with limited data and short -term time series.

OUT OF CALL – EXTERNAL FUNDS