Performance of quantitative precipitation estimation using radar, raingauge and satellite in a hydrological application for the Iguassu Basin, Brazil

More than 90% of the electric power generation in Brazil is produced in hydropower plants, largely dependent on water availability in reservoirs and basins. Therefore evaluation of the quantitative precipitation estimation is fundamental to determine water volume available for power generation. To characterize and evaluate the precipitation in the region on many spatial and temporal scales over different regimes, a hydrometeorological system is used, comprising of a S-Band Doppler weather radar, satellite information and a network of 37 automatic hydrological stations (with hourly observations of precipitation and streamflow) and 38 meteorological stations (providing hourly observations of precipitation among other weather conditions), throughout Parana state, operated by SIMEPAR, as well as measurements from other raingauge networks, with a total of more than 100 points of measurements in an area of approximately 200,000km². A statistical objective analysis scheme (SOAS) was used to merge radar, raingauge and satellite. This approach integrates the rainfall estimates and measurements to improve the rainfall accumulation field such that the error of the analyzed rainfall field becomes less than the least error among gauge, radar and satellite. To study the amplitude and phase of the precipitation in river basins, spatial averages in each sub-basin areas were used, in order to characterize surface and river draining. The quantitative precipitation estimates were applied in a hydrological model to evaluate the identification of the meteorological systems and their impacts in the streamflow in the Iguassu basin. This paper presents the results of this evaluation, with a set of observations covering the spring and summer season of 2004/2005, 2005/2006 and 2006/2007. The results indicate that radar rainfall estimates provide better spatial and temporal resolution, while satellite estimates provide information in a larger area and raingauge measurements contribute to the adjustment of the amplitude of the precipitation estimation. The precipitation estimates applied to a hydrological model showed an improvement in the forecasting of streamflow variations, specially in situations of heavy precipitation, mostly due to the improvement of average precipitation estimation in the river basin.