Data: 17/09/2014 à 19/09/2014
Local: São Paulo - Brazil
A Quality Analysis in Hydrological Forecasting. Lessons Learnt From the Basque Country Flood Early Warning System (Spain) (PAP014836)
Ignacio Villanueva Lacabrera, David Ocio Moreno, Angel Eraso Alberdi, Naiara Pérez Guruchet, Christian Stocker
Flood forecasting and early warning systems
Flood early warning systems are increasingly becoming a common tool to reduce flood risks in wide parts of a certain region or country. Their low investment needs, short implementation times and significant effectiveness make them a desirable protection measure in comparison with other structural solutions. Nevertheless, in order to fully exploit their potential there are certain requisites that must be fulfilled, such as the availability of accurate real-time hydro-meteorological data, the development of well-calibrated hydrological models, the existence of detailed meteorological forecasts, the use of data assimilation algorithms and the implementation of a reliable integration software that ensures a proper automation of tasks. Advanced systems are now including weather ensembles and radar nowcasting as a way to improve hydrological forecasts and take into account the related uncertainty. However, before getting into more complex calculations, a special attention should be paid to the current quality of the system in an effort to identify the elements that mainly contribute to its performance. Such an exercise has been made in the Basque Country Flood Early Warning System. The hydrology of the region has torrential features, both due the occurrence of very intense and persistent convective storms and due to the fact that rivers are short and with steeply slopes, which means they have small concentration times. Therefore, a correct rainfall observation, a reliable short-term rainfall forecast, the availability of adequate antecedent moisture conditions, a valid model response and a consistent data assimilation procedure are key factors that need to be analysed and cross validated. Here we present the results of this analysis and the alternatives proposed to overcome the detected pitfalls, which in overall will contribute to build a better simulation core before other developments area implemented. In addition, this new developments are also outlined and their potential effect on the system discussed.