88-Monitoring Network Optimization for Flood Forecasting and Warning Purposes.

Abstract

The hydrological-hydraulic risk mitigation as well as the flood forecasting and warning activity require the knowledge of the main hydrological quantities and, hence, cannot prescind from observations. The distributed or semidistributed hydrological models need detailed measurements and the new technologies (radar, remote sensing), providing spatially distributed information, have to be validated through ground acquisition. Therefore, a reliable hydrometeorological network is fundamental for Civil Protection activities, particularly for flood-prone areas where significant flooding events occurred in the past. The aim of the paper is to identify reliable procedures for streamflow, rainfall and thermometric networks optimization considering as case study the hydrometeorological network operating in an inlet Italian region (8400 km2). As regards the streamgauge network, different criteria were adopted. At first, the real-time monitoring of each tributary with an area greater than 100 km2 was guaranteed. Then, further stage control locations were defined considering the historical flood damages and, finally, the monitoring of incoming floods and releases from artificial reservoirs was finalized. The raingauge network optimization was investigated considering the precipitation spatial variability for different temporal aggregation scales. Using monthly data, geostatistical techniques allowed to identify the areas where the precipitation field was poorly represented by the existing network which, hence, was densified and checked at the daily time scale. Finally, the thermometric network was analyzed through the Kriging approach with external drift for considering the variability with altitude and redundant stations were identified using a cross-correlation analysis at hourly time scale. The proposed procedures provided a suitable and modular proposal for monitoring network optimization.