Research On Rain Guage Network In Real-time Flood Forecasting System

For the real-time flood forecasting system, it is desired to determine the minimum number of ram gauging stations within the forecasting precision restriction. The density and distribution of rain gauging stations will directly influence the precision of flood forecasting, while the flood forecasting method will affect the layout of gauging stations network, the interrelationship between which has drawn attention during the determination of gauging stations network for a flood forecasting system. Especially for a real-time data collection system, with the increase of available data and the enhancement of technique measures as well as the capability of forecasting, real-time correction of system and self-adapted tracking method, research on the effect of real-time correction to density of rain gauging stations will not only decrease the density of gauging stations but also increase the reliability of gauging station determination.In this paper, it is studied by an ideal system the objective criteria for the design of rain gauging stations network which takes many aspects including flood volume, flood peak, number of certified floods and rain gauging stations etc.. The result shows that the objective function combined with number of gauging stations and the precision indexes bears explicit definition and is applicable, because it takes into account the forecasting precision, number of gauging stations and some other important conditions. Then the objective function is applied to five difference watersheds for design of rain gauging stations network, and is compared with result of other method. The optimized number of rain gauging stations determined by the objective function is equivalent to the gauging density formula of Jiangxi Province with error of area average rain e=15%, while the objective function is more widely suitable for no limitations in watershed area and characteristics. Moreover, the relationship between density of rain gauging stations and area of watershed. Finally, making use of the observed discharge information, real-time correcting method is brought in rainfall network design. AR model is used to revise the simulated process of forty floods in Qilijie basin in Min River .It is concluded from comparison that the flood forecast project with real-time correcting method is of higher precision and thus useful to reduce the rainfall network density .The same conclusion are obtained by using the method in other four watersheds. All the conclusions above are of great significant to rain gauge network designing in the future.