Operational Flood Forecasting As A Web-Service

The paper presents flood forecasting and simulation system applied to a river flood analysis and risk prediction. An advanced river flood monitoring, modelling and forecasting approach is introduced. It extends the traditional approach based on modelling of river physical processes by integration of different types of models and technologies such as input data clustering and filtering, digital maps of a relief and river terrain, data crowdsourcing, heterogeneous data processing, hydrological models for time scale modelling water flows and geo-simulation, inundation visualization and duly warning on flooding. The service-focused architecture of system of monitoring and anticipatory forecasting of floods and weak connectivity of program modules peculiar to it allows to convert developed system to a format of the "cloudy" application implemented as service. The paper presents the implementation using BPEL language implements the business process work hydrological model, and describes how to use the BPEL can be realized by using CASE representations continuous and independent operation of several different hydrological models and automatically provide the user with the results of an optimal model for the present time. Developed servicefocused architecture can be used for other infrastructure projects. Group of authors focuses on programmathematical and algorithmic support of research projects.