Font Size: 

During typhoon approaching Taiwan, it tends to bring the threat of heavy rains and results in flood inundation and other compounded disasters in lowlands. Real time observations and forecasts are helpful to make appropriate decisions for disaster reduction and response under changing conditions. However, the procedures of data collection and model simulation are complicated and time-consuming to complete the task. Establishing an instantaneous and effective flood warning system becomes a vital issue. An operational information platform is developed in this study by linking various datasets with numerical models for flood inundation forecasting.
The framework of platform can be sketched as a process flow with four components, including data, model, display, and management. Data integration, model computing, data visualization, and task application are the major functions supported by cyberinfrastructure with high performance computing and high capacity data storage. Through adopting a common exchange mechanism, data from a variety of sources can be transferred and stored in the platform. For a flood inundation model, rainfall is one of the most important factors. Therefore, this operational platform regularly gathers the latest precipitation forecasts as input data over the internet. If the precipitation reaches a given threshold, numerical model will be executed to simulate the pluvial flood inundation for the next few hours. Then the simulation results containing the inundation extent and depth are converted into maps. The spatiotemporal changes of rainfall and flood inundation can be displayed and browsed through visualization tools. The whole process of inundation forecasting can be conducted at operational platform periodically and automatically. In addition, inundation maps can be overlaid with other spatial data, such as demographics, pump location, and transportation network. These data analysis and applications will assist decision maker in disaster assessment and resource allocation.
Based on actual examination, a lot of time and manpower can be saved by the utilization of high performance information platform. Thus, instant disaster warning information can be provided and a real time decision can be made. Furthermore, because the observational, forecasting, and simulated data are all stored in the platform, it is simple and useful to make a comparison between different circumstances. For example, a comparative analysis between archived pre-warning data and real disaster situation can help to improve future disaster prevention strategies.

operational platform; flood inundation forecasting; disaster prevention