Real-time wireless inundation monitoring and warning system


This project integrates people from hydrology and computer science for the researches of hydrological measurement, hydrological forecasting and wireless communication to generate the real-time inundation monitoring and forecasting system. And the short term (1-3hour) inundation prediction will be provided for early warning. The study area is located in Dianbao Creek. The low-lying lands in western Taiwan are always suffering from the inundation. It is difficult to simulate the range and depth of inundation due to the lack of complete water-level measurements. Especially, the inundation simulation model requires highly stability of calculation. Thus, it is not easy to forecast the real-time inundation. Thus, a high density of water-level monitoring network is necessary to reveal the real-time inundation. This project integrates wireless sensor network and water-level gauges to generate a synchronous inundation water-level measure network. For the hardware, this project is going to make the water-level gauge by ourselves. The wireless sensor network device will be attached to the water-level gauge to form the wireless water-level measure network and tested in the field. For the software, a communication testbed will build for seeking the balance among cost, power efficiency and measure quality to enhance the strength of the communication. 

This project integrates the real-time measurement of the inundation water-level measure network, QPESUMS radar rainfall and rainfall observation to build the inundation forecasting model. We will discover the relationship between the rainfall of upstream and the real-time water-level of downstream through support vector machine regression method. Thus, the project can provide the forecasting of real-time inundation water-level. Finally, we can provide all information for building the real-time water-level monitoring and forecasting system.