Assessment of Hydrological Database for Automated Flood-alert System—Case Study: Shonai River, Kasugai

Abstract

This study uses database infrastructure, hydrologic models, and geospatial analysis to assess flood inundation areas and their distribution in the Shonai sub-basin, Kasugai City, in western Japan. The proposed method utilizes near-real time data; hence, efficient database processing is an important requirement for the assessment tool. The river hydrological models utilize geometric input flow characteristics and river discharge volume to generate an inundation map. Discharge volume is estimated owing to the lack of data on discharge measurements. Distributions are presented in maps of inundation-susceptibility using an interactive web-based system. The simulated results from the hydrological model indicate that the topographic characteristics (slope and elevation) of the watershed significantly affect the depth of inundation. Hydrological factors (e.g., rainfall and water level) also significantly affect changes in the rate of river discharge. The centralized geo-database can be processed on-the-fly, and it is believed that this approach is reliable and well suited to such a flood alert system. The significant points of this work are to determine a necessary but unknown parameter (discharge volume) via existing data resources, and—ultimately—to make informative mapping data available in a timely manner. The use of a near-real time system ensures that the public and relevant agencies can assess emergency information and initiate appropriate and prompt responses. 