To more quickly assess the risk of urban flooding during heavy rainfall, artificial intelligence (AI) and high-performance computing technologies will be utilized. The integration of advanced technologies, from high-resolution rainfall analysis to ultra-fast flood simulations and the use of closed-circuit television (CCTV) for flood situation recognition, is key to improving flood forecasting.
The Han River Flood Control Center, part of the Ministry of Climate, Energy and Environment, announced on August 10 that it will sign a memorandum of understanding with the Korea Institute of Science and Technology Information (KISTI) on August 11 at KISTI's headquarters in Yuseong-gu, Daejeon, to enhance the urban flood forecasting system.
Under this agreement, the two organizations will combine the Han River Flood Control Center's operational experience in urban flood forecasting with KISTI's AI and high-performance computing technologies. The goal is to apply and expand the use of advanced technologies developed in research to real-world urban flood forecasting and response.
Initially, the urban flood forecasting system of the Han River Flood Control Center will be linked with KISTI's problem-solving support platform. AI and high-performance computing infrastructure and intelligent technologies will be deployed on-site, and a collaborative system will be established for the demonstration and verification of urban flood response technologies.
KISTI has been developing technologies such as AI-based high-resolution rainfall analysis and ultra-fast simulations using high-performance computing. It has also developed technology to recognize flooding situations from CCTV footage using explainable artificial intelligence (XAI).
XAI is a technology that allows AI to explain the processes it used to arrive at specific results, making it easier for humans to understand. In fields like disaster response, where quick and accurate judgments are essential, this technology can assist on-site personnel in reviewing and utilizing AI analysis results.
The two organizations plan to gradually integrate these technologies into the urban flood forecasting system. They will verify prediction accuracy, stability, and on-site applicability using actual observational data and past heavy rainfall cases, thereby supporting the judgment of flood forecasters.
As localized heavy rainfall events have become more frequent, the importance of precise flood predictions and rapid situation assessments at the urban level has increased. Once AI and high-performance computing technologies are established on-site, they are expected to aid forecasters in analyzing various observational data and assessing flood risks.
The Han River Flood Control Center is currently operating urban flood forecasts in the Gangnam Station and Sindebang Station areas in accordance with the Urban River Basin Flood Damage Prevention Act.
Urban flood forecasting is a system that comprehensively analyzes real-time observational data, including rainfall and flood prediction results, as well as data from rivers, sewage systems, and road surfaces, to proactively alert about flood risks and quickly disseminate on-site conditions.
Kim Gu-beom, head of the Han River Flood Control Center, stated, "For urban flood forecasting to operate reliably on-site, accurate predictions, rapid situation recognition, and the professional judgment of flood forecasters must be organically linked. We will continue to verify the applicability of advanced scientific technologies and enhance the effectiveness of urban flood forecasting."
Lee Sik, head of KISTI, emphasized, "To respond to the climate crisis, it is crucial that science, technology, policy, and on-site operations are closely connected. We will actively collaborate to secure technologies and outcomes that can be utilized in real-world applications."
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.