Flood Area Prediction System: A New Feature of GISTDA’s Disaster Platform for Flood Depth Information and Preparedness

Flood Area Prediction System: A New Feature of GISTDA's Disaster Platform for Flood Depth Information and Preparedness

   The lessons learned from the major flooding crises in Chiang Rai, Chiang Mai, and Hat Yai over the last year have highlighted a critical issue in the information available to the public. Many people lacked access to detailed data on flood depth and the rate at which floodwaters were rising. As a result, some residents moved their belongings onto tables in an attempt to protect them, only to find that the water eventually submerged the entire first floor. These incidents underscore a critical reality — traditional warning systems alone are no longer sufficient.

   The Geo-informatics and Space Technology Development Agency (Public Organization), or GISTDA, has developed a flood area prediction model system. This is a new feature of the disaster platform to show depth data to people and transform post-event disaster reporting into predictive flood forecasting , helping people prepare and cope more efficiently.

   The Natural Disaster Management Division, GISTDA, has enhanced disaster management by developing the disaster platform (https://disaster.gistda.or.th) as a centralized disaster information platform , including flooding, drought, and air pollution. It is a tool that supports relevant government agencies, and people can view the overall situation efficiently and in an up-to-date manner. It serves as an important decision-support platform for national disaster prevention and mitigation.

   During the major flooding events in Chiang Rai and Chiang Mai in 2025, Even though flood warnings were issued at the sub-district level, they did not identify the boundaries of risk areas or potential damage in other areas. As a result, people and local government agencies did not have access to flood-depth data to make informed decisions, such as how high the water level would be at a specific location or how quickly the water level would rise. As a result, preparedness and emergency response efforts were unable to keep pace with the rapidly changing situation. with the actual situation. Due to the limitations of current SAR satellite observations, flooding in urban areas with many buildings cannot be effectively detected. Radar signals are reflected by buildings instead of the water surface. For more efficient coping and to reduce the limitations mentioned above, GISTDA has developed a flood area prediction model system and will conduct a pilot test in flood-prone areas of Hat Yai District, Songkhla Province, in late 2025.

   The background technology of the process mentioned above is the development of a flow condition model and a 3D flood representation system. Multiple geospatial datasets are integrated to develop the model, together with the Digital Elevation Model (DEM) — a significant factor in identifying high and low areas to define water flow. River networks, hydrological data, water levels, and flow rates in rivers are collected from designated streamflow monitoring stations in each area, along with flood area data from satellite imagery.

   The model is developed differently from conventional forecasting methods. It utilizes real-time water levels from key hydrological stations as primary data for simulation and forecasting — instead of relying only on rainfall prediction. This may reduce the high levels of uncertainty and error that occur in some areas, resulting in forecasts that more closely align with actual water flow conditions.

   Urban areas have complex geography and infrastructure. GISTDA utilizes drone technology (UAV) to conduct high-resolution topographic surveys in greater detail and generate a high-resolution DEM capable of accurately representing overland flow paths water flow routes in urban areas, including flows through roads, alleys, and other urban features, making risk area analysis and damage prediction more accurate.

   GISTDA also utilizes optical satellite imagery to validate the model when weather conditions are favorable, and the Earth’s surface can be clearly observed. It compares flood areas detected from satellite imagery with the results from the model to confirm the location, boundaries, and distribution of flooding. This helps improve the accuracy and efficiency of flood prediction for management systems and spatial planning.

   One of the key features of the system is that it can display a 3D map, allowing users to visualize simulated flood depths together with realistic 3D building models. Users can estimate how high floodwaters may reach at their location, enabling them to make more appropriate decisions on evacuation or moving belongings to safety during floods. GISTDA is currently testing this system in Chiang Mai and Nakhon Si Thammarat provinces.

   This system has features to meet the needs of professional users, such as spatial impact analysis, which can overlay population density data and agricultural areas for damage assessment. Historical flood records covering the past 12 years and can be utilized for city planning and long-term prevention. The system also provides various service formats, including web-based maps, downloadable files for further use, damage reports at the sub-district and district levels in PDF format, and API services to help other agencies connect and integrate the data into their own systems.

   The effectiveness of this approach was clearly demonstrated during the major flooding in Hat Yai last year. GISTDA used this model data to create a water-depth map in Hat Yai to support decision-making by local government agencies. Even though the system was still under development at that time, it proved to be very effective in real-world use.

   In the future, GISTDA will continue to expand the system to cover major flood-prone areas across Thailand. It will cover more than 100 communities and further improve the quality of the model and high-resolution DEM data in major cities, enabling more accurate and efficient flood prediction.

   Although no major flooding has occurred so far this year, the impacts of climate change are becoming more severe each year. The flood area prediction model system on the Disaster Platform is not only a solution for today; it is also an innovation that will help transform water management from reactive to proactive. With an accurate water-level forecasting system, authorities can better prepare before floodwaters arrive.