TEMPO.CO, Jakarta – The National Aeronautics and Space Administration’s (NASA) Temporal Artifact and Continuous Learning System (TACLS) project serves as the latest example of using artificial intelligence (AI) for flash flood prediction systems. WALHI’s Head of Urban Campaigns and Spatial Policy, Wahyu Eka Sutawan, said early warning systems that traditionally relied on sensors are now moving towards machine learning models, an essential step in disaster mitigation.
“This shift is strategically necessary to change the flood management paradigm from reactive to proactive, predictive and evidence-based,” he said. tempo Tuesday, July 7, 2026.
What are the benefits of AI-powered alarm systems?
Wahyu said traditional flood analysis has an accuracy of 70 to 80 percent. AI-based models can increase this accuracy to 90%.
Additionally, traditional systems take time to process data updates, leading to delayed response times. It is also believed that physical and empirical models are unable to capture the complex relationships between land cover changes, river dynamics, rainfall patterns, and the effects of climate change on flood potential.
He further said that artificial intelligence will provide earlier warning through modeling of water level rise that integrates various data sources such as satellite imagery, in-situ sensors and weather information. A longer warning period increases the likelihood that governments and communities will be able to prepare mitigation and evacuation measures.
Another advantage of AI in early warning systems is the ability to process large amounts of data simultaneously. Artificial intelligence can process Internet of Things (IoT) sensors, satellite imagery such as Sentinel-1 SAR, and historical weather data collected over many years.
In addition to enhancing predictive capabilities, explainable artificial intelligence (XAI) can also reveal factors that influence the likelihood of flooding, such as land use change, river sedimentation, and extreme rainfall intensity. Digital twin technology and hydrodynamic models can also handle different flood scenarios, supporting mitigation planning and evacuation routes.
Scope of AI utilization in flood warning systems
WALHI believes that the use of AI in disaster management should not be limited to time and location predictions, but should also extend to identifying factors that increase flood risk. With smart systems, lawmakers should be able to predict the effects of changes in forest cover, damage to river basins, changes in infiltration areas, expansion of mines and plantations, and development that ignores carrying capacity.
“If this technology is used only to speed up evacuations, it will only increase the country’s ability to adapt to disasters without solving the root of the problem. AI should not only generate early warnings, but also serve as the basis for policy evaluation, enforcement against environmental damage, and prevention of potentially harmful permits that increase disaster risk,” Wahyu said.
It has also been featured in tempoPremium report of: Bagaimana AI Menpredixi Banjir Levi Precisi (How AI More Accurately Predicts Floods), NASA’s TACLS machine learning was trained using 30 years of historical data from the Global Network of Satellites (GNSS). The results of that analysis were translated into a visual model that can be interpreted by climate researchers.
“The system is Enhance existing methods to reduce the time it takes for human analysts to determine whether to issue a flash flood warning. ” As announced by NASA.
The entire TACLS analysis process is expected to be rapid and run in near real-time. After being tested using initial data from past extreme weather events, specifically from the period 2017 to 2023, TACLS was proven to be able to predict 93% of flash flood incidents.
Will Indonesia introduce an AI-based early warning system?
In the United States, the National Weather Service has begun integrating TACLS into its flash flood prediction system in Southern California. The TACLS software, including training data, is provided as open source for use by other researchers.
Guswant, head of the Meteorology, Climatology and Geophysical Agency (BMKG), said the agency is also promoting the digitalization of its early flood warning system. However, the implementation of AI-based systems in Indonesia is not without challenges. “One of the challenges is data quality and availability,” he said.
He said system integration is also a major challenge. AI systems cannot operate independently. These need to be connected to systems used by the BMKG, the National Disaster Management Agency (BNPB), and local authorities. This synchronization is necessary so that the information generated is used consistently by all agencies involved in disaster management.
“Synchronization is needed to prevent AI systems from operating in isolation from existing systems,” he said.
AI and hydrological models also require strong computational support. These predictive models require high-capacity servers to process large amounts of data and run simulations quickly. Human resource capabilities will also determine the success of AI implementation in early flood warning systems.
read: Droughts and floods: How El Niño causes global extreme weather events
click here Get the latest news from Tempo on Google News
