Early-warning systems are essential for mitigating the impacts of natural disasters by providing timely alerts that enable preparation and damage reduction. The United Nations Early Warnings for All Initiative aims to ensure global coverage by 2027, yet as of 2023, only 52% of nations have access, with least-developed countries and small island states significantly underserved. To bridge this gap, artificial intelligence (AI) is increasingly being utilized to enhance the efficiency, accuracy, and user-friendliness of these systems. AI applications include advanced weather forecasting, wildfire detection, flood mapping, and infrastructure monitoring. However, the integration of AI presents risks such as data biases and lack of interoperability due to the absence of international standards. Establishing globally agreed-upon best practices is crucial to ensure responsible and equitable deployment of AI in disaster management. Collaborative efforts by the International Telecommunication Union and partners are underway to develop these standards, promoting trustworthy AI solutions that enhance global resilience to natural hazards.
Early-warning systems are vital for reducing the impact of natural disasters by providing timely alerts that allow for preparation and damage mitigation. The United Nations Early Warnings for All Initiative aims to protect everyone with such systems by 2027. However, as of 2023, only 52% of nations have access, with least-developed countries and small island states at 46% and 39%, respectively.
To bridge this gap, researchers, the private sector, and governments are increasingly utilizing artificial intelligence (AI) technologies. AI can make early warnings more efficient, accurate, and user-friendly, while also addressing geographical disparities in disaster preparedness. Companies like Google DeepMind, Huawei, and Nvidia have developed AI-based weather forecasting models that outperform traditional tools in speed and precision. AI is also effective in predicting small-scale events like thunderstorms and tornadoes, enhancing the ability to issue timely warnings.
AI applications extend to wildfire detection, flood mapping, and infrastructure monitoring. Firms such as Pano AI and Fireball Information Technologies use AI to detect smoke from various sources, aiding in wildfire warnings. Satellite imagery combined with AI accurately maps flood extents, even those obscured by clouds. Additionally, AI helps monitor critical infrastructure, such as telecommunications and transportation systems, ensuring they remain functional during disasters. AI chatbots and translation tools, developed by the US National Weather Service and UNESCO, improve communication of warnings across different languages and regions.
Despite these advancements, the integration of AI in disaster management poses risks, including data biases favoring wealthier regions with more radar systems. Moreover, the lack of international standards for AI-infused disaster tools can lead to issues like incompatibility and non-interoperability, hindering continuous early-warning coverage across borders.
Establishing internationally agreed best practices is essential to mitigate these risks. Standards should cover data collection, algorithm training, testing, and usage to ensure responsible and trustworthy AI. The European Union’s AI Act categorizes AI in early-warning systems as ‘high risk,’ imposing strict regulations. However, globally comprehensive standards are still needed. The UNESCO Recommendation on the Ethics of Artificial Intelligence (2021) and the UN AI advisory body’s 2024 report “Governing AI for Humanity” provide foundational guidelines, but more work is required.
The International Telecommunication Union (ITU), in partnership with the World Meteorological Organization and the UN Environment Programme, has led the Focus Group on AI for Natural Disaster Management. From 2020 to 2024, this group brought together experts to explore AI’s potential and lay the groundwork for standards. Their efforts include a road map of existing standards, a glossary of terms, technical reports, workshops, and hackathons, addressing data interoperability, AI training, transparency, and trust.
Looking ahead, the focus group is transitioning to an ITU-led Global Initiative on Resilience to Natural Hazards through AI Solutions, launching in November. This initiative aims to identify new AI use cases, update technical reports, develop proof-of-concept studies, and enhance capacity sharing. Collaboration with the UN Framework Convention on Climate Change (UNFCCC) at the COP 29 conference will further integrate AI standards into global environmental strategies.
For AI-based early-warning systems to be effective and equitable, they must be trustworthy, interpretable, and transparent. Establishing robust international standards is crucial to ensure AI enhances global resilience to natural hazards, preventing disparities and protecting vulnerable populations worldwide.
Sources:
Nature
https://www.nature.com/articles/d41586-024-03149-z .
Provided by the IKCEST Disaster Risk Reduction Knowledge Service System
Comment list ( 0 )
