Weather forecasting is on the verge of becoming quicker, more cost-effective, and precise thanks to artificial intelligence (AI). This revolutionary technology might replace expensive supercomputers and change global forecasting, as emphasized by experts at a conference in Abu Dhabi.
Speaking exclusively to Khaleej Times during the AI for Weather Prediction: Advances, Challenges & Future Outlook conference, World Meteorological Organization (WMO) President, Dr. Abdulla Al Mandous, noted, “We are here because the UAE government, led by Sheikh Mansour Bin Zayed Al Nahyan, Vice-President, Deputy Prime Minister, and Chairman of the Presidential Court, urged us to leverage AI for enhancing weather forecasting.
“Traditional physical models are costly, and many countries with limited resources can’t afford them. We’ve explored AI’s accuracy, making this conference crucial for discussing advancements, challenges, and future prospects in weather prediction.”
Stay informed with the latest updates. Follow KT on WhatsApp Channels.
AI set to replace supercomputers
The National Centre of Meteorology (NCM) currently depends on supercomputers worth millions to operate traditional numerical weather prediction models.
AI-based forecasting tools, however, promise greater accessibility and affordability, particularly for developing nations.
“AI is expected to be more economical compared to traditional models that require supercomputers, manpower, and electricity. For instance, the NCM supercomputer costs over $10 million. Yet, if AI can offer precise numerical predictions, it will save substantial funds and empower many countries to forecast weather effectively for public safety.”
The expert clarified that global forecasting has progressed thanks to powerful institutions like the European Centre for Medium-Range Weather Forecasts (ECMWF) in Europe and the National Oceanic and Atmospheric Administration (NOAA) in the US, supported by satellite data and advanced computational systems.
Challenges ahead
Challenges remain. “Many countries lack the infrastructure to install weather stations or receive satellite data, leading to a significant data gap worldwide. Furthermore, the atmosphere’s heterogeneity can affect predictions, and sudden severe weather events can decrease forecast accuracy. Additionally, climate change poses a substantial challenge to producing reliable forecasts with physical models.”
Despite these challenges, AI is being successfully piloted in various countries. “Currently, AI is in its initial stages at WMO, but we have pilot projects in places like Norway and Malawi. Malawi, unable to afford costly supercomputers or expertise, can now produce three-day forecasts using computers under $5,000. This demonstrates AI’s potential to enable cost-effective forecasting globally.”
Looking forward, Mandous envisions integrating AI with traditional models for more robust hybrid systems. “Our goal is to develop a hybrid system combining physical numerical weather models with AI. AI can identify cases missed by physical models due to its knowledge of historical data. We aim to incorporate diverse data sources from other agencies into the models. Importantly, we seek international cooperation through WMO, collaborating with UN bodies and governmental organizations.”
He emphasized that AI could significantly enhance the accuracy and resilience of forecasts amidst climate change and severe weather.
“AI prediction models are cheaper, more accurate, and can detect sudden weather changes. Trained on such events, their predictive capacity surpasses that of physical models.”
