India is currently exploring the utilization of artificial intelligence (AI) to improve environmental designs and enhance weather forecasting accuracy. A notable climate official has underscored the importance of this endeavor, particularly in response to the escalating incidents of heavy rainfall, floods, and droughts across the nation.
Over the past few years, India has experienced a surge in extreme weather conditions due to challenges within the climate system linked to global warming. Reports from the Centre for Science and Environment indicate that these occurrences resulted in nearly 3,000 fatalities in 2023 alone.
On a global scale, climate organizations are increasingly turning to AI for its potential to streamline costs and boost forecast precision. Supported by Google, a model has exhibited superior performance compared to conventional methods, prompting the British Met Office to acknowledge AI’s capability to “revolutionize” weather forecasting.
The Role of AI in Weather Forecasting Advancements in India
Accurate weather predictions are of utmost importance for India, given its massive population of 1.4 billion, a significant portion of whom live in poverty. As the world’s second-largest producer of staple crops like corn, wheat, and sugar, India relies heavily on dependable weather forecasts.
Presently, the India Meteorological Department (IMD) relies on supercomputer-driven mathematical models for their forecasts. By integrating AI into their extensive research network, the IMD aims to produce higher-quality predictions at a reduced cost.
K.S. Hosalikar, the head of climate research and services at IMD, revealed to Reuters that the agency is actively developing AI-based models and advisories to enhance forecast accuracy.
IMD has already leveraged AI for issuing alerts related to diseases and heatwaves. To enhance data collection for more refined modeling, there are plans to expand wind observatories down to the town level, as mentioned by Hosalikar.
Similarly, the U.S. government has expressed interest in combining AI with traditional forecasting models, establishing a dedicated center to explore this integration through workshops and conferences.
Saurabh Rathore, an associate professor at the Indian Institute of Technology in Delhi, has highlighted the cost-efficiency of AI models. These models can operate effectively on standard desktop computers without the need for expensive supercomputers, unlike traditional methods.
Experts emphasize the significance of increased data availability to fully leverage AI capabilities. Parthasarathi Mukhopadhyay from the Indian Institute of Tropical Meteorology stressed the necessity of high-resolution spatial and temporal data for AI models to enhance the accuracy of location-specific forecasts.
Global Perspectives on AI Applications in Wind Forecasting
China has also embraced AI in weather forecasting, particularly in predicting climate-related disasters such as heatwaves, heavy rainfall, and thunderstorms. The Shanghai Artificial Intelligence Laboratory collaborated with China’s National Meteorological Center to develop the AI-powered Fengwu weather model, enhancing storm predictions.
Comparative analyses with existing models like the European Centre for Medium-Range Weather Forecasts (ECMWF) and the US National Centers for Environmental Prediction (NCEP) have demonstrated Fengwu’s superior accuracy in forecasting typhoons such as Doksuri and Khanun.
Released in April by Chinese institutions, the Fengwu model employs advanced bidirectional and deep learning technologies to provide high-resolution meteorological forecasts for up to ten days. This model can generate global climate projections for the next ten days in just 30 seconds, a significant advancement over traditional supercomputer-dependent models.
While AI shows promise in climate prediction, further developments are deemed necessary to extend forecast timelines to the street level, aiming for district-level accuracy.
The AI-powered Fengwu model can deliver 10-day weather forecasts in just 10 seconds.
Innovative AI models like Fengwu and Pangu Weather, developed in China, are enhancing the effectiveness of climate modeling. These AI models are anticipated to complement traditional physical models, offering valuable insights for industries, disaster prevention, energy security, and Earth science research.
Although conventional numerical weather prediction methods are effective, their progress is hindered by computational constraints and model complexity. AI-based prediction strategies are viewed as a cost-effective solution to overcome these challenges.
Pangu Weather, utilizing a 3D neural network and historical formation strategy, processes intricate weather data. Developed by Huawei Cloud and featured in Nature, this model surpasses some European and American weather centers and is accessible online to users worldwide.
Recent successes in typhoon predictions showcase the capabilities of Fengwu and Pangu models. China’s Central Meteorological Observatory plans to further integrate AI into weather forecasting for enhanced international services, fostering collaborations among universities and research institutions.
Private Sector Engagement in AI Weather Services
Businesses across Asia, including Thailand and Vietnam, are increasingly leveraging AI to safeguard customers from weather-related disasters. Following a damaging flash flood in 2021, an electronics shop in Thailand’s Bangpoo industrial park subscribed to Weathernews’ captain estimates service, providing real-time forecasts that significantly enhance the Thai Meteorological Department’s regional predictions.
Weathernews’ AI system collects and analyzes data to deliver precise forecasts, enabling businesses to take proactive measures in response to impending weather changes. Collaborations with local authorities to install radar systems aim to elevate projection accuracy in Thailand.
Weathernews expanded its AI-based modeling service to Thailand in March and Vietnam in June, becoming the first in Asia to offer such products. With plans to increase clientele and revenue in these regions, the company aims to address disaster management challenges prevalent in densely populated Asian countries.
Chihito Kusabiraki, the company’s president, envisions a significant increase in export revenue and emphasizes the importance of advanced forecasting services in Asia, a region prone to natural disasters. The Asian Development Bank reports that developing Asia accounted for a substantial portion of global disaster victims and financial losses from natural disasters, underscoring the critical need for advanced forecasting services.
In 2021, weather and water-related incidents in Asia resulted in $35.6 billion in damages. The Philippines, ranked high in the WorldRiskIndex for crisis vulnerability, highlights the region’s susceptibility to such events.
Companies like Atmo and Ninecosmos are developing tailored weather solutions in Asia, focusing on early warning systems for storms and hurricanes and providing transport route advice based on climate conditions.
Tailoring cost-effective solutions for emerging Asian markets is crucial for the success of weather-related services in the region, ensuring accessibility and relevance to diverse economic contexts.
