Revolutionizing Weather Forecasting for Farmers
About the partnership
MBZUAI partners with the
AI for Climate (AICE) initiative at the
University of Chicago to advance state-of-the-art forecasting research and democratize access to high-quality weather forecasts in low- and middle-income countries (LMICs). Working in close collaboration with partners in LMICs and the private sector, we aim to develop tailored, farmer-centered, AI-driven weather forecasts, with a specific focus on agriculturally-relevant forecast products.
Our goal is not only to contribute to revolutionizing weather forecasting with cutting-edge AI research but also to ensure the adoption of these tools. We are designing a training program to transfer skills and build capacity within national meteorological and hydrological services (NMHS) and agricultural institutions. We aim to engage with the operational forecasting community, enabling them to produce cost-effective, regionally tailored weather forecasts based on their own data and meteorological expertise.
This initiative is part of the Weather Forecasts for Farmers innovation package under
AIM for Scale, an international initiative to mobilise investment to increase access to accurate, farmer-relevant, and timely weather and seasonal forecasts.
Project components
AI-based weather forecasting
We are developing and operationalizing AI-based 1–10 day weather forecasts for important crops in low- and middle-income countries. Our work includes:
- Identifying key variables relevant to AI model customization, such as optimizing agricultural decisions and increasing resilience to extreme heat.
- Determining the needs for nested models of precipitation and other weather variables.
- Examining the spatial variability of predictions at different resolutions and identifying areas where new observations might be needed.
- Evaluating the accuracy of tailored models based on ground truth data
Capacity building
To ensure sustainability, local relevance, and long-term impact, our efforts aim to build capacity within NMHS and agricultural institutions staff in 30 low- and middle-income countries. Our teams are designing and implementing a tailored training program on generating AI-based weather forecasts. Our goal is to institutionalize a training curriculum and program to equip local experts to generate and disseminate AI-driven weather forecasts using their own data and expertise.
Advancing AI Forecasting Research
The third component focuses on advancing state-of-the-art AI weather forecasting methods. This work includes conducting research on the forefront of the field, for example, on generating probabilistic AI forecasts at longer scales (subseasonal to seasonal “S2S” forecasting). We will work in partnership with other institutions and the private sector to develop and test agriculturally-relevant forecast products and create the spaces to build awareness, foster trust in these methods, and guide the research agenda