$1b global package wants to boost weather services for farmers in impoverished regions
Announced at COP29 by the Agriculture Innovation Mission for Climate (AIM for Climate) – a joint initiative between the UAE and the Bill & Melinda Gates Foundation – the move aims to help national meteorological and hydrological services produce high-quality, farmer-centered forecasts using AI-supported weather forecasting.
The AIM for Scale plan is essentially to mobilise investments to improve both the dissemination and generation of high-quality forecasts that can help farmers adapt to climate change, which makes weather patterns less predictable.
Most of the funding will be disbursed by multilateral development banks to country governments in the form of concessional loans, following a process to assess the needs and capabilities of each country.
With support from the Asian Development Bank, the Inter-American Development Bank, and the World Bank, governments will get help to set up systems for the dissemination of weather information to farmers, either through new or existing advisory channels (including digital systems).
‘Transformative potential’ of AI-supported weather forecasting
The plan also aims to “democratise” access to AI-based forecasting techniques to improve forecasts in low- and middle-income countries.
AI models can produce forecasts of variables that farmers need and when they need them, while generating accurate forecasts at a fraction of the cost and computing power of physics-based methods.
For example, it is estimated that NVIDIA’s FourCastNet and Google DeepMind's GraphCast are around 100,000 times faster and more accurate for 1 to 10-day forecasts than physics-based models.
AI models can also be optimised to enable improved prediction of variables for adaptation decisions at key times in the agricultural calendar.
But these benefits remain as-yet untapped in low- and middle-income countries.
AIM for Scale is therefore supporting partnerships like those between the Mohamed Bin Zayed University of Artificial Intelligence in Abu Dhabi and the University of Chicago’s Human-Centered Weather Forecasts Initiative.
This project is creating AI-based 1 to 10-day forecasts for crops in 10 countries and designing a training course on AI-based techniques for meteorological services for the agriculture ministries in 30 low-and-middle-income countries.
It also designing advance AI models for subseasonal-to-seasonal (S2S) forecasts (ones that extend beyond a short-term weather forecast) with a specific focus on developing and testing agriculturally relevant forecast products for these countries.
The partnership will “push the research frontier to generate forecasts that meet farmers’ needs,” says Imara Salas, associate director at the University of Chacago’s Development Innovation Lab, and “build local capacity to sustain access to actionable weather services over time.”
Increasing resiliance and incomes
The World Meteorological Organisation secretary-general, Professor Celeste Saulo, says: “More and better data leads to better weather forecasts, early warning systems and climate information services for agriculture and other vital economic sectors. Closing basic data gaps will also help inform AI models.”
Saulo adds that “the agriculture sector is undoubtedly one of the most vulnerable sectors to climate variability and change. Additional partnerships are needed to ensure that farmers are involved in the coproduction of weather and climate services which will enhance resilience and adaptation in the agriculture sector.”
Inter-American Development Bank’s executive vice president Jordan Schwartz, adds: “Food insecurity and hunger have actually worsened in Latin America and the Caribbean over the past decade. Increasing agricultural productivity, and small farmers’ output in particular, will be a key part in reversing this trend.
“Providing more accurate and relevant weather forecasting to small farmers will improve decision making around planting, harvesting and fertilizer use, leading to higher incomes and poverty reduction.”
