For farmers, every plant decision entails risks and many of those risks increase with climate change. One of the most consistent is the weather that can damage the yield of crop and resources of existence. For example, a delayed monsoon can force a rice farmer in South Asia to replant or change crops, lose both time and income.
Access to reliable, timely weather forecasts can help farmers to prepare for the coming weeks, find the best time to plant or determine how much fertilizer will be needed, resulting in better Wash revenues and lower costs.
Nevertheless, in many countries with low and middle income, accurate weather forecasts remain out of reach, limited by the high technological costs and infrastructure requirements of traditional prediction models.
A new wave of AI-driven weather forecast models can change that.
By using artificial intelligence, these models can provide accurate, localized predictions at a fraction of the calculation costs of conventional physics-based models. This makes it possible for national meteorological authorities in developing countries to offer farmers the timely, localized information about changing rainfall patterns that farmers need.
The challenge is to get this technology where it is needed.
Why AI now predicts things
The physics -based weather forecast models used by large meteorological centers around the world are powerful but expensive. They simulate atmospheric physics to predict weather conditions, but they require an expensive computer infrastructure. The costs set them out of reach for most developing countries.
Moreover, these models are mainly developed by and optimized for northern countries. They tend to concentrate on moderate, high -income areas and pay less attention to the tropics, where many countries with low and middle income are located.
A big shift in weather models started in 2022 As industry and university researchers developed Deep learning models that can generate accurate short and medium-sized predictions for locations around the world for two weeks ahead.
These models worked various orders of size faster at speeds than on physics -based models, and they could run on laptops instead of supercomputers. Newer models, such as I-OFATER And Tombhave linked or even better performed Leading physics -based systems for some predictions, such as temperature.
AI-driven models require less computing power than the traditional systems.
Although physics-based systems may need thousands of CPU hours to perform a single prediction cycle, modern AI models can do that Use a single GPU in a few minutes Once the model is trained. This is because the intensive part of the AI model training, which teaches relationships in the climate from data that can use learned relationships to produce a prediction without further extensive calculation – that is an important shortcut. The physics -based models, on the other hand, must calculate physics for each variable in every place and time for every prediction produced.
Although training these models out on physics-based model data, considerable investments require in advance, as soon as the AI is trained, the model can generate large ensemble forecasts-with multiple prediction runs at A fraction of the computer costs of models -based models.
Even the expensive step of training an AI weather model shows considerable computational savings. One study showed that the early model FourCastnet could be trained on a supercomputer within about an hour. That made his time to present a prediction thousands of times Faster than state-of-the-art, on physics-based models.
The result of all these progress: high -resolution predictions worldwide within a few seconds on a single laptop or desktop computer.
Research is also progressing rapidly to expand the use of AI for predicts weeks to months aheadWhat farmers helps to make plant choices. AI models are already being tested for improving extreme weather forecast, as for before Extratropic cyclones And abnormal rain.
Coordinating predictions for Real-World decisions
Although AI weather models offer impressive technical possibilities, they are not a plug-and-play solutions. Their impact depends on how well they are calibrated for local weather, benchmarkt against real agricultural conditions and tailored to the actual decisions that farmers should make, such as what and when they have to plant, or when drought is probably.
To unlock the full potential, AI prediction must be connected to the people whose decisions are intended to handle it.
That is why groups such as Strive for scaleA collaboration with which we work as Researchers in public policy And sustainabilityHelp Governments to develop AI tools that meet the needs in practice, including training users and tuning predictions with farmers’ needs. International development institutions and the world meteorological organization are also working on Expand access to AI prediction models In countries with a low and average income.
AI predictions can be tailored to context-specific farm needs, such as identifying optimum plant windows, predicting dry spells or planning pest control. Distributing those predictions via SMS messages, radio, expansion agents or mobile apps can then help to reach farmers who can benefit. This is especially the case when the messages themselves are constantly tested and improved to ensure that they meet the needs of the farmers.
A Recent study in India Discovered that when farmers there received more accurate monsoon prognoses, they better informed decisions about what and how much they had to plant – or not planting at all – in better investment results and reduced risks.
A new era in climate adjustment
AI -weather forecast has reached a crucial moment. Tools that were experimental only five years ago are now integrated into Government weather forecast systems. But technology alone will not change lives.
With support, countries with low and middle income can build up the capacity to generate, evaluate and act and provide valuable information to farmers who have long lacked the weather services.
Paul Winters is a professor in sustainable development on the University of Notre Dame.
Amir Jina is a assistant professor of public policy at the University of Chicago.
This article has been re -published from The conversation Under a Creative Commons license. Read the Original article.
#weather #forecast #reached #turning #point #valuable #farmers