Extreme temperatures, torrential rains, sudden drought. The climate change made the extreme weather events more and more frequent and difficult to predict. But if in rich countries there are infrastructures that can anticipate, at least in part, these events and activate emergency plans, in many areas of the world – especially in developing countries – the lack of early alert systems It continues to put millions of lives at risk.
A new study of theCambridge Universityin collaboration with theAlan Turing Institute, Microsoft Research and the European Center for Medium -term weather forecasts (ECMWF), proposes to exploit artificial intelligence to create a fast, reliable and above all accessible meteorological forecasting system.
The project is called Aardvark Weather And it is based on an artificial intelligence model which, according to researchers, is able to overcome traditional supercomputer in performance, while using only the calculation power of a common laptop. A revolution for the meteorological sector, this is the hope, which could democratize access to advanced weather forecast And save lives in the most exposed territories today.
In the preliminary version of the study-the peer-reviewed article has been accepted for publication on Nature but is still subject to final review-it reads that the AI-based system manages to process forecasts quickly and precision. This type of technology underline the authors of the study, could provide targeted local forecasts, such as those relating to extreme temperatures for agricultural crops in Africa or wind speed in European wind farms.
Although the study is still in the pre-publication phase and susceptible to changes during editorial production, the message is already clear: the Combination between machine learning and meteorological data It could radically change the way we foresee time.
Today, most detailed weather forecasts are based on complex and expensive numerical models, managed by supercomputer that perform millions of calculations in a few hours. These tools are available to a limited number of countries, those equipped with advanced technological resources and highly specialized staff.
On the contrary, Aardvark Weather wants to represent an opportunity to fill this gap. His strong point, the researchers know, lies in the lightness of the system, which allows use also in contexts where access to technology is limited. According to the authors of the research, each country – even the most isolated or not developed – could adopt this model to create an early alert system capable of anticipating natural disasters, from floods to hurricanes.
According to the data of theWorld Meteorological Organizationless than half of African countries has a working meteorological alert system. The lack of timely information often leads to tragic budgets in terms of human lives, destroyed harvests and damaged infrastructures. Aardvark Weather, with its scalability and low cost, wants to represent a historic turning point for these areas.
Of course, they still remain challenges to face. The integration of these systems with existing meteorological networks, the validation of data in real time and the formation of local staff are some of the aspects that must be deepened before a large -scale adoption.
View this post on Instagram
