The scope of **[artificial intelligence](https://noticiasambientales.com/ciencia/crean-una-manera-sostenible-de-monitorear-el-clima/)** is becoming increasingly surprising. This time, the famous technological tool will be part of **[a weather prediction system](https://noticiasambientales.com/ciencia/crean-una-manera-sostenible-de-monitorear-el-clima/)** promising more accurate forecasts and ten times faster, using less processing power.
According to the article by researchers from the University of Cambridge, in collaboration with the Alan Turing Institute, the European Centre for Medium-Range Weather Forecasts, and Microsoft Research, published in the journal Nature, the system, called Aardvark Weather, **will be able to make faster and more accurate weather predictions**.
Currently, weather forecasts are generated through a complex set of stages, which **take several hours to process** on supercomputers dedicated to this task. On the other hand, the daily use, development, implementation, and maintenance of these complex systems require a lot of time and a large number of experts dedicated to them.
As a result, several studies conducted by Microsoft, Google, and Huawei have shown that **it is possible to replace the numerical solver**, which calculates the evolution of the climate over time, with AI. This change results in quick and simple predictions, which is why it is already **being implemented at the European Centre for Medium-Range Weather Forecasts**.
## What is the new weather prediction model like?
With this new system, researchers chose to **replace the entire meteorological prediction process with the unique and simple machine learning model**, which incorporates satellite observations, weather stations, and other mechanisms that would help generate forecasts, both global and local.
Characterized by having a completely AI-based approach, **Aardvark offers the possibility of making predictions in just a few minutes**, although the most surprising thing is that it can be done from a desktop computer. In fact, with only 10% of the input data from existing systems, this discovery outperforms the United States weather system, which uses information from dozens of weather models, along with the analysis of different meteorologists.
## Aardvark, in full development
According to the researchers who designed Aardvark, the learning approach used by the system can be adapted to other more complex prediction problems, such as **tornadoes, hurricanes, and forest fires**. They even claim that it will also be able to measure air quality, ocean dynamics, and **[sea ice prediction](https://noticiasambientales.com/ciencia/un-nuevo-estudio-revela-la-relacion-del-deshielo-en-el-artico-y-los-ciclones/)**.
Furthermore, they also ensure that **this system can help transform weather prediction**, both in developed and developing countries.
## How can AI help in weather prediction?
AI algorithms, especially those based on machine learning and neural networks, can:
– **Perform much more accurate climate models:** these systems have the ability to analyze and process large amounts of data, such as satellite images, atmospheric measurements, and historical data, in real-time.
– **Detect trends:** thanks to AI neural networks, different complex patterns can be identified through large datasets.
– **Integrate multidimensional data:** AI is capable of combining different sources of information, such as satellites, ocean buoys, and weather stations, creating a holistic view of the climate system.
Source: CienciaPlus.
**[Do you already know our YouTube channel? Subscribe!](https://www.youtube.com/channel/UCrYLIWysAyv4m833zGAQ02g)**
