NASA is developing a system to predict solar storms half an hour before they hit Earth
Arabia Weather - Scientists often talk about the dangers of solar storms. Recently, interest in research related to specific types of artificial intelligence has increased, and this topic appears prominently in technological discussions.
A NASA team is currently working to integrate artificial intelligence models with solar storm data, with the goal of developing an early warning system that can provide about 30 minutes advance warning before a potential solar storm reaches a certain area.
NASA is developing a system to predict solar storms half an hour before they hit Earth
This time limit is related to the fact that light, part of which is radio signals, can travel more quickly than solar material emitted from the Sun during solar storms.
Power outages can last for hours in some cases, as happened in the incident that affected Quebec about 35 years ago. Extreme solar events, such as the Carrington event that occurred more than 150 years ago, could cause massive damage to electrical infrastructure and communications networks if they occurred today.
Despite these challenges, scientists did not sit idly by. Many satellites are used to monitor the sun and detect solar flares, and this monitoring provides NASA with the necessary information to predict the occurrence of solar storms and take preventive measures.
As any AI researcher will tell us, a predictive model must be told what to predict
Knowing there's an upcoming solar storm is only part of the equation; you also need to know what kind of impact to expect on Earth. So the researchers collected data from ground stations, some of which had already been affected by previous storms monitored by satellites.
Scientists have begun training a deep learning model, a term that is currently in vogue, called “dagger,” and which has reasonable specifications compared to existing predictive algorithms that have tried to achieve the same purpose, including increased speed.
According to the researchers' statements, the algorithm can predict the intensity and direction of a solar storm within seconds, and is able to make accurate forecasts periodically every minute.
Previous attempts using algorithms were time-consuming, getting to the point where they provided almost no warning time before a solar storm reached Earth. Part of the challenge was the difficulty of pinpointing where a storm might hit on any part of the Earth's surface.
This is a prominent feature of the Dagger system, as it performs rapid predictions for every surface of the Earth. By combining rapid forecasts with the ability to apply them on a planetary scale, Dagger is an important step toward improving prediction and accurate response to potential solar storm risks.
In due course, the system will release an open source platform to collect large amounts of data, as the 11-year solar cycle heads towards its peak in 2025. This will give service and communications companies enough time to integrate Dagger into warning systems before peak weather conditions occur. Solar.
There may not be alarms similar to the tornado warning sirens we find in the American Midwest, but at least the individuals involved will have a quicker understanding of the danger than they did in the past.
Read also:
Light pollution...a global threat to living organisms
Source: sciencealert
Browse on the official website