Our sun is done through its energetic activity model every 11 years, but the technologies that scientists use to observe progress at a faster rate.
This is one of the take -out messages of a new study, which maintains that artificial intelligence (AI) can fill the growing gap between more recent and older solar data and help scientists discover the neglected aspects of the long -term evolution of our star.
The new generations of telescopes and solar instruments constantly offer unprecedented views the sun. These advances, which allow scientists to capture complex details solar eruptions And map the magnetic fields of the sun with increasing precision, are crucial to understanding its complex processes and driving new discoveries. However, the data collected by each new instrument, while offering superior quality, is often incompatible with the data of the oldest due to variations in resolution, calibration and quality, which makes it difficult to study how the sun evolves over the decades, supports the new study.
The new AI -based approach overcomes these limits by identifying models and relationships in data sets from different solar instruments and data types, translating them into a common normalized format. This provides scientists with a richer and more coherent archive of solar observations for their research, in particular for long -term analyzes of sundaclyRare events and studies that require combining data from several instruments, according to study authors.
In relation: The sun of the earth: facts on age, size and the history of the sun
“AI cannot replace observations, but that can help us make the most of the data we have already collected,” said Robert Jarolim, who develops advanced algorithms to treat statement. “It is the real power of this approach.”
The AI method developed by Jarolim and its team can mainly translate the observations from one instrument to another, even if these instruments never worked at the same time. This makes their data -based approach applicable to many sets of astrophysical imaging data, suggests the new study.
The team obtained this thanks to a two -step process involving neural networks, a type of automatic learning The algorithm modeled in a loose way on the human brain. First, a neural network takes high quality images of an instrument and simulates degraded images as if they had been taken by a different instrument of lower quality. This allows AI to learn the “damage” or the systematic differences introduced by the instruments.
A second neural network is then formed to take these artificially degraded images and “cancel” the degradation, which makes them look like new high quality images again. In doing so, he learns to correct the differences between the two instruments, according to the new study.
Once the AI has learned to “repair” artificially degraded images, the second neural network can be used to improve resolution and reduce noise in real low -quality images collected by older instruments in a way that does not deform or delete the real physical characteristics of the sun which were present in the original data.
This AI framework allows older data to effectively benefit from the capacities of new instruments, allowing scientists to provide less detailed historical observations to the quality of modern data, according to the press release.
“This project shows how modern IT can breathe new life into historical data,” said the co-author of the Tatiana Podladchikova study of the Skolkovo Institute of Sciences and Technology in Russia, in the same declaration. “Our work goes beyond the improvement of old images – it is a question of creating a universal language to study the evolution of the sun over time.”
The researchers applied this technique to the data collected by various space telescopes out of two Solar cyclescovering a little more than two decades. According to the declaration, the approach has improved the details of solar images with a complete disc, reduces the vagueness and distortion in the soil observations caused by atmospheric noise, and even the magnetic fields estimated on the other side of the sun.
The team also applied the method to a solar spot (Noaa 11106) which was followed for about a week in September 2010. According to the results, the AI produced “magnetic images” sharper and more detailed from the solar spot which allowed scientists to see its magnetic structure more effectively than with the original data collected by two years. Solar and helospheric observatoryA joint effort from NASA and the European Space Agency.
“In the end, we build a future where each observation, past or future, can speak the same scientific language,” Podladchikova said in the press release.
This research is described in a paper Posted on April 2 in the journal Nature Communications.