IBM and NASA develop a digital twin to predict the future solar energy storms

The most sun Complex puzzles can soon be solved thanks artificial intelligence. On August 20, IBM and Nassa The launch of Syria, a Foundation For the sun. After it has been trained in large data collections of solar activity, this artificial intelligence tool aims to deepen the understanding of humanity of solar weather and precisely predicting solar energy – of electromagnetic radiation emitted by our star who threatens both astronauts in orbit and Communication infrastructure on Earth.

Syria has been trained with nine years of data collected by the NASA solar energy observatory (SDO), a tool that prevents the sun since 2010, with high -resolution images every 12 seconds. SDO picks up sun notes in various wavelengths of the various electromagnetic electromagnetisms to estimate the temperature of the star layers. It also requires accurate measurements of the magnetic field of the sun – the data necessary to understand how energy moves through the star, and to predict solar storms.

Historically, the interpretation of this wide amount of various and complex data was a challenge to helicopters. To address this challenge, IBM says SURYA developers use SDO data to create a digital twin – a dynamic virtual identical version of the star that is updated when taking new data, which can be processed and studied more easily.

The process started to unify the various data formats fed by the model, allowing it to be processed constantly. After that, a long-range vision transformer-the AI ​​structure that allows detailed analysis of highly high-precision images and identify relationships between its components, regardless of its distance.

The performance of the model is improved using a mechanism called Special Gating, which reduces memory use by up to 5 percent by filtering noise in data, thus increasing the quality of the processed information.

More precise predictions in less time

Its developers say that this design gives Syria a great advantage: unlike other algorithms that require intensive signs of the data that are fed to them, Syria can learn directly from the initial data. This allows her to quickly adapt to various tasks and provide reliable results in less time.

During the test, Syria showed its diversity in combining data from other tools, such as Parker Solar Probe The soho, the soho, is another spacecraft that observes the sun. Syria has also proven that it is effective in various predictive functions, including predicting glow and solar wind speed.

According to IBM, traditional prediction models can only be predicted after one hour based on signs discovered in specific areas of the sun. On the contrary, “Syria presented two hours to use visual information. The model is believed to be the first to provide a warning of this type. In the early test of the model, the team said they achieved a 16 percent improvement in the accuracy of solar energy glow, a noticeable improvement in the current methods,” the company said in A. statement.

NASA confirms that although the model was designed to study the helicopter physics, its structure is adaptable to various fields, from planetary science to Earth’s monitoring. Kevin Murphy, director of data science at NASA, said in A. statement. “This model enables the broader understanding of how solar activity affects the critical systems and techniques on which we all depend here on Earth.”

The danger posed by abnormal solar activity is not simple. The main solar storm can directly affect global communications, electrical networks, collapse, disturb the mobility in GPS, satellite operations, internet connections, and radio transport.

Andres Monuz-Jarmilo, a solar energy physicist at the Southern West Research Institute in San Antonio, Texas, and a leader in the project, confirmed that Syria’s goal is to increase the main deadline for these possible scenarios. “We want to give the Earth the longest possible time. Our hope is that the model may teach all the critical processes behind the development of our star through time so that we can extract practical visions.”

This story was originally appeared on Wireless En español It was translated from Spanish.