Enough problems already exist on this planet, so no new problems need arise from the Sun. Unfortunately, this merciless star cannot be destroyed yet. But NASA may at least soon be able to let us know when his one of its murderous flares will throw the Earth system into chaos.
Understanding and predicting space weather is a big part of NASA’s work. There is no air there, so no one can hear you yelling, “Wow, what about this radiation!” It therefore relies on a series of satellites to detect and relay this vital data.
One such measurement is the solar wind, the “continuous flow of matter from the sun.” Even NASA can’t say good things about it! Normally this current is absorbed or dissipated by the magnetosphere, but in the event of a solar storm it can become powerful enough to overwhelm local defenses. .
When this happens, electronics can be set to fritz because these charged particles can flip bits and corrupt volatile memory such as RAM and solid-state storage.NASA says it was blown up during the Carrington Event of 1859, the largest solar storm on record, and even the telegraph station was unsafe.
We can’t stop these stellar events from happening, but if we knew they were coming, we might be better prepared. We are already here. But how can we predict such rare and chaotic events?
NASA’s SOHO satellite is overwhelmed by the 2003 solar storm. Image credit: NASA
A joint Department of Energy project of NASA, the US Geological Survey, and the Frontier Development Lab is investigating this question, and the answer is exactly what you’d expect: machine learning.
The team collected data on solar flares from multiple satellites monitoring the sun, as well as data from ground stations monitoring geomagnetic disturbances (called perturbations) that impact technology. bottom. The deep learning model they designed identifies patterns in how the former leads to the latter, and the resulting system he calls DAGGER. D.UmaNinjaGGelectromagnetic wave ppicturertuR.felt.
Yes it is a stretch. But it seems to work.
Using geomagnetic storms that hit Earth in 2011 and 2015 as test data, the team found that DAGGER was able to quickly and accurately predict the effects of geomagnetic storms across the globe. This combines the strengths of the previous approaches and avoids the weaknesses. NASA said:
Previous forecasting models used AI to create local geomagnetic forecasts for specific locations on Earth. Other models without AI provided less timely global predictions. DAGGER is the first to combine AI rapid analytics with real-world measurements from space and across the planet to generate fast, accurate and frequently updated predictions for sites around the world.
It may take a while before you see a solar alert on your phone telling you to stop or your car is stuck (this will never actually happen…probably). It may shut down unexpectedly. A few minutes of warning is better than nothing!
By the way, the paper describing the DAGGER model can be read open source in this issue of the journal Space Weather.
