Written by Stephen Luntz
Compiler: Chen Zhaozhao
Image source: NASA/SDO
Flares like this can pose a threat to important technical systems, but now operators are given at least half an hour of warning.
Efforts to predict when solar eruptions will affect Earth are imprecise and frustrating, but NASA expects them to get a little better. An artificial intelligence system is now processing data from satellites to warn of solar storms powerful enough to damage vital infrastructure.
If you've ever wondered what you would do the day before a global catastrophe, you might want to cut your timeline short just to be prepared. It's not that NASA hopes solar storms will destroy Earth or anything so dramatic that it can now predict with a 30-minute warning, but they do involve the (small) risk of events that could put our civilization under a lot of stress.
Perhaps not surprisingly, scientists still have little ability to predict when a major solar storm will occur. In fact, our ability to predict the average activity level of a solar cycle is still poor. However, we expect that once storms occur, it will be easy to know if and when they will hit Earth's magnetic field, triggering geomagnetically induced currents (GICs). GICs can wreak havoc on anything long, thin, and metallic – including electrical technology, oil pipelines, and railroads.
This ability gives us ample opportunity to be ready in the 2-3 days it takes for charged particles associated with coronal mass ejections (CMEs) to cross Earth-Sun distances and affect our world. However, conversely, even after we witnessed a huge flare, the prediction hit rate was still very misguided. For example, we were blown away by the epic aurora on March 23 – one expected that night's aurora to be a gentle setting for the big events the next night.
Missing the miracle of the polar light can be sad, but it's a trifle compared to not being prepared for the consequences of an event that could crash the internet and disrupt the power grid, at least temporarily.
The pursuit of warnings of events like this led NASA, the U.S. Geological Survey and the U.S. Department of Energy to jointly develop a deep learning program to identify patterns of solar wind activity. THE DEEP LEARNING GEOMAGNETIC PERTURBATION COMPUTER MODEL, NOW NAMED DAGGER, IS THE RESULT OF THIS. DAGGER'S CAPABILITIES HAVE BEEN DESCRIBED IN A RECENT PAPER.
However, it depends on solar wind activity close to Earth, rather than providing only a 30-minute warning when it first leaves the sun, updated every minute.
Vishal Upendran of the Indian Centre for Interuniversity Astronomy and Astrophysics said in a statement: "With this AI, it is now possible to make fast and accurate global predictions and inform decisions in the event of solar storms, thereby minimizing – and even preventing – disruption to modern society."
THE AUTHORS VERIFIED DAGGER'S CAPABILITIES BY PROVIDING DATA FROM AUGUST 2011 AND MARCH 2015 PRIOR TO GEOMAGNETIC STORMS. IN BOTH CASES, DAGGER PREDICTED NOT ONLY THE IMPACT, BUT ALSO THE MAGNITUDE OF THE IMPACT AND WHERE IT WOULD BE FELT ON EARTH.
If you don't remember these events, it's because none of them were that catastrophic. HOWEVER, WE DON'T HAVE DETAILED SOLAR STORM DATA FROM 1989 – let alone 1859 – TO TEST HOW WELL DAGGER PERFORMS IN THE FACE OF A TRULY SEVERE OUTBURST FROM THE SUN.
However, the advantage of deep learning is that the more data it processes, the better its predictions become. BEFORE A BIG STORM, EVERY NON-THREATENING STORM WE EXPERIENCE IMPROVES DAGGER'S ABILITY TO PERFORM AT CRITICAL MOMENTS.
COMBINED WITH THE IMPROVED ABILITY TO OBSERVE THE SOLAR WIND AT ITS SOURCE, DAGGER COULD BE A STEPPING STONE TO A LONGER WARNING PERIOD. Perhaps by then, in addition to giving system operators a chance to put their charge into safe mode, the rest of us could also implement our doomsday party plan.
RESEARCH INTO DAGGER'S PERFORMANCE WAS PUBLISHED IN THE JOURNAL SPACE WEATHER.
Source:
https://www.iflscience.com/nasas-dagger-should-give-30-minutes-warning-of-catastrophic-solar-storms-68937
Written by Stephen Luntz
Stephen Lentz graduated from the University of Melbourne with degrees in Science (Physics) and Arts (English Literature and History and Philosophy of Science), followed by a Postgraduate Diploma in Science Communication from the Australian National University. In addition to designing exhibits for the Science Museum, the diploma includes performances for thousands of students in rural Australia about the science of flight and collision and the uses of liquid nitrogen. Before becoming a writer for IFLScience.com, Stephen wrote for Science Australia for 15 years. His column was shortlisted for the Queensland Prime Minister's Science Writing Award and formed the basis for the book Forensics, Fossils and Fruit Bats: A Field Guide for Australian Scientists. The book explores the careers and motivations of scientists who do a great deal of modern research, but they rarely receive attention to the few who have won major awards. On the other hand, he also interviewed three Nobel Prize winners for his articles.