High-Level Project Summary
Did you know that suddenly every electronic display and even energy available on earth could be shut down from night to day? That’s one of the devastating effects that another solar event, like the Carrington Event of 1859, could have on today's society. No one wants that to happen, right? That’s why that to prevent that from happening, the Schrödinger’s Cats team proposed to use Machine Learning, a kind of Artificial intelligence, to predict when such an event could happen, helping scientists and politicians make better decisions to protect society’s safety and economics based on this data.
Link to Final Project
Link to Project "Demo"
Detailed Project Description
The heat of the Sun on Earth is essential for life, but, sometimes, this heat can be quite dangerous for us. On August 28, 1859, several spots were observed in the sun. But what are sunspots? Sunspots are large areas of reduced temperature on the surface of the Sun. They are formed in regions where there are intense magnetic activities and act as a warning of big solar eruptions that can throw a large amount of gas and energized particles. But what happens if these particles come directly to our planet? The particles are deflected by the Earth's magnetic field and end up hitting an atmosphere in the polar regions, forming auroras. However, the impact of these particles compacts and distorts our magnetic field. This fact usually doesn't have major implications, but during the most intense storms, it can have big ones. Here we can already understand that particle flow alters our magnetic field, leaving our planet defenseless.
The first major solar flare was seen on September 1st, by Richard Christopher Carrington and Rich Hodgson in 1859. The earth was unprotected because of the first particle streams that hit two days before. That eruption generated one of the largest geomagnetic storms ever registered, which caused the aurora to be seen almost all over the planet, even in tropical regions. Named the Carrington Event, this solar storm caused people to wake up at dawn and see the sky lit up by polar auroras almost until they started to prepare breakfast. On that date, even though few electrical systems were used, devices began to collapse, poles began to spark and telegraphs had their functioning impaired.
This way, because of the few electrical systems, there was not so much damage. However, today we are surrounded by devices that only work because of electricity, and almost all things we do involve some kind and to some extent technology and the internet.
To prevent this kind of event from happening nowadays and a solar storm from damaging our satellites, GPS, and mediums of communication besides millionaire losses, our group developed a machine learning algorithm to correctly track changes in the peak speed of the solar wind and to provide early warning of the next potential Carrington-like event.
Further details of the code we created can be found on our GitHub page!
Space Agency Data
In order to train and model our machine learning algorithm to help track changes in the peak solar wind speed and provide an early warning of the next potential Carrington-like event, we realized we need to get some data that were related to that and that would make it possible to make a model. In the first step, established that we were capable to create machine learning to realize the relation between the Wind magnetic field and the DSCOVR magnetic field plus Wind ion parameters. And this relation would help to reach the principal objective. So, we search for and took these data in NASA Resources. For that, we used the Coordinated Data Analysis Web (CDAWeb) directories to get all of them. In summary, we get:
- BW(t) - Wind magnetic field (wind/mfi/mfi_h2/2021)
- BD(t) - DSCOVR (dscovr/h0/mag/2021)
- Wind ion parameters (density, velocity, thermal velocity) (wind/swe/swe_h1/2021)
And we just consider the first 6 months of 2021.
Used data:
https://cdaweb.gsfc.nasa.gov/pub/data/wind/mfi/mfi_h2/2021
Hackathon Journey
Our motivation for solving this challenge was the terrible damage that phenomena like the Carrington Event could cause us today affecting society directly, which is why we are presenting a solution that will allow early detection of changes in peak solar wind speed and provide early warning of the next potential Carrington-like event. As young science students, we decided to try to solve this problem in a simple and concise way, using all the knowledge we had previously and trying to acquire different ways of innovating. During the development of the project, we learned different programming features in machine learning and data collection, in addition to developing great teamwork. The process became fun and enjoyable, after all, working with things we like and helping the world at the same time is something we all seek to do. At first, we had difficulty collecting data, but we were looking for solutions to this problem throughout the days. Therefore, a union of things and people made us get here and complete our challenge. We would like to thank the challenge’s mentors, who helped and motivated us during the project.
References
Written References:
Solar phenomena. (2022, August 8). In Wikipedia. https://en.wikipedia.org/wiki/Solar_phenomena
Coronal mass ejection. (2022, September 30). In Wikipedia. https://en.wikipedia.org/wiki/Coronal_mass_ejection
Stellar corona. (2022, September 24). In Wikipedia. https://en.wikipedia.org/wiki/Stellar_corona
Carrington Event: The biggest solar event ever registered. (2021, November 18). On the Olhar Digital website. https://olhardigital.com.br/2021/11/18/colunistas/evento-de-carrington-a-maior-tempestade-solar-ja-registrada/
Could a solar storm destroy the earth? (2022, September 1). In Socientifica website. https://socientifica.com.br/uma-tempestade-solar-poderia-destruir-a-terra/
Carrington Event: the most impressive event in solar physics. (2021, February 3). On the Portal Fei website. https://portal.fei.edu.br/noticia/237/evento-de-carrington-o-acontecimento-mais-impressionante-na-fisica-solar
Images and Video Frames:
Model of 1859 Carrington Event. On the NASA website: http://www.nasa.gov/content/goddard/model-of-1859-carrington-event
Music:
Gaming Room by EnjoyMusic at https://enjoymusic.ai
Tags
#safety #space #security #sun #solar #event #phenomena #internet #connection #machinelearning #artificial intelligence #algorithm #planet #code #carringtonevent #solarstorm #geomagnetic #sun #earthmagneticfield