Predicting Carrington events before it's too late!

High-Level Project Summary

We have developed through the creation of a data model using information published by NASA of electromagnetic emission data from the sun. Our way to "solve" the challenge is to anticipate these events, being able to warn the population to minimize the consequences of an electromagnetic storm.This is possible because this radiation would take approximately 17 hours to affect the earth. It is therefore important to anticipate these events because it would help us to minimize the effects caused on our planet.

Detailed Project Description

First, we downloaded the data provided by NASA to generate a data model that allows us to detect anomalies in the sun's magnetic field emission in time. 


Once we have made the data model, the operation is the continuous injection of the readings made by the NASA SOHO probe by downloading this published data. 


In case of detecting some kind of anomaly, it would be visually perceived that something is happening, and would start the warning to governmental administrations to take the necessary measures to minimize the effects. The benefits obtained by these warnings is to avoid the economic loss that could be caused by the electronic breakdown of millions of electronic devices that would stop working due to these magnetic fluctuations. 


For all this we have used machine learning technology for the processing of the datasets provided. The main programming language used was Python. We have used Visual Studio Code and JupiterLab as IDE. 

Space Agency Data

We have extracted the data from this source provided for the challenge. From which we have extracted data to generate a data model to be able to represent and interpret the possibility of predicting a future geomagnetic storm that may affect the earth. 

Hackathon Journey

We found it a race against the clock, an incredible experience. We have learned that a data model involves a lot of processing hours and hardware power. 


On the human side, working as a team and managing time to meet the milestones. The complexity of the project to face a real challenge and to help us learn, which we believe we have achieved. Using advanced data science engineering techniques.


The challenges posed did not encourage us to carry out enough research to reach an optimal solution. And also the communication with other groups on the discord platform, as well as the help of the staff available for it. 


We would like to thank the organization for maintaining an ideal working environment and the colleagues from other groups who have assisted us to solve technical problems or doubts that we have been having. 

References

Tags

#CarringtonEvent #Madrid #Spain #2022 #MachineLearning