High-Level Project Summary
We used a Neural network called long-short term memory to predict the next Carrington event with High accuracy. We specially used this particular neural network because it is famous for solving Times series forecasting issues. Thats why we felt that it is a good fit for our project. This project is very important to prevent the next Carrington event by giving an early notice before the phenomenon happen to avoid lots of devastating consequences. We collected lots of variables which has a major effect in the solar wind behaviour and measurment with the help of datasets from multiple space agencies specially nasa related with some filtering and understanding of these variables.
Link to Final Project
Link to Project "Demo"
Detailed Project Description
we aspire to is not just to win, but to benefit, protect and preserve the universe
Therefore, we tried to do our best to come up with the best idea that could protect our planet, and we hope that our project has inspired you, even in a small way, to protect the planet the project was supposed to predict Geometric storm in a few hours with high accuracy it will give an earlier notice about when will the next Carrington event happen Which has devastating effects telecommunications in our planet by inducing a geomagnetic storm on earth And the earlier notice will give us time to take measures towards such devastating events It use a neural network model like LSTM or high board mother CML LSTM To predict total series forecasting on solving data the coding language is python and the tools are tensorflow – kenvas – pandas – sklearn – numpy.
Space Agency Data
I used Wind craft data from NASA's website to have a better picture on how to measure changes that promote the suspicion of a peak of solar wind velocity and other changes that play an important role in predicting the next carrington event.
Hackathon Journey
On 7th of august 2022, we started assembling the team. It was member\ Rewan idea. Then Ola agreed to join, and after that Ahmed, Sondos and Kareem joined the team, by the time we realised that we joined and enjoyed it too.
It was a bit confusing to choose the name of the team. Ola suggested naming the tame "ASKROY", Each letter is the beginning of a team member's name. Youssef was a member of the team but unfortunately, he couldn't join.
Then Rewan suggested changing the name of the team, Kareem suggested " sun tamers", and the others agreed, so, here we are, sun tamers!
We used to have at least a weekly meeting. We didn't know each other, only chance brought us together. Is it luck or fate? I don't know but it's a good thing for sure.
We were worried as it is our time to join the Hackathon. A mixture of worry and excitement.
we arrived in Cairo on 30 September, we were all from different governorates far from Cairo, it deserves.
The organization was very good and we got in easily, put on the T-shirt and entered the opening hall, we listened to the opening, it was amazing.
The volunteers showed us the location of the room where we were going to spend the hackathon, and after a while, we went to perform Friday prayers and then went back to work again.
At ten in the evening of the first day, we finished our work for the first day and went to sleep, and then went back to work at seven in the morning of the second day.
We submitted the project at eleven in the morning of the second day, and although we know perfectly well that we were not neglected and that we did our best, we felt the remorse and guilt accompanying every work that one cares for and in which he puts a lot of effort.
This experience was really useful and effective for us, we learned many skills and improved others, and we made new friends with kind people, Now we have a clearer view of what we want to be when we grow up, it was an irreplaceable experience, we hope it will happen again.
What inspired us to choose the Carrington event was our desire to do something outside of the things we had done previously, it seemed to us valuable to help save the planet, we are for the planet just five out of billions, and the possibility of five people saving billions like them, isn't it tempting?
Our approach to this project was not to provide a 100% correct imperative solution but rather a step forward that contributes to the ultimate solution to the problem.
We divided the tasks between the team members( writing-design-programming-scientific research)
We had some setbacks with time with the amount of data needed to get predictions with high accuracy
And the setbacks with time were treated by changing the mechanism on how we will solve our problem and what tools and neural networks will work faster and by doing more data exploration and filtering.
There are a lot of people we want to thank, our families, friends, the Nasa Agency, The AUC and all the organizers and sponsors.
We did our best and wish it is good enough to make us and judges and the public satisfied.
References
Resources (hardware & software, Data, references):
https://www.spaceappschallenge.org/space-apps-challenge-2022-example-resource-save-the-earth-from-another-carrington-event/
https://www.ngdc.noaa.gov/dscovr/portal/#/
https://cdaweb.gsfc.nasa.gov/pub/data/wind/mfi/mfi_h2/2022/
https://omniweb.gsfc.nasa.gov/ow.html
https://www.swpc.noaa.gov/sites/default/files/images/u33/DSCOVR-Instrumentation-Space-Weather-Week-April-2015_szabo.pdf
https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2020JA028228
https://agupubs.onlinelibrary.wiley.com/doi/full/10.1002/2017SW001764
https://www.frontiersin.org/articles/10.3389/fspas.2020.550874/full
https://royalsocietypublishing.org/doi/10.1098/rsta.2020.0097
https://www.kaggle.com/code/mineshjethva/forecasting-geomagnetic-storm-index/notebook
Tags
#Neural_Networks #Sun #Space #Particle_physics #LSTM