High-Level Project Summary
In 1859, British astronomer Richard Carrington saw a blast of white light on the surface of the sun. This was the Carrington Event, as scientists now call it, and it is the largest recorded solar storm ever recorded. It nearly shook the world with considerable destruction, which included sparking and fires in telegraphic stations, some of the telescopes melted and produced powerful auroral displays that were reported worldwide.Solar storms constantly occur every 27 days, and there's a high probability that another Carrington Event can happen sooner or later.Is Earth ready for another Carrington Event?If the Carrington Event, the biggest solar storm ever noted, occurred right now.
Link to Final Project
Link to Project "Demo"
Detailed Project Description
The Challenge requires the participants to develop a machine learning (ML) algorithm or neural network pipeline for the DSCOVR spacecraft's FC instrument to track and follow the changes in the peak solar wind speed so that DSCOVR can continue to provide early warnings of the next potential Carrington-like event.
Cybit understands that this is a real problem and has collaborated with its team to come up with ,havea solution for it.
We gathered useful datasets of solar and geomagnetic storms of different intensities. They processed them and considered the data with their attributes. The attributes included density, speed, temperature and more.
After the data was finely processed the results were plotted on the basis of model solar winds with respect to the attributes.
The Charts which are plotted via dataset are essentially significant for the visualization, of dataset and to understand that factors like density and temperature , a prominent role in the frequency of speed to the relation of density and temperature and the plots are telling us when the speed is high and low.
The model which works behind all this logic is Regression, Ridge and Losso algorithm.
The data of solar winds achieved from the instuments is first cleansed with garbage data, further processed and analyzed with respect to considerate factors. Linear Regression, Ridge and Lasso with are applied and followed for it which gives us the perfect solution to save the earth from next Carrington Event.
Space Agency Data
For this challenge, we have used 'NASA and NOAA Satellites Solar-Wind Dataset from Kaggle. we used this dataset because it has all the labels, satellite position and sunspots and solar winds. the length of data was vast and it was for us to apply algos on it data was in csv format which gave us more independence to manipulate iit have all the required field that can be
usefull for the modellinhg and analysis and luckily it gave us just the result that we needed.
The data is composed of solar wind measurements collected from two satellites: NASA's Advanced Composition Explorer (ACE) and NOAA's Deep Space Climate Observatory (DSCOVR).
But for our work as we are instructed we have used data from DSCOVR .
In the given dataset we got many attributes that got to be dealt in different ways, but first we take our numeric attribute and plot them to understand our data set we considered attributes like "bx_gse", "bx_gsm", "bt", "density", "speed" and "temperature"
Hackathon Journey
"Being a student is easy. Learning requires actual work" - William Crawford
On a fine Sunday, we went to college for our extra class. We were excited to know and learn something new today and fortunately, we weren't disappointed. At the end of our classes, one of our classmates announced to all of us about the NASA Space App Challenge.
our team decided to participate in it. We made our team, Cybit, and started our pre-research for the challenge as they say knowledge is power.
We knew that NASAA has sponsored in our city and this event will be properly addressed.
In our city, Karachi, Salim Habib University sponsored this event. Our team went there on 1st October and observed what we were being told and guided.
We really admired NASA's efforts to save Earth which is remarkably admirable.
We came home from the event and started working on our challenge as we really got inspired by NASA's mission to save our planet. This is our planet and we must take responsibility for it.
We further did our research and extracted useful data for our analysis. We plotted the data and applied machine learning algorithms to create a solution and we did achieve it.
Team Cybit firmly believes that if we extend more of our limitations, we will definitely create the perfect major solution to save the Earth from the next Carrington Event.
"The Earth is all that we have in common"- Wendell Berry
References
Tags
#AI #nasa #ML #MAchineLearning #software #Software #data #DATA #deep_learning #spaceapps #science #carrington #earth #sun #geomagnetic_storm