DEEPTHINKING

https://github.com/AlexLoayza129/carrington-algorithm

https://docs.google.com/presentation/d/13IA89jwo6ZxBlPWVMJBwsb9Y5BP8wupPXjWpkRKJB9g/edit#slide=id.g14097106877_0_197

What exactly does Carrington's Algorithm do?

The Carrington's Algorithm is a neural network that predicts peak solar wind velocity, magnetic fields, proton density, and thermal speed measures utilizing existing data from two orbiting spacecrafts, the DSCVR and WIND. These parameters are then given through a graphic representation. This graphic aid allows easier analysis of data, and enables monitoring and keeping track of parameters, such as the solar wind velocity, in a more effective manner. Thus making it easier to forecast any early signs of events due to increase solar activity. Eg. The Carrington Event of 1859.

How does it work?

Below you will find a quick flow diagram that highlights how we took on this challenge!

What benefits does it have?

Whilst it is a simple model, the benefits Carrington's Algorithm possess are out of this world. As an example, one can consider the possibility for improvement this neural network has. Given a temporal period of 3 months worth of data to work on the neural network already shows promising ranges of accuracy. Thus when given more data, it is speculative that the values given by CA would be of optimal accuracy to the 'ground truth' values. Furthermore, the model is simple, easy to use, and through the interactive website, one can visualize the data in a more comprehensible, and approachable way. Easier to read makes it easier to understand. Hence, combining these two. Carrington's Algorithm has the opportunity to bring incredible precision to the prediction of parameters such as the solar wind speed, and then offer a visual to allow monitoring and analysis easier for the user.

What do you hope to achieve?

Here at DeepThinking, our aim is to promote the implementation of machine learning for the better future of humanity. Through this challenge we attempted to create a tool that would allow the early detection of a solar event that could painfully impact the society we live today. We understand that the space we live in does not constitute of peace and tranquility, but rather of a turbulent darkness, and that due to the way society has apprehended a dependency to technology so greatly, events such as Carrington's would be disastrous. If we depend in technology, then the least we could do is use it to protect us. This project aims to distribute inspiration in the understanding that whilst acknowledging the fragility of our kind is true, the way that technology has progressed exponentially over the years, one must not underestimate it, never less under use it. This type of challenge opens eyes to appreciating how the technology we have available today, could indeed help humanity tomorrow.

What tools, coding languages, hardware, or software did you use to develop your project?

For the development of the neural network this project used Jupyter Notebook with Python3.

The development of the logo, font, and visual design was done via Figma

The data that was used through this project was primarily from the Goddard Space Flight Centre-The Coordinated Data Analysis Web. This is an open source provided by NASA. The following links could be used to access the information:

DSCOVR, and WIND, BW(t) and BD(t):

https://cdaweb.gsfc.nasa.gov/pub/data/wind/mfi/mfi_h2/2022/

https://cdaweb.gsfc.nasa.gov/pub/data/dscovr/h0/mag/2022/

Wind Ion Parameters:

 https://cdaweb.gsfc.nasa.gov/pub/data/wind/swe/swe_h1/2022/

This experience has been truly enlightening. As participants of the NASA Space Apps Challenge 2022, we have truly enjoyed being able to take on an incredibly complex challenge, and thinking our-of-the-box approaches. The opportunity was also given to meet new people, from different backgrounds, and different cultures. Learning from each other is always extremely motivating, and creating this sense of open-mindedness towards new ideas is always an element that should be cherished. Our team was composed from scratch, out of people with a wide range of different skills, but one common passion, space. We decided to take on this challenge, as we started to dig-in deeply into what the Carrington Event was, the data available, and the interpretation of the challenge. Our approach consisted in creating an open flow of communication within the participants of the team. None of our team members are subject matter experts, and in fact only one member had previous experience with ML. Hence, when it came to understanding the challenge itself, and the resources given, we established an open line of communication that would allow each to have a space to speak what their interpretation was and then it would be held under discussion. We faced many problems through this challenge, the lack of experience with handling large amounts of data made it difficult to analyse it rapidly. This took about 3/4ths of the time given in the challenge to resolve. In the end, we solved it by mutually agreeing that due to lack of time and experience we should just create the database manually, from data given, and synchronize it manually without the implementation of a DTW. We are aware that this approach might not deliver the best results, but nonetheless we are grateful for the learning opportunity this experience has been, and are proud of the outcome.

https://cdaweb.gsfc.nasa.gov/pub/data/wind/mfi/mfi_h2/2022/

https://cdaweb.gsfc.nasa.gov/pub/data/dscovr/h0/mag/2022/

https://cdaweb.gsfc.nasa.gov/pub/data/wind/swe/swe_h1/2022/

#Software #Carrington #Sun #Earth #Graphics