FLARECAST
Flare Likelihood and Region Eruption Forecasting
The fully automated system for predicting solar flares and space weather
ROLE OF THE INSTITUTE FOR DATA SCIENCE
> Computer science: software architecture, machine learning, algorithms
> Science communication: dialogue with industry, public engagement, citizen science, media
Project lead at I4DS: Hanna Sathiapal, Marco Soldati
Partners: international consortium
Funding: EU Horizon 2020
Project duration: 2015-2018
Keywords: machine learning, solar flare prediction, space weather forecasting
 
SUMMARY
FLARECAST is an automated forecasting system for solar flares – high-energy eruptions on the surface of the sun. Solar flares may trigger solar storms affecting technology on and near the earth. Until now, space weather forecasting is done based on observations and judgements of experienced scientists. This project develops algorithms based on machine learning methodologies for analysing solar active regions automatically and assessing the risk of an eruption.
PEOPLE @I4DS WORKING ON FLARECAST
Orell Bühler
Apprentice Software Emgineering
until 2018
Florian Bruggisser
Coputer Science Engineer
until 2017
Samuel von Stachelski
Computer Science Engineer
until 2017
Colin Klausner
Software Engineering Apprentice
until 2018
Dario Vischi
Computer Science Engineer
until 2017
OPEN RESOURCES AND RESULTS
VISUALS AND AUDIO
A solar flare in different wavelengths. Credits: NASA/SDO/Wiessinger
Active regions on the sun detected by the SMART algorithm. Credit: Paul A. Higgins
Solar Flare of April 17, 2016. Credits: NASA’s Goddard Space Flight Center/SDO/Genna Duberstein
Different people working together towards a common goal. Faces and voices of the international FLARECAST Team. Credit Hanna Sathiapal