FLARECAST · Project

Automated Solar Flare Forecasting to Protect Satellites, Power Grids, and Aviation

environmentTestedTRL 5Thin data (2/5)

The Sun sometimes throws massive tantrums — bursts of radiation and magnetized plasma that can knock out satellites, scramble GPS signals, and even damage power grids on Earth. The problem is we've been pretty bad at predicting when these outbursts will happen. FLARECAST built an automated system that watches sunspot regions in real time, measures their magnetic properties, and uses machine learning to predict whether a solar flare is coming — like a weather forecast, but for space weather.

By the numbers

consortium partners across multiple disciplines

countries represented in the consortium

total project deliverables produced

The business problem

What needed solving

Solar flares and coronal mass ejections can disable satellites, disrupt GPS and communications, damage power grid transformers, and increase radiation exposure on polar flight routes — often with little warning. Current prediction methods lack the automation and physical grounding needed for reliable, real-time commercial use. Companies operating space-dependent or radiation-sensitive infrastructure need better advance warning to protect their assets.

The solution

What was built

FLARECAST built an automated solar flare prediction system that extracts physical properties from solar images in near-real-time and feeds them into machine learning algorithms. Key deliverables include a validated prediction database with a finalized schema (27 deliverables total), covering the full pipeline from image processing to flare probability output.

Audience

Who needs this

Commercial satellite operators (SES, Eutelsat, Inmarsat)Power grid transmission operators (Terna, Elia, National Grid)Airlines with polar routes (Finnair, SAS, Cathay Pacific)Space insurance underwriters (Lloyd's space syndicate)National space weather service providers

Business applications

Who can put this to work

Satellite Operations & Telecommunications

enterprise

Target: Commercial satellite operators and telecom providers

If you are a satellite operator dealing with unexpected service outages caused by solar storms — this project developed an automated flare prediction system using machine learning on solar magnetogram data that can give advance warning before damaging radiation hits your fleet. With 27 deliverables including a validated prediction database, the system could help you schedule protective maneuvers and reduce costly downtime.

Electric Utilities & Grid Management

enterprise

Target: Power grid operators and transmission companies

If you are a grid operator worried about geomagnetically induced currents frying your transformers during solar storms — FLARECAST built a near-real-time forecasting service that monitors active solar regions and predicts flare likelihood. This gives your operations team actionable lead time to implement protective switching procedures before a geomagnetic storm arrives.

Aviation & Aerospace

enterprise

Target: Airlines operating polar routes and aviation safety authorities

If you are an airline routing flights over polar regions where solar radiation exposure spikes during flare events — this project created the first autonomous, physically motivated flare forecasting system validated with 9 research partners across 6 countries. Accurate flare predictions let you reroute flights proactively instead of reacting after radiation alerts are issued.

Frequently asked

Quick answers

What would it cost to integrate this flare prediction system into our operations?

The project data does not include budget figures or licensing costs. FLARECAST was developed as a research infrastructure tool by 9 academic and research partners. Any commercial licensing or integration pricing would need to be negotiated directly with the coordinator at the Academy of Athens.

Can this system scale to serve multiple commercial clients simultaneously?

FLARECAST was designed as a near-real-time forecasting service processing solar magnetogram and white-light images automatically. The prediction database schema went through at least two iterations (prototypal and final release), suggesting it was built for structured, scalable data access. However, commercial-scale deployment was not demonstrated during the project.

Who owns the intellectual property and how is it licensed?

The project was funded under the RIA (Research and Innovation Action) scheme, which typically means IP stays with the consortium partners. With 9 partners across 6 countries and zero industry partners, licensing discussions would likely involve multiple academic institutions. Contact the Academy of Athens as coordinator for IP terms.

How accurate are the flare predictions compared to existing methods?

FLARECAST used state-of-the-art prediction methods including statistical analysis, unsupervised clustering, and supervised learning, validated using forecast verification measures. The project describes itself as the first quantitative, physically motivated and autonomous forecasting system of its kind. Specific accuracy percentages are not available in the project data.

Is this system currently operational or would we need to build from scratch?

Based on available project data, FLARECAST developed a working prediction database with a finalized schema and validated algorithms during its 2015-2017 run. The project aimed to launch a near-real-time service, but post-project operational status would need to be verified with the consortium. The project website flarecast.eu may have current status information.

What data inputs does the system need to function?

The system requires solar magnetogram and white-light images, which it processes using advanced image-processing techniques to automatically extract active region properties like magnetic flux, shear, complexity, helicity, and energy proxies. These data sources are publicly available from solar observatories.

Are there regulatory requirements for space weather forecasting services?

Space weather services increasingly fall under critical infrastructure protection regulations in the EU and US. FLARECAST was developed under the PROTEC topic (Protection of European assets in and from space), indicating alignment with EU space safety priorities. Specific regulatory compliance details are not in the project data.

Consortium

Who built it

FLARECAST is a purely academic and research consortium — 9 partners from 6 countries (Switzerland, Greece, France, Ireland, Italy, UK) with zero industry participants and zero SMEs. The consortium includes 6 universities and 2 research organizations, coordinated by the Academy of Athens. This is a strength for scientific credibility but a gap for commercial readiness. Any company looking to adopt this technology would be the first commercial user, meaning both opportunity to shape the product and risk of being an early adopter without established commercial support structures.

ACADEMY OF ATHENSCoordinator · EL
THE PROVOST, FELLOWS, FOUNDATION SCHOLARS & THE OTHER MEMBERS OF BOARD, OF THE COLLEGE OF THE HOLY & UNDIVIDED TRINITY OF QUEEN ELIZABETH NEAR DUBLINparticipant · IE
FACHHOCHSCHULE NORDWESTSCHWEIZ FHNWparticipant · CH
MET OFFICEparticipant · UK
UNIVERSITY OF NORTHUMBRIA AT NEWCASTLEparticipant · UK
CONSIGLIO NAZIONALE DELLE RICERCHEparticipant · IT
UNIVERSITE PARIS-SACLAYparticipant · FR
UNIVERSITA DEGLI STUDI DI GENOVAparticipant · IT
CENTRE NATIONAL DE LA RECHERCHE SCIENTIFIQUE CNRSparticipant · FR

How to reach the team

Academy of Athens (Greece) — reach out to the solar physics or applied mathematics department

Order a report on this consortium See ACADEMY OF ATHENS profile →

Next steps

Talk to the team behind this work.

SciTransfer can connect you with the FLARECAST team and help assess whether their prediction technology fits your space weather risk management needs.

Order a report on this topic More in Environment & Climate