CEDAR email: Live Solar Wind Prediction Tournament and Prize Fund.

[Tom Penlington]

Dear CEDAR List Moderator,

Please find below an announcement regarding the FDL Benchlab Solar Wind
Prediction 2026 AI Tournament, organized in partnership with  *NASA ,*CU
Boulder and Queen Mary University of London.

We would be grateful if you could approve this message for distribution to
the CEDAR mailing list or in any other relevant channels. This is an
exciting opportunity and we believe members of your community would be
interested in participating or following along with the live leaderboard.
Please find more info here: benchlab.live <http://www.benchlab.live/>

Please reach out if you need any more information or if I can support in
any way, I look forward to hearing from you,

Tom
[image: 7ad6d68a-b85f-4b61-8372-56b5f34f6117.jpg]

Dear Colleagues,

Predicting the ambient solar wind three days ahead remains one of the most
significant open problems in heliophysics. This is why we are engaging the
community to drive SOTA in operational conditions.

We are pleased to share the FDL Benchlab Solar Wind Prediction 2026 AI
Tournament, a live operational AI competition designed to test whether
machine learning systems can produce robust 72-hour ambient solar-wind
forecasts under realistic forward-time constraints.

Benchlab is a Frontier Development Lab initiative in partnership with CU
Boulder, Queen Mary University of London and NASA. (benchlab.live
<http://www.benchlab.live>)

Key information:


   -

   True Operational Conditions: Unlike traditional static competitions,
   this tournament tests models inside a live, constraint-driven operational
   inference loop. Your workflows will run continuously to generate
   forecasts against future, unseen ACE solar-wind data.



   -

   Prizes & Impact: Compete for a share of a $20,000 cash prize fund.
   Winners will co-author a joint technical paper with the organizing
   committee, and the top-performing model will be integrated directly into
   the live operational suite.


We encourage interdisciplinary teams to combine your domain expertise in
solar physics with data science to build robust prediction architectures.

Pre-register your team here:
https://fdl.ai/benchlab-solar-wind-prediction-2026

[image: image.png]

For full competition details, technical requirements, and to learn more
about the Benchlab ecosystem, visit the competition website
<https://fdl.ai/benchlab-solar-wind-prediction-2026>. Join us for the
official briefing September 1, 2026.

Best regards,

Enrico Camporeale <enrico.camporeale at qmul.ac.uk> and Nina Bonaventura
<nina.bonaventura at trillium.tech>

________________________________
*Tom Penlington*
*Project Assistant*
*Trillium Technologies*
E:  tom.penlington at trillium.tech  <tom.penlington at trillium.tech>
W: https://trillium.tech
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.ucar.edu/pipermail/cedar_email/attachments/20260730/63f27d9a/attachment-0001.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image.png
Type: image/png
Size: 1877718 bytes
Desc: not available
URL: <https://mailman.ucar.edu/pipermail/cedar_email/attachments/20260730/63f27d9a/attachment-0001.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 7ad6d68a-b85f-4b61-8372-56b5f34f6117.jpg
Type: image/jpeg
Size: 82277 bytes
Desc: not available
URL: <https://mailman.ucar.edu/pipermail/cedar_email/attachments/20260730/63f27d9a/attachment-0001.jpg>