iiinova — where your digital self keeps evolving

Reposted by Marcus van Lier-Walqui

arseblog @arseblog.com · 19/05/2026

Never in doubt ❤️ ❤️

Reposted by Marcus van Lier-Walqui

arseblog @arseblog.com · 08/04/2025

What a night, what a performance from this far from 100% team, one of the great European nights at the Emirates. Declan Rice has produced two of the greatest free kicks you'll ever see at this level, and we 100% deserved that third goal too. Incredible. This team deserves so much love. #arsenal

1573453393

Marcus van Lier-Walqui @mvanlierwalq.bsky.social · 12/01/2025

Huge for del Grosso! Wish it was broadcast

000

Reposted by Marcus van Lier-Walqui

Gavin Schmidt @climateofgavin.bsky.social · 12/01/2025

With people blaming the LA fires on lesbian firefighters, space lasers, Oprah, Canadians, Mexicans, homeless people, and DEI initiatives, focusing your complaints on people justifiably mentioning climate change as a factor is… a choice.

820128

Marcus van Lier-Walqui @mvanlierwalq.bsky.social · 12/01/2025

In honor of Belgian Cyclocross National Championship

140

Marcus van Lier-Walqui @mvanlierwalq.bsky.social · 21/12/2024

Schreiber drive side dismount has been paying off there tho

010

Reposted by Marcus van Lier-Walqui

LEAP-STC @leapstc.bsky.social · 11/12/2024

LEAP #MomentumFellow23 Mohammed Erfani presents "A33J-08: Spatiotemporal ML Models for Emulation of Atmospheric Composition" TODAY @ 15:30p, Salon H! buff.ly/4gtYhh1 @karadlamb.bsky.social + @mvanlierwalq.bsky.social #LEAPEducation #AGU24 #NSF #CUSEAS @agu.org @columbiaclimate.bsky.social

052

Reposted by Marcus van Lier-Walqui

LEAP-STC @leapstc.bsky.social · 11/12/2024

Take a stroll to see "A31G-1810: Discovering a Model for Depositional Ice Growth from Observations Using Neural Ordinary Differential Equations + Symbolic Regression" (by Kara Lamb), TODAY at #AGU24! buff.ly/4ffc7CZ #NSF #CUSEAS @agu.org @columbiaclimate.bsky.social

buff.ly

Discovering a Model for Depositional Ice Growth from Observations Using Neural Ordinary Differential Equations and Symbolic Regression

Depositional ice growth is an important process for the growth and evolution of...

032

Reposted by Marcus van Lier-Walqui

LEAP-STC @leapstc.bsky.social · 11/12/2024

"How Well Can Power Laws Describe Microphysical Processes in Bulk Schemes?" Arthur Hu, @mvanlierwalq.bsky.social + Kaitlyn Loftus have ideas for you this morning in their #AGU24 poster (A31B-1693)! buff.ly/49sVAtT #NSF #CUSEAS @agu.org @columbiaclimate.bsky.social

buff.ly

How well can power laws describe microphysical processes in bulk schemes?

Understanding and accurately simulating cloud microphysics is crucial for impro...

031

Reposted by Marcus van Lier-Walqui

LEAP-STC @leapstc.bsky.social · 09/12/2024

LEAP closes out Day 1 w/ @mvanlierwalq.bsky.social presenting "GC14E-08: Challenges and Opportunities for Unified Top-down + Bottom-up Constraint of ESM Parameterizations" - 15:12pm in Salon C! agu.confex.com/agu/agu24/me... #AGU24 #NSF @columbiaclimate.bsky.social @columbiauniversity.bsky.social

agu.confex.com

Challenges and opportunities for unified top-down and bottom-up constraint of ESM parameterizations

I will present work using perturbed parameter ensembles (PPEs) of models at mul...

021

Marcus van Lier-Walqui @mvanlierwalq.bsky.social · 04/12/2024

Gets at an interesting issue. In trad ESMs, large scale interactions & processes are explicitly modeled via cloud parameterizations, but these large scale features are emergent properties of microphysics in km-scale models, which that may be more strongly nonlinear & harder to investigate via PPEs

010

Marcus van Lier-Walqui @mvanlierwalq.bsky.social · 25/11/2024

but if it's warm rain processes (rather than cold/mixed phase), then rain drops will tend to be smaller. However, in that case, rain water content will be systematically underestimated by reflectivity alone.... so the rain will seem wetter than radar would indicate!

110

Marcus van Lier-Walqui @mvanlierwalq.bsky.social · 25/11/2024

MacKay's book on Information Theory, Inference, and Learning Algorithms is one of my all-time favorites! RIP indeed.

020

Marcus van Lier-Walqui @mvanlierwalq.bsky.social · 25/11/2024

The criticism here also applies to the performative aspects of DEIA that I find annoying insomuch as they fail to address any of the structural issues that imperil the most vulnerable people (e.g. early career scientists who lack stability, money, and privilege) in our field

000

Marcus van Lier-Walqui @mvanlierwalq.bsky.social · 25/11/2024

This brilliant Guardian piece by Nesrine Malik cuts through the chaff like a hot knife through half-melted ice cream. "'Doing the work' focused on interpersonal dynamics of racism rather than on structural ones" www.theguardian.com/commentisfre...

theguardian.com

‘Woke’ didn’t lose the US election: the patrician class who hijacked identity politics did | Nesrine Malik

Why is this simple explanation being so widely embraced? Because it does not require a commitment to real, structural change, says Guardian columnist Nesrine Malik

100

Marcus van Lier-Walqui @mvanlierwalq.bsky.social · 18/11/2024

I remember being fascinated (in high school!) by JET and the promise of ITER. Would sincerely love to see this tech succeed and prove wrong my worries that this buzz is an oversell.

000

Marcus van Lier-Walqui @mvanlierwalq.bsky.social · 18/11/2024

progress! 60 years ago we were 30 years away.

110

Marcus van Lier-Walqui @mvanlierwalq.bsky.social · 18/11/2024

I'm sure many have seen this wonderful presentation by Grace Hopper; it's worth underlining her point (multiple times, with permanent marker) the need to listen to young scientists, advocate for their careers, and get out of the way when they can lead. www.youtube.com/watch?v=si9i...

youtube.com

Capt. Grace Hopper on Future Possibilities: Data, Hardware, Software, and People (Part One, 1982)

YouTube video by National Security Agency

020

Marcus van Lier-Walqui @mvanlierwalq.bsky.social · 18/11/2024

A few unfortunately glib arguments made to make a point. for ex, even if it's not surprising that airports disrupt sleep in nearby neighborhoods, it's still important to measure how severe & how widespread the disruption is. Some serious points buried in this sort of unserious baiting.

000

Marcus van Lier-Walqui @mvanlierwalq.bsky.social · 14/11/2024

Some takeaways: 1. It works! We tune 45 parameters w appx 700 ensemble members & 36 satellite metrics 2. Strong evidence that *virtually all* parameters are important for tuning 3. Observational uncertainty & its definition it plays a big role 4. We coin the term "CPE" Calibrated Physics Ensemble

110

Marcus van Lier-Walqui @mvanlierwalq.bsky.social · 14/11/2024

Am hesitant to get back into social media after my blissful year+ twitter hiatus, but here's a preprint of Greg Elsaesser and my (and others) work on using Bayesian inference, machine learning, and satellite observations to tune the NASA ModelE Earth System Model essopenarchive.org/users/538881...

essopenarchive.org

Using Machine Learning to Generate a GISS ModelE Calibrated Physics Ensemble (CPE)

A document by Gregory Elsaesser. Click on the document to view its contents.

131

Marcus van Lier-Walqui @mvanlierwalq.bsky.social · 20/10/2023

better to commit finite computational resources to big perturbed-parameter ensembles to improve GCM physics, or invest in km-scale global sims? (assuming we can't do both in the near-term)

000