Lightnews — Scholar-powered news

Brenhin Keller

@brenhinkeller.bsky.social

1.3K followers 800 following 120 posts

Posts Media Videos Starter Packs

Pinned

Brenhin Keller @brenhinkeller.bsky.social · Dec 22

New paper on glacial erosion rates and the Sadler effect that came out of a @dartmouthears.bsky.social grad seminar class just out now in Sci Adv! doi.org/10.1126/scia...

Limits to timescale dependence in erosion rates: Quantifying glacial and fluvial erosion across timescales

Globally, glacial erosion rates outpace fluvial, with limits to the biases introduced by measuring across different timescales.

doi.org

2 12 37

Reposted by Brenhin Keller

Brendan O’Neill @permafrostee.bsky.social · Sep 4

Here’s a really nice alluvial fan I thought you should see

6 17 140

Brenhin Keller @brenhinkeller.bsky.social · Aug 16

It's kind of wild that the ultramafics are competent enough to make up one of the highest peaks in the region!

Brenhin Keller @brenhinkeller.bsky.social · Aug 15

Ah, that's probably the same ultramafics that make up Monte Disgrazia, just above/east of the roof contact of the Bergell pluton - doi.org/10.1029/97JB01510

Fossil crust‐to‐mantle transition, Val Malenco (Italian Alps)

An exhumed, undisturbed fossil lower crust to upper mantle section is preserved in Val Malenco, Italian Alps, and is now exposed along the boundary between Penninic and Austroalpine nappes. Lower-cru....

doi.org

1 3

Reposted by Brenhin Keller

Kenneth Becker @geophotographer.org · Aug 14

Check out these photos of the Tracy Arm Landslide to help get an idea of just how big it is.

📷 © Jacek Maselko and posted by AEC, taken at 3,500ft elevation on Tuesday, August 12th.

AEC page
earthquake.alaska.edu

AEC Facebook post
www.facebook.com/share/1Fzmxn...
🧪⚒️🌊
#alaska
#tsunami
#landslide

The image depicts a mountainous landscape with steep cliffs prominently visible in the foreground. The rock faces are textured with varying shades of gray, brown, and areas of green vegetation. A small body of water with a frothy surface lies at the base of the cliffs. In the background, a range of snow-capped mountains stretches across the horizon under a sky filled with thin, layered clouds. The overall scene has a rugged and expansive natural beauty.

The image depicts a rugged mountain landscape featuring towering, snow-capped peaks under partly cloudy skies. The mountains are steep and rocky with patches of green vegetation on the lower slopes. A mixture of brown and gray rock is visible, contrasting with the white snow. At the base of the mountains lies a glacial area, showing large, sharp ice formations with a distinct blue tint. A small lake of muddy water is surrounded by chunks of ice. The foreground features green foothills. The overall setting is vast and remote, evoking a sense of untouched wilderness.

The image depicts a mountainous landscape with steep, rocky peaks partially covered in snow. A glacier is visible in a valley between the mountains, flowing towards a body of water peppered with chunks of ice. The water is a mix of murky brown and icy blue, suggesting it is glacial runoff. The foreground is a contrast of vibrant green vegetation, displaying a dense forest at the bottom. The sky above is partly cloudy. The photo captures the natural beauty and ruggedness of the terrain.

The image captures an aerial view of a vast glacial landscape. The foreground shows a rugged, uneven expanse of gray and white ice with visible crevasses, characteristic of a glacier. To the left, a wing of the aircraft is partially visible, indicating the perspective of the photo. Beyond the glacier, steep, rocky mountains rise dramatically, their surfaces marked with patches of green vegetation and some snow at the peaks. In the background, another glacier winds through a valley, bordered by more mountainous terrain. The sky is mostly clear, allowing sunlight to illuminate the scene.

1 5 16

Reposted by Brenhin Keller

Dr. Brendan Anderson @fossilsndcoffee.bsky.social · Jun 17

For Day50 of photoshopping @lastweektonight.com's John Oliver with fossils in the hopes that he saves the Paleontological Research Institution

Celebrating 90yrs of mollusk research at PRI! John is joined by a number of other H. sapiens who love science (and fun)-tag them all! 🧪🦑⚒️🐌 #fossils #savePRI

1 7 38

Brenhin Keller @brenhinkeller.bsky.social · Apr 30

I'm hoping that's sarcastic and these don't exist?

1 2

Reposted by Brenhin Keller

David Clark @clarkeocrinus.bsky.social · Apr 18

I sometimes take advantage of the analytical equipment at my real job. This is a Cambrian arthropod (probably not a Beckwithia) from the Wheeler shale of Utah. I used the XRF (X-ray fluorescence) to highlight iron and phosphorus. This is a 500 Ma digestive system.

#FossilFriday

3 7 40

Reposted by Brenhin Keller

Dr. Lynne Elkins @ljelkins.bsky.social · Apr 16

I am pleased to announce that my department (EAS at Nebraska) is hiring a teaching postdoc for AY 2025-2026 in the area of Earth Materials! I'm not on the committee but am spreading the word to my mineralogy/petrology networks on behalf of the committee. See ad for more details.

Postdoctoral Teaching Associate

The Department of Earth and Atmospheric Sciences (https://eas.unl.edu) at the University of Nebraska-Lincoln (UNL) has an opening for a Postdoctoral Teaching Associate in Earth Sciences. We are seekin...

employment.unl.edu

1 8 11

Brenhin Keller @brenhinkeller.bsky.social · Apr 4

That's wild

Brenhin Keller @brenhinkeller.bsky.social · Apr 4

That's with juliac I assume? Not too bad!

1 2

Brenhin Keller @brenhinkeller.bsky.social · Apr 4

That sounds like a confusing choice

Reposted by Brenhin Keller

Jeff Greene @jeffgreene.bsky.social · Mar 29

Unsurprisingly, @tressiemcphd.bsky.social is spot-on in this analysis: “A.I.’s most revolutionary potential is helping experts apply their expertise better and faster. But for that to work, there has to be experts.”
www.nytimes.com/2025/03/29/o...

Opinion | Actually, A.I. Is Pretty Mid (Gift Article)

A.I. is just what we need in the post-fact era: less research and more predicting what we want to hear.

www.nytimes.com

6 47 130

Reposted by Brenhin Keller

Alistair Hamill @ahamill.bsky.social · Mar 29

1/ An ongoing thread on the devastating M7.7 Mandalay (Myanmar) earthquake, 28 March 2025.

This GIS map is from USGS Pager.

#geographyteacher

5 15 24

Reposted by Brenhin Keller

Sarah Hörst @planetdr.bsky.social · Mar 19

We are assuming this was a planetary scientist traveling to the Lunar and Planetary Science Conference

Paul Cohen @paulecohen.bsky.social · Mar 19

Le Monde reporting that a French scientist traveling to Houston to attend a conference was denied entry to US after a search of his phone & computer revealed messages critical of Trump's science cuts, "which [says CPB] conveyed hatred of Trump & could be qualified as terrorism". Computer confiscated

10 47 170

Reposted by Brenhin Keller

Ke Gao I 高珂 @goco.bsky.social · Mar 14

Who’s guilty of the discrepancy between minerals and whole-rock chemistry in plutons? Melt extraction might be innocent—mineral reequilibration is likely the real culprit! Fresh out on Pi Day by me, Brenhin @brenhinkeller.bsky.social, & Changqian! doi.org/10.1029/2024...

Amphibole Compositions Record Cold Post‐Emplacement Reequilibration in Plutons

Melts predicted from mineral compositions in arc plutonic rocks do not match bulk compositions, a discrepancy absent in volcanic rocks Neither bulk crystallization nor melt extraction can fully e...

doi.org

1 2

Reposted by Brenhin Keller

Scott Dance @byscottdance.com · Mar 14

New: A NOAA lab in Hawaii that is connected to the longest-running observation of global greenhouse gas concentrations is slated for closure in August, according to a list of lease terminations Democratic members of Congress shared with @washingtonpost.com.

Trump moves to close government lab that tracks planet-warming pollution

The lab is connected to the Mauna Loa Observatory, where scientists gather data to produce the Keeling Curve, a chart on the daily status of atmospheric carbon dioxide concentrations.

www.washingtonpost.com

33 350 470

Reposted by Brenhin Keller

Bernat Font @fontb.bsky.social · Mar 12

Playing a bit with our WaterLily.jl #CFD #JuliaLang solver and homogeneous isotropic decaying turbulence 🌀 128^3 grid running on my local #GPU ⚡ visualizing in real-time with @makie.org
github.com/WaterLily-jl...

6 26

Brenhin Keller @brenhinkeller.bsky.social · Feb 14

Ah so it's in skyfeed and was just set to the maximum search time (seven days), though it looks like they do now have a "remember posts" option that one can add as an extra "block" so we can try that

Reposted by Brenhin Keller

Tidier.jl @tidierjl.bsky.social · Feb 9

Coming soon to TidierFiles.jl for #JuliaLang: reading Google sheets as DataFrames.

5 18

Reposted by Brenhin Keller

Carl Zimmer @carlzimmer.com · Jan 30

National Science Foundation suspends salary payments, leaving researchers unable to pay their bills www.statnews.com/2025/01/30/t...
Story by @ericboodman.bsky.social

National Science Foundation suspends salary payments, leaving researchers unable to pay their bills

An NSF online payment system remained down after the federal funding freeze was lifted, leaving early-career scientists scrambling to pay bills

www.statnews.com

59 1K 1.5K

Brenhin Keller @brenhinkeller.bsky.social · Jan 31

Possibly apocryphal, but I remember hearing something along the lines that in the very early days Jeff & co at some point considered something more like this as the default syntax for Julia. Maybe not too shocking given that Sussman was on Jeff's thesis committee 😄

Brenhin Keller @brenhinkeller.bsky.social · Jan 31

One of my favorite Julia factoids is that in addition to LispSyntax.jl there's also a not-much-discussed built-in S-expresssion syntax for Julia which is normally only used as an alternate way of viewing Julia `Expr`s but can be hacked into a REPL mode in twenty-some lines (code in alt text):

$to_expr(x) = x to_expr(t::Tuple) = Expr(to_expr.(t)...) lisparse(x) = to_expr(eval(Meta.parse(x))) function lispmode() printstyled("\nlisp> ", color=:magenta, bold=true) l = readline() # READ while l !== "(:exit)" try # So we don't get thrown out of the mode result = eval(lisparse(l)) # EVAL if isa(result, Expr) # PRINT (Meta.show_sexpr(result); println()) elseif isa(result, Tuple) Meta.show_sexpr(:($(result...),)); println() else display(result) end catch e display(e) end printstyled("\nlisp> ", color=:magenta, bold=true) l = readline() # READ end end$

Reposted by Brenhin Keller

Adrian Hill @adrianhill.de · Jan 30

You think Jacobian and Hessian matrices are prohibitively expensive to compute on your problem? Our latest preprint with @gdalle.bsky.social might change your mind!
arxiv.org/abs/2501.17737
🧵1/8

Figure comparing automatic differentiation (AD) and automatic sparse differentiation (ASD).

(a) Given a function f, AD backends return a function computing vector-Jacobian products (VJPs). (b) Standard AD computes Jacobians row-by-row by evaluating VJPs with all standard basis vectors. (c) ASD reduces the number of VJP evaluations by first detecting a sparsity pattern of non-zero values, coloring orthogonal rows in the pattern and simultaneously evaluating VJPs of orthogonal rows. The concepts shown in this figure directly translate to forward-mode, which computes Jacobians column-by-column instead of row-by-row.

5 29 140

Reposted by Brenhin Keller

Tao Wen 文涛 @jaywen.bsky.social · Jan 23

Nominations for the M. Lee Allison Award for excellence in Geoinformatics are open. This award is granted annually by GSA Geoinformatics and Data Science Division. Review of nominations will begin on March 1st, 2025. Please spread the word! Here is the nomination form: forms.office.com/r/CEBjb4kWZk

2 3

Reposted by Brenhin Keller

Dr. Kenna Harmony Rubin @kennarubin.bsky.social · Feb 14

Fresh & dense picrite, fresh aphyric 'a'a, basalts with infilled vesicles, and basalts with sharp flow contacts. just a few of the #hawaiian lavas cored well below modern sea level on #IODP #exp389 are showing a surprising range of #volcanic textures... all in the first site we've examined thus far.

2 10