Miguel D. Mahecha
@miguelmahecha.bsky.social
1.1K followers 440 following 43 posts
Earth system data sciences Prof. @UniLeipzig @RSC4Earth related to @UFZ_de @iDiv @SCADS.AI, worried about climate extremes, biodiversity, the human-environment nexus and the 🌎's future in general. Editor @EnvDataScience
Posts Media Videos Starter Packs
Reposted by Miguel D. Mahecha
scadsai.bsky.social
#DataWeekLeipzig 2025 has started!
In the opening keynote @miguelmahecha.bsky.social spoke about "Environmental research in a data-rich age" and discussed how #AI can help to supplement missing information in the field of environmental data.

Find the stream here:
👉 2025.dataweek.de/live.html
Reposted by Miguel D. Mahecha
soechting.bsky.social
Major update to Lexcube.org - our interactive Earth System Data Cube visualization tool & my PhD project!

➡️ What’s new?
🌍 Region borders (or any GeoJSON) overlaid in the visualization!
🌎 Record GIF/MP4 animations!
🌏 Progress & dataset boundary indicators!

...and many more improvements 👇
Reposted by Miguel D. Mahecha
bigearthdata1.bsky.social
📢Interactive Earth system data cube visualization in Jupyter notebooks by Maximilian Söchting (@soechting.bsky.social), Miguel D. Mahecha (@miguelmahecha.bsky.social) et al.
👉https://doi.org/10.1080/20964471.2025.2471646
#opensource #3D #datacube #visualization #Jupyter #geoscience #remotesensing
miguelmahecha.bsky.social
You’ll work on spatiotemporal anomaly detection and use explainable AI to link impacts to weather, soil, and human land use—paving the way for better predictions.

One of 34 funded PhD topics at ScaDS.AI Dresden/Leipzig, spanning ML, time series, ethics, and more.

More details below! Check T9.3!
miguelmahecha.bsky.social
🌍 PhD student wanted in AI for Climate Extremes 🌿

Heatwaves, droughts, and heavy rain are intensifying—threatening vegetation, forestry, and agriculture. Join us to develop weakly supervised ML to detect and explain climate impacts in satellite data.
scadsai.bsky.social
@scadsai.bsky.social offers 34 open topics for Research Associates / PhD Students (f/m/x) within the Graduate School, together with mentors and host institutions. 👩‍🎓 In the upcoming days, we will introduce you to all of those areas and the corresponding topics.

👉 scads.ai/about-us/job...
An advertisement from ScaDS.AI Dresden Leipzig featuring an opportunity for 34 Research Associate and PhD Student positions (f/m/x). The focus areas include knowledge representation and inference, mathematical foundations of AI and representation learning, scalable machine learning (ML) and LLM inference, time series analysis and reinforcement learning, visualization and causal inference, and ML ethics for protein design and chemical reactions. Applications are due by April 9, 2025.
Reposted by Miguel D. Mahecha
holzheustefan.bsky.social
Vorschau auf die Druckversion des #S4FAppell an die Politik

Zeichnung noch möglich bis Montag 24.03. 23:59 Uhr

Öffentliche Übergabe des Appells:
Dienstag, 25.03. 09:30 Uhr, Nordende Friedrich-Ebert-Platz

www.bayceer.uni-bayreuth.de/s4f/de/top/o...
miguelmahecha.bsky.social
Supported primarily by the DeepESDL project @esa.int advised by @ancaanghelea.bsky.social in cooperation with with @brockmannconsult.bsky.social but also supported by @nfdi4earth.bsky.social and @belspo.bsky.social BELSPO via the HERMES projects led by Diego Miralles
miguelmahecha.bsky.social
WHAT? YOU DON'T WANT TO USE IT IN JUPYTER? 😊 No worries! Just visit hashtag#Lexcube www.lexcube.org

This work was led by @soechting.bsky.social with support from Gerik Scheuermann, @dmlmont.bsky.social and myself. I am very proud of my strong team!!!
LexCube - Leipzig Explorer of Earth Data Cubes
LexCube is an interactive visualization for large-scale earth data sets. Created at Leipzig University by Maximilian Söchting.
www.lexcube.org
miguelmahecha.bsky.social
Introducing #Lexcube4Jupyter
An open-source tool for interactive 3D data cube visualization, seamlessly integrated into scientific workflows. It enables:
✅ Efficient, memory-aware handling of large datasets
✅ Direct integration into Python workflows (e.g., Jupyter Notebooks)
miguelmahecha.bsky.social
We believe that interactive (!) visualisation is key to making high-dimensional Earth system data more accessible, interpretable, and actionable. With this you can achieve
✅ Real-time exploration of model outputs & residuals
✅ Quick data quality checks & anomaly detection
miguelmahecha.bsky.social
Excited to share the latest paper emerging from the fantastic PhD thesis of @soechting.bsky.social! Interactive Earth System Data Cube visualization in Jupyter Notebooks! www.tandfonline.com/doi/full/10....
Reposted by Miguel D. Mahecha
isp-uv-es.bsky.social
🚨 New paper out in @naturecomms.bsky.social

We review how #AI is transforming modeling & understanding extreme weather & climate events like floods, heatwaves, and wildfires🌍🔥💨

🔗Read it! isp-uv.short.gy/AIforExtreme...
miguelmahecha.bsky.social
This is one outcome of the DeepExtremes project funded by @esa.int excellently guided by @Anca Anghelea! Thank you!

The idea of the paper in a nutshell was already presented by @dmlmont.bsky.social (2025) in Earth System Data Cubes: Avenues for advancing Earth system research. EnvDataSci, 3, e27
miguelmahecha.bsky.social
From the @mpi-bgc.bsky.social we have @vitusbenson.bsky.social Fabian Gans @clair3rbn.bsky.social and Melanie Weynants and from @brockmannconsult.bsky.social we have the amazing Tonio Fincke!
miguelmahecha.bsky.social
🌿 The dataset focuses on sampling areas impacted by climate extremes, across diverse vegetation types. By prioritizing compound heatwave and drought events, it provides a globally representative and reproducible resource for studying these phenomena!
miguelmahecha.bsky.social
💡 DeepExtremeCubes is a new database designed to address this gap. 📊 Key features of DeepExtremeCubes: 40,000 globally sampled mini data cubes (2.5 x 2.5 km) with Sentinel-2 L2A for vegetation monitoring, ERA5-Land variables, and ancillary data, including land cover and topography maps.
miguelmahecha.bsky.social
🌍 Climate extremes, such as compound heatwaves and droughts, are increasing in both frequency and intensity, yet predicting their impacts on terrestrial ecosystems remains a challenge. Machine learning holds promise, but traditional datasets often struggle to represent these rare events effectively.
Reposted by Miguel D. Mahecha
elifoezmen.bsky.social
Point of no return.
Diesen Tag und diese Nachricht werde ich den Rest meines Lebens nicht vergessen. Denn heute habe ich eingesehen, dass man mir und meinesgleichen in meinem Geburts- und Heimatland niemals eine sichere Heimat bieten wird.
search.app/kscX74bMF3Qj...
Merz gibt deutliche Botschaft: „Aberkennung der deutschen Staatsbürgerschaft“
Merz zeigt sich unzufrieden mit den Einbürgerungsrichtlinien der Ampel. Bei einem Sieg bei der Bundestagswahl hat der CDU-Chef vor, einiges grundlegend zu verändern.
search.app
Reposted by Miguel D. Mahecha
petergleick.bsky.social
If you've never donated to Wikipedia, now's the time. Elon Musk is attacking it because it's one of the only sources of honest, curated information that he can't control.
www.newsweek.com/elon-musk-ta...
Elon Musk takes aim at Wikipedia
The billionaire has become one of the most prominent supporters of President-elect Donald Trump.
www.newsweek.com