About

New Grant NASA Early Career Investigator Program (ECIP) Physics-Aware Multimodal Foundation Model for Earth's Evolving Cryosphere · $300,000 · 3 years
New Paper FloodDuo: A Dual Foundation-Encoder Architecture for Generalizable High-Resolution Flood Mapping ISPRS Journal of Photogrammetry and Remote Sensing · Nov 2026

I build AI models that turn satellite imagery into knowledge about Earth’s water. My work sits at the intersection of computer vision, remote sensing, and machine learning. I love designing models, large-scale data pipelines, and open tools that can read vast archives of optical and radar imagery and map what is changing on the ground, quickly and reliably. Much of my current focus is on geospatial foundation models: what they actually learn, where they fail, and how to make them efficient enough for real-world use.

The engineering serves a set of science questions I keep coming back to. How fast are glaciers retreating, and where are glacial lakes growing into hazards? Where and how often do floods happen, and who is exposed? What do these changes mean for water resources and sea level rise in a warming climate? The cryosphere and floods are two parts of the same water story, and both unfold in places that are remote, cloudy, or changing too fast to monitor by hand. That is exactly where AI can help most.

I am a postdoctoral researcher at the University of Wisconsin-Madison in the Tellman Lab, where I also help teach AI-for-Earth and sustainability courses. Before this, I was a postdoc at Ohio State University, where I helped build the first global map of potential natural hydrogen sites, and I earned my Ph.D. through AcSIR and the German Aerospace Center (DLR), studying glaciers, glacial lakes, and the floods they can cause. I believe in open science, so my code, data, and trained models are public on GitHub and Hugging Face.

Academic Achievements

  1. 2026 NASA Early Career Investigator Program in Earth Science (ECIP) grant for “Physics-Aware Multimodal Foundation Model for Earth’s Evolving Cryosphere” ($300,000 over three years).
  2. 2025 Young Scientist travel Award to attend Debris Cover Glacier Workshop, ISTA Austria (1000$)
  3. 2024 Professional development/travel award: Office of Postdoctoral affair, OSU USA. (700$).
  4. 2019 IASC Early Career Scientist travel award for attending ISMASS workshop, Reykjavik. (1500 $).
  5. 2019 DAAD-Bi-nationally supervised doctoral degree fellowship (~24000 Euro).
  6. 2019 Best poster award “Third” at 5th International YES Congress, Berlin (Certificate and Book).
  7. 2019- DST-International Travel Support to attend 5th International YES Congress, Berlin (~1500 $).
  8. 2018- Innovation in Science Pursuit for Inspired Research (INSPIRE), DST-India (~13,000 $)

News

  1. November 2026: awarded a NASA Early Career Investigator Program in Earth Science (ECIP) grant for “Physics-Aware Multimodal Foundation Model for Earth’s Evolving Cryosphere” ($300,000 over three years).
  2. November 2026: our paper “FloodDuo: A Dual Foundation-Encoder Architecture for Generalizable High-Resolution Flood Mapping” is now published in ISPRS Journal of Photogrammetry and Remote Sensing, available here. Code and the pretrained model are on GitHub and Hugging Face.
  3. July 2026: our paper “Glacial-Lake-Bench: A Global Multi-Sensor Benchmark Dataset for Evaluating Deep Learning Models for Glacial Lake Mapping” is now published in Earth System Science Data (ESSD), available here.
  4. February 2026: our latest study is now available on IEEE Xplore (official publication version), available here.
  5. January 2026: our workshop paper in WACV (CV4EO-Workshop) introduces Prithvi-CAFE for improved flood inundation mapping, available here.
  6. January 2026: our workshop paper in WACV (GeoCV-Workshop) presents GLACIA, a multimodal large language model framework for glacial lake segmentation with positional reasoning, available here.
  7. November 2025: our preprint on benchmarking geo-foundation models for flood inundation mapping across Sentinel-1, Sentinel-2, and PlanetScope is available here.
  8. February 2025: read our new article published in Water Resources Research about declining groundwater storage in the Indus Basin, available here.
  9. February 2025: read our new article published in Science of Remote Sensing about improved and consistent GRACE/GRACE-FO data using machine learning, available here.
  10. September 2024: read our new article published in Remote Sensing of Environment about the most updated glacial lake inventory in HMA using deep learning, available here.
  11. July 2024: read our new article published in Remote Sensing Applications: Society and Environment about increasing glacial lake outburst flood risk in Sikkim, available here.
  12. May 2024: joined Social-Pixel Lab at the University of Arizona as a Postdoctoral Research Associate.