Denys Godwin

Prithvi Geospatial Foundation Model

3 min read

Active: 2023 - 2026

Summary

Prithvi is a family of open-source AI foundation models developed through a collaboration between NASA and IBM, with contributions from partners including Oak Ridge National Laboratory, the Jülich Supercomputing Centre, and Clark University. Named after the Sanskrit word for Earth, Prithvi was first released in August 2023 as NASA’s first open-source geospatial AI foundation model.

My role in each round of model releases was in benchmarking the model against existing real-world applications. For the first round, I compared Prithvi 1.0 against a Conditional Generative Adversarial Network (CGAN) for cloud gap imputation (Link to paper →). For the second round, I compared Prithvi EO 2.0 300M and 600M against the highest-performing model on the BioMassters competition benchmark (Link to paper →).

Online Stories

AGU recognizes researchers from Clark's Center for Geospatial Analytics

2025 AGU Open Science Recognition Prize for work on Prithvi Geospatial Foundation Model.

Read the story →

Clark team receives NASA Marshall Group Achievement Award

Awarded August 2024 for work on Prithvi Geospatial Foundation Model.

Read the story →

Clark geospatial team partners with NASA, IBM to harness AI technology

Start of collaboration for Prithvi Geospatial Foundation Model.

Read the story →

Publications

Peer-Reviewed Journal Articles

  1. Szwarcman, D., Roy, S., Fraccaro, P., Gíslason, Þ. E., Blumenstiel, B., Ghosal, R., De Oliveira, P. H., De Sousa Almeida, J. L., Sedona, R., Kang, Y., Chakraborty, S., Wang, S., Gomes, C., Kumar, A., Gaur, V., Truong, M., Godwin, D., Khallaghi, S., Lee, H., … Moreno, J. B. (2026). Prithvi-EO-2.0: A Versatile Multitemporal Foundation Model for Earth Observation Applications. IEEE Transactions on Geoscience and Remote Sensing, 64, 1–20. https://doi.org/10.1109/TGRS.2025.3642610

Peer-Reviewed Conference Papers

  1. Godwin, D., Li, H., Cecil, M., & Alemohammad, H. (2024). Seeing through the clouds: Cloud gap imputation with Prithvi foundation model. 2nd ICLR Workshop on Machine Learning for Remote Sensing, Vienna, Austria.

Pre-Print Publications

  1. Jakubik, J., Roy, S., Phillips, C. E., Fraccaro, P., Godwin, D., Zadrozny, B., Szwarcman, D., Gomes, C., Nyirjesy, G., Edwards, B., Kimura, D., Simumba, N., Chu, L., Mukkavilli, S. K., Lambhate, D., Das, K., Bangalore, R., Oliveira, D., Muszynski, M., … Ramachandran, R. (2023). Foundation Models for Generalist Geospatial Artificial Intelligence (No. arXiv:2310.18660). arXiv. https://doi.org/10.48550/arXiv.2310.18660

Open-Access Datasets

  1. Godwin, D., Li, H., & Alemohammad, H. (2024). Multi-Temporal Cloud Gap Imputation With HLS Data Across CONUS (Version v1.0) [Dataset]. Zenodo. https://doi.org/10.5281/ZENODO.11281740

Conference Presentations

  1. Godwin, D., Khallagi, S., Balogun, R., Yao, Y., Roy, S., Ramachandran, R., Alemohammad, H. (2025, December). GELOS: A Benchmark Dataset for Geospatial Exploration of Latent Observation Space [Poster presentation]. AGU Annual Meeting 2025, New Orleans, LA, USA.
  2. Godwin, D. (2024, March). A Homunculus of Place: Tactile Cartographies of Internal Worlds. AAG Annual Meeting 2025, Detroit, MI, USA.
  3. Godwin, D., Li, H., Cecil, M., & Alemohammad, H. (2023, December). Comparative Analysis of Vision Transformer and CGAN Models for Cloud Gap Filling in Time Series of Satellite Images. AGU Annual Meeting 2023, San Francisco, CA, USA.

Graph View

Backlinks