GeoAI Arctic Mapping Challenge

Welcome to the GeoAI Arctic Mapping Challenge, a competition at the intersection of Geoscience and Artificial Intelligence. We invite you to develop innovative AI models to map retrogressive thaw slumps (RTS), a critical indicator of climate change in the Arctic.

newspaper News

  • (09/09/25) Competition Kick-off! We are excited to announce the official launch of the GeoAI Arctic Mapping Challenge.

info About the Challenge

Retrogressive thaw slumps (RTS) are landslides that occur in icy permafrost terrain. As the climate warms, these slumps are becoming more frequent and are dramatically reshaping the Arctic landscape. They have significant environmental impacts, including altering hydrology, affecting ecosystems, and accelerating the release of greenhouse gases.

Mapping RTS is challenging due to their small size, subtle appearance, and dynamic evolution over time. This competition invites you to leverage deep learning and GeoAI to automatically detect and map these features. By participating, you will contribute to advancing permafrost science and improving our understanding of Arctic change.

handshake Collaborators

groups Partners

library_books References

The GeoAI Arctic Mapping Challenge builds upon a RTS dataset and cutting-edge multimodal GeoAI models. We encourage participants to cite the following works when using the dataset or baseline model:

Dataset Source

The dataset was curated by Yang et al. (2023), who developed a deep learning approach for mapping RTS from satellite imagery. This dataset forms the foundation of the competition.

  • Yang, Yili, Brendan M. Rogers, Greg Fiske, Jennifer Watts, Stefano Potter, Tiffany Windholz, Andrew Mullen, Ingmar Nitze, and Susan M. Natali. “Mapping retrogressive thaw slumps using deep neural networks.” Remote Sensing of Environment 288 (2023): 113495. https://doi.org/10.1016/j.rse.2023.113495
@article{yang2023mapping,
  title={Mapping retrogressive thaw slumps using deep neural networks},
  author={Yang, Yili and Rogers, Brendan M and Fiske, Greg and Watts, Jennifer and Potter, Stefano and Windholz, Tiffany and Mullen, Andrew and Nitze, Ingmar and Natali, Susan M},
  journal={Remote Sensing of Environment},
  volume={288},
  pages={113495},
  year={2023},
  publisher={Elsevier}
}

Our Multimodal GeoAI Model

We provide a sample notebook using the baseline model from Li et al. (2025), a multi-scale vision transformer-based multimodal GeoAI model that integrates multi-spectral imagery and topographic features for high-accuracy RTS mapping in Arctic permafrost regions.

  • Li, Wenwen, Chia-Yu Hsu, Sizhe Wang, Zhining Gu, Yili Yang, Brendan M. Rogers, and Anna Liljedahl. “A multi-scale vision transformer-based multimodal GeoAI model for mapping Arctic permafrost thaw.” IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing (2025). https://doi.org/10.1109/JSTARS.2025.3564310
@article{li2025multi,
  title={A multi-scale vision transformer-based multimodal GeoAI model for mapping Arctic permafrost thaw},
  author={Li, Wenwen and Hsu, Chia-Yu and Wang, Sizhe and Gu, Zhining and Yang, Yili and Rogers, Brendan M and Liljedahl, Anna},
  journal={IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing},
  year={2025},
  publisher={IEEE}
}

image Dataset Samples

The dataset consists of high-resolution, multi-spectral satellite imagery paired with segmentation masks outlining RTS boundaries. While the dataset provides multi-band inputs (including non-visible spectral channels), the sample images below are displayed using RGB composites for visualization purposes only.

Input: RGB composite example Target: RTS segmentation