About the dataset

This dataset contains 3,330 satellite imagery samples (1,352 landslide, 1,978 non-landslide) for landslide segmentation in Pakistan's Gilgit-Baltistan region, developed for the PKLandSeg study (IEEE Geoscience and Remote Sensing Letters, DOI: 10.1109/LGRS.2026.3673080).

Each sample includes RGB imagery (PlanetScope, 3m resolution), Copernicus DEM elevation data, NDVI, slope, and a binary landslide annotation mask, all at 512x512 pixel resolution. The dataset is split into fixed train (2,177), validation (660), and test (493) sets for reproducibility.

Annotations were produced through a two-phase AI-assisted workflow: an initial set of 1,072 sparse expert annotations was used to train a model that generated probability maps, which annotators then used to refine and expand the final labels. Dataset quality was independently validated using DeepLabV3+, achieving a TTA F1-score of 0.578.

RGB imagery is released as JPEG in compliance with Planet's Education and Research Program data sharing guidelines; DEM, NDVI, and slope layers are unrestricted derivative/external data. See the included README.md for full folder structure, methodology, licensing, and known limitations.

Recommended Citation

Please cite the following publication when using this dataset:

S. B. Hammad, E. Naseer, M. A. Siddique and R. Ahmed, "PKLandSeg: Annotation-Guided Hybrid CNN-Vision Transformer for Landslide Segmentation," in IEEE Geoscience and Remote Sensing Letters, vol. 23, pp. 1-5, 2026, Art no. 5001505.

For methodology details (e.g., annotation workflow, model architecture, performance metrics), please refer to the above paper.

Download Data

To download the dataset, please visit: https://zenodo.org/records/21065148 

About the dataset

This dataset contains 8,808 glacial lake boundaries (total area: ~136 km²) mapped from PlanetScope satellite imagery (September 2020) across the Hindu Kush-Himalaya (HKH) region of Pakistan. The lakes were delineated manually with careful inspection, as part of a broader effort to monitor cryospheric hazards (e.g., GLOF risks) in the HKH. This work is supported by the National Geographic Society and Microsoft through its "AI for Earth Innovation" program. The imagery used in this work is a Planet Labs product - 2023 PBC., and has been accessed freely under their education and research support program. The inventory is provided as ESRI shapefiles (inside a zipped folder). The CRS is EPSG:4326. The column "Shape_Area" in the attributes table provides lake areas in square meters.

Funded by: National Geographic Society

Download Data

To download the dataset, please visit https://zenodo.org/records/15476013

About the dataset

Glacial lake outburst floods (GLOFs) are a major threat to the local communities and important infrastructures in the high mountain regions. Early detection of glacial lakes can prevent these disastrous events. Towards this end, we collected Sentinel 2 true color scenes of High-Mountain Asia (HMA) region using glacial lakes inventory of this region. It covers an area of 2080.12 km²  with nearly 30,121 glacial lakes. After data collection, we retained 1200 cloud free true color images and manually generated their ground truth masks. The train and test sets contain 1000 and 200 images, respectively. The dataset covers lakes with different shapes, sizes and radiometric signatures. 

Funded by: National Geographic Society and Microsoft

Download Data

To download the dataset, please visit https://ieee-dataport.org/documents/glacial-lakes-detection-dataset

All datasets on this page are provided strictly for academic and non-commercial research purposes. Users must cite the original dataset using its assigned Digital Object Identifier (DOI), and, where applicable, also reference the related publication. Redistribution is not allowed; please link directly to the official DOI. Use in commercial or governmental projects requires prior written permission.