Rocky Mountain Snowpack: Snow Crystal, Profile and Core Imagery from Colorado Snowpits
- Provider
- Rocky Mountain Digerati (RMDig.ai)
- License
- CC-BY-4.0 (source)
- Size
- 8,076 files
- Format
- JSON Lines, PNG (in sampled files)
- Stored at
- Hugging Face (https://huggingface.co/datasets/RMDig/rocky_mountain_snowpack)
- Access
- Free and public. No account with the data provider is needed.
- Category
- Environmental Science
- Source
- https://www.rmdig.ai
About this dataset
Field imagery of snowpack layers for snow-science and avalanche-forecasting machine learning, collected in the Colorado Rocky Mountains from 7 snowpits dug between January 2025 and February 2026. Each naturally separating segment of a snow core is photographed three ways: magnified snow-crystal images, non-magnified snowpack profile images and images of the core segment taken with a corer, each re-sampled up to 25 times. There are 4,040 preprocessed samples and 3,195 raw images (983 PNG files in the repository, about 2.7 GB), with per-segment metadata: site, column, core and segment ids, avalanches spotted, wind loading, snowpack and core depth, slope face and angle, air and core temperature, and extended column test (ECT) results. Samples carry time and location so weather or avalanche-danger forecasts can be joined as labels. There is no official split; because many rows show the same core and pit, the card advises leave-one-site-out cross-validation. Load with the datasets library or read the PNGs and JSONL metadata directly. Licensed CC BY 4.0 by Rocky Mountain Digerati (LICENSE.txt).
How to load this dataset in Python
from datasets import load_dataset
ds = load_dataset("RMDig/rocky_mountain_snowpack")
print(ds)Data structure
File types in sampled files: JSON Lines, PNG
Files
Example file: metadata/preprocessed.jsonl
Analyze this dataset without downloading it
On Scigantic, Rocky Mountain Snowpack: Snow Crystal, Profile and Core Imagery from Colorado Snowpits opens in a hosted JupyterLab notebook with the data already mounted, so you can read files where they are stored instead of copying them first. Each notebook includes an AI assistant that can see the mounted files and write and run analysis code. Open the dataset page to launch a notebook, or browse the full catalog.
This page is also available as Markdown and as JSON.
Related Environmental Science datasets
- Carbon and Permafrost Vulnerability to Wildfire v1.0
- Non-Floodplain Wetlands
- WWF Terrestrial Ecoregions of the World (2017)
- DEA Burnt Area Characteristic Layers (Sentinel-2 Near Real-Time, Provisional, Australia)
- New Zealand Wildlife Thermal Imaging (Cacophony Project)
- Phenology Normal Hawaii: PhenoCam Green Chromatic Coordinate Time Series of Individual Tree Crowns at NEON PUUM
- Snapshot Safari South Africa and Kenya Reserves (Karoo, Kgalagadi, Enonkishu, Camdeboo, Mountain Zebra, Kruger)
- The Wilds bioacoustic monitors: passive acoustic recordings, Ohio, summer 2025