TreeSatAI Time Series: Tree Species Labels with Aerial Imagery and Yearly Sentinel-1/2 Series (Lower Saxony)
- Provider
- IGN France (Astruc G. et al.), extending TreeSatAI by Ahlswede et al.
- License
- CC-BY-SA-4.0 (source)
- Size
- 9 files
- Format
- PNG (in sampled files)
- Stored at
- Hugging Face (https://huggingface.co/datasets/IGNF/TreeSatAI-Time-Series)
- Access
- Free and public. No account with the data provider is needed.
- DOI
- 10.5194/essd-15-681-2023
- Category
- Earth Observation & Geospatial
- Source
- https://arxiv.org/abs/2404.08351
About this dataset
Multi-label tree species classification over 50,381 forest patches of 60 m x 60 m in Lower Saxony, Germany, with labels for 20 European tree species (15 genera) from state forest administration data. It extends the TreeSatAI Benchmark Archive of Ahlswede et al. (ESSD 2023), which gave one Sentinel-1 and one Sentinel-2 image per patch, by adding all Sentinel-1 ascending and descending backscatter (VV, VH) and Sentinel-2 L2A reflectance (10 bands with cloud and snow probability masks) acquired over a year, aligned with the year of the 0.2 m RGB-infrared aerial patch (2017 for earlier flights). Zips hold aerial, sentinel (single-date), sentinel-ts (HDF5 per patch with product names), labels (species presence and proportion), geojson patch locations and train, val and test splits. Measured: 9 files, about 22.8 GB. It was introduced with OmniSat (ECCV 2024) and used by MAESTRO; read the time series with h5py. Licensed CC BY-SA 4.0 per the card; cite Ahlswede et al. 2023 and Astruc et al. 2024.
How to load this dataset in Python
from huggingface_hub import hf_hub_download
path = hf_hub_download(repo_id="IGNF/TreeSatAI-Time-Series", filename="TreesatAI-TS-fig.png", repo_type="dataset")
print(path)Data structure
File types in sampled files: PNG
Files
Example file: TreesatAI-TS-fig.png
Analyze this dataset without downloading it
On Scigantic, TreeSatAI Time Series: Tree Species Labels with Aerial Imagery and Yearly Sentinel-1/2 Series (Lower Saxony) 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 Earth Observation & Geospatial datasets
- Tokyo 13999 district in FlatCityBuf
- Digital Earth Africa Fractional Cover
- Satellite - Ocean Colour - NOAA20 - 1 day - Diffuse attenuation coefficient (k490)
- USGS 3DEP Alaska IfSAR Digital Surface Models (DSM), The National Map
- NREL NSRDB 2km-10min data for the all of the Western Hemisphere (GOES full-disc) starting with 2018 in HDF5
- Satellite - Sea surface temperature - Level 3 - Single sensor - 1 day - Day and night time - Southern Ocean
- HIRLAM Weather Model
- OPERA Land Surface Disturbance Alert from Harmonized Landsat Sentinel-2 product (Version 1)