metadata

license: cc-by-sa-4.0
tags:
  - earth-observation
  - remote-sensing
  - sentinel-2
  - multi-spectral
  - satellite
size_categories:
  - 1M<n<10M
dataset_info:
  features:
    - name: product_id
      dtype: string
    - name: grid_cell
      dtype: string
    - name: product_datetime
      dtype: string
    - name: thumbnail
      dtype: image
    - name: B01
      dtype: binary
    - name: B02
      dtype: binary
    - name: B03
      dtype: binary
    - name: B04
      dtype: binary
    - name: B05
      dtype: binary
    - name: B06
      dtype: binary
    - name: B07
      dtype: binary
    - name: B08
      dtype: binary
    - name: B8A
      dtype: binary
    - name: B09
      dtype: binary
    - name: B11
      dtype: binary
    - name: B12
      dtype: binary
    - name: cloud_mask
      dtype: binary

Core-S2L2A (🥧 Pre-Release)

Contains a global coverage of Sentinel-2 (Level 1C) patches, each of size 1,068 x 1,068 pixels.

Source	Sensing Type	Number of Patches	Patch Size	Total Pixels
Sentinel-2 Level-2A	Optical Multispectral	2,245,886	1,068 x 1,068 (10 m)	> 2.564 Trillion

Content

Column	Details	Resolution
B01	Coastal aerosol, 442.7 nm (S2A), 442.3 nm (S2B)	60m
B02	Blue, 492.4 nm (S2A), 492.1 nm (S2B)	10m
B03	Green, 559.8 nm (S2A), 559.0 nm (S2B)	10m
B04	Red, 664.6 nm (S2A), 665.0 nm (S2B)	10m
B05	Vegetation red edge, 704.1 nm (S2A), 703.8 nm (S2B)	20m
B06	Vegetation red edge, 740.5 nm (S2A), 739.1 nm (S2B)	20m
B07	Vegetation red edge, 782.8 nm (S2A), 779.7 nm (S2B)	20m
B08	NIR, 832.8 nm (S2A), 833.0 nm (S2B)	10m
B8A	Narrow NIR, 864.7 nm (S2A), 864.0 nm (S2B)	20m
B09	Water vapour, 945.1 nm (S2A), 943.2 nm (S2B)	60m
B11	SWIR, 1613.7 nm (S2A), 1610.4 nm (S2B)	20m
B12	SWIR, 2202.4 nm (S2A), 2185.7 nm (S2B)	20m
thumbnail	RGB composite [B04, B03, B02] saved as png	10m

Spatial Coverage

This is a global monotemporal dataset. Nearly every piece of Earth captured by Sentinel-2 is contained at least once in this dataset (and only once, excluding some marginal overlaps).

The following figure demonstrates the spatial coverage (only black pixels are absent):

Example Use

Interface scripts will be released soon.

Here's a sneak peek with a thumbnail image:

from fsspec.parquet import open_parquet_file
import pyarrow.parquet as pq
from io import BytesIO
from PIL import Image

PARQUET_FILE = 'part_03900' # parquet number
ROW_INDEX = 42 # row number (about 500 per parquet)

url = "https://huggingface.co/datasets/Major-TOM/Core-S2L2A/resolve/main/images/{}.parquet".format(PARQUET_FILE)
with open_parquet_file(url,columns = ["thumbnail"]) as f:
    with pq.ParquetFile(f) as pf:
        first_row_group = pf.read_row_group(ROW_INDEX, columns=['thumbnail'])

stream = BytesIO(first_row_group['thumbnail'][0].as_py())
image = Image.open(stream)

Cite

@inproceedings{Major_TOM,
  title={Major TOM: Expandable Datasets for Earth Observation}, 
  author={Alistair Francis and Mikolaj Czerkawski},
  year={2024},
  eprint={2402.12095},
  archivePrefix={arXiv},
  primaryClass={cs.CV}
}