Downloading with CDSE

[1]:

# install with
# !pip install terragon-downloader[cdse]

[1]:

import terragon
import geopandas as gpd
import xarray as xr
from utils import visualize_sat_images, list_of_dicts_to_tree
import rasterio

[3]:

# we will use the sample data to download a minicube
# other datasets can be found here: https://dataspace.copernicus.eu/explore-data/data-collections
gdf = gpd.read_file("data/TUM_OTN.geojson")
gdf = gdf.to_crs("EPSG:32632")
arguments = dict(
    shp=gdf,
    collection="SENTINEL-2",
    start_date="2021-01-01",
    end_date="2021-01-05",
    bands=["B02", "B03", "B04"],
    resolution=10,
    download_folder="tests/download/",
    filter={
        "processingLevel": {"eq": "S2MSI2A"}  # we will use a filter to get only level 2 products
    },
)
gdf.crs

[3]:

<Projected CRS: EPSG:32632>
Name: WGS 84 / UTM zone 32N
Axis Info [cartesian]:
- E[east]: Easting (metre)
- N[north]: Northing (metre)
Area of Use:
- name: Between 6°E and 12°E, northern hemisphere between equator and 84°N, onshore and offshore. Algeria. Austria. Cameroon. Denmark. Equatorial Guinea. France. Gabon. Germany. Italy. Libya. Liechtenstein. Monaco. Netherlands. Niger. Nigeria. Norway. Sao Tome and Principe. Svalbard. Sweden. Switzerland. Tunisia. Vatican City State.
- bounds: (6.0, 0.0, 12.0, 84.0)
Coordinate Operation:
- name: UTM zone 32N
- method: Transverse Mercator
Datum: World Geodetic System 1984 ensemble
- Ellipsoid: WGS 84
- Prime Meridian: Greenwich

credentials

[2]:

# we use the s3 bucket to download the data: https://documentation.dataspace.copernicus.eu/APIs/S3.html
# credentials can be created here: https://eodata-s3keysmanager.dataspace.copernicus.eu/panel/s3-credentials
credentials = {"aws_access_key_id": "<your-id>", "aws_secret_access_key": "<your-key>"}
tg = terragon.init("cdse", credentials=credentials)

Finding collections

[5]:

# explore the CDSE data on their website: https://dataspace.copernicus.eu/data-collections, https://browser.dataspace.copernicus.eu/
collections = tg.retrieve_collections(
    query={"id": "Sentinel.1.*"}, fields=["id", "title"]
)  # filter field with regex and only return wished fields
list_of_dicts_to_tree(collections)

[5]:

download (with eodata AWS S3 backend)

[6]:

# we can now download the data, in the background we search the folder of the corresponding tile for the correct files
# and use rasterio with its virtual files to open and clip the extend of the files directly,
# it is possible to download data with up to 4 concurrent workers per account, see here: https://documentation.dataspace.copernicus.eu/Quotas.html
# make sure to have the right packages installed (depending on the data you want to download), for example jp2 needs libgdal-jp2openjpeg and .nc needs libgdal-hdf5
ds = tg.create(**arguments, num_workers=4)
# or
# items = tg.search(**arguments, num_workers=4)
# ds = tg.download(items)
ds

[7]:

ds = ds.where(ds != 65535)  # remove no data values
visualize_sat_images(ds, gdf, list(reversed(arguments["bands"])))

../_images/demo_files_demo_cdse_10_0.png

special arguments for cdse

[8]:

# filter_asset_path: filtering asset paths with regex
# sometimes the same file names exists with a different folder path, for example for Sentinel-2 the image data is in the folder IMG_DATA
# defaults are already set for COP-DEM and Sentinel-2
ds = tg.create(**arguments, num_workers=4, filter_asset_path={"SENTINEL-2": ".*/IMG_DATA/.*"})

[9]:

# use_virtual_rasterio_file: downloading the tile files instead of using the virtual rasterio file - default: bool = True
# in combination with rm_tmp_files=True, the tile files will be deleted after the download
ds = tg.create(**arguments, use_virtual_rasterio_file=False, rm_tmp_files=True)

[10]:

# resampling: using a different resampling method in the reprojection process - default: rasterio.enums.Resampling = rasterio.enums.Resampling.nearest
ds = tg.create(**arguments, resampling=rasterio.enums.Resampling.bilinear)

filtering meta data

[11]:

# it is also possible to filter for more meta data from the tiles,
# we match them with the meta data from the stac items: https://documentation.dataspace.copernicus.eu/APIs/STAC.html
# we support four types of filters: eq, ueq, lt, gt, in (equal, unequal, less than, greater than, python "in" with list)
# for example we can use this to filter for cloud cover:
arguments["filter"].update({"cloudCover": {"lt": 90}})
ds = tg.create(**arguments)
ds = ds.where(ds != 65535)  # remove no data values
visualize_sat_images(ds, gdf, list(reversed(arguments["bands"])))

../_images/demo_files_demo_cdse_16_0.png