CDSETool

About CDSETool

This script downloads copernicus data from the Copernicus Data Space Ecosystem

Quick start

from cdsetool.query import query_features, shape_to_wkt
from cdsetool.credentials import Credentials
from cdsetool.download import download_features
from cdsetool.monitor import StatusMonitor
from datetime import date

features = query_features(
    "SENTINEL-1",
    {
        "contentDateStartGe": "2020-12-20",
        "contentDateStartLe": date(2020, 12, 25),
        "processingLevel": "LEVEL1",
        "productType": "IW_GRDH_1S",
        "geometry": shape_to_wkt("path/to/shapefile.shp"),
    },
)

list(
    download_features(
        features,
        "path/to/output/folder/",
        {
            "concurrency": 4,
            "monitor": StatusMonitor(),
            "credentials": Credentials("username", "password"),
        },
    )
)

Or use the CLI:

cdsetool query search SENTINEL-2 --search-term contentDateStartGe=2020-01-01 --search-term contentDateStartLe=2020-01-10 --search-term productType=S2MSI2A

cdsetool download SENTINEL-2 PATH/TO/DIR --concurrency 4 --search-term contentDateStartGe=2020-01-01 --search-term contentDateStartLe=2020-01-10 --search-term productType=S2MSI2A

Table of Contents

• CDSETool
  • About CDSETool
  • Quick Start
  • Table of Contents
  • Installation
  • Usage
    • Querying features
      • Querying by shapes
      • Querying by lists of parameters
      • Querying by dates
      • Listing search terms
    • Downloading features
      • Authenticating
      • Concurrently downloading features
      • Sequentially downloading features
  • Roadmap
  • Contributing
  • LICENSE

Installation

Install cdsetool using pip:

pip install cdsetool==0.2.13

Usage

Querying features

Querying is always done in batches, returning len(results) <= top records each time. A local buffer is filled and gradually emptied as results are yielded. When the buffer is empty, more results will be requested and the process repeated until no more results are available, or the iterator is discarded.

Important: The API has a pagination limit of 10,000 results per query. If your query returns more results, you'll need to narrow your search criteria (e.g., use smaller date ranges).

Since downloading features is the most common use-case, query_features assumes that the query will run till the end. Because of this, the batch size is set to 1000, which is the size limit set by CDSE.

from cdsetool.query import query_features

collection = "SENTINEL-2"
search_terms = {
    "top": 100,  # batch size, between 1 and 1000 (default: 1000)
    "contentDateStartGe": "2024-01-01",
    "productType": "S2MSI1C"
}

# wait for a single batch to finish, yield results immediately
for feature in query_features(collection, search_terms):
    # do something with feature

# wait for all batch requests to complete, returning list
features = list(query_features(collection, search_terms))

# manually iterate
iterator = query_features(collection, search_terms)

featureA = next(iterator)
featureB = next(iterator)
# ...

Query Options

Product Format

Query results are returned directly from the Copernicus API. Each product has the following structure:

product = {
    "Id": "uuid-string",
    "Name": "S2A_MSIL2A_20240110T105421_...",
    "Collection": "SENTINEL-2",
    "ContentDate": {"Start": "2024-01-10T10:54:21Z", "End": "..."},
    "Online": True,
    "ContentLength": 1043654649,
    "GeoFootprint": {"type": "Polygon", "coordinates": [...]},
    "Attributes": [
        {"Name": "productType", "Value": "S2MSI2A"},
        {"Name": "cloudCover", "Value": 5.2},
        ...
    ]
}

Expand Product Attributes

By default, query results do not include product attributes (productType, cloudCover, platform, instrument, etc.). To include product attributes, you need to request this from the server using the option expand_attributes and can then access the attributes using the function get_product_attribute():

from cdsetool.query import query_features, get_product_attribute

features = query_features(
    "SENTINEL-2",
    {"contentDateStartGe": "2024-01-01"},
    options={"expand_attributes": True}
)
feature = features[0]

# Access basic properties directly
print(feature["Name"])  # Product name
print(feature["Id"])    # Product UUID

# Access attributes using helper function
cloud_cover = get_product_attribute(feature, "cloudCover")
product_type = get_product_attribute(feature, "productType")

Querying by geometry

To query by shapes, you must first convert your shape to Well Known Text (WKT). The included shape_to_wkt can solve this.

from cdsetool.query import query_features, shape_to_wkt

geometry = shape_to_wkt("path/to/shape.shp")
features = query_features("SENTINEL-3", {"geometry": geometry})

Querying by lists of parameters

Most search terms only accept a single argument. To query by a list of arguments, loop the arguments and pass them one by one to the query function.

from cdsetool.query import query_features

tile_ids = ["32TPT", "32UPU", "32UPU", "31RFL", "37XDA"]

for tile_id in tile_ids:
    features = query_features("SENTINEL-2", {"tileId": tile_id})
    for feature in features:
        # do things with feature

Querying by dates

Its quite common to query for features created before, after or between dates.

Search terms support comparison operator suffixes:

Suffix	Meaning	Example
Eq	equals (=)	contentDateStartEq
Gt	greater than (>)	contentDateStartGt
Ge	greater than or equal (>=)	contentDateStartGe
Lt	less than (<)	contentDateStartLt
Le	less than or equal (<=)	contentDateStartLe

Eq can be applied on any field but the other suffixes can only be applied to numeric and date fields.

Interval syntax is only allowed on the base name, not on suffixed variants, and only on numeric and date fields.

Interval notation	Suffixes to combine	Meaning
[a, b]	Ge + Le	a <= value <= b (closed)
(a, b)	Gt + Lt	a < value < b (open)
[a, b)	Ge + Lt	a <= value < b (half-open)
(a, b]	Gt + Le	a < value <= b (half-open)

from cdsetool.query import query_features
from datetime import date, datetime

date_from = date(2020, 1, 1) # or datetime(2020, 1, 1, 23, 59, 59, 123456) or "2020-01-01" or "2020-01-01T23:59:59.123456Z"
date_to = date(2020, 12, 31)

features = query_features("SENTINEL-2", {"contentDateStartGe": date_from, "contentDateStartLe": date_to, "cloudCover": "[0, 30]"})

Listing search terms

To get a list of all search terms for a given collection, you may either use the describe_collection function or use the CLI:

from cdsetool.query import describe_collection

search_terms = describe_collection("SENTINEL-2").keys()
print(search_terms)

And the CLI:

$ cdsetool query search-terms SENTINEL-2

Downloading features

Authenticating

An account is required to download features from the Copernicus distribution service.

To authenticate using an account, instantiate Credentials and pass your username and password

from cdsetool.credentials import Credentials

username = "[email protected]"
password = "password123"
credentials = Credentials(username, password)

Alternatively, Credentials can pull from ~/.netrc when username and password are left blank.

# ~/.netrc
machine https://identity.dataspace.copernicus.eu/auth/realms/CDSE/protocol/openid-connect/token
login [email protected]
password password123

# main.py
from cdsetool.credentials import Credentials

credentials = Credentials()

The credentials object may then be passed to a download function. If left out, the download functions will default to using .netrc.

credentials = Credentials()

download_features(features, "/some/download/path", {"credentials": credentials})

Credentials can be validated using the validate_credentials function which will return a boolean.

from cdsetool.credentials import validate_credentials

validate_credentials(username='user', password='password')

If None are passed to username and password, validate_credentials will validate .netrc

Concurrently downloading features

CDSETool provides a method for concurrently downloading features. The concurrency level should match your accounts privileges. See CDSE quotas

The downloaded feature ids are yielded, so its required to await the results.

from cdsetool.query import query_features
from cdsetool.download import download_features

features = query_features("SENTINEL-2", {"contentDateStartGe": "2024-01-01", "contentDateStartLe": "2024-01-10"})

download_path = "/path/to/download/folder"
downloads = download_features(features, download_path, {"concurrency": 4})

for id in downloads:
    print(f"feature {id} downloaded")

# or

list(downloads)

Sequentially downloading features

Its possible to download features sequentially in a single thread if desired.

from cdsetool.query import query_features
from cdsetool.download import download_feature

features = query_features("SENTINEL-2", {"contentDateStartGe": "2024-01-01", "contentDateStartLe": "2024-01-10"})

download_path = "/path/to/download/folder"
for feature in features:
    download_feature(feature, download_path)

Download specific files within features

It's possible to download specific files within products bundles using Unix filename pattern matching.

It can be used in CDSETool:

• Through the filter_pattern option of download_features and download_feature:

  from cdsetool.query import query_features
  from cdsetool.download import download_features
  
  features = query_features("SENTINEL-2", {"contentDateStartGe": "2024-01-01", "contentDateStartLe": "2024-01-10"})
  
  download_path = "/path/to/download/folder"
  filter_pattern = "*TCI.jp2"
  downloads = download_features(features, download_path, {"filter_pattern": filter_pattern})
  
  for id in downloads:
      print(f"feature {id} downloaded")
  

• Or through the CLI:

  cdsetool download SENTINEL-2 PATH/TO/DIR --filter-pattern *TCI.jp2 --concurrency 4 --search-term contentDateStartGe=2024-01-01 --search-term contentDateStartLe=2024-01-10 --search-term productType=S2MSI2A
  

Roadmap

• Query schema validation
• High-level API
  • Query features
  • Download features
    • Download single feature
    • Download list of features
    • Download by ID
    • Download by URL
• Command-Line Interface
  • Update to match the high-level API
  • Better --help messages
  • Quickstart guide in README.md
• Test suite
  • Query
  • Credentials
  • Download
  • Monitor
• Strategy for handling HTTP and connection errors

Contributing

Any contributions you make are greatly appreciated.

If you have a suggestion that would make this better, please fork the repo and create a pull request. You can also simply open an issue with the tag "enhancement".

1. Fork the Project
2. Create your Feature Branch (git checkout -b feature/cool-new-feature)
3. Commit your Changes (git commit -m 'Add some feature')
4. Push to the Branch (git push origin feature/cool-new-feature)
5. Open a Pull Request

LICENSE

Distributed under the MIT License. See LICENSE for more information.