rstac

R Client Library for SpatioTemporal Asset Catalog (rstac)

STAC is a specification of files and web services used to describe geospatial information assets. The specification can be consulted at https://stacspec.org/.

The R client library for STAC (rstac) was designed to fully support STAC API v1.0.0. It also supports earlier versions (>= v0.8.0).

Installation

# install via CRAN 
install.packages("rstac")

Development version

To install the development version of rstac, run the following command:

remotes::install_github("brazil-data-cube/rstac")

Load the rstac package:

library(rstac)

Usage

rstac supports the following STAC endpoints:

STAC endpoints	rstac functions	API version
/	stac()	>= 0.9.0
/stac	stac()	< 0.9.0
/collections	collections()	>= 0.9.0
/collections/{collectionId}	collections(collection_id)	>= 0.9.0
/collections/{collectionId}/items	items()	>= 0.9.0
/collections/{collectionId}/items/{itemId}	items(feature_id)	>= 0.9.0
/search	stac_search()	>= 0.9.0
/stac/search	stac_search()	< 0.9.0
/conformance	conformance()	>= 0.9.0
/collections/{collectionId}/queryables	queryables()	>= 1.0.0

These functions can be used to retrieve information from a STAC API service. The code below creates a stac object and lists the available collections of the STAC API of the Brazil Data Cube project of the Brazilian National Space Research Institute (INPE).

s_obj <- stac("https://data.inpe.br/bdc/stac/v1/")

get_request(s_obj)
#> ###Catalog
#> - id: INPE
#> - description: 
#> This is the landing page for the INPE STAC server. The SpatioTemporal Asset Catalogs (STAC) provide a standardized way to expose collections of spatial temporal data. Here you will find collections of data provided by projects and areas of INPE.
#> - field(s): type, title, description, id, stac_version, links, conformsTo

The variable s_obj stores the information needed to connect to the Brazil Data Cube STAC web service. The get_request method makes an HTTP GET request and retrieves a STAC Catalog document from the server. Each links entry refers to an available collection that can be accessed via the STAC API.

In the code below, we get some STAC items from the CB4-16D-2 collection that intersects the bounding box passed to the bbox parameter. To do this, we call the stac_search function, which implements the STAC /search endpoint. The returned document is a STAC Item Collection (a GeoJSON containing a feature collection).

it_obj <- s_obj %>%
  stac_search(collections = "CB4-16D-2",
              bbox = c(-47.02148, -17.35063, -42.53906, -12.98314),
              limit = 100) %>% 
  get_request()

it_obj
#> ###Items
#> - matched feature(s): 0
#> - features (0 item(s) / 0 not fetched):
#> - assets: 
#> - item's fields:

rstac uses the httr package to manage HTTP requests, allowing the use of tokens from the authorization protocols OAuth 1.0 or 2.0 as well as other configuration options. In the code below, we present an example of how to pass a token in an HTTP request header.

it_obj <- s_obj %>%
  stac_search(collections = "CB4-16D-2",
              bbox = c(-47.02148, -17.35063, -42.53906, -12.98314)) %>%
  get_request(add_headers("x-api-key" = "MY-TOKEN"))

In addition to the functions mentioned above, the rstac package provides additional functions for handling items and bulk-downloading assets.

Item functions

rstac provides functions that facilitate interaction with STAC data. In the example below, we get how many items matched the search criteria:

# it_obj variable from the last code example
it_obj %>% 
  items_matched()
#> [1] 0

However, items_length() counts only the items currently stored in it_obj. If this value is smaller than items_matched(), more items can be fetched from the STAC service:

it_obj %>% 
  items_length()
#> [1] 0

# fetch all items from server 
# (and store them back in `it_obj`)
it_obj <- it_obj %>% 
  items_fetch(progress = FALSE) 

it_obj %>%
  items_length()
#> [1] 0

Download assets

All we got in the previous example was metadata for STAC Items, including links to geospatial data called assets. To download all assets in a STAC Item Collection, you can use assets_download(), which returns an updated STAC Item Collection referring to the downloaded assets. The code below downloads the thumbnail assets (.png files) of up to 10 items stored in it_obj.

download_items <- it_obj %>%
  assets_download(assets_name = "thumbnail", items_max = 10)

CQL2 query filter

rstac also supports advanced query filtering using the Common Query Language (CQL2). Users can write complex filter expressions using R code in an easy and natural way. For a complete list of supported operators and helper functions, see ?ext_filter.

s_obj <- stac("https://planetarycomputer.microsoft.com/api/stac/v1")

it_obj <- s_obj %>% 
  ext_filter(
    collection == "sentinel-2-l2a" && `s2:vegetation_percentage` >= 50 &&
      `eo:cloud_cover` <= 10 && `s2:mgrs_tile` == "20LKP" && 
      anyinteracts(datetime, interval("2020-06-01", "2020-09-30"))
  ) %>%
  post_request()

Getting help

You can get a full explanation of each STAC (v1.0.0) endpoint at STAC API spec. Detailed documentation with examples on how to use each endpoint and other functions available in the rstac package can be obtained by typing ?rstac in the R console.

Citation

To cite rstac in publications, use:

R. Simoes, F. C. de Souza, M. Zaglia, G. R. de Queiroz, R. D. C. dos Santos and K. R. Ferreira, “Rstac: An R Package to Access Spatiotemporal Asset Catalog Satellite Imagery,” 2021 IEEE International Geoscience and Remote Sensing Symposium IGARSS, 2021, pp. 7674-7677, doi: 10.1109/IGARSS47720.2021.9553518.

Acknowledgments for financial support

We acknowledge and thank the project funders that provided financial and material support:

• Amazon Fund, established by the Brazilian government with financial contribution from Norway, through the project contract between the Brazilian Development Bank (BNDES) and the Foundation for Science, Technology and Space Applications (FUNCATE), for the establishment of the Brazil Data Cube, process 17.2.0536.1.

• Radiant Earth Foundation and STAC Project Steering Committee for advancing the STAC ecosystem.

• OpenGeoHub Foundation and the European Commission (EC) through the project Open-Earth-Monitor Cyberinfrastructure: Environmental information to support EU’s Green Deal (1 Jun. 2022 – 31 May 2026 - 101059548)

How to contribute

The rstac package was implemented based on an extensible architecture, so feel free to contribute by implementing new STAC API extensions/fragments based on the STAC API specifications.

1. Fork the project by creating a fork.
2. Create a file inside the R/ directory called ext_{extension_name}.R.
3. In the code, specify a subclass name (e.g., my_subclass) for your extension and use it when calling rstac_query(). You also need to implement the following S3 generic methods for your subclass: before_request(), after_response(), and parse_params(). With these S3 generic methods, you can define how parameters are submitted in HTTP requests and how returned documents are parsed. See the implemented ext_filter API extension as an example.
4. Make a Pull Request against the most recent development branch.