Skip to content

Latest commit

 

History

History
71 lines (49 loc) · 4.9 KB

Data_Sources_Info.md

File metadata and controls

71 lines (49 loc) · 4.9 KB
Raw

Data Sources Info

ERA5 Global Climate Model Data Accessors

Copernicus Data Store API - CDSDataAccessor

API Documentation

Note: To use the CDS API, one must save a "dot-file" with their API key on it after making and logging into their account. Specific instructions are available here. One can ignore the instructions about installing cdsapi as that is included in this library's environment.yml file.

Single levels (hourly) datasets - info

  • reanalysis-era5-single-levels
  • reanalysis-era5-single-levels-preliminary-back-extension

Single levels (monthly) datasets - info (not supported as of 4/5/2023)

  • reanalysis-era5-single-levels-monthly-means
  • reanalysis-era5-single-levels-monthly-means-preliminary-back-extension

Pressure levels datasets - info (not supported as of 4/5/2023)

  • reanalysis-era5-pressure-levels
  • reanalysis-era5-pressure-levels-monthly-means
  • reanalysis-era5-pressure-levels-preliminary-back-extension
  • reanalysis-era5-pressure-levels-monthly-means-preliminary-back-extension

ERA5-land datasets - info

  • reanalysis-era5-land
  • reanalysis-era5-land-monthly-means

Planet OS AWS bucket - AWSDataAccessor

ERA-5 single/surface level data is also available on an AWS s3 bucket. However, there are far fewer variables available, and the data goes back to 1979 (as opposed to 1959).

Single levels (hourly) datasets - info

  • reanalysis-era5-single-levels

Note that the variables are named differently. To see the variable names available one can run the following function and be returned a list:

from xarray_data_accessor import DataAccessorFactory

DataAccessorFactory.supported_variables(
    data_accessor_name='AWSDataAccessor',
    dataset_name='reanalysis-era5-single-levels',
)

A crosswalk between CDS and AWS variables names can be found here.

A note on CDSDataAccessor vs AWSDataAccessor

While reading data from AWS can be much faster than the CDS API (especially for large time ranges), loading the AWS data to disk is much slower!

We recommend using AWSDataAccessor for data visualization and xarray native workflows. However, if you want to sample the data and convert into a pandas data frame (i.e., via xarray_data_accessor.get_data_tables()), using CDSDataAccessor will be significantly faster.

Another relevant difference is that AWS ERA5 data is returned along a uniform 0.25 decimal degree grid (i.e., 0.25, 0.5, 0.75,...) while CDS returns a grid with 0.25 increments as well, but centered based on the bounding box.

NASA DataAccessors

Note: For all NASA DataAccessors one must have an active EarthData Account, and pass in your username/password via the following get_xarray_dataset() keyword argument authorization={'username': 'example_username', 'password': 'example_password'}.

NASA/USGS LP DAAC DataPool - NASA_LPDAAC_Accessor

Organization information.

NASADEM_NC - info

  • This dataset provides a global 30m Digital Elevation Model derived from the Shuttle Radar Topography Mission (SRTM). Access to the QA/QC layers documented in the attached link ("NUM" and "SWB") are not currently supported, the only associated variable is "DEM" providing elevation relative to sea level in meters.

NASADEM_SC - info

  • This dataset provides DEM by-products. See the "Layers" section of the dataset documentation for details.

Note: Both of the NASADEM datasets have no time dimension. However, one must still provide a start_time/end_time argument to get_xarray_dataset(). The time provided will not effect the data pulled.

GLanCE30 - info

  • This dataset provides a (mostly) global 7 class Land Cover (LC) grid at 30m / yearly resolution. Note that there are areas without coverage. The "LC" variable provides the main grid, however other documented variables can be accessed that track land cover class changes over time.
  • Note: All GLanCE30 data is collected on July 1st (7/1), therefore if your date range does not pass over July no data will be returned!