Skip to content

Latest commit

 

History

History
102 lines (62 loc) · 7.62 KB

WeatherModels.md

File metadata and controls

102 lines (62 loc) · 7.62 KB

Accessing weather model data

RAiDER has built-in support for a number of different weather models. RAiDER provides all the interfacing to data servers required to access data for the different weather models, although some weather models require a license agreement and accounts to be set-up. Instructions for accessing data, including license-limited data, are provided below. It is the user's responsibility to accept license agreements for whatever model is desired.

In addition, RAiDER provides functionality for adding additional weather models. See the RAiDER-docs repository page on how to do this. We would love to expand the suite of supported models, and welcome any contributions. Please see the contributing guidelines or reach out through an issue ticket for help.

1. Usage

::: RAiDER.processWM.prepareWeatherModel options: show_root_heading: true heading_level: 3

Potential download failure

ERA-5/ERA-I products require access to the ESA Copernicus servers. GMAO and MERRA-2 products require access to the NASA Earthdata servers. If you are unable to download products, ensure that you have registered and have downloaded the public API key, and accepted/added the license/application for type of product you wish to download as detailed below.

2. NOAA weather models (HRRR)

High-resolution rapid refresh (HRRR) weather model data products are generated by NOAA for the coninental US (CONUS) but not archived beyond three days. However, a public archive is available at the University of Utah. This archive does not require a license agreement. This model has the highest spatial resolution available in RAiDER, with a horizontal grid spacing of about 3 km, and is provided in a Lambert conformal conic projection.

3. ECMWF weather models (ERA5, ERA5T, ERAI, HRES)

The Copernicus Climate Data Store (CDS) provides access to the European Centre for Medium-Range Weather Forecasts (ECMWF) provides a number of different weather models, including ERA5 and ERA5T reanalysis models.

The ECMWF provides access to both reanalysis and real-time prediction models. You can read more information about their reanalysis models here and real-time model here. ECMWF models are global, with horizontal resolution of about 30 km for ERA-I, ERA-5, and ERA-5T, and 6 km for Hi-RES. All of these models come in a global projection (EPSG 4326, WGS-84).

Accessing ERA5 and ERA5T weather reanalysis products from Copernicus CDS

  1. Create an account on the Copernicus servers here

  2. Confirm your email, etc.

  3. Install the public API key and client as instructed here:

    a. Copy the URL and API key from the webpage into a file in your home directory name ~/.cdsapirc

     url: https://cds.climate.copernicus.eu/api/v2
 key: your_key_here

    Note: the key represents the API key obtained upon the registration of CDS API, and should be replaced with the user's own information.

    b. Install the CDS API using pip

     pip install cdsapi

    Note: this step has been included in the conda install of RAiDER, thus can be omitted if one uses the recommended conda install of RAiDER

  4. You must accept the license for each product you wish to download.

Accessing ERAI, HRES

ECMWF requires a license agreement to be able to access, download, and use their products. Instructions for completing this process is below.

  1. Create an account on the ECMWF servers here. The ERA-I model is open-access, while HRES requires a special licence agreement.

  2. Confirm your email, etc.

  3. Install the public API key and client as instructed here:

    a. Copy the URL and API key from the webpage into a file in your home directory name ~/.ecmwfapirc

     {
     "url"   : "https://api.ecmwf.int/v1",
     "key"   : your key here,
     "email" : your email here
 }

    Note: the email that is used to register the user account, and the key represents the API key obtained upon the registration of ECMWF API, and should be replaced with the user's own information.

    b. Install the ECMWF API using pip:

     pip install ecmwf-api-client`

    Note: this step has been included in the conda install of RAiDER, thus can be omitted if one uses the recommended conda install of RAiDER

4. NASA weather models (GMAO, MERRA2)

  1. The Global Modeling and Assimilation Office (GMAO) at NASA generates reanalysis weather models. GMAO products can also be accessed without a license agreement through the pyDAP interface implemented in RAiDER. GMAO has a horizontal grid spacing of approximately 33 km, and its projection is EPSG code 4326 (WGS-84).

  2. The Modern-Era Retrospective analysis for Research and Applications, Version 2 (MERRA-2) provides data beginning in 1980. MERRA-2 is also produced by NASA and has a spatial resolution of about 50 km and a global projection (EPSG 4326, WGS-84).

Reference: The Modern-Era Retrospective Analysis for Research and Applications, Version 2 (MERRA-2), Ronald Gelaro, et al., 2017, J. Clim., doi: 10.1175/JCLI-D-16-0758.1

Accessing NASA weather model data

  1. Create an account on the NASA's Earthdata website here

  2. Confirm your email, etc.

  3. Copy the login username and password to a file in your home directory name ~/.netrc

      machine urs.earthdata.nasa.gov
          login <USERNAME>
          password <PASSWORD>

    Note: the username and password represent the user's username and password.

  4. Add the application NASA GESDISC DATA ARCHIVE by clicking on the Applications->Authorized Apps on the menu after logging into your Earthdata profile, and then scrolling down to the application NASA GESDISC DATA ARCHIVE to approve it. This seems not required for GMAO for now, but recommended to do so for all OpenDAP-based weather models.

  5. Install the OpenDAP using pip:

      pip install pydap==3.2.1

    Note: this step has been included in the conda install of RAiDER, thus can be omitted if one uses the recommended conda install of RAiDER

    Note: PyDAP v3.2.1 is required for now (thus specified in the above pip install command) because the latest v3.2.2 (as of now) has a known bug in accessing and slicing the GMAO data. This bug is expected to be fixed in newer versions of PyDAP.