Users can access and download subsets of the authorized datasets using a web interface to select the region and variables of interest (not available for all datasets). Alternatively, some packages have been developed in order to remotely access subsets of the datasets from scientific scientific computing environments (mainly from R and Matlab):

Web interface to explore and download data

The User Data Gateway (UDG) builds on a THREDDS data server which provides different web interfaces to explore and access the datasets. For instance, the NetCDF Subset Service allows to select (and download as a NetCDF file) a subset of the data (some variables in a particular region for a particular time slice).

In this example, we illustrate the use of this service to select geo-temporal data subsets from the ERA-Interim (predictors for downscaling) dataset (available for VALUE and CORDEX-ESD users). We start by clicking in the catalog link of the corresponding dataset:

Then resulting web page (shown in the figure below) allows selecting a number of variables (by clicking in the corresponding checkbox), geographical domain (by choosing the boundary coordinates of the spatial subset), and period (by specifying the initial and end date of the time subset). Note that the horizontal (and temporal) stride options allow selecting a sub-sampled spatial (and temporal) subsets, e.g. 3 = every third point (or time steps). Note that the 3D predictors are labelled as "VARLEVEL", where VAR is the name of the variable and LEVEL denotes the vertical coordinate (in mb); e.g. T850 refers to 850mb temperature.

After selecting the subset of interest, the button "Submit" allows sending the request resulting in a NetCDF file. Note that this service is not efficient to download the whole global dataset.

Data access (remote access via R)

  • The loadeR package allows accessing observational and model simulation datasets using a single command.
  • The loadeR.ECOMS package extends loaderR focusing on the seasonal forecast data used in the ECOMS initiative.

The use of these packages is strongly recommended for seasonal forecasting products, since the use of two temporal axis (for initialization and run time, respectively) makes it difficult to select the desired information from the standard web tool for data downloading.

Last modified 4 years ago Last modified on May 11, 2016 1:38:48 PM

Attachments (1)

Download all attachments as: .zip