CLIMAAX datasets

Note

These pages are under construction.

Workflows as presented in this handbook mostly rely on public datasets hosted, e.g., by the CDS. However, some workflows have specific data requirements (in terms of variables or resolution) that cannot easily be satisfied with public datasets or simply require large amounts of data as input. In these cases, preprocessed datasets and data samples can be provided by CLIMAAX developers and utilized in the workflows to lower the threshold of entry for users and serve as examples for the datasets required for running the climate risk assessment workflows. Such data is hosted on the CLIMAAX cloud storage and described here.

This illustration is created by Scriberia with The Turing Way community. Used under a CC-BY 4.0 licence. DOI: 10.5281/zenodo.3332807

Available Datasets

• Drought risk data (NUTS3)

• Precipitation hazard data (Europe)

• Wildfire training and vulnerability data (Catalonia)

• Partial mirror of ECLIPS-2.0 (Europe)

How to access

pooch (recommended)

We provide file registries for use with the pooch Python package.

1. Set up a download manager for a CLIMAAX dataset

   import pooch
   
   climaax_data = pooch.create(
       base_url="https://object-store.os-api.cci1.ecmwf.int/climaax/<DATASET-ID>/",
       path="." # set your download location
   )
   

   where <DATASET-ID> is a placeholder for the identifier of the accessed dataset.

2. Download the registry.txt file from the corresponding dataset page. Load the pooch registry for the dataset (adapt the file path to your download location):

   climaax_data.load_registry("registry.txt")
   

3. Individual files from the dataset can be downloaded with:

   climaax_data.fetch("path/to/file")
   

   and the registry can be used to easily fetch an entire dataset:

   for path in climaax_data.registry:
       climaax_data.fetch(path)
   

   Pooch will download the files into a folder structure as laid out by the dataset.

file listing

Each dataset page contains a listing of all files included in the dataset. Individual files can be downloaded via the links in the listings from a web browser or with tools such as wget and curl.