Get the data to be used

We are going to use monthly averaged surface variables from ERA5.

Four variables:

  • 2m_temperature

  • sea_surface_temperature

  • mean_sea_level_pressure

  • total_precipitation

We also want a land-sea mask (for plotting only). Use a land-surface only variable from ERA5-land for this (we only need one month).

We can download all this from the awesome Copernicus Climate Data Store

You will need to register with the CDS before downloading. (You will also need to install the CDS API - but you’ve already done that, it’s in the conda environment).

The script to do the whole download (about 8Gb, will take a few hours) is download_all_data.sh in the get_data directory. It only downloads data where it is not already on disc (so you can just run it again if it fails part way through).

#!/usr/bin/bash

# Run all the data downloads

# Retrieves about 8Gb data.
# Will take a few hours - depends on CDS load.

(cd ERA5 && ./get_data_for_period_ERA5.py | parallel -j 1)

(cd land_mask && ./get_land_mask_from_ERA5_land.py)