Skip to content

Latest commit

 

History

History
21 lines (15 loc) · 1.45 KB

README.md

File metadata and controls

21 lines (15 loc) · 1.45 KB

Climate Informatics 2023: 'A locally time-invariant metric for climate model ensemble predictions of extreme risk'

Code and data associated with the submission 'A locally time-invariant metric for climate model ensemble predictions of extreme risk'.

Data download

Data used to produce the results presented in tha paper are provided in the folder 'data'. These are time-series taken from one realisation per model for the CMIP6 members GFDL-ESM4, IPSL-CM6A-LR, MPI-ESM1-2-HR, MRI-ESM2-0, and UKESM1-0-LL, for nine cities Paris, Chicago, Sydney, Tokyo, Kolkata, Kinshasa, Shenzhen and Santo Domingo. The same time-series from an observational reanalysis data, W5E5, are also provided.

-- Notes: I'm using Python 3.11.5

  1. Some non-exhaustive dependencies I've had to install:
pip3 install esgf-pyclient
pip3 install geopy
pip3 install xclim
pip3 install netcdf4
  1. Download one of the nc dataset files from the reanalysis reference dataset W5E5 Data set. The smallest one is this one.
  2. python3 data_download.py or simply use data_download.ipynb. For convenience, the resulting datasets you would get from running this (using the example smallest dataset in step 2.), has already been pushed to the repo in the directory data/.