Data

This repository contains two kinds of data. There are files resulting from running the Climate Variability Diagnostics Package, which gives a single file per simulation stored in NetCDF format, and then some spreadsheets collating together individual numbers from across the ensemble (stored in data_frames as csv files).

Data Frames

These spreadsheets are in comma separated value format. They can be downloaded and read in using software such as Excel. To read into Python, you can use pandas

import pandas as pd

df = pd.read_csv(’file.csv’)

The file called ESGF_doi.csv contains the digital object identifiers for every simulation on the Earth System Grid Federation. Ideally, these should be cited by every paper using the simulation data and hopefully collating them here makes this easier.

If a simulation’s value is listed as -999.00 then is should be treated as a missing number. Generally this is becuase the simulation has not been performed, although a small number result from the required output variable not being available

Climate Modes

These tables collate together indices used to track different modes of (interannual) climate variability.

The modes defined by timeseries are summarised by the temporal standard deviation of the index

The modes defined by principal compenents are summarised by the spatial standard deviation of the pattern

Monsoon Domains

Interannual time series of the areal extent and area-averaged rainfall of the global monsoon domain, as well as each of the 7 regional monsoon domains were computed for each simulation. The means and standard deviation of each time series are collated into various spreadsheets. The 7 regional monsoon domains are abbreviated as:

---
header-rows: 1
---
* - Acronym 
  - Domain
* - AUSMC
  - Australian and the Maritime Continent
* - EAS
  - Eastern Asia
* - NAF
  - Northern Africa
* - NAMS
  - North America Monsoon System
* - SAF
  - Southern Africa
* - SAMS
  - South America Monsoon System
* - SAS
  - Southern Asia

AR5 Regions

The means and standards of monthly resolution area-averaged time series for every region used in IPCC Assessment Report 5. See http://www.ipcc-data.org/guidelines/pages/ar5_regions.html for definitions of the regions and explanations of their abbreviations.

Common Measures

Some metrics that people often compute from the simulations are summarised:

Temperature Changes in Latitude Bands

The change in temperature in 30o latitude bands over land, ocean and both are collated in this directory. This files have been collected from individual publications and ported into the similar format. They make the same assumptions about changing land sea masks. These files were used as input to IPCC WG1 AR6 Fig 3.2b

NetCDF files

The repository comes bundled with a compressed file (tar.gz). These files contain a small subset of the available fields- just the spatial annual mean, JJA and DJF surface temperature and precipitation. It can be uncompressed with tar -xf PMIP4_tas_pr_data.tar.gz. This only provides a small selection of the available fields - bundling everything makes for an unwielding repository.

You can download files with all the available variables in them from http://www2.geog.ucl.ac.uk/~p2f-v/PMIP4-PMIP3_allfields.tar.gz. To download unpack this data type the following in a terminal (such as is opened by JupyterLab in the docker image)

cd data_netcdf/ wget http://www2.geog.ucl.ac.uk/~p2f-v/PMIP4-PMIP3_allfields.tar.gz tar -xf PMIP4-PMIP3_allfields.tar.gz

Source of CMIP6-PMIP4 model output

Nearly all the data associated with this repository originates from the Earth System Grid Federation. The full listing of the files that sit within the curated ESGF replica at UCL is available via the pmip4 GitHub repositories. Additional data has been provided by:

  • Deepak Chandan (University of Toronto, CCSM4-UofT)

  • Charlie Williams (University of Bristol, HadGEM3-GC31-LR)

  • Vittoria Guarino (British Antarctic Survey, HadGEM3-GC31-LR)

  • https://psl.noaa.gov/data/20thC_Rean/ for the 20th Century Reanalysis (Compo et al, 2011)