Chapter 2 CMIP6 Data

About:

This section of the repository contains code detailing how the CMIP6 data was processed to prepare any bias-corrections. This readme exists to orient users through what files are connected and need to be run to maintain upstream-downstream connections.

2.1 Support Function Sources:

For both python and R Workflows the data processing steps were broken down into discrete steps and written as individual functions.

R functions for processing bias correction steps can be found here

Python functions used to download and process cmip data can be found here

2.2 Accessing Data

CMIP6 data was downloaded using climate data operator (CDO) command line tools using the following specifications:

2.3 Preparing Date Keys for CMIP6 Models

For some reason the R packages that support NetCDF file handling do not like the calendar structure used by climate models. To still use R as a processing tool the dates for each individual CMIP6 model were extracted and placed into a lookup table that could be referenced to lookup the date values when needed.

These lookup tables include the source model, the variable of interest, and the proper dates that go with them. To prepare the lookup table we used a python3 script CMIP6 Variable Date Keys. This file was stepped for each of the four variables to create lookup tables for dates and also to flag models that had inconsistent structures/dates/variables.