Postprocessing data

A wide range of tools exist for postprocessing and analyzing data with techniques and methods exist. One of the first things you have to decide is how to store your files.

In the diagnostics notebooks we have have provided examples of how to use both history files and timeseries files, described below.

History vs. Timeseries files

When you run the CESM model the default output is history files, or files for a single timestep that include all variables for a given component and time frequency. However, most CESM community experiment data will be provided as timeseries files, or files that are a single variable over many timesteps. It is important you understand how to use both types of files, and for you to know that for some tasks (e.g. debugging) you should be using history files instead of timeseries files. However, it is much more efficient to store timeseries files because the overall size is smaller once the files have been processed into timeseries format.

The current recommendation is to use the new CUPiD diagnostics system to convert CESM history files into time series. You can try it yourself on CESM tutorial simulation data by running through the CUPiD Notebook under the diagnostics section.