Zonal and Meridional Means#
This functional diagnostic takes two models designated by CONTROL and EXPERIMENT and compares them via a number of
analyses. Optionally a number of observational datasets can be added for processing. There are three types of standard analysis:
lat_lon, meridional_mean and zonal_mean. Each of these diagnostics can be run on a separate basis (each an entry to diagnostics/scripts).
The lat_lon analysis produces the following plots: a simple global plot for each variable for each dataset, a global plot for the
difference between CONTROL and EXPERIMENT, a global plot for the difference between CONTROL and each of the observational datasets.
The meridional_mean and zonal_mean produce variable vs coordinate (
longitude) with both
in each plot, for the entire duration of time specified and also, if the user wishes, for each season (seasonal means): winter DJF, spring MAM, summer JJA, autumn SON (by setting
seasonal_analysis: true in the recipe).
At least regridding on a common grid for all model and observational datasets should be performed in preprocessing (if datasets are on different grids). Also note that it is allowed to use the same dataset (with varying parameters like experiment or ensemble or mip) for both CONTROL and EXPERIMENT (as long as at least one data parameter is different).
Available recipes and diagnostics#
Recipes are stored in recipes/
Diagnostics are stored in diag_scripts/
Required settings for script
title: title of the analysis, user defined;
control_model: control dataset name e.g. UKESM1-0-LL;
exper_model: experiment dataset name e.g. IPSL-CM6A-LR;
observational_datasets: list of at least one element; if no OBS wanted comment out; e.g. [‘ERA-Interim’];
analysis_type: use any of: lat_lon, meridional_mean, zonal_mean;
seasonal_analysis: boolean, if seasonal means are needed e.g.
save_cubes: boolean, save each of the plotted cubes in
Observations and reformat scripts#
Note: (1) obs4MIPs or OBS or ana4mips can be used.
it is important to note that all observational data should go through the same preprocessing as model data
none, basic technical analysis