The pyLamb package is a set of Python scripts and functions made to evaluate global climate models in terms of atmospheric circulation patterns on multiple time-scales and also explores the low frequency variability seen in observations. To this end, the 2 main scripts <interpolate_xesmf.py> and <makecalcs_parallel.py> are provided that must be run sequentially in order to classify continuous 6-hourly sea-level pressure patterns into 27 discrete the Lamb Weather Types (LWTs, Jones et al. 1993) at any grid-box in northern and southern Hemisphere extratropics in reanalysis and GCM data. The package is designed to process netCDF data from ESGF of virtually any GCM participating in CMIP5 and 6 and additionally comes with an extensive metadata archive on these GCMs. The scripts have the following tasks and should be run in the following order

1. interpolate_xesmf.py: interpolates GCM or reanalysis data from the grid provided by ESGF to a common 2.5 grid covering either the Northern or Southern Hemisphere extratropics between 30 and 70 degress North or South. Converts non-standard GCM calendars to a standard calendar and keeps the original metadata provided by the ESGF files. Interpolation methods from the xesmf module can be applied. Should be run with 6-hourly psl data as input. Save results in netCDF format.

2. makecalcs_parallel.py: assigns a single Lamb Weather Type to each time instant and grid-box of the netCDF files generated in 1. The script can be run in parallel mode, which significantly speeds it up. Saves the results in netCDF format keeping the original metadata from the ESGF files passed throug 1. With some small modifications can also calclate the 5 circulaation inidices / intermediate variables of the LWT approach.

3. analysis_hist.py: is the main script to calculate several climatologcial aspects of the point-wise LWT time series in the models and verify them against reanalysis data. The focus is here put on verifying one member per GCM, taking a large number of different GCMs into acccount. The applied methods and metrics are explained in https://doi.org/10.5194/gmd-15-1375-2022 In addtion, GCM dependencies are estimated by means of error pattern correlation as described in https://doi.org/10.1029/2022GL101446

4. analysis_ensemble.py: As 3, but for ensembles of a reduced number of GCMs

5. analysis_functions.py: Contains the functions used in the scripts

6. get_composites.py: Retrieves and plots average SLP composites for a given dataset, location (or city) and LWT

7. get_corr.py: subscript of analysis_hist.py, used to calculate correlations between atmosphere and ocean resolution paramters and mean / median GCM performance

8. to be continued...

Within these scripts, a large number of input parameters can be set that are commented to understand their meaning or purpose. This is followed by a line stating "EXECUTE" from where on no futher user options appear, i.e. the script is run with the parameters set before.

Credits

This ongoing research work is being funded by the Ministry for Ecological Transition and Demographic Challenge (MITECO) and the European Commission NextGenerationEU (Regulation EU 2020/2094), through CSIC's Interdisciplinary Thematic Platform Clima (PTI-Clima).