Datasets used in analysis for publication "Reducing tropical cyclone activity in global climate models by evaporative suppression"

This dataset hosts data supporting the analysis performed in the publication titled "Reducing tropical cyclone activity in global climate models by evaporative suppression", submitted to Journal of Geophysical Research: Atmosphere. This dataset includes processed output from a suite of global climate models developed by the NOAA Geophysical Fluid Dynamics Laboratory (GFDL): HiRAM, AM2.5, and CM2.5-FLOR. Note that AM2.5 is the atmospheric model used in CM2.5-FLOR.

Links and instructions for each model are provided below:

HiRAM: https://www.gfdl.noaa.gov/hiram-quickstart/

CM2.5-FLOR: https://www.gfdl.noaa.gov/cm2-5-and-flor-quickstart/

The data was generated by performing a total of 6 experiments, with 2 experiments performed per global climate model. For every pair of experiments per model, one was performed using a control configuration of the model, while the other was performed using an experimental configuration, termed "SWISHE". The control configurations are those available in the links provided above for each model. The SWISHE configuration is hosted at: https://github.com/mr-gabrielrios/research/tree/main/swishe. by performing model integrations of 50 years for the atmosphere-only models (HiRAM, AM2.5) and 200 years for the coupled atmosphere-ocean model (CM2.5-FLOR). See the publication for details regarding the SWISHE algorithm. 

Tropical cyclones were tracked in each model using the methodology in Harris et al. (2016), https://doi.org/10.1175/JCLI-D-15-0389.1. An unmaintained repository hosting the described method is hosted at: https://github.com/lharris4/GFDL_quick_tracks.

The files included in this dataset are described below:

TC_density.s101_e150.models-HIRAM_AM2.5_FLOR.experiments-CTL1990s-CTL1990s_swishe.storm_type-TS-C15w.binned_5deg.pkl: tropical cyclone activity density binned into 5 x 5 degree windows for each global climate model and each experiment over 50 year subsamples of the model integrations. This is a pickle file and can be opened in Python (see for https://docs.python.org/3/library/pickle.html instructions). This dataset is used to generate Figure 1 in the publication.

SWISHE_density.s101_e150.models-HIRAM_AM2.5_FLOR.experiments-CTL1990s-CTL1990s_swishe.storm_type-TS-C15w.binned_5deg.pkl: density of SWISHE application binned into 5 x 5 degree windows for each global climate model and each experiment over 50 year subsamples of the model integrations. This is a pickle file and can be opened in Python (see for https://docs.python.org/3/library/pickle.html instructions). This dataset is used to generate Figure 1 in the publication.

TC_track_data.s101_e150.models-HIRAM_AM2.5_FLOR.experiments-CTL1990s-CTL1990s_swishe.storm_type-TS-C15w.generated-20241030.pkl: tropical cyclone track data for each global climate model and each experiment over 50 year subsamples of the model integrations. Track data was generated using the methodology of Harris et al. (2016). This is a pickle file and can be opened in Python (see for https://docs.python.org/3/library/pickle.html instructions). This dataset is used to generate Table 1, Figure 2, Figure 3, and Figure 4 in the publication.

TC_snapshots-AM2.5_FLOR_HIRAM-TS.tar.gz: snapshots of global climate model output fields corresponding to individual tropical cyclones for each global climate model and each experiment over 50 year subsamples of the model integrations. These snapshots were composited to obtain mean properties of tropical cyclones in each model configuration. This tarball contains netCDF files with data for individual tropical cyclones. This dataset is used to generate Figure 5, Figure 6, Figure 7, and Figure 8 in the publication.