This collection consists of NetCDF files that contain intermediate products derived from data produced by the French Research Institute for Exploitation of the Sea (IFREMER) and the Remote Sensing Systems corresponding to all experimental analysis by Colosi et al. (2021). The collection has seven main directories: ifremer_swh: All files under this directory tree correspond to significant wave height (SWH) data gridded on a 1 degree longitude by 1 degree latitude bins from 66 degrees North to 66 degrees South at daily time intervals. Data is derived from the IFREMER cross-calibrated along-track satellite altimetry SWH product found at ftp://ftp.ifremer.fr/ifremer/cersat/products/swath/altimeters/waves (open in a Mac Finder or Windows Explorer window and connect as "Guest"). The gridded variables available in the netCDF file are: nobs: Number of Observations swh: Significant Wave Height The files are named as followed: IFREMER_binned_alt_swh_{year}.nc where year ranges from 1993 to 2015. For example, the binned SWH for days from January 1st, 1993 to December 31, 1993 is saved in the file IFREMER_binned_alt_swh_1993.nc ccmp2_wsp: All files under this directory tree correspond to wind speed data averaged in 1 degree longitude by 1 degree latitude bins from 66 degrees North to 66 degrees South as well as in daily time intervals. Data is derived from the Cross Calibrated Multi-Platform version 2 (CCMP2) wind vector analysis found at http://www.remss.com/measurements/ccmp/ . The deresolved variable available in the netCDF file is: wsp: Wind Speed The files are named as followed: CCMP2_deresolved_wsp_{year}.nc where year ranges from 1993 to 2015. For example, the deresolved WSP for days from January 1st, 1993 to December 31, 1993 is saved in the file CCMP2_deresolved_wsp_1993.nc ww3_swh: All files under this directory tree correspond to the SWH output from the WAVEWATCH III (WW3) wave model forced by Climate Forecast System Reanalysis (CFSR) winds which has been averaged in 1 degree longitude by 1 degree latitude bins from 66 degrees North to 66 degrees South as well as in daily time intervals. All WW3 hindcast data may be found at ftp://ftp.ifremer.fr/ifremer/ww3/HINDCAST (open in a Mac Finder or Windows Explorer window and connect as "Guest"). The deresolved variable available in the netCDF file is: hs: Significant Wave Height The files are named as followed: ww3_deresolved_hs_{year}.nc where year ranges from 1993 to 2015. For example, the deresolved swh for days from January 1st, 1993 to December 31, 1993 is saved in the file ww3_deresolved_hs_1993.nc ww3_wsp: All files under this directory tree correspond to the WSP output from CFSR winds which has been averaged in 1 degree longitude by 1 degree latitude bins from 66 degrees North to 66 degrees South as well as in daily time intervals. The deresolved variable available in the netCDF file is: wsp: Wind Speed The files are named as followed: ww3_deresolved_wsp_{year}.nc where year ranges from 1993 to 2015. For example, the deresolved wsp for days from January 1st, 1993 to December 31, 1993 is saved in the file ww3_deresolved_wsp_1993.nc decor_scales: All files under this directory tree correspond to decorrelation time scales for observed and modeled SWH and WSP computed from integrals of the lagged covariance for each month. The decorrelation time scale variable available is: decor_scale: Monthly Decorrelation time scales The files are named as followed: {product}_{field}_decor_time_scale.nc where product may be either CCMP2, IFREMER, or WW3 and field refers to the physical variable that the decorrelation time scale is computed which is either swh or wsp. For example, the WW3 decorrelation time scales for swh is saved in the file ww3_swh_decor_time_scale.nc lsf_parameters: All files under this directory tree correspond to phase, amplitude, and fraction of variance explained parameters of the weighted least-squares fit to the mean, annual, and semi-annual cycles for observed and modeled SWH and WSP. The least squares fit variables available are: a_amp: Annual Cycle Amplitude a_phase: Annual Cycle Phase s_amp: Semi-Annual Cycle Amplitude s_phase: Semi-Annual Cycle Phase fve: Fraction of Variance Explained by Model a_amp_unc: Annual Cycle Amplitude Uncertainty a_phase_unc: Annual Cycle Phase Uncertainty s_amp_unc: Semi-Annual Cycle Amplitude Uncertainty s_phase_unc: Semi-Annual Cycle Phase Uncertainty The files are named as followed: {product}_{field}_lsf_parameters.nc where product may be either CCMP2, IFREMER, or WW3 and field may be either swh or wsp. For example, the WW3 least squares parameters for swh is saved in the file ww3_swh_lsf_parameters.nc prob_swell: All files under this directory tree correspond to probability of swell computed using modeled WSP and peak frequency or mean period. The probability of swell variables available are: monthly_prob_swell: Monthly Progression of Probability of Swell seasonal_prob_swell: Seasonal Progression of Probability of Swell The files are named as followed: WW3_probability_swell_3year.nc WW3_probability_swell_t01_3year.nc WW3_probability_swell_t0m1_3year.nc WW3_probability_swell_t01.nc where files ending in 3year correspond to probability of swell data computed with only 3 years of data from 2013 through 2015. The file not ending with 3year has probability of swell computed using peak frequency data from 1993 through 2015 . Files with t01 refers to the probability of swell computed using the mean period T_{m0,1}. Files with t0m1 refers to the probability of swell computed using the mean period T_{m0,-1}. To reproduce all figures from Colosi et al. (2021), you should first download the source code from the repository https://github.com/lcolosi/WaveClimatology. The directories from this data collection (ifremer_swh, ccmp2_wsp, ww3_swh, ww3_wsp, decor_scale, lsf_parameters, prob_swell) should then be downloaded and moved to a directory called data in the root directory of the project, like so: WaveClimatology ├── data │ ├── ifremer_swh │ ├── ccmp2_wsp │ ├── ww3_swh │ ├── ww3_wsp │ ├── decor_scale │ ├── lsf_parameters │ └── prob_swell ├── figs ├── src └── tools With the directory tree above, one should be able to run all Python scripts in src and tools and completely reproduce the analyses from Colosi et al. (2021).