FILE DOCUMENT This document explains the contents of data files stored in the repository of the Finnish Meteorological Institute, supporting the study "Extreme cold days and spells in Northern Europe at 0.5-2.0°C global warming levels" by Kimmo Ruosteenoja and Kirsti Jylhä, submitted to "International Journal of Climatology". The file format of all the data files is "netcdf" (.nc). Mean temperatures and standard deviations ========================================= temp_mean+standard_deviation.nc The file contains 20-year monthly means and temporal standard deviations of the daily mean temperatures at four global warming levels (multi-model means). Data variables: temp_mean: monthly mean temperatures temp_stdev: temporal standard deviations of temperature Coordinate variables: lon: longitude lat: latitude lev: the number of month (lev=1: January, lev=2, February,...) time: 0.5°C (t=1), 1.0°C (t=2), 1.5°C (t=3), 2.0°C (t=4) global warming levels The files were used to create Figs. 1 and 11 of the paper. Percentile values ================= percentile_p1_25models.nc percentile_p2.5_25models.nc percentile_p5_25models.nc percentile_p10_25models.nc The files contains the 1th, 2.5th, 5th and 10th percentiles of December-February daily mean temperatures of the period 1961-2000. Percentile values are given separately for the 25 GCMs, formally under the time variable. However, for models 9 and 10, multimodel means are given instead. Data variable: tas - temperature percentile in °C Coordinate variables: lon: longitude lat: latitude z: not used t: the ordinal number of the model (see Table 1 of the manuscript) These files were used to construct Figures S1 and S2 and can be used to derive the climate indices. Maximum extremity index from the E-OBS analyses =============================================== extremity_index_max_eobs_1962-2020.nc The file contains the extremity index of the strongest cold spell of the winter in 1962-2020, derived from the observational E-OBS analyses. Data variable: ex_max_p10 - extremity index above the T(p10) threshold temperature Coordinate variables: lon: longitude lat: latitude z: not used t: the year; e.g. 1987 refers to the winter of 1986/87. This file was used to construct Fig. 2. Geographical distributions of the indices at four warming levels ================================================================ cold_day_indices_25gcm_mean_p10.nc cold_day_indices_25gcm_mean_m10.nc The files contains the geographical distributions of the 20-year mean values of both climate indices; 25 GCM means at the 0.5, 1.0, 1.5 and 2.0°C global warming levels. In the first file, the threshold for a cold day is the 10th percentile of winter mean temperatures; in the second file, a constant threshold temperature of -10°C is used. Data variables: count_days_mean - count of cold days, 25 GCM mean extr_ind_mean - extremity index, 25 GCM mean Coordinate variables: lon: longitude lat: latitude z: not used t: the warming level (t=1 - 0.5°C; t=2 - 1°C; t=3 - 1.5°C; t=4 - 2°C) These files were used to construct Figs. 5, 6, 12 and S7. Spatial means of the indices at four warming levels =================================================== cold_day_indices_25gcm_mean+range_spatial_mean.nc cold_day_indices_25gcm_median_spatial_mean.nc The files contains the spatial means of both climate indices over the entire domain; the 25 GCM means, medians and the upper and lower estimates, at the 0.5, 1.0, 1.5 and 2.0°C global warming levels. Data variable in file 1: count_days_mean - count of cold days, 25 GCM mean extr_ind_mean - extremity index, 25 GCM mean count_days_low - count of cold days, 2. lowest within 25 GCM extr_ind_low - extremity index, 2. lowest within 25 GCM count_days_high - count of cold days, 2. highest within 25 GCM extr_ind_high - extremity index, 2. highest within 25 GCM Data variable in file 2: count_days_md - count of cold days, 25 GCM median extr_ind_md - extremity index, 25 GCM median Coordinate variables: x, y and z: not used t: the warming level (t=1 - 0.5°C; t=2 - 1°C; t=3 - 1.5°C; t=4 - 2°C) These files were used to construct Figure 4. Return levels of the extremity index at the 0.5°C level ======================================================= return_levels_large_extr_ind.nc The 10 and 100 year return levels of the extremity index of the strongest cold spell of winter under the 0.5°C warming level. Data variables: returnlev_10yr - 10 year return level returnlev_100yr - 100 year return level Coordinate variables: lon: longitude lat: latitude z: not used t: not used This file was used to construct Fig. 9. Spatial means of the probabilities of extreme cold spells ========================================================= probabilities_of_occurrence_large_extrind.nc The file contains spatial means of the probability for the occurrence of a maximum cold spell with an extremity index larger than that occurring once in 2-100 years at the 0.5°C warming level. Probabilities are calculated as means over the entire domain at the 0.5, 1.0, 1.5 and 2.0°C global warming levels. Data variable: probab_extr_ind - probability for exceeding the baseline return level Coordinate variables: x, y: not used z: the return levels at the 0.5°C warming level; z=1 - 2 years; z=2 - 4 years; z=3 - 10 years; z=4 - 20 years; z=5 - 40 years; z=6 - 100 years t: the warming level (t=1 - 0.5°C; t=2 - 1°C; t=3 - 1.5°C; t=4 - 2°C) These files were used to construct Figure 10.