Data for "Increasing Frequency and Persistence of the Summertime Greenland High Regime Not Captured by a Seasonal Prediction Model Very Large Ensemble"

hese data support the work presented in Lee & Polvani (2025) "Increasing Frequency and Persistence of the Summertime Greenland High Regime Not Captured by a Seasonal Prediction Model Very Large Ensemble", published in Geophysical Research Letters: https://doi.org/10.1029/2025GL119421

This work is based on the year-round weather regime classification introduced by Lee et al. (2023) and Lee & Messori (2024), with slight modifications as outlined in Lee & Polvani (2025).

era5_1981_2024_year_round_north_america_weather_regimes_running_mean.nc

The ERA5 daily weather regime classification for 1981–2024 and associated parameters. The weather regime index (WRI) is also supplied here following Lee & Messori (2024).

Dimensions:

latitude = 61
longitude = 151
time = 16060
eof_number = 12
regime = 5
trend_params = 2
doy = 365
Variables:

latitude(latitude)
longitude(longitude)
time(time) (units = "days since 1970-01-01")
z_anoms_detrend(time, latitude, longitude); units = "gpm" ; long_name = "Detrended Z500 anomalies 1981-2024 ERA5 climate"
z_norm(time, latitude, longitude) ; units = "std" ; long_name = "Z500 anomalies normalized by seasonal cycle of area-averaged standard deviation"
z_climo(doy, latitude, longitude) ; units = "gpm" ; long_name = "Z500 climatology 1981-2024 ERA5"
trend(doy, trend_params) ; units = "m per day since 1981-01-01" ; long_name = "60d smoothed linear trend slope & intercept,1981-2024"
norm_factor(doy) ; long_name = "Variance normalization factor"
pc(time, eof_number) ; units = "raw pc" ; long_name = "pc time series"
eof(eof_number, latitude, longitude) ; long_name = "raw EOFs"
eigs(eof_number) ; long_name = "eigenvalues"
var_frac(eof_number) ; long_name = "variance fraction"
regime(time) ; long_name = "Daily regime attribution"
WRI(regime, time) ; units = "std" ; long_name = "Weather regime index"
WRI_mean(regime) ; long_name = "WRI mean"
WRI_std(regime) ; long_name = "WRI standard deviation"
cluster_mean_norm(regime, latitude, longitude) ; units = "std" ; long_name = "Cluster-mean normalized Z500 anomalies"
double cluster_mean(regime, latitude, longitude) ; units = "gpm" ; long_name = "Cluster-mean Z500 anomalies"
double cluster_centroids(regime, eof_number) ; long_name = "Cluster centroids in PC space"
double dist_to_regime_centroid(time) ; long_name = "Distance to centroid of assigned regime"
seas5_may_1981_2024_jja_year_round_north_america_weather_regimes_monte_carlo_10000_member.nc

The weather regime classification for 1 June–31 August using a Monte Carlo/random sampling of ECMWF's SEAS5 hindcasts & forecasts from 1981–2024. Variables are defined similarly to the ERA5 dataset above.

Dimensions:

lead_time = 92
year = 44
ens_member = 10000
cluster = 5
pc = 12
trend_parameter = 2
Variables:

year(year)
lead_time(lead_time) (units = "days since 1 May")
ens_member(ens_member)
regime(ens_member, year, lead_time) # Daily regime attribution
wri(ens_member, year, lead_time, cluster) # Weather regime index
pcs(ens_member, year, lead_time, pc) # PCs obtained from projecting onto ERA5 EOFs
norm_factor(ens_member, lead_time) # variance normalisation factor
trend(ens_member, lead_time, trend_parameter) # Z500 trend in each ensemble member (slope + intercept)
member_id(ens_member, year) # this corresponds to the true ensemble member selected in each random sampling