This experiment provides output from a six-member ensemble simulation using the UA-ICON (Upper Atmosphere ICON) general circulation model, version 2.6.6. The model setup uses a high-top configuration with 120 vertical levels extending to ~150 km altitude and a horizontal resolution of ICON R2B4 (~160 km), allowing detailed representation of processes in the troposphere, stratosphere, mesosphere, and lower thermosphere. The experiment uses a time step of 360 seconds and includes a comprehensive physics package (ICON numerical weather prediction, NWP, package), notably gravity wave (GW) parameterizations for both subgrid-scale orographic (SSO) and non-orographic (NO) sources. Radiative transfer is handled via the ecRad radiation scheme.
The experiment includes six ensemble members to account for internal atmospheric variability. Initial conditions are based on ERA5 climatology (1979–2022). Ensemble 1 uses the mean January 1 state from ERA5 data (1979–2022), while ensembles 2–6 each exclude one year (1984, 1992, 2000, 2008, or 2016) to introduce slight variations in the initial conditions. Each ensemble simulation spans 30 years, with the first year treated as spin-up and excluded from output, resulting in a data range from 1991-01-01 to 2019-12-31 (arbitrarily numbered dates). All simulations are conducted under seasonally repeating boundary conditions to represent a stationary present-day climate. Sea surface temperature and sea ice are based on ERA5 climatology (1979–2022), greenhouse gas concentrations follow CMIP6 historical means (1979–2020), and ozone climatology is derived from MACC and GEMS datasets.
The key experimental perturbation is an artificial tenfold enhancement of the stratospheric SSO drag component within the Northwest America region (NA: 30°–60°N, 100°–130°W). The scaling factor of 10 was determined experimentally, ensuring that the enhanced drag remains within the range of natural variability and preserves realistic dynamical forcing. It is the only applied external perturbation in this experiment. This approach is designed to isolate the influence of this known GW hotspot on atmospheric dynamics and circulation patterns. In this experiment, Unlike the other ensemble members, which span 30 years, ensemble member 5 terminated early in 2005 due to numerical instability. the output of the ensemble is archived from 1991 onward, excluding the spin-up year. The dataset is well-suited for studying Northwest America stratospheric GW forcing effects on stratospheric variability, polar vortex dynamics, and for quantifying signal-to-noise characteristics via ensemble analysis.
Daily-mean atmospheric fields are stored as monthly NetCDF files over the global domain on model levels. Variables include 3D fields (temperature, winds, pressure, vertical velocity), GW drag tendencies (SSO and NO), and surface values such as 2-meter temperature and surface pressure.
