This dataset contains the output of three fully coupled seasonal forecast experiments performed as a case study of summer 2022, using the same model setup as ECWMF seasonal forecast system 5 (SEAS5), as presented in the paper Patterson, M., Befort, D., O'Reilly, C., Weisheimer, A. "The ECMWF SEAS5 seasonal forecast of the hot and dry European summer of 2022" Quarterly Journal of the Royal Meteorological Society, https://doi.org/10.1002/qj.4851

​

​The three experiments are:

1) CONTROL

A coupled hindcast ensemble with perturbed initial conditions for the 2022 summer, identical to the operational hindcast with the start date 1st May, but for 200 members rather than 51.

2) ATMOS-IC-2022

Similar to the CONTROL with 2022 initial conditions for the atmosphere and land surface and 2022 concentrations of carbon dioxide, but with ocean initial conditions taken from a year in 1981-2021. Five perturbed initial condition states taken for each year hence, 41 x 5 = 205 members.

3) OCEAN-IC-2022

Similar to CONTROL with 2022 initial conditions for the ocean, but with initial atmosphere, land surface and carbon dioxide conditions taken from a year in 1981-2021. Like ATMOS-IC-2022, 41 x 5 = 205 members.

​

The OCEAN-IC-2022 experiments were performed by taking the setup for hindcasts 1981-2021 and swapping the ocean initial conditions for 2022. The time dimension therefore corresponds to the time of the base hindcast taken from 1981-2021. Conversely, the ATMOS-IC-2022 were performed by running the 2022 hindcast and swapping the ocean initial conditions for other years. The time dimension for ATMOS-IC-2022 therefore is identical for all members. Hindcasts are all started on 1st May and run for four months.

The data stored is monthly-mean output from the experiments with the following directory structure

for variables on pressure levels:

[experiment_name]/pl/[experiment_name]_[ensemble_member_number]_pl_[level]hPa.nc

and for those on a single level or at the surface:

[experiment_name]/sfc/[experiment_name]_[ensemble_member_number]_sfc.nc

The files each contain multiple variables. Variables on pressure levels are stored at 700hPa,500hPa and 250hPa levels.

Pressure level variables stored are:

u (zonal wind), v (meridional wind), t (air temperature), z (geopotential), q (specific humidity)

Surface variables stored are:

sst (sea surface temperature), swvl1 (Volumetric soil water layer 1), swvl2 (Volumetric soil water layer 2), swvl3 (Volumetric soil water layer 3), swvl4 (Volumetric soil water layer 4), msl (air pressure at mean sea level), t2m (2m temperature), tprate (mean total precipitation rate).