WMO Lead Centre for
Annual-to-Decadal Climate Prediction

System Configuration Information for Global Producing Centers

	CCCMA	BSC	MOHC	DWD
When was the current decadal prediction system implemented?	2012	June 2011	Oct 2015	2020
Is it a coupled forecast system?	Yes	Yes	Yes	Yes
Atmospheric model and resolution	CanCM4: T63 (~300km), 35 levels	IFS: T159 (~125 km), 62 levels	HadGEM3-GC2: N216 (~60km), 85 levels	ECHAM6: T127(~200km), 47 levels
Ocean model and resolution (if applicable)	1.4° longitude ´ 0.94° latitude, 40 levels	NEMO: ORCA 1ᵒ grid, 42 levels	NEMO: ORCA 0.25ᵒ grid, 75 levels	MPIOM: 1.5ᵒ average, 40 levels
Sea-ice model (if applicable)	Cavitating fluid rheology and an energy balance model	LIM2	CICE	Within MPIOM
Source of atmospheric initial conditions	6 hourly atmospheric wind, temperature and specific humidity	ECMWF (FULLPOS software)	6 hourly atmospheric winds and temperature from ERA40/ERA-Interim	Full-field nudged atmosphere with Ensemble Kalman filtering using ERA40/ERA5
Source of ocean and ice initial conditions	Interpolated monthly/daily sea surface temperature, sea ice concentration. Sea ice thickness nudged to values from a statistical model. Offline variational assimilation of 3D ocean temperature	Interpolation of ORAS4 data	Monthly potential temperature and salinity from MOSORA and monthly sea-ice concentration from HadISST	Assimiliated ocean state using Ensemble Kalman Filtering with ocean temperature and salinity from EN4 observations and daily sea-ice concentration from NSIDC
Forecast Start Month	January	November	November	November
Hindcast period	1960 – 2010	1960-2017	1960-2015	1960 - 2020
Ensemble size for the hindcasts	10	10	10	16
How is the hindcast ensemble configured?	Separate coupled assimilation for each ensemble member.	Perturbations of atmosphere and ocean initial conditions.	Stochastic physics perturbations	Lagged-day initialization
Ensemble size for the forecast	10	10	20	10
How is the forecast ensemble configured?	Same as hindcasts	Same as hindcasts	Same as hindcasts	Same as hindcasts
Length of forecasts	10 years	10 years	5 years	10 years
How are the forecast anomalies constructed?	Lead-time dependent climatology with trend correction for near-surface temperatures	Lead-time dependent climatology	Lead-time dependent climatology	Lead-time dependent climatology
URL where forecast(maps) are displayed	Not available	Not available	https://www.metoffice.gov.uk/ research/ climate/ seasonal-to-decadal/long-range/decadal-fc/index	https://www.fona-miklip.de/decadal-forecast/forecasts-archive/
Point of Contact	reinel.sospedra-alfonso@canada.ca	louis-philippe.caron@bsc.es	nick.dunstone@metoffice.gov.uk	klaus.pankatz@dwd.de

System Configuration Information for Contributing Centers

	DMI/SMHI	MIROC	CSIRO
When was the current decadal prediction system implemented?	2019	2019	January 2020
Is it a coupled forecast system?	Yes	Yes	Yes
Atmospheric model and resolution	IFS: T255 (~80km), 91 levels (model top: 0.01 hPa)	MIROC6-AGCM: T85 (~1.4°), 81 levels	AM2: 2◦ latitude x 2.5◦ longitude, 24 levels
Ocean model and resolution (if applicable)	NEMO3.6: ORCA 1°-grid, 75 levels	COCO: 1° with enhanced resolution in the tropics, 62 levels	MOM5: ACCESS grid (1° resolution, enhanced resolution in the tropics and Southern Ocean, 50 levels)
Sea-ice model (if applicable)	LIM3	Multi-category EVP sea-ice model as a submodule of COCO	Sea Ice Simulator (SIS)
Source of atmospheric initial conditions	Full fields of ERA40/ERA-Interim/ERA5	6 hourly atmospheric winds, temperature and specific humidity from JRA55	Coupled ensemble Kalman filter assimilation of atmosphere, ocean and sea ice observations
Source of ocean and ice initial conditions	Addition of interpolated ORAS5-anomalies (ORAS5 backward extension used before 1979) onto model climatology	Data assimilation based on simplified IAU for monthly ocean temperature, salinity, and sea-ice concentration	Coupled ensemble Kalman filter assimilation of atmosphere, ocean and sea ice observations
Forecast Start Month	January	January	November
Hindcast period	1960-2018	1960-2018	1960-2018
Ensemble size for the hindcasts	15	10	10
How is the hindcast ensemble configured?	1 unperturbed and 2 perturbed atmospheric temperature initial states	Separate coupled assimilation for each ensemble member	Initial conditions are picked from members of a climate re-analysis.
Ensemble size for the forecast	15	10	10
How is the forecast ensemble configured?	Same as hindcasts	Same as hindcasts	Same as hindcasts
Length of forecasts	10 years + 2 months	10 years	10 years
How are the forecast anomalies constructed?	Data provided as full fields, bias- and drift- correction performed by WMO LC-GPC	Lead-time dependent climatology	Lead-time dependent climatology
URL where forecast(maps) are displayed	Not available	Not available	Not available
Point of Contact	shuting@dmi.dk, tim.kruschke@smhi.se	kimoto@aori.u-tokyo.ac.jp	richard.matear@csiro.au

Global Producing Centres

Logos of Global Producing Centres

Contributing Centres

Logos of Contributing Centres