GloSea5: Met Office seasonal prediction system
GloSea5 is the seasonal prediction system developed and run operationally at the Met Office.
GloSea5 stands for Met Office Global Seasonal forecasting system version 5 (MacLachlan et al 2015, Scaife et al 2014). It became operational in July 2013, replacing Met Office seasonal prediction system: GloSea4 which had been operational since September 2009.
GloSea5 is an ensemble prediction system built around the high resolution version of the Met Office climate prediction model: HadGEM3 family atmosphere-ocean coupled climate model. The upgrade from GloSea4 is primarily in the increased horizontal resolution: GloSea5 uses the N216 version (0.8 degrees in latitude and 0.5 degrees in longitude, which is approx. 50 km in mid-latitudes, in the horizontal) for the atmosphere, and the ORCA0.25 grid (0.25 degrees) for the ocean. The vertical resolution, like that of GloSea4, is 85 levels for the atmosphere and 75 levels for the ocean.
Each forecast requires initial ocean, land and atmosphere conditions. The land and atmosphere conditions are specified from atmospheric analyses that are produced separately for weather prediction purposes. The ocean initial conditions are taken from ocean analyses generated specifically for seasonal forecasting, using the 3-dimensional variational ocean data assimilation based on the multi-institution NEMOVAR project (Mogensen et al, 2009 describes a similar implementation of this method).
GloSea5 has two components: the forecast itself and an associated set of hindcasts, also called historical re-forecasts, used for calibration purposes and for skill assessment. In the case of GloSea5 the hindcast covers the period 1993 - 2015. Both forecast and hindcast are performed using the same configuration of the GloSea5 ensemble prediction system but, obviously, with different initial conditions.
A lagged initialisation approach, with all simulations being initialised daily (four ensemble members initialised every day in the case of the forecast - two run for 60 days and two for 195 days - and three members initialised on fixed calendar dates - 1st, 9th, 17th and 25th - for the hindcast) is followed to represent the uncertainties in the initial conditions. In a similar manner to the Met Office short- and medium-range ensembles, model uncertainties are represented through the use of stochastic physics schemes (Bowler et al 2008; Shutts, 2005). All climate forcings (aerosols, methane, CO2 concentrations, etc.) are set to observed values for the period 1960-2000 and follow the emissions scenario A1B afterwards. Ozone is fixed to observed climatological values and includes a seasonal cycle. The model contains no flux corrections or relaxations to climatology.
Every month, a 42-member ensemble seasonal forecast for the next six months is generated by combining and bias correcting all forecast members available from the most recent three weeks.
- Bowler, N.E., Arribas, A., Mylne, K.R., Robertson, K.B. and Beare, S.E., 2008: The MOGREPS short-range ensemble prediction system. Quarterly Journal of the Royal Meteorological Society, 134, 703-722, doi:10.1002/qj.234.
- MacLachlan C., A. Arribas, K.A. Peterson, A. Maidens, D. Fereday, A.A. Scaife, M. Gordon, M. Vellinga, A. Williams, R. E. Comer, J. Camp and P. Xavier, 2015. Description of GloSea5: the Met Office high resolution seasonal forecast system. Q. J. R. Met. Soc., DOI: 10.1002/qj.2396.
- Mogensen K.S., M.A. Balmaseda, A. Weaver, M.J. Martin, A. Vidard. NEMOVAR: A variational data assimilation system for the NEMO ocean model. ECMWF newsletter, summer 2009.
- Scaife A.A., A. Arribas, E. Blockley, A. Brookshaw, R. T. Clark, N. Dunstone, R. Eade, D. Fereday, C. K. Folland, M. Gordon, L. Hermanson, J. R. Knight, D. J. Lea, C. MacLachlan, A. Maidens, M. Martin, A. K. Peterson, D. Smith, M. Vellinga, E. Wallace, J. Waters and A. Williams, 2014. Skilful Long Range Prediction of European and North American Winters. Geophys. Res. Lett., 41, 2514-2519, DOI:10.1002/2014GL059637.
- Shutts, G., 2005: A kinetic energy backscatter algorithm for use in ensemble prediction systems. Quarterly Journal of the Royal Meteorological Society. 131, 3079-3102, doi:10.1256/qj.04.106.