Initial conditions and ENSO prediction using a coupled ocean-atmosphere model

Article English OPEN
Larow, T. E. ; Krishnamurti, T. N. (2011)

A coupled ocean-atmosphere initialization scheme using Newtonian relaxation has beendeveloped for the Florida State University coupled ocean-atmosphere global general circulationmodel. The initialization scheme is used to initialize the coupled model for seasonal forecastingthe boreal summers of 1987 and 1988. The atmosphere model is a modified version of theFlorida State University global spectral model, resolution T-42. The ocean general circulationmodel consists of a slightly modified version of the Hamburg’s climate group model describedin Latif (1987) and Latif et al. (1993). The coupling is synchronous with information exchangedevery two model hours. Using ECMWF atmospheric daily analysis and observed monthlymean SSTs, two, 1-year, time-dependent, Newtonian relaxation were performed using thecoupled model prior to conducting the seasonal forecasts. The coupled initializations wereconducted from 1 June 1986 to 1 June 1987 and from 1 June 1987 to 1 June 1988. Newtonianrelaxation was applied to the prognostic atmospheric vorticity, divergence, temperature anddew point depression equations. In the ocean model the relaxation was applied to the surfacetemperature. Two, 10-member ensemble integrations were conducted to examine the impact ofthe coupled initialization on the seasonal forecasts. The initial conditions used for the ensemblesare the ocean’s final state after the initialization and the atmospheric initial conditions areECMWF analysis. Examination of the SST root mean square error and anomaly correlationsbetween observed and forecasted SSTs in the Nin˜o-3 and Nin˜o-4 regions for the 2 seasonalforecasts, show closer agreement between the initialized forecast than two, 10-member noninitializedensemble forecasts. The main conclusion here is that a single forecast with the coupledinitialization outperforms, in SST anomaly prediction, against each of the control forecasts(members of the ensemble) which do not include such an initialization, indicating possibleimportance for the inclusion of the atmosphere during the coupled initialization.DOI: 10.1034/j.1600-0870.1998.00006.x
  • References (36)
    36 references, page 1 of 4

    Anderson, J. and Gudgel, R. 1997. Impact of atmospheric initial conditions on seasonal predictions with a coupled ocean-atmosphere model. Proceedings of the 21st Annual Climate diagnostic Workshop. Department of Commerce/NOAA, 61-62.

    Arakawa, A. and Lamb, V. R. 1977. Computational design of the basic dynamical processes for the UCAL general circulation model. J. Comput. Phys. 16, 173-263.

    Barnett, T. P., Latif, M., Kirk, E. and Roeckner, E. 1991. On ENSO Physics. J. Climate 4, 487-515.

    Businger, J. A., Wyngaard, J. C., Izumi, Y. and Bradly, E. F. 1971. Flux profile relationship in the atmospheric surface layer. J. Atmos. Sci. 28, 181-189.

    Chang, L. W. 1978. Determination of surface flux of sensible heat, latent heat and momentum utilizing the bulk Richardson number. Papers in Meteor. Res. 1, 16-24.

    Chen, D., Zebiak, S. E., Busalacchi, A. J. and Cane, M. A. 1995. An improved procedure for El Ni n˜o forecasting. Implications for predictability. Science 269, 1699-1702.

    Daley, R. and Puri, K. 1980. Four-dimensional data assimilation and the slow manifold. Mon. Wea. Rev. 108, 85-99.

    Derber, J. D. and Rosati, A. 1989. A global oceanic data assimilation system. J. Phys. Oceanogr. 19, 1333-1347.

    Ghil, M. and Malanotte-Rizzoli, P. 1991. Data assimilation in meteorology and oceanography. Advances in Geophysics 33, 141-266.

    Hellerman, S. and Rosenstein, M. 1983. Normal monthly wind stress over the world ocean with error estimates. J. Phys. Oceanogr. 13, 1093-1104.

  • Metrics
    No metrics available
Share - Bookmark