Seasonal Forecasts of the Extratropical Pacific Ocean
contributed Guillermo Auad, Arthur J. Miller, John O. Roads, John N. Ritchie
Experiment Climate Prediction Center Scripps Institution of Oceanography UCSD, 0224 La Jolla, CA 92093-0224
1. Background
Large-scale, seasonal oceanic anomalies in the extratropical regions are generally controlled by anomalous atmospheric forcing (e.g., Miller and Roads, 1990; Cayan, 1992). The accuracy of seasonal forecasts of the extratropical ocean therefore will usually depend on the accuracy of atmospheric seasonal forecasts. Unfortunately, atmospheric forecast skill is generally poor on seasonal time scales (e.g., Roads et al., 2001). Moreover, because extratropical oceanic conditions are very persistent, dynamical forecast skill levels that are superior to persistence forecasts are very difficult to achieve even for short-term forecasting (e.g., Miller et al., 1995). On the other hand, teleconnections from anomalous tropical conditions (e.g., Alexander, 1992) may be able to add some level of skill to forecasts of the extratropical oceanic forcing. Also, some oceanic variations can be controlled by processes that are established by ocean dynamics, such as currents due to intrinsic oceanic instability processes or to Rossby wave propagation, that may have predictability time scales that are longer than those of the atmosphere. Subsurface conditions, such as thermocline depth and geostrophic currents, and their consequent effects on SST may also be easier to predict than surface conditions, such as mixed layer depth or Ekman currents, and the atmospherically controlled part of SST variability.
Presently there are several groups around the world that engage in continuous real-time forecasting of large-scale and mesoscale extratropical oceanographic conditions. The NOAA Coastal Ocean Forecast System (http://chartmaker.ncd.noaa.gov/csdl/COFS/v32/cofs32.html ; see Aikman et al., 1996) predicts northwestern Atlantic Ocean conditions 24 hours in advance The NRL Experimental Real-Time North Pacific Ocean Nowcast/Forecast System (http://www7320.nrlssc.navy.mil/npacnfs_www/NPACNFS.html) predicts North Pacific Ocean conditions 72 hours in advance. The Forecasting Ocean-Atmosphere Model (http://www.meto.govt.uk/sec5/OA/FOAM/FOAM.html) predicts global ocean conditions 5 days in advance; Bell et al., 2000). The Mediterranean Forecasting System Pilot Project predicts Mediterranean Sea conditions 10 days in advance (http://www.cineca.it/mfspp).
ECPC has embarked on a project to develop seasonal forecasts of the large-scale (non-eddy-resolving) extratropical Pacific Ocean conditions. We are exploring the relative importance of oceanic initial conditions, atmospheric forcing forecast skill and dynamical evolution of the ocean on the forecasts. Seasonal forecasts are being displayed in real-time on the web (http://ecpc.ucsd.edu/ocean/) and the latest forecasts are discussed below. Skill levels are being studied and will be reported in the near future.
The ocean model is an isopycnal model (10 layers with nearly constant potential density) fully coupled to a bulk surface mixed layer model (Oberhuber, 1993). The model has been tested in a variety of ocean hindcasting scenarios (e.g., Miller et al., 1994; Auad et al., 2001) and it is here used with 1.5 degree resolution (enhanced to 0.67 degrees north-south resolution near the equator) in the following Pacific Ocean forecast framework. Mean seasonal cycle forcing from momentum, heat, fresh-water and TKE fluxes is specified a priori (see Auad et al., 2001, for a complete discussion). Anomalous forcing of momentum, heat, fresh-water and TKE fluxes are specified from 3-month forecasts of the ECPC GSM (Roads et al., 2001; Roads et al., this issue). Initial ocean conditions are taken from a continuously updated hindcast using atmospheric conditions taken from NCEP analyses up to the time of the launch of the forecast.
2. Winter 2000-2001 Forecasts
Figure 1 shows the SST anomalies hindcast for fall 2000 and forecast for winter (DJF) 2000-2001. In both panels, the eastern tropical and subtropical Pacific is colder than normal, while the western North Pacific north of 30N cools during the winter. In both panels, a warm tongue of superficial water is seen off the southwestern U.S. coast. The western tropical Pacific shows slightly warmer SST than normal. The sampling rate of the observations that are used to constrain the atmospheric initial conditions is low in the Southern Ocean so that the model solutions in this area are not very reliable.
Figure 2 shows the mixed-layer depth (MLD) anomalies hindcast for fall 2000 and forecast for winter (DJF) 2000-2001. The deepest anomalies for the fall season are observed in the western and central equatorial Pacific and off the Mexican coast at about 25N. In general, the rest of the basin shows MLD shallower than normal. In the winter forecast, the North Pacific north of 20N exhibits deeper than normal MLD, especially on its western side, while the tropical band shows its most prominent changes in the central region where the MLD is shallower than normal. Wind stresses used in our forecast run have larger errors than the ones used in the hindcast run which could explain some of the larger differences in the plots. However, the amplitude of the changes observed between both seasons is considered to be within the expected range for both tropical and extratropical regions.
Figure 3 shows the heat storage (HS; temperature integrated to 400m) anomalies hindcast for fall 2000 and forecast for winter (DJF) 2000-2001. Hindcast and forecast HS show similar patterns for fall and winter although the anomalies are of slightly lower amplitude in the latter. In both seasons the North Pacific is generally colder than normal with the exception of some central regions. The eastern tropical Pacific is forecast to remain colder than normal.
Figure 4 shows equatorial cross-sections of temperature anomalies hindcast for fall 2000 and forecast for winter (DJF) 2000-2001. They show maximum warm perturbations west of the dateline at 150 m for both seasons. In winter, the central Pacific is forecast to cool and the strength of the western warm anomaly is forecast to decrease. The eastern tropical Pacific is forecast to remain colder than normal throughout the thermocline.
Improvements to the Pacific Ocean forecasting system are now being sought. Since no information from observed oceanic conditions is presently being used in the initialization of the ocean model, we are exploring ways of using the NASA and/or NCEP ocean analysis to improve the oceanic initial conditions. Since skill levels of the atmospheric forcing forecast deteriorate with time, we are determining how much skill would have been achievable if accurate atmospheric forcing was specified; these results will then guide improvements in model physics. Since tropical forecasts will necessarily be deficient due to the lack of coupled ocean-atmosphere dynamics in the equatorial region, we are beginning to explore the efficacy of coupling the ECPC GSM to the ocean model. This may improve tropical ocean forecasts and potentially improve extratropical forecasts in regions where feedbacks to the atmosphere may be important such as the Kuroshio-Oyashio Extension region (e.g., Schneider et al., 2001).
References
Aikman, F. III, G. L. Mellor, T. Ezer, D. Sheinin, P. Chen, L. Breaker, K. Bosley, and D. B. Rao, 1996: Towards an Operational Nowcast/Forecast System for the U.S. East Coast. Modern Approaches to Data Assimilation in Ocean Modeling. P. Malanotte-Rizzoli, ed., Elsevier Oceanography Series, 61, 347-376.
Alexander, M. A., 1992: Midlatitude atmosphere ocean interaction during El Nino. 1. The North Pacific Ocean. J. Climate, 5, 944-958.
Auad, G., A. J. Miller, J. O. Roads and D. R. Cayan, 2001: Pacific Ocean wind stresses and surface heat fluxes from NCEP, COADS and FSU: Cross-statistics and ocean model responses. J. Geophys. Res., sub judice.
Bell, M. J., R. M. Forbes, and A. Hines, 2000: Assessment of the FOAM global data assimilation system for real-time operational forecasting. J. Mar. Syst., 25, 1-22.
Cayan, D. R., 1992: Latent and sensible heat flux anomalies over the northern oceans: Driving the sea surface temperature. J. Phys. Oceanogr., 22, 859-881.
Miller, A. J. and J. O. Roads, 1990: A simplified coupled model of extended-range predictability. J. Climate, 3, 523-542.
Miller, A. J., D. R. Cayan, T. P. Barnett, N. E. Graham and J. M. Oberhuber, 1994: Interdecadal variability of the Pacific Ocean: Model response to observed heat flux and wind stress anomalies. Clim. Dyn., 9, 287-302.
Miller, A. J., P.-M. Poulain, A. R. Robinson, H. G. Arango, W. G. Leslie and A. Warn-Varnas, 1995: Quantitative skill of quasigeostrophic forecasts of a baroclinically unstable Iceland-Faroe Front. J. Geophys. Res., 100, 10,833-10,849.
Oberhuber, J. M., 1993: Simulation of the Atlantic circulation with a coupled sea ice - mixed layer isopycnal general circulation model. Part I: model description. J. Phys. Oceangr., 23, 808-829.
Roads, J., S. Chen, and F. Fujioka, 2001: ECPC's Weekly to Seasonal Global Forecasts. Bull. Amer. Meteor. Soc., in press.
Schneider, N., A. J. Miller and D. W. Pierce, 2001: Anatomy of North Pacific decadal variability. J. Climate, sub judice.
Fig. 1 Anomalous Mixed Layer Depth (MLD) hindcast for the 2000 fall (top panel) and forecast for the 2000/2001 winter (bottom panel). Contour interval is 18 m.
Fig. 2 Anomalous sea surface temperature (SST) hindcast for the 2000 fall (top panel) and forecast for the 2000/2001 winter (bottom panel). Contour interval is 0.4 oC.
Fig. 3 Anomalous equatorial SSTs hindcast for the 2000 fall (top panel) and forecast for the 2000/2001 winter (bottom panel). Contour interval is 0.4 oC.
Fig. 4 Anomalous heat storage hindcast for the 2000 fall (top panel) and forecast for the 2000/2001 winter (bottom panel). Contour interval is 5 m oC.
