A comprehensive coupled ocean-atmosphere model has been developed for use for long-lead climate forecasts in the Coupled Model Branch of the Environmental Modeling Center (EMC) at NOAA's National Centers for Environmental Prediction (NCEP) (Ji et al. 1994a,b). This prediction system, which uses an ocean data assimilation system (Ji et al. 1995), is described in an article in this issue of the Bulletin (Barnston et al.) that presents NCEP's forecasts for tropical Pacific SST. Here, the prediction system is taken one step farther: After obtaining the time series of the predicted tropical Pacific SST fields out to a few seasons in advance, these SST fields are used as a lower boundary condition for integrations of the atmospheric GCM (AGCM) out to comparable lead times. The AGCM forecast is the ensemble mean of 18 individual integrations, each using a different atmospheric initial condition. The nature of the model's mid-latitude atmospheric response to the ENSO state is discussed in a specification setting in Livezey et al. (1997), where the skill over the U.S. is found to be moderate during the northern winter during warm and cold ENSO episodes. Earlier work of the same nature demonstrated that over the Pacific/North American (PNA) region, skill in specifying geopotential heights, temperature and (to a lesser extent) precipitation are usable during the cold half of the year when a non-neutral ENSO condition exists (Livezey et al. 1996). On the other hand, skill during non-ENSO years is not significantly different from zero. This ENSO-dependency of expected skill is explainable on the basis of signal-to-noise ratio considerations in the extratropics with respect to ENSO - related anomalous tropical Pacific SST boundary forcing (Kumar and Hoerling 1998).

Here, we present AGCM forecasts for 3-month mean North American surface temperature and precipitation out to 8 months lead time, where lead time refers to the time between the forecast time and the middle month of the 3-month target period. An assessment of the expected forecast skill is based on a combination of (1) skill in the forecast of the tropical Pacific SST boundary condition and (2) skill of the simultaneous simulations (specifications) of the atmosphere, given a perfect SST forecast ("AMIP" skill). Specifically, the final correlation skill is the product of the two constituent skills at each location for the season and lead time being forecast. The skill of the SST forecasts (see the article on the NCEP SST forecasts in this issue of the Bulletin) for the first four of the eight lead times is shown in Table 1, based on the data set of forecasts and observations for the Nino 3.4 region over the 1982-97 period. The expected forecast skill over the U.S., given a perfect SST forecast, is described in Livezey et al. (1996) and is not shown by geographical region in this presentation. The specification skill, averaged over the U.S. for winter and expressed as temporal correlation coefficients, are in the neighborhood of 0.4 for 700 mb height and for surface temperature, and about 0.1 to 0.2 for precipitation, when ENSO is active (i.e., tropical Pacific SST at least 1 standard deviation away from mean). When multiplied by the imperfect correlation skill for predicting tropical Pacific SST, the areal average skill drops accordingly. All in all, the final skill is approximately comparable to that of the statistical CCA forecasts, also produced at CPC (Barnston et al. 1994); however, in specific circumstances the skill of the two forecasting approaches may differ substantially. In the regions that experience strong ENSO impacts (e.g. around the Great Lakes and southern Canada, in the vicinity of the Gulf of Mexico and the Southeast); however, the winter skill is locally moderately high during ENSO episodes at lead times of up to 6 months. Thus, the user of the surface forecasts presented here is urged to exercise caution in accepting the forecasts, considering the lead time and the region of interest. Additionally, spatial displacement biases may cause "centers of action" to appear slightly out of their appropriate location. A model output statistics (MOS) correction scheme (as in Smith et al. 1997) might be able to reduce these spatial biases, as well as systematic forecast amplitude errors.

Table 1. Expected skill (correlation X100) of the NCEP coupled model in predicting the SST anomaly in the Nino 3.4 region at 4 lead times for 12 overlapping 3-month seasons. A 1-month lead is, for example, a forecast for JFM 1998 made at the end of December 1997.

The current forecasts for North American surface temperature and precipitation are shown in Figs. 1 and 2, respectively. While warmth is forecast for much of North America in Jun-Jul-Aug, negative temperature anomalies appear over large pockets of the U.S. in the latter half of the summer, continuing throughout the remainder of 1998. Only northern Canada and Alaska are forecast to remain near normal to warm throughout the coming eight months. It should be noted that the expected skill of the forecasts for later in 1998 are not high enough to regard the forecasts with much practical confidence. They should therefore be viewed mainly academically or experimentally. However, when the certainty of the tropical Pacific SST forecast (i.e. the ENSO condition) is high, as it was in summer 1997 for forecasts of the approaching 1997-98 winter, more confidence can be placed in the AGCM forecasts. The skill of the tropical Pacific SST (in particular, the Nino 3.4 region) forecast at the 1 year lead time, averaged over all seasons, is roughly 0.5, compared with the mid-0.70s for 4 month lead SST forecasts (average across the bottom row of Table 1 above).

The precipitation forecasts call for enhanced rainfall in much of the western and central interior of the U.S. and southern Canada through northern fall of 1998. Precipitation deficits are expected in the southeastern U.S. summer and fall, and in Mexico throughout most of the remainder of 1998. Dryness in southwestern Canada/northwestern U.S. is a consistent feature of the forecasts from Aug-Sep-Oct to Jan-Feb-Mar 1999. In similar fashion to temperature, the precipitation forecasts for the latter half of 1998 should be considered very cautiously at this time. This is especially true in view of the uncertainty surrounding the ENSO condition expected for late 1998: This model is now predicting moderate La Nina conditions developing for fall 1998, continuing and perhaps strengthening into the following winter. While other models, both dynamical and statistical, are also predicting a cold phase of ENSO, the magnitude of the episode is still uncertain at this time and could range anywhere from mild to very strong. The anomalies of the subsurface sea temperatures that are expected to appear in the SST in the next few seasons (under the condition that the presently observed normal low-level easterly trades continue) are at least moderate in strength, indicating a likelihood of at least a moderate La Nina.

References

Barnston, A.G., H.M. van den Dool, S.E. Zebiak, T.P. Barnett, M. Ji, D.R. Rodenhuis, M.A. Cane, A. Leetmaa, N.E. Graham, C.F. Ropelewski, V.E. Kousky, E. A. O'Lenic and R.E. Livezey, 1994: Long-lead seasonal forecasts--Where do we stand? Bull. Amer. Meteor. Soc., 75, 2097-2114.

Ji, M., A. Kumar and A. Leetmaa, 1994a: A multi-season climate forecast system at the National Meteorological Center. Bull. Am. Meteor. Soc., 75, 569-577.

Ji, M., A. Kumar and A. Leetmaa, 1994b: An experimental coupled forecast system at the National Meteorological Center: Some early results. Tellus, 46A, 398-418.

Ji, M., A. Leetmaa and J. Derber, 1995: An ocean analysis system for seasonal to interannual climate studies. Mon. Wea. Rev., 123, 460-481.

Kumar, A. and M.P. Hoerling, 1998: Annual cycle of Pacific/North American seasonal predictability associated with different phases of ENSO. J. Climate, 11, in press.

Livezey, R.E., M. Masutani and M. Ji, 1996: SST-forced seasonal simulation and prediction skill for versions of the NCEP/MRF model. Bull. Am. Meteor. Soc., 77, 507-517.

Livezey, R.E., M. Masutani, A. Leetmaa, H. Rui, M. Ji and A. Kumar, 1997: Teleconnective response of the Pacific-North American region atmosphere to large central equatorial Pacific SST anomalies. J. Climate, 10, 1787-1820.

Smith, T.M., and R. Livezey, 1998: GCM systematic error correction and specification of the seasonal mean Pacific/North American region atmosphere from global SSTs. J. Climate, 11, in press.

Figure Captions

Fig. 1. North American surface air temperature anomaly forecasts (^cC or ^oK) of the two-tiered NCEP model (coupled model for the SST forecast, then AGCM using the SST forecast as boundary condition) for the coming 8 overlapping 3-month periods. The mean of an 18-member ensemble of GCM integration result is used.

Fig. 2. As in Fig. 1, except for precipitation.