Understanding the Causes of the Excessive Cold-tongue in Coupled GCMs

PI: De-Zheng Sun

NOAA-CIRES/Climate Diagnostics Center
NOAA/OAR/CDC-(R/CDC1)
325 Broadway
Boulder, CO 80303 USA

Phone: (303) 497-6272

Fax: (303) 497-6449

Email: ds@cdc.noaa.gov

The equatorial cold-tongue is a salient feature in the sea surface temperature (SST) distribution in the tropical Pacific. It corresponds to a region with dry subsiding air and thereby serves as a "radiator fin". The intensity of the cold-tongue also determines the heat flux into the ocean. Therefore, the cold-tongue plays a fundamental role in heat balance of the ocean-atmosphere system in the tropics and beyond. Moreover, the intensity of the equatorial cold-tongue varies substantially on the seasonal, interannual, and longer time-scales. These variations affect climate world-wide including the climate over the continental US.

The physics controlling the intensity of the equatorial cold-tongue, however, has not been fully understood. This lack of understanding is reflected in the inability of current coupled GCMs to simulate correctly the intensity of the equatorial cold-tongue. A cold-bias in the equatorial Pacific SST is a common problem in coupled GCMs without the use of flux adjustment, and a major contributor to the double ITCZ syndrome in the these GCMs. The NCAR CCSM, arguably the most frequently used climate system model by climate researchers world-wide, also suffers severely from this problem.

The almost ubiquitous presence of the excessive cold-tongue in the coupled models may also offer a unique opportunity to understand the nature of the problem and thereby to better understand the dynamics and thermodynamics controlling the intensity of the equatorial cold-tongue. This is because hypotheses developed from a particular model can be readily tested against the output from another coupled model. The Coupled Model Inter-comparison Project (CMIP) and the earlier Atmospheric Model Inter-comparison Project (AMIP) have in fact long envisioned this approach and accordingly coordinated groups to conduct runs with the same or very similar boundary conditions so that inter-comparison studies can be maximumly facilitated (Meehl et al. 2000).

In light of a recent analysis of the feedbacks over the cold-tongue region in the NCAR CCM3--the atmospheric component (Sun et al. 2003), we hypothesize that the cloud and water vapor feedbacks play a fundamental role in determining the intensity of the cold-tongue and that the excessive cold-tongue in many coupled GCMs may be partially due to common errors in the direct and indirect effects of water vapor and cloud forcing. More specifically, if a model overestimates the water vapor feedback and underestimates the negative feedback from the short-wave forcing of clouds, the net atmospheric feedback may be less negative than the observed value. Consequently, even if ocean feedback is perfectly simulated, a drift in the coupled model will be less constrained by the model atmosphere than by the real atmosphere. Clearly, this hypothesis can be tested to a significant degree by comparing the feedbacks in other coupled models with the feedbacks in the NCAR CCSM. In any case, the analysis of the feedbacks in different coupled models may help us to gain confidence in the projection of global warming by these models because these feedbacks also control the sensitivity of the climate system to an external perturbation.

With AMIP runs, we will first quantify the radiative and dynamical feedbacks over the equatorial Pacific cold-tongue region, using the El Nino warming as the forcing signal. The observational data will be from ERBE, CERES, ISCCP, and the NCEP reanalysis. We will then examine the equilibrium tropical Pacific climate of the corresponding coupled models to check whether there is a general correspondence between the severity of the cold-bias in the equatorial Pacific SST and the degree of errors in the water vapor and cloud feedbacks. In conjunction with this analysis, we will examine the surface winds, the upper ocean thermal structure, and the upper ocean circulation in the tropical Pacific with a focus on the aspects that are directly relevant to the intensity of the equatorial Pacific cold-tongue.

We will need the following model outputs:

Fluxes at the TOA and at the surface from AMIP2 runs
Surface wind-stress from the AMIP2 runs
Upper ocean temperature from the forced ocean runs and/or coupled runs. We will be only looking at the subsurface temperature down to the 400 m and over the equatorial Pacific.

References

Meehl, Gerald A., Boer, George J., Covey, Curt, Latif, Mojib, Stouffer, Ronald J., 2000: The Coupled Model Intercomparison Project (CMIP). Bulletin of the American Meteorological Society 81: 313-318

Gates, W.L., 1992: AMIP: The Atmospheric Model Intercomparison Project. Bull. Amer. Meteor. Soc., 73, 1962-1970

Sun, D.Z., J. Fasullo, T. Zhang, and A. Roubicek, 2003: A Note On the Radiative and Dynamical Feedbacks over the Equatorial Cold-tongue J. Climate, accepted ( http://www.cdc.noaa.gov/~ds/dspapers/jc2003-2 or http://www.cdc.noaa.gov/~ds/publications.shtml)