Ning Zeng, J. David Neelin, Chia Chou, Johnny Wei-Bing Lin, and Hui Su
Department of Atmospheric Sciences and
Institute of Geophysics and Planetary Physics
University of California, Los Angeles
We review the development of a new type of model for simulation and theory of the tropical atmospheric component of climate variations. These models are referred to as ``quasi-equilibrium tropical circulation models'', or QTCMs, because they make use of approximations associated with quasi-equilibrium (QE) convective parameterizations originally proposed by Arakawa and Schubert. The formulation uses a Galerkin framework in the vertical, but with basis functions tailored to quasi-equilibrium deep convective physics via analytical solutions. The first of these (QTCM1) retains a single vertical structure of temperature and humidity. For a balanced treatment of dynamics and sub-grid scale physics, a physics parameterization package of intermediate complexity is developed. This includes a linearized longwave radiation scheme, a simple cloud prediction method, simple shortwave radiation schemes, and an intermediate land-surface model.
The QTCM1 climatology has a reasonable spatial pattern and seasonal evolution of the tropical convergence zones, including over land regions. The outgoing longwave radiation and net surface heat flux are simulated satisfactorily. The intraseasonal oscillation is largely excited by midlatitude disturbances and is influenced by the evaporation-wind feedback mechanism. The Asian monsoon is slightly weak but depicts the northward progression of the monsoon onset, and a monsoon wind shear index exhibits interannual variability associated with observed SST. The extent and position of the main El Ni\~no/Southern Oscillation rainfall anomalies are simulated, as well as a number of the observed tropical and subtropical teleconnections. The seasonal cycle and interannual variability of the Amazon water budget illustrate reasonable simulation of the hydrologic cycle. While the results are imperfect with respect to observations, many aspects are comparable to GCMs of the previous generation. Considering the complexity of these simulated phenomena, the model is computationally light and easy to diagnose. Sensitivity experiments show that the convective time scale that controls departures from QE has little impact on the climatology, although it can influence fast waves. The model helps to demonstrate the usefulness of QE thinking not only as a powerful concept, but also as a basis for quantitative simulation of tropical climate and variability.