Equatorial Pacific

Figure 3. Summary diagram for the convective—radiative-mixing processes in the tropical ocean—atmosphere.

intraseasonal oscillations in the warm pool region illustrated in Fig. 3. Because the precipitation rate is greater than the evaporation rate over the equatorial western Pacific warm pool, whereas it is smaller over the eastern Pacific, a barrier layer appears year round within the warm pool (e.g. Lukas and Linstrom, 1991; Ando and McPhaden, 1997; Vialard and Delecluse, 1998ab). The barrier layer insulates the penetrated solar energy in the barrier layer from the mixed layer. As an intraseasonal oscillation propagates eastward to the warm pool, the strong westerly wind burst causes strong vertical mixing in the upper ocean that destroys the barrier layer and entrains cold water beneath into the mixed layer. The rapid decay of the in-traseasonal oscillation near the dateline means that there is no impact on the barrier layer there and vertical entrainment would entrain the warm water beneath into the mixed layer. These entrainment processes together with the surface air-sea interaction processes illustrated in Fig. 3 may result in a cool SST anomaly over the warm pool and a warm SST anomaly near the dateline. The SST anomalies in turn induce an eastward extension of atmospheric intraseasonal oscillations, and reduce easterly trade winds by weakening the overall zonal SST gradient. The air-sea interaction might strengthen and prolong the intraseasonal oscillation and play an important role in the development stage of ENSO, which could be phase-locked to the annual cycle.

The above discussion proposes some key climate processes to be addressed by more advanced models with more physical representations of convective, radiative, and mixing processes. In addition, it is important to utilize new observations like satellite measurements to advance cloud-climate feedback study in climate models (e.g. Lau et al., 2005). There are also some critical issues related to convective-radiative-microphysical processes not discussed in this review article, and they deserve further investigation. For example, turbulence mixing in the boundary layer is still crudely parametrized in CRMs and needs to be improved. There are different time scales in the cloud-microphysical processes (condensation versus coalescence) and the probabilistic nature of microphysical processes that should be included in the micro-physics parametrization (e.g. Chen and Liu, 2004). The cloud-aerosol interaction, which is not considered in most CRM's, needs to be studied comprehensively in the future.

0 0

Post a comment