Effects of subtropical Stratocumulus clouds on coupled simulations

The present work examines the sensitivity of climate simulations done with a CGCM with respect to the incidence and radiative properties of marine stratocumulus (Scu) in the southeastern part of the Pacific and Atlantic Oceans. The simulations last for 30 years, and are computed with the UCLA AGCM coupled with the MIT OGCM. The simulations performed are: a control simulation, a set of three simulations that prescribe in selected regions values of liquid water path (LWP) that would correspond to Scu clouds, and another set of three simulations that prescribe null optical thickness associated to Scu clouds in the same regions considered in the previous set. These simulations are called control, ScuPac, ScuAtl, ScuPac&Atl, NoScuPac, NoScuAtl and NoScuPac&Atl.

ScuPac prescribes an uniform value of LWP at the PBL top in the region that spans from the Equator to 20ºS, and from 80ºW to the South America coast, The prescribed monthly LWP values vary month to month, in a way that allows to reproduce the mean annual cycle of net short wave radiation at the ocean surface, as obtained from NASA SRB analysis, averaged in the same area in which the optical thickness is prescribed. In the rest of the globe, the incidence of Scu clouds and its optical properties are computed by the coupled model. The ScuAtl simulation is analogous to the ScuPac, but prescribes the LWP associated with Scu clouds in the region that spans from the equator to 20ºS, and from 5ºW to the coast of Africa. The ScuPac&Atl prescribes the Scu optical thickness simultaneously at both regions referred above.

The NoScuPac prescribes null LWP at the same region in which the ScuPac simulation prescribes its monthly LWP values. In this way, while ScuPac prescribes Scu properties in order to produce realistic effects on the short wave radiation at the ocean surface in the southeastern tropical Pacific, NoScuPac ignores the radiative effects of the Scu clouds in this region. The differences between the ScuPac and NoScuPac simulations allow to asses the effects of Scu clouds at the southeastern Pacific on this coupled model. In analogous way, the NoScuAtl and NoScuPac&Atl simulations ignore the radiative effects of Scu clouds in the regions in which the ScuAtl and ScuPac&Atl simulations prescribe them.

The results of the control simulation are consistent with those of previous studies that used this coupled model, so we shall refer to some of these results, but we will not show them in this abstract. The Scu incidence has local maximums at the regions offshore Peru (southeastern tropical Pacific), Namibia (southeastern tropical Atlantic), California (subtropical North Pacific) and Canarias Islands (subtropical North Atlantic). Scu incidence is overestimated at the southeastern tropical Atlantic, while at the southeastern Pacific it is realistic during austral winter and spring, and is underestimated during austral winter and summer. The precipitation field obtained from the control simulation shows that the Pacific and Atlantic ITCZ are centered north of the equator, in a manner consistent with the observational data. The Southern Pacific Convergence zone is more zonally oriented and equatorward shifted than in the analysis of observed data, which implies a moderate “double ITCZ” bias. As in previous studies we this coupled model, the control simulation shows ENSO like interannual variability.

Next we focus on the differences between simulations that prescribe non zero LWP and zero LWP in each basin. The SST difference between ScuPac minus NoScuPac simulations (not shown) has a La Niña like type of pattern, with a negative (cold) difference that reaches the western equatorial Pacific. The difference of the mean annual precipitation fields (upper pannel of the left column in the Figure) indicates an ITCZ shifted to the north in the ScuPac simulation, compared to the NoScuPac one. The ScuPac minus NoScuPac difference on stream function at 200hPa (upper panel of the right column in the Figure) shows cyclones approximately symmetric respect to the equator in the eastern tropical Pacific, which are consistent with a Gill type of response to equatorial heating. In the tropical Atlantic, the results indicate that the difference of wind at the Equator and at the 200hPa level is easterly, and so it's opposite to that found in the equatorial Pacific. There are also differences at extratropical latitudes, which are consistent with barotropic waves.

Differences of ScuAtl minus NoScuAtl (middle panels at both columns of the Figure) show at the Atlantic basin similar patterns as those found in the Pacific in the ScuPac and NoScuPac experiments. Besides this, the 200 hPa difference of stream function at 200hPa indicates easterly winds in the equatorial Pacific and in western equatorial South America.

The ScuPac&Atl minus NoScuPac&Atl differences (lower panels of the Figure) are similar to the addition of the experiments that consider each basin separately.

In conclusion, the Scu clouds affect quite strongly the SST and precipitation patterns at the tropical oceans, and through them, the global tropical and extratropical atmospheric circulation. We also find that the Scu incidence both at the Pacific and the Atlantic basins have the potential to affect the other basin, which is consistent with conclusions of previous studies. The ongoing work on these results focuses on the effect of the suppression of regional radiative effects of Scu prescribed in the NoScuPac and in the NoScuPac&Atl simulations on the simulated interannual variability of ocean and atmospheric circulation at the tropical Pacific.