Chapter 10 Gcm Sensitivity To 1982–83 Equatorial Pacific Sea-Surface Temperature Anomalies

Abstract

Abstract Three control and anomaly simulation pairs run with the Goddard Laboratory for Atmospheric Sciences (GLAS) climate model have been analyzed in order to investigate the atmospheric response to the 1982–83 tropical sea-surface temperature anomalies. The observed 1982–83 SST anomalies obtained from the Climate Analysis Center were applied to two separate 75-day control simulations, starting on 16 Dec. 1982 and 16 Dec. 1979, respectively, and a third 60-day control simulation starting on 1 Jan. 1975. In each experiment the equatorial Pacific precipitation increased significantly in a wide band stretching from just east of the dateline to the South American coast, in agreement with observed outgoing longwave radiation (OLR) anomalies. West of this region the precipitation was reduced in the anomaly simulations. As in previous GCM experiments, the major contributor to the tropical precipitation changes was the low-level moisture convergence. The largest evaporation differences were around 4 mm per day and occurred over the regions of highest SST in the anomaly simulations. The tropical sea-level pressure field showed a marked Southern Oscillation pattern, with a magnitude of roughly 2 mb and a node at the dateline. There was a strong (∼ 10ms −1 ) increase in the equatorial eastern Pacific 850mb westerlies as well as a large (∼ 20ms −1 ) easterly wind anomaly at 200 mb. The latter anomaly was flanked by strong (∼ 20ms −1 ) westerly anomalies at roughly 30°S and 30°N. In agreement with earlier simulations with composite SST anomalies, the tropical precipitation anomalies for 1982–83 were also closely related to the extent of very warm (≥ 29°C) sea-surface waters. Each experiment had anomalous anticyclonic circulations aloft straddling the equator in the eastern Pacific, although they were weaker and more eastward than those observed. The extra-tropical response varied between the three experiments, as well as between months of a given experiment. Over North America the ensemble average anomaly minus control 300 mb geopotential height difference field resembled the observed February or March anomaly field more than the typical PNA-like pattern. Other extra-tropical responses were difficult to interpret, although they were clearly equivalent barotropic in structure and showed a much stronger dependence on initial conditions than was noted for the tropics.