The climate sensitivity of CCSM3 is studied for two past climate forcings, the Last Glacial Maximum (LGM) and the mid-Holocene. The LGM, approximately 21,000 years ago, is a glacial period with large changes in the greenhouse gases, sea level, and ice sheets.

Preindustrial (PI) simulations of the Community Climate System Model version 3 (CCSM3) at two resolutions, a moderate and a low resolution, are described and compared to the standard controls for present-day (PD) simulations. Because of computational efficiency, the moderate- and low-resolution versions of CCSM3 may be appropriate for climate change studies requiring simulations of the order of hundreds to thousands of years. The PI simulations provide the basis for comparison for proxy records that represent average late Holocene conditions. When forced with PI trace gases, aerosols, and solar irradiance estimates, both resolutions have a global cooling of 1.2°–1.3°C, increased sea ice in both hemispheres, and less precipitation near the equator and at midlatitudes as compared to simulations using PD forcing. The response to PI forcings differs in the two resolutions for North Atlantic meridional overturning circulation (MOC), the Antarctic Circumpolar Current (ACC), and ENSO. The moderate-resolution CCSM3 has enhanced ACC, North Atlantic MOC, and tropical Pacific ENSO variability for PI forcings as compared to PD. The low-resolution CCSM3 with more extensive sea ice and colder climate at high northern latitudes in the PD simulation shows less sensitivity of the North Atlantic MOC to PI forcing. ENSO variability and the strength of the ACC do not increase with PI forcing in the low-resolution CCSM3.

No abstract

Observations in South America based on paleoenvironmental data for the Last Glacial Maximum (LGM) that shows a generally lower temperatures and reduced precipitation of the region are compared to simulation results from the paleoclimate version of the National Center for Atmospheric Research coupled climate system model. Analyses of the LGM wind simulation for the tropical Atlantic show that the convergence zone does not extend all the way into South American continent during the Austral summer. This would have prevented moisture inflow into the adjacent continental area of equatorial northeastern Brazil. Interpretations of paleoclimate proxy records from this region are consistent with this scenario. In the subtropics, LGM westerlies were weaker and data show characteristics of more humid conditions. At higher latitudes model simulations suggest intensification of the westerlies which, in combination with colder sea surface temperatures would imply less moisture influx into the South American continent from the Atlantic sector. Low lake levels in high southern latitudes and general aridity would support the model findings.

[1] Time series from the PIRATA 10-minute measurements are investigated in order to study the upper-ocean thermal field time-scales of variability using the serial covariance power spectra method and wavelet analysis. This is one of the few studies that analyze in time and space the PIRATA buoy data in the region. Analysis of the depth of the 20°C isotherm show distinct patterns for the northwestern Atlantic at 8N -38W and the southeastern Atlantic, at 10S-10W. In the west, warm surface waters are accompanied by a shallower thermocline while in the east warm surface waters are associated with its deepening, consistent with Houghton [1991]. At higher frequencies the dominant signals in sea surface temperature are the diurnal and semi-diurnal cycles. Spectral analysis of the depth of the 20°C isotherm show a predominance of the semi-diurnal period. In the eastern Atlantic there are two additional dominant bands in the spectra: 10-14 days, 25-27 days and for the depth of the 20°C isotherm there is also a lower frequency band at 20-40 days associated with tropical instability waves at the equator. The first band can be associated with atmospherically forced motions and the lower frequency band with barotropic instability waves.