HR: 08:30h
AN: PP31A-03    [PDF]
TI: Sensitivity to Uncertainties and Change in the Tropical and Subtropical Ocean During the Last Glacial Maximum: Reassessment of the CLIMAP LGM.
AU: * Pisias, N G
EM: pisias@coas.oregonstate.eud
AF: College of Oceanic and Atmospheric Science Oregon State Univ, 104 Ocean Admin Bldg, Corvallis, OR 97331-5503 United States
AU: Hostetler, S
EM: steve@coas.oregonstate.edu
AF: US Geological Survey, Dept. of Geosciences 104 Wilkinson Hall, Corvallis, OR 97331-8564 United States
AU: Mix, A C
EM: mix@coas.oregonstate.edu
AF: College of Oceanic and Atmospheric Science Oregon State Univ, 104 Ocean Admin Bldg, Corvallis, OR 97331-5503 United States
AB: The CLIMAP (1981) reconstructions of sea surface temperature (SST) of the Last Glacial Maximum (LGM) ocean have stimulated significant research and controversy. Some geochemical tracers, and the presence of glaciers in areas such as Hawaii and New Guinea, suggest that the CLIMAP SSTs may be biased in the tropics and subtropics. Here we assemble available data and use a statistical approach to correct for hypothetical biases. We generated new global monthly SST estimates at the LGM as boundary conditions for general circulation model simulations to evaluate sensitivity of global climate to these uncertainties in SST reconstructions. The revised SST estimates are based on the premise that the faunal and floral gradients that underlie the CLIMAP reconstructions reflect ocean temperature gradients and frontal positions, but the transfer function that translates these faunal gradients to SST have biases. We developed a strategy to systematically correct for two biases: 1) low sensitivity of faunal SST estimates at the warmest sites and 2) underestimation of SST where the pycnocline is strong within the euphotic zone. The hypothetical corrections to the published SST fields (August and February) can be expressed by the following equation: SST$_{Adjusted}$ = (1+{\bf a})*SST$_{CLIMAP}$ + {\bf b}*(density$_{0m}$-density$_{100m}$) + {\bf c} Parameter {\bf a} adjusts temperatures for bias as a function of temperature, {\bf b} corrects for bias associated with steepness of pycnocline and {\bf c} shifts the mean SST for the reconstruction. For density gradients, we use modern atlas values of density at 0 and 100m water depth. This equation was applied to the 2$^{o}$ x 2$^{o}$ grid of CLIMAP SST fields. Experiments using GENESIS have been run to examine the effects of the warm-pool and pycnocline biases. Results show significant improvement in comparisons of surface temperatures to independent climate estimates from oceanic and terrestrial records. Preliminary analyses of the model runs suggest show that at latitudes greater than 35$^{o}$ the climate response to changes in the warm pool or in the subtropical gyres are similar, whereas there is marked improvement in the model fit to data in the tropics with changes in the warm pool temperatures.
DE: 4267 Paleoceanography
SC: Paleoceanography and Paleoclimatology [PP]
MN: 2003 Fall Meeting