Simulations of Permian Climate and Comparisons with Climate‐Sensitive Sediments

• Published in: The Journal of geology 2002-01, Vol.110 (1), p.33-55
• Main Authors: Gibbs, Mark T., Rees, P. McAllister, Kutzbach, John E., Ziegler, Alfred M., Behling, Pat J., Rowley, David B.
• Format: Article
• Language: English
• Subjects: Atmospherics; Climate change; Climate models; Coal; Modeling; Oceans; Paleoclimatology; Paleogeography; Precipitation; Prehistoric era; Sediments; Simulation; Simulations; Summer
• ISSN: 0022-1376; 1537-5269
• DOI: 10.1086/324204

We use a climate model to simulate two intervals of Permian climate: the Sakmarian (ca. 280 Ma), at the end of the major Permo‐Carboniferous glaciation, and the Wordian (ca. 265 Ma). We explore the climate sensitivity to various levels of atmospheric CO2concentration and to changes in geography and topography between the two periods. The model simulates large seasonality and high aridity in the continental interiors of both hemispheres for both periods. The northern summer monsoon weakens and the southern monsoon strengthens between the Sakmarian and the Wordian, owing to changes in geography and topography. The northern middle and high latitudes cool in winter, between the Sakmarian and Wordian, associated with northward shift of the continents. This high‐latitude cooling strengthens the winter westerlies and shifts the maximum storm‐track precipitation south. In the Southern Hemisphere, the winter westerlies weaken from the Sakmarian to the Wordian. Starting the simulations with no permanent ice fields (i.e., by assuming that the late Sakmarian postdates deglaciation) and imposing increased levels of atmospheric CO2four times the present level, we find no tendency for reinitiation of major glaciation. Some permanent snow fields do develop in high southern latitudes, but these are primarily at high elevation. However, the combination of low CO2levels (such as present‐day levels) and a cold summer orbital configuration produces expanded areas of permanent snow. The results are based on statistics derived from the final 5 yr of 20‐yr simulations. Paleoenvironmental indicators such as coal, evaporite, phosphate, and eolian sand deposits agree qualitatively with the simulated climate. The extreme cold simulated in high latitudes is inconsistent with estimates of high‐latitude conditions. Either the interpretation of observations is incorrect, the model is incorrect, or both; a possible model deficiency that leads to cold conditions in high latitudes is the relatively weak ocean‐heat transport simulated by the heat diffusion parameterization of the upper ocean model.