Detection and explanation of spatiotemporal patterns in Late Cenozoic palaeoclimate change relevant to Earth surface processes

Abstract

Abstract. Detecting and explaining differences between palaeoclimates can provide\nvaluable insights for Earth scientists investigating processes that are\naffected by climate change over geologic time. In this study, we describe\nand explain spatiotemporal patterns in palaeoclimate change that are\nrelevant to Earth surface scientists. We apply a combination of multivariate\ncluster and discriminant analysis techniques to a set of high-resolution\npalaeoclimate simulations. The simulations were conducted with the ECHAM5\nclimate model and consistent setup. A pre-industrial (PI) climate\nsimulation serves as the control experiment, which is compared to a suite of\nsimulations of Late Cenozoic climates, namely a Mid-Holocene (MH, approximately 6.5\u2009ka), Last Glacial Maximum (LGM, approximately 21\u2009ka) and Pliocene (PLIO, approximately 3\u2009Ma)\nclimate. For each of the study regions (western South America, Europe,\nSouth Asia and southern Alaska), differences in climate are subjected to\ngeographical clustering to identify dominant modes of climate change and\ntheir spatial extent for each time slice comparison (PI–MH, PI–LGM and\nPI–PLIO). The selection of climate variables for the cluster analysis is\nmade on the basis of their relevance to Earth surface processes and includes\n2\u2009m air temperature, 2\u2009m air temperature amplitude, consecutive freezing days,\nfreeze–thaw days, maximum precipitation, consecutive wet days, consecutive\ndry days, zonal wind speed and meridional wind speed. We then apply a\ntwo-class multivariate discriminant analysis to simulation pairs PI–MH,\nPI–LGM and PI–PLIO to evaluate and explain the discriminability between\nclimates within each of the anomaly clusters. Changes in ice cover create\nthe most distinct and stable patterns of climate change, and create the best\ndiscriminability between climates in western Patagonia. The distinct nature\nof European palaeoclimates is statistically explained mostly by changes in\n2\u2009m air temperature (MH, LGM, PLIO), consecutive freezing days (LGM) and\nconsecutive wet days (PLIO). These factors typically contribute\n30\u2009%–50\u2009%, 10\u2009%–40\u2009% and 10\u2009%–30\u2009%, respectively, to climate\ndiscriminability. Finally, our results identify regions particularly prone\nto changes in precipitation-induced erosion and temperature-dependent\nphysical weathering.