Models and Techniques That Project an Arid Greenhouse Future Also Project a Wet Last Glacial Maximum, Contrary to Pollen and Dust Evidence

Abstract:

Precipitation (P) and Penman-Monteith potential evapotranspiration (PET) from global climate models (GCMs) have been used to infer that Earth's land areas will dry out under future greenhouse warming outside of the high latitudes. This has been argued using simulated declines in both the aridity index P/PET (by the present author among others) and the Palmer drought index, driven by warming-powered PET increases. However, this picture is at odds with the broad paleoclimate tenet that greenhouse eras in fact appear "wet" on land while cold intervals appear "dry."

Here, we show that the same GCMs which project widespread P/PET declines for the greenhouse future also project fairly widespread P/PET increases (i.e. "wetting") for the last glacial maximum (LGM), when CO₂ was half present levels and snow and ice were extensive. Yet, global pollen and dust records of the LGM suggest mostly "drier"-looking vegetation patterns than today, as we also review here. Thus, either the GCMs' P and/or PET responses to past global change are flawed, or, P/PET change is not relevant for the vegetation response to CO₂-driven climate change. Either way, this calls into question the ecological relevance of the above "drying-out" conclusions. We also show parallel results for the Palmer index, and investigate whether the P/PET results seem any more relevant for lakes and other abiotic wetness indicators than for vegetation.