ALBERT

Description: The nature of the warm climates of the Cretaceous has been enigmatic since the first numerical climate models were run in the late 1970s. Quantitative simulations of the paleoclimate have consistently failed to agree with information from plant and animal fossils and climate sensitive sediments. The ‘cold continental interior paradox’ (first described by DeConto et al. in Barrera E, Johnson C (eds) Evolution of the Cretaceous Ocean/climate system, vol 332. Geological Society of America Special Paper, Boulder, pp 391–406, 1999), has been an enigma, with extensive continental interiors, especially in northeast Asia, modeled as below freezing in spite of plant and other evidence to the contrary. We reconsider the paleoelevations of specific areas, particularly along the northeastern Siberian continental margin, where paleofloras indeed indicate higher temperatures than suggested by current climate models. Evidence for significant masses of ice on land during even the otherwise warmest times of the Cretaceous is solved by reinterpretation of the δ18O record of fossil plankton. The signal interpreted as an increase in ice volume on land is the same as the signal for an increase in the volume of groundwater reservoirs on land. The problem of a warm Arctic, where fossil floras indicate that they never experienced freezing conditions in winter, could not be solved by numerical simulations using higher CO2 equivalent greenhouse gas concentrations. We propose a solution by assuming that paleoelevations were less than today and that there were much more extensive wetlands (lakes, meandering rivers, swamps, bogs) on the continents than previously assumed. Using ~ 8 × CO2 equivalent greenhouse gas concentrations and assuming 50–75% water surfaces providing water vapor as a supplementary greenhouse gas on the continents reduces the meridional temperature gradients. Under these conditions the equatorial to polar region temperature gradients produce conditions compatible with fossil and sedimentological evidence.