A Comparison of "Ice-House" (Modern) and "Hot-House" (Maastrichtian) Drainage Systems: the Implications of Large-Scale Changes in the Surface Hydrological Scheme

A GIS analysis of modern and Maastrichtian (Late Cretaceous) drainage systems has been made in order to investigate the potential differences between the surface hydrology of "ice-house" and "hot-house" worlds and how this might be reflected in the geological record. Because of the importance of CO₂ concentrations for generating "hot-house" climates this study also has implications for potential future changes in the climate system. For the modern system we have utilized global maps of observed river systems, the Hydro1K digital dataset, observations of freshwater and sediment fluxes from recording stations, and modern day climate models and observations. For the Maastrichtian we have compiled a detailed global paleogeographic map and geological database (based on earlier work by the Paleogeographic Atlas Project, University of Chicago) that has been used to generate a paleo-DEM using the suite of hydrological tools in ArcGIS, complete with reconstructed river systems and drainage basins. This forms the primary boundary condition for a coupled ocean-atmosphere experiment using the HadCM3 model, with atmospheric CO₂ set at 4 x pre-industrial levels. The results indicate a Maastrichtian world dominated by high sea surface temperatures (as high as 30-35 C in the tropics), and a consequently greatly enhanced hydrological cycle when compared with the Present. Globally, modeled Maastrichtian precipitation and evaporation are 1.5x that for the Present, with a 2.5x increase in total runoff. These changes are not evenly distributed, either spatially or seasonally, and therefore a detailed consideration of the paleogeography and paleo-drainage is essential, as these changes have a major influence on the distribution of vegetation and freshwater and sediment fluxes. For example, the Maastrichtian Tethyan monsoon, though less intense than noted for other modeled Mesozoic intervals, nonetheless dominates the seasonal distribution of precipitation and runoff over Saharan and northeastern Africa. Seasonally high, modeled freshwater fluxes from the Hoggar Massif (northern hemisphere Summer and Fall) drain south into the Iullemmeden Basin, where they augment persistent runoff from the southern Saharan areas including the proto-Niger drainage. The modeled vegetation and weathering regime of the surrounding hinterland is dominated by everwet tropical forests and intensive chemical weathering, consistent with interpretations from sedimentological and palaeontological observations: the dominance of carbonaceous-rich silts and clays, lack of evaporite minerals and lack of coarse immature clastics. We speculate that changes in the distribution of the seasonal wet-everwet climate regimes due to Milankovitch forcing may account for the cyclicity observed in the Maastrichtian stratigraphy of this region. Along the North African margin the picture is very different with low rates of runoff, high evaporation rates and aridity. This aridity is enhanced locally by the atmospheric consequences of offshore oceanic upwelling. The large differences in the surface hydrology of the Earth between the Maastrichtian "hot-house" and Present-day "ice-house" worlds clearly indicates that we must be prepared to model regimes that may in some areas be very different from the present day. Variations in the distribution and intensity of rainfall may trigger rapid changes in vegetation cover, groundwater levels and activity of soil infauna such as termites, which in turn would greatly affect terrestrial sediment flux and carbon flux responses.