Changes in south-east African tropical cyclone landfall frequency over the past 161 years

Abstract

Variations in the sediment input to the Namaqualand mudbelt during the Holocene are assessed using an integrative terrestrial to marine, source to sink approach. Geochemical and Sr and Nd isotopic signatures are used to distinguish fluvial sediment source areas. Relative to the sediments of the Olifants River, craton outcrops in the northern Orange River catchment have a more radiogenic Sr and a more unradiogenic Nd isotopic signature. Furthermore, upper Orange River sediments are rich in heavier elements such as Ti and Fe derived from the chemical weathering of Drakensberg flood basalt. Suspension load signatures change along the Orange River's westward transit as northern catchments contribute physical weathering products from the Fish and Molopo River catchment area. Marine cores offshore of the Olifants (GeoB8323-2) and Orange (GeoB8331-4) River mouths show pulses of increased contribution of Olifants River and upper Orange River input, respectively. These pulses coincide with intervals of increased terrestrial organic matter flux and increased paleo-production at the respective core sites. We attribute this to an increase in fluvial activity and vegetation cover in the adjacent catchments during more humid climate conditions. The contrast in the timing of these wet phases in the catchment areas reflects the bipolar behavior of the South African summer and winter rainfall zones. While rainfall in the Orange River catchment is related to southward shifts in the ICTZ, rainfall in the Olifants catchment is linked to northward shifts in Southern Hemisphere Westerly storm tracks. The later may also have increased southern Benguela upwelling in the past by reducing the shedding of Agulhas eddies into the Atlantic. The high-resolution records of latitudinal shifts in these atmospheric circulation systems correspond to late Holocene centennial–millennial scale climate variability evident in Antarctic ice core records. The mudbelt cores indicate that phases of high summer rainfall zone and low winter rainfall zone humidity (at ca. 2.8 and 1 ka BP) may be synchronous with Antarctic warming events. On the other hand, dry conditions in the summer rainfall zone along with wet conditions in the winter rainfall zone (at ca 3.3, 2 and 0.5 ka BP) may be associated with Antarctic cooling events.