Evidence for suppressed mid-Holocene northeastern Australian monsoon variability from coral luminescence

Abstract

Summer monsoon rainfall in northeastern (NE) Australia exhibits substantial interannual variability resulting in highly variable river flows. The occurrence and magnitude of these seasonal river flows are reliably recorded in modern inshore corals as luminescent lines. Here we present reconstructed annual river flows for two~120 year mid-Holocene windows based on luminescence measurements from five cores obtained from three separate coral colonies. We were able to cross-date the luminescence signatures in four cores from two of the colonies, providing confidence in the derived reconstruction. Present-day NE Australian rainfall and river flow are sensitive to El Niño-Southern Oscillation (ENSO) variability, with La Niña (El Niño) events typically associated with wetter (drier) monsoon seasons. Thus, our replicated and annually resolved coral records provide valuable insights into the northern Australian summer monsoon and ENSO variability at a key period (6 ka) when greenhouse gas levels and ice sheet cover were comparable to the preindustrial period but orbital forcing was different. Average modern and mid-Holocene growth characteristics were very similar, suggesting that sea surface temperatures off NE Australia at 6 kyr were also close to present values. The reconstructed river flow record suggests, however, that the mid-Holocene Australian summer monsoon was weaker, less variable from year to year (possibly indicative of reduced ENSO variability), and characterized by more within-season flood pulses than present. In contrast to today, the delivery of moisture appears to have been dominated by eastward propagating convective coupled waves associated with the Madden-Julian Oscillation.

River response and climate change in the Wet Tropics, northeast Queensland Australia

“…This wet-dry pattern in the Wet Tropics is consistent with documented changes in the amplitude of ENSO cycles over the last several millennia (Lough et al, 2014;Nott and Forsyth, 2012). While there remain some key limitations for reconstructing hydroclimate, the multi-proxy approach employed in this study enhances our long-term understanding of flood variability in the Wet Tropics.…”

“…Notably, interpretations of changes in rainfall and vegetation have been used to infer a ~60 % reduction in rainfall relative to the present during the glacial period after 45 ka (Kershaw, 1978). Changes in both the frequency and intensity of tropical cyclones (Haig et al, 2014) and late Holocene river discharges (Lough et al, 2014) also point to considerable climatic variability throughout the region. A key question, therefore, is how such changes in discharge and associated sediment supply were reflected in the nature and timing of catchment and river response?…”

“…Prior to 1870, the discharge reconstruction was deemed less reliable than the cyclone reconstruction of Haig et al (2014) owing to the absence of coral proxies derived from reefs within the flood plume extent of Wet Tropics rivers for this time period. This highlights the impact of past coral-based, palaeoclimate research that is biased towards reefs offshore from the Burdekin River, the largest river in the GBR catchment (Isdale et al, 1994;Isdale et al, 1998;McCulloch et al, 2003;Lough, 2007;Lough, 2011a;Lough et al, 2014). In the Wet Tropics, the largest floods within the past 700 years were concluded to have occurred during the period 1680-1820 in association with the highest cyclone activity.…”

“…This model represents an important conceptual development in floodplain geomorphology given the significant effect of ENSO on river flooding globally (Ward et al, 2010). Further, it is well established that since 4 ka, ENSO cycles have been a key driver of hydroclimatic variability in Australia (Shulmeister and Lees, 1995;Donders et al, 2007) and more specifically, North Queensland (Lough, 2011;Lough, 2014). The ENSO phenonmenon is therefore potentially a major factor influencing floodplain formation.…”

“…While there remain some key limitations for reconstructing hydroclimate, the multi-proxy approach employed in this study enhances our long-term understanding of flood variability in the Wet Tropics. The Wet Tropics in the northeast of Queensland has yielded a diverse range of palaeoenvironmental proxies and comprises the continent's most detailed records of Quaternary climate change (Haberle, 2005;Kershaw, 1976;Kershaw, 1978;Kershaw and Nanson, 1993;Moss and Kershaw, 2007;Tibby and Haberle, 2007;Turney et al, 2004;Turney et al, 2006;Haig et al, 2014;Lough et al, 2014). Individually these records point to considerable climatic variability over the past 30 kyr, however, the regional record from the humid tropics has never been systematically This study revealed that fluvial response in Wet Tropics catchments appears regionally synchronous with aggradation during two discrete phases spanning 30-13 ka and 8-5 ka.…”