H13I-1686
The Changing Nature of Drought Risk in South-east Australia Over the Past Two Millennia

Monday, 14 December 2015
Poster Hall (Moscone South)
Anthony Kiem1, Michelle W Ho2 and Danielle Verdon-Kidd1, (1)University of Newcastle, Callaghan, Australia, (2)Columbia University of New York, Palisades, NY, United States
Abstract:
The Murray-Darling Basin (MDB) is one of the most important food and fibre regions in Australia, producing one-third of the national food supply and exporting produce to many other countries. In total, the Basin contains about 40% of Australia’s farms and 70% of Australia’s irrigated land area. However, the MDB is also one of the most spatially and temporally variable river systems in the world, with severe droughts a regular occurrence over the ~100 years of instrumental record and decadal-scale droughts (e.g. “Federation” (~1895-1902), “World War II” (~1937-1945) and “Millennium” or “Big Dry” (~1997-2010) droughts) matched by flood dominated epochs (e.g. 1950s, 1970s). The accurate estimation of drought risk in the MDB is hampered by relatively short instrumental records and also by the complexity of the region’s climate teleconnections with several large-scale ocean-atmospheric processes in the Pacific (El Niño Southern Oscillation, Interdecadal Pacific Oscillation), the Indian (Indian Ocean Dipole) and Southern Oceans (Southern Annular Mode). Climate-sensitive paleoclimate records provide an opportunity to resolve hydroclimatic variability over long time periods prior to the availability of instrumental records and therefore offer the potential for improved quantification of risks associated with hydroclimatic extremes. However, the MDB, as with many regions in Australia, currently lacks suitable in situ proxies necessary to do this.

Therefore, remote paleoclimate rainfall proxies in the Australasian region spanning are used to develop new reconstructions of MDB rainfall over the Common Era (CE) (i.e. approximately the past 2000 years). The nature of MDB dry epochs from 749BCE to 1981CE are then compared with the frequency and duration of droughts recorded in instrumental records (i.e. approximately the past 100 years). Importantly, the results show that the probability of decadal scale droughts is three times greater than instrumental records suggest.