A new study of Law Dome Ice cores tells us that droughts are common in Australia, and that there appears to be eight mega-droughts over the last thousand years, including one that lasted a whopping 39 years from 1174- 1212AD. By their reckoning the 12th Century in Australia was a shocker with 80% of it spent in drought conditions. Things weren’t so bad from 1260 – 1860, at least, as far as they can tell. The researchers are convinced theirs is the first millennial-length Australian drought record. It does seem significant.
The researchers, sensibly, think we might want to pay attention to the Pacific cycles and store a bit more water. Without fanfare the paper also suggests that droughts were worse in medieval times.
“this work suggests Australia may also have experienced mega-droughts during the Medieval period that have no modern analog. Therefore, management of water infrastructure in eastern Australia needs to account for decadal-scale droughts being a normal feature of the hydrological cycle.”
h/t to Paul Homewood at Notalotofpeopleknowthat
The ABC reported this largely as a water management story, without asking whether their past stories that blamed CO2 for droughts were less likely to be true. They also didn’t talk about how natural climate change could be worse than the current “man-made” variety. Nor did they discuss the uncomfortable idea that if there were naturally more droughts when CO2 was ideal, perhaps the droughts now might be natural ones too? At least The Australian and Graham Lloyd put the emphasis on the idea that current droughts are par for the course.
Figure 4 (see Caption in footnote to blog).
So how does a study find droughts in Australia with a hole drilled in Antarctic ice?
A very good question — and though the study is corroborated with historic studies, and instruments in Australia as much as it can be, the bottom line is that we really need millenial proxies from Australia, but at the moment, this is the best we have. Law Dome is a spot in East Antarctica that is almost due south of Perth, Western Australia. Researchers looked at sea salt in the ice to estimate wind speed, and in this case, it means specifically winds from the Indian Ocean sector of the Southern Ocean. Those winds correlate with the strength and position of the Antarctic High and circumpolar trough (among other things). Law Dome has such good resolution they can even identify summer patterns from winter ones. The age uncertainty is ±1 y at 1000AD. It’s just flat out remarkable. Other researchers have used these sea salts as a proxy for Eastern Australia rain.
The paper also reconstructs the IPO — the Interdecadal Pacific Oscillation (the big cousin of the PDO*) and also talks about the effect of Antarctic Rossby waves on rainfall in Australia. They used a combination of annual snowfall and sea salt to reconstruct the IPO. They combine the sea-salt rainfall proxy with the reconstructed IPO cycles to estimate when the droughts hit, and confirmed as many of these as they could with other studies.Things were dry during the medieval warm period, then wetter during the little ice age:
When the previously published LD summer sea salt rainfall proxy and the new IPO reconstructions are combined, the three major epochs (two dry – AD 1000–1260, AD 1920-2009 and one wet – AD 1260–1860) previously identified in Vance et al. [2013] are confirmed. Eight IPO positive phase mega-droughts are observed, five during the first dry epoch (Fig. 4e), and six prior to AD 1320. The longest mega-drought, of 39 y duration (AD 1174–1212), occurred at the end of a century of pronounced IPO positive drought conditions, with 80 of the 111 y period AD 1102–1212 in drought.
ABSTRACT
The Interdecadal Pacific Oscillation (IPO) influences multidecadal drought risk across the Pacific, but there are no millennial-length, high resolution IPO reconstructions for quantifying long-term drought risk. In Australia, drought risk increases in positive phases of the IPO, yet few suitable rainfall proxies and short (∼100 y) instrumental records mean large uncertainties remain around drought frequency and duration. Likewise, it is unknown whether megadroughts have occurred in Australia’s past. In this study, an atmospheric teleconnection in the Indian Ocean mid-latitudes linking East Antarctica and Australia is exploited to produce the first accurate, annually dated millenniallength IPO reconstruction from the Law Dome (East Antarctica) ice core. Combined with an eastern Australian rainfall proxy from Law Dome, the first millennial-length Australian mega-drought (>5 y duration) reconstruction is presented. Eight mega-droughts are identified including one 39 y drought (AD 1174–1212), which occurred during an unprecedented century of aridity (AD 1102–1212).
This part seems important. They appear to be quietly suggesting that Australia was “warm” in the medieval warm period, then using this to say that droughts might get worse under future warming. They don’t entertain the thought that past warm periods and droughts have nothing to do with CO2.
Nonetheless, it is clear that this study identifies far fewer mega-droughts during the long, wetter middle epoch of AD 1260–1860. It is possible that the link between IPO positive phases and increased drought risk for eastern Australia is temperature dependent. If the medieval and recent dry epochs were warmer than the wet epoch of AD 1260–1860 (as suggested by a recent continental-scale temperature reconstruction [2k Consortium, 2013]) the IPO positive drought relationship may intensify under future warming.
This study won’t pick up droughts that happen in IPO negative times (La Nina eras)
We note that we present an IPO positive subset of the drought history of eastern Australia. However, droughts can occur at any time, a prime example being the ‘Big Dry’ or ’Millennium’ drought (1997-2009). This drought is thought to be of exceptional length and severity [Gergis et al., 2012],
I think they are ever so quietly telling Gergis off.
CAPTION
Figure 4. (a) LD summer sea salt record-Australian Water Availability Project (AWAP)
annual (Jan-Dec) rainfall correlation for IPO positive phases (1924–41, 1979–97) (significance at
contour r=0.325, p < 0:05, unreliable data masked [Gallant et al., 2013]). M, S, T,Wdenote long
record rainfall stations [Vance et al., 2013]. (b) 13 y window sliding correlations for LD summer
sea salt-rainfall station data (p <0.0001–0.01) with blue banding showing the IPO positive phases
used to calculate AWAP and station data correlations. The instrumental IPO series is shown
(red) with 0.5 threshold (blue). (c) Timeseries of rainfall stations (black) and LD summer sea
salts (pink). (d) Independent reconstructions of the IPO (blue – Decision Tree, red – Piece-wise
linear) with IPO positive phases (>0.5 for both reconstructions) highlighted (blue banding). (e)
Annual LD summer sea salt timeseries [Vance et al., 2013] (grey) with 13 y Guassian smooth
(thick black) and drought periods (> 5 y duration, >0.5 for both IPO reconstructions) identified
(pink banding).
REFERENCE
Vance et al, Interdecadal Pacifi c variability and eastern Australian mega-droughts over the last millennium (2014) American Geophysical Union, doi: 10.1002/2014GL062447
* The IPO v the PDO: From Vance et al ” The Interdecadal Pacific Oscillation (IPO) and the Pacific Decadal Oscillation (PDO)
are somewhat interchangeably used to describe low frequency Pacific Ocean sea surface temperature anomalies (SSTA). However the IPO differs from the PDO in that it describes the Pacific basin-wide bi-hemispheric expression of the North Pacific derived PDO [Power et al., 1999; Parker et al., 2007] Negative phases of both modes reflect La Ni˜na-like
SSTA, while positive phases appear El Ni˜no-like, and phase changes occur every 20-30 y [Mantua et al., 1997; Power et al., 1999].