The lag between tropical expansion and rainfall shifts – is it to disappear?

Synoptic chart for Sunday. Note the subtropical high pressure belt located around 48°S, around eight degrees south of its historic summer position

For over a decade now, I have noticed a major gap between the observed expansion of the tropics (Seidel et. al, 2008) and observed poleward shifts in rainfall in the southern hemisphere. A combination of observed measurements of the poleward edge of the Hadley Circulation suggests an expansion of eight degrees since the 1950s. Whilst as the following table for a representative set of stations in southern Australia and Central Chile shows, rain belts have shifted poleward – to disastrous effect for the water supplies of Perth – the shifts have not been nearly so marked as would be predicted from simply moving each locality seven to eight degrees equatorward:

Station	Virgin mean rainfall (beginning of record to 1974)	Mean rainfall predicted from simple 7.5˚ poleward tropical expansion	Percentage decline vis-à-vis virgin mean rainfall	Actual lowest observed rainfall
Santiago de Chile (MJJA)	272.0 mm	15 mm	94.48%	22.1 mm*
				37.2 mm
Concepción (annual)	1230.3 mm	121.6 mm	90.12%	598.6 mm
				599.3 mm
Valdivia (annual)	2393.5 mm	275 mm	88.51%	1033.1 mm
Perth (annual)	882.0 mm	233.2 mm	73.56%	466.6 mm
Perth (MJJA)	624.9 mm	158.0 mm	74.72%	260.2 mm
Collie (MJJA)	667.0 mm	140.8 mm	78.89%	275.9 mm
				296.3 mm
Manjimup (annual)	1055.6 mm	248.5 mm	72.73%	549.0 mm
Horsham Polkemmet Road (annual)	450.8 mm	185.0 mm	58.96%	181.1 mm

An asterisk (*) indicates that the record low rainfall occurred before 1974. MJJA (May-June-July-August) refers to the May to August period that constitutes the rainy season in Southwest Western Australia and Central Chile.

As we can see, the annual rainfall one would expect from a 7.5˚ poleward shift in all climate belts remains less than the driest observed year for all selected stations except Horsham Polkemmet Road (BOM 079023). Even there the driest observed year – 1982 – is only a few millimetres drier than the estimated mean. In addition, the stations used to model Wimmera rainfall under a 7.5˚ poleward shift in climate belts suggest it likely that the median would be under 181 millimetres even with a mean of 185 millimetres.

One major problem is that Central Chile rainfall was historically limited much more by unfavourable land-ocean temperature gradients than by the descending limb of the Hadley Cell. The dryness of the El Niño years of 2014, 2015, and 2018 suggest, however, that such is emphatically not the case beyond the 2010 “magic gate”. The implication is that current ongoing expansion of the Hadley Cell sets a rigid ceiling upon Central Chile rainfall in a manner absent even during the 2000s, when Santiago exceeded its maximum 2010s MJJA rainfall of 209.9 millimetres in four non-El Niño years (2000, 2001, 2005 and 2008). Another problem is that the topography and coastal shape 7.5˚ closer to the equator differ from those surrounding the stations listed, although I was careful to choose those stations least likely to be controlled by differences of this type.

What is revealing about the last three months – in which Melbourne has seen only 33 millimetres with little hope for more in the foreseeable future – is that the subtropical anticyclone has been located as far south as 48˚S (see synoptic map at top). If we combine the subtropical anticyclone’s historic summer position and the known expansion of the Hadley Cell since the 1950s, 48 degrees South is almost precisely where we would expect the summer subtropical anticyclone to be today. This has suggested to me that we will be observing a “catch up” of rainfall belts with the observed shifts of the Hadley Cell since the 1950s (Seidel et. al, 2008, Liu et. al, 2012, plus personal communication). Should this be correct, rainfalls in Southwest Western Australia, southeastern Australia and Central Chile will, beginning this year, show dramatic declines below 2010s averages. These 2010s averages are already 30 percent below virgin averages before man-made greenhouse emissions expanded the Hadley Cell, and 50 percent less in the Santiago region.

Given widespread predictions of another record El Niño in 2019, annual rainfalls in southern Australia of one-half or even one-quarter existing record lows appear even at this early stage a probable outcome if we study the above tables. Even if positive Indian Ocean Dipole and negative Southern Oscillations are less persistent than some models (e.g. Chie et. al. 2008) suggest, there is still a likelihood that the frontal rain belt will be wholly shifted beyond any part of mainland Australia by the “catch up” noted in the previous paragraph.

The implications for public and private farming policy of a 7.5˚ or larger shift in annual rainfalls are stark. The winter rain belt would become wholly extinct, and with it rainfed winter grain crops – a complete 2019/2020 crop failure throughout southern Australia already appears plausible. Irrigated crops would also likely disappear. The rainfall declines modelled at the beginning of this article would certainly mean zero median annual runoff (Chiew et. al. 2006) for every river in Australia’s historic winter rainfall zone.

What governments would do confronted with this situation and powerful agribusinesses demanding bailouts from certain severe financial losses is not worth imagining. Expensive schemes to redirect runoff from other parts of Australia, or desalination and pipelines, would create still more disastrous after-effects in greenhouse gas emissions and disturbance to sensitive and unique river ecologies. Nonetheless, I still think it plausible that agribusiness possesses sufficient power to gain such bailouts, tragic as they would be not only for Australia’s ecology but for the remainder of the globe. It would even speak ill of Australia if it were to abandon farming because of catastrophic climate change rather than as a result of recognising it as inherently unsustainable on our uniquely ancient soils.

Rainfall methodology:

To estimate rainfall in Central Chile and Southern Australia under a 7.5˚ of latitude poleward shift of rain belts:

1. rainfall stations with the most similar topography and coastal aspect to land 7.5˚ northward were selected
2. rainfall for stations 7.5˚ northward and the most similar coastal aspect for the period before the first “magic gate” in 1975/1976 was entered for the stations in (1)
1. for stations in southwestern WA, stations in BOM District 6 (West Gascoyne) were chosen, and for Collie (009628) and Manjimup (009573) stations comparably distant from the coast were used
2. for Collie (009628) and Manjimup (009573) the estimated rainfall under a 7.5˚ poleward shift was increased by 10 percent to account for the Darling Scarp orographic effect.
3. for Perth, data from Carnarvon (006062 and 006011) were used without alteration
4. for Horsham Polkemmet Road, stations in BOM district 46 (Western—Far Northwest) were used
5. for stations in Chile, stations on the coast or nearby 7.5˚ northward were used

References:

• Chandler, Mark A.; Rind, David and Rüdy, Reto; ‘Pangaean climate during the Early Jurassic: GCM simulations and the sedimentary record of paleoclimate’; Geological Society of America Bulletin, v. 104 (May 1992), p. 543-559
• Chie Ihara; Yochana Kushnir and Mark A. Cane; ‘Warming Trend of the Indian Ocean SST and Indian Ocean Dipole from 1880 to 2004’; Journal of Climate, vol. 21 (2008), pp. 2035-2046
• Chiew, Francis; Peel, Murray; McMahon, Thomas Aquinas and Siriwardena, Lionel (2006); ‘Precipitation elasticity of streamflow in catchments across the world’; [Harry Lins, Richard Vogel, Mike Bonell, Wolfgang Grabs et al.; WMO/UNESCO WCP-Water, FRIEND 2006, Havana Cuba, 26 November-1 December 2006]
• Hochman, Zvi; Gobbett, David L.; and Horan, Heidi; ‘Climate trends account for stalled wheat yields in Australia since 1990’; Global Change Biology (2017); published by CSIRO Agriculture and Food
• J. Liu, M. Song, Y. Hu and X. Ren; ‘Changes in the strength and width of the Hadley Circulation since 1871’; Climates of the Past; vol. 8 (2012); pp. 1169-1175
• Seidel, Dian J. Qiang Fu; Randel, William J. and Reichler, Thomas J.; ‘Widening of the tropical belt in a changing climate’; Nature Geoscience, vol. 1 (January 2008), pp. 21-24