Rapid warming in SE Australia challenges plans to adapt gradually

Studies of prehistoric climate change in Victoria’s western lakes imply that future changes might not be smooth. Dacre Smith's painting of Lake Gnotuk, from Views of Victoria in the steps of von Guerard.

Step changes in warming of a few tenths to 1°C can produce rapid changes in risks such as extreme heat and fire danger. Yet, adaptation-planning that follows the dominant model of smooth climate change makes gradual adjustments to keep up with small changes in extremes. In these circumstances, a rapid change can catch sensitive systems out. Poorly planned responses may also lead to maladaptation.

The IPCC has just released the full Special Report on Extreme Weather Events (SREX: The Summary for Policymakers was released late last year). The report looks at what the science tells us about the likely future incidence of extreme weather. It asks how societies can be better prepared to deal with such events.

The incidence of extreme weather in the SREX report follows the science in the IPCC’s Fourth Assessment Report where anthropogenic climate change is assumed to be smooth. Rapid changes are assumed to be due to natural variability and are thought to be random.

Averaged output from climate models are usually presented as lines of best fit. When the output from many models is combined to project future changes, it is sensibly presented as curves. IPCC temperature projections are a typical example.

It is important to remember that these smooth curves are a generalisation. When output from many individual climate model runs are drawn on a graph, coloured spaghetti diagrams that combine change with random variability are produced. Under climate change the real world will experience a similarly noisy future. The climate change component is known as the signal and the random variability component as the noise.

However, climatologists disagree as to whether the enhanced greenhouse effect leads to smooth or non-linear changes in climate. The two main alternatives are:

  1. Anthropogenic climate change is smooth and all observed noise is climate variability (the dominant view).

  2. Anthropogenic change processes interact with climate variability to produce a non-linear climate signal.

Recent abrupt warming in south-eastern Australia bringing heat waves and historically unprecedented catastrophic fires, led to a re-assessment of these two alternatives. This built on earlier work looking at prehistoric climate change in Western Victorian lakes and historical rainfall changes. This work concluded that climate could switch from one regime to another, implying that future climate change may not be all that smooth.

Lake Keilambete. Photomontage by Roger Jones

Techniques used in these earlier studies were adapted to detect step changes in annual time series of temperature and rainfall. This was combined with methods to attribute changes in temperature to internal climate variability and externally forced climate change.

In a stationary climate dominated by natural variability, rainfall explains variations in maximum temperature and over time the remaining errors are random. Likewise, maximum temperature explains variations in minimum temperature. If the errors are non-random, the system is experiencing an external influence.

The climate of south-eastern Australia was stationary from 1910–1967. In 1968, minimum temperature shifted up relative to maximum temperature by 0.7°C and in 1973, maximum temperature shifted upwards relative to rainfall by 0.5°C. The latter was not picked up in observed maximum temperatures because higher rainfall in the early 1970s masked the change.

In 1997, observed maximum temperature increased by 0.8°C. Of this, 0.3°C was due to a decrease in rainfall associated with the “big dry” and 0.5°C was due to external warming.

Instead of being a trend, temperature in south-eastern Australia has shifted upwards in two episodes. The first episode was in 1968–73 and the second in 1997. Similarly timed shifts were found in average temperature for the latitudes 24–44°S and for the southern hemisphere. The upward shift in temperature in 1997 also coincided with a global increase of 0.3°C. All shifts were statistically significant.

Eleven model simulations for the same region showed a similar pattern of changes. Climate was stationary through the first part of the 20th century followed by an upward shift in minimum temperature. Maximum temperature stepped up at the same time or soon after. Increasing greenhouse gas emissions throughout the 21st century produce a temperature pattern that looks like an upwards curving staircase. Simulated shifts for maximum temperature range from 0.5–1.8°C and for minimum temperature range from 0.3–1.0°C, occurring every 15–40 years.

The main reason for step-like increases in warming is most probably the ocean, which absorbs about 85% of the energy from greenhouses gases as stored heat. This heat builds up and is re-emitted periodically into the atmosphere. The timing of these changes suggests that climate variability is playing a part in the storage and release process.

Ocean storage of heat is the main reason why atmospheric warming takes decades to centuries to respond fully to changes in greenhouse gas concentrations in the atmosphere. The evidence is that this heat is being released in bursts. Circumstantial evidence suggests that rainfall may also be changing in response to these shifts, as has occurred in south-western Western Australia, but more research is required.