Australasian Science: Australia's authority on science since 1938

Droughts? Floods? Or Will We Run Out of Fuel First?

By James Ward & Simon Beecham

Does the impending arrival of “peak carbon” mean that alarming climate change scenarios need to be revised downwards?

For Australia’s water resource managers, one of the hottest topics is the impact of long-term climate change. Are Queensland’s floods and cyclones a sign of the future? Will there be more droughts in the Murray–Darling Basin?

The stakes are certainly high when it comes to the impacts of long-term climate change on water resources, and some of the predictions are indeed dire. But have you ever considered that there may not be enough carbon fuel available to burn to generate such severe changes in climate? Most people, it seems, have not.

Unfortunately, when studying the changing water cycle, hydrologists are always faced with tremendous uncertainty. With a system this complex there is a great deal of natural chaos, leading to Australia’s natural variability in rainfall (our land of “droughts and flooding rains”).

Because of this natural uncertainty, the computer models we use to assess future climate change impacts from greenhouse gas emissions often do not agree with each other. This is especially true when it comes to future changes in rainfall patterns. In fact, some models predict a decrease in rainfall while others predict an increase. And, the more severe the climate change scenario, the greater the disagreement between the models.

One of the major unknowns we have when it comes to modelling our future climate is guessing what the planet’s future greenhouse gas emissions are going to be. To deal with this problem, in 2000 the Intergovernmental Panel on Climate Change’s Special Report on Emissions Scenarios set out a wide range of different future greenhouse gas emission pathways to 2100. There were 40 scenarios in all, but these were simplified to six “marker scenarios” – two high emissions, two medium and two low.

When the report was released in 2000, oil and coal were both much cheaper than they are today, and it was generally thought that there were plenty of fossil fuel resources available. Without global action to cut carbon emissions, it seemed reasonable to expect continued high emissions growth to 2100 and beyond.

As a result, it has become normal for modellers to adopt the medium and high emissions scenarios for their long-term projections. The high scenario is often called “business as usual” because it represents a general continuation of the last century of steady growth in fossil fuel consumption.

For water resources predictions, business as usual scenarios effectively amplify the existing uncertainty in model results. In other words, the problem of models disagreeing (e.g. over whether the future climate will be wetter or drier) becomes far worse when the models are forced by high greenhouse gas emissions.

In CSIRO’s 2007 report Climate Change in Australia, the results of many different global climate models were assessed for low, medium and high emissions scenarios, and the spread of model output was shown as percentiles. As an example, take the predicted change in annual rainfall for south-east Queensland (Fig. 1), home of the recent catastrophic floods. Under low emissions, predictions for 2060–2080 range from a decrease of around 10–20% (top) to an increase of around 5–10% (bottom) compared with the baseline 20-year period 1980–99. Meanwhile, under high emissions the spread is effectively doubled: the driest models predict a decline of 20–40% while the wettest models give an increase of 10–20%. High emissions clearly increase uncertainty, leaving water resources managers understandably perplexed.

At this point, it should be starting to become clear that if we begin to run short on fossil fuels sooner than we thought, then the climate may not change as much as if we continue to burn carbon at ever-increasing rates.

The “peak fossil fuel production” phenomenon is was first identified by geologist M. King Hubbert in the 1950s. Hubbert proposed that the production rate of a finite resource is likely to exhibit a collective rise, then a peak, followed by a collective decline. Peak oil production has been observed in many countries including the US, Norway and Australia, validating Hubbert’s hypothesis.

In fact, the concept is not limited to oil, but affects any non-renewable commodity. For instance, coal production in the UK exhibited a classic peak about 100 years ago, and has been in fairly steady decline ever since.

Despite its long history, the “peak fossil fuel” phenomenon was hardly being discussed in 2000, when energy was abundant and cheap. Since then the oil price has rocketed from about $20 to more than $100 per barrel, and there has been growing scientific interest in the potential imminent peak (and decline) of the world’s oil supply.

The UK Energy Research Centre reviewed more than 500 studies of peak oil, and concluded that the peak in production is most likely to occur before 2030, with forecasts of a later peak “at best optimistic and at worst implausible”.

Increasing evidence emerging in the scientific literature suggests that all fossil fuel resources – oil, coal and gas, conventional and unconventional – are in far shorter supply than has previously been assumed. Just as oil is now forecast to peak and decline in the next decade or two, similarly coal and gas are forecast to peak and decline before 2050.

Figure 2 shows the range of “realistic” future fossil fuel scenarios compared with the IPCC’s low, medium and high emissions scenarios. It looks increasingly as though “business” (as we know it) may not be permitted to continue “as usual”. There simply may not be sufficient fuel to continue driving our engines of growth.

In many ways this ought to be good news for our future climate. For those of us concerned with making long-term projections of likely changes to water resources (e.g. floods or droughts), it should indeed come as very good news to identify any way to tie down a more certain greenhouse gas emissions future, especially if that involves lower emissions.

As well as reducing the severity of the possible human-induced changes to droughts and floods, overall uncertainty around our predictions is also greatly reduced. There may still be significant changes at global and/or regional scales, but it appears that the most extreme projected impacts (produced under business as usual assumptions) can probably be revised downward.

Ironically, the mainstream climate science community remains highly sceptical of the notion that fossil fuel shortages will limit future greenhouse gas emissions and climate change. A frequently cited reason for their scepticism is the potential for conversion to more emissions-intensive and “unconventional” fossil fuels – tar sands, shale gas, coal-to-liquids and methane hydrates – in response to declining conventional production even though ongoing research is finding that unconventional fuels are unlikely to plug the gap left by declining conventional production.

Other reasons for scepticism derive from the huge uncertainties in the global carbon cycle, especially the potential for positive feedbacks (e.g. melting permafrost contributing extra methane, and thus exacerbating global warming) that may lead to a significant global temperature rise, even under low emissions. Incredibly, fossil fuel resource constraints are even dismissed by some as a “climate change denialist” tactic!

Without discounting the importance of improving our understanding of global climate change, at this point it is of at least equal concern that the world begins preparing for life on the downhill side of Hubbert’s fossil fuel peak: what would the social and economic impacts of declining energy production look like, especially with a world population that is still growing?

An appropriate metaphor is the Titanic running out of fuel before hitting the iceberg – the worst part of the crisis (in this case, climate change) may be averted, but being stranded with no fuel in the Arctic Ocean presents its own set of challenges.

James Ward is a Lecturer in Water and Environmental Engineering at the University of South Australia. Simon Beecham is Professor of Sustainable Water Resources and Head of the School of Natural and Built Environments at the University of South Australia.