I’m sometimes asked to describe what science is. Well, there are many definitions and philosophical positions which cover this question, but to me, as a working scientist, one stands out above all others as relevant to what I do. Science constrains uncertainty. Or, to put it in a slightly longer form, science is the method that allows humans to put realistic bounds on our understanding of how the world (the universe) works and the natural laws it obeys. Although there is almost no problem in science that can be explained fully, and few ideas can be proven absolutely, science is still among our most effective intellectual tools. From the technological sophistication of our modern society, to our appreciation of the hidden mechanisms of evolution or quantum mechanics, science tells us what is possible (and plausible), but not what is, or must be.
One scientific problem for which we can never have definitive proof is the cause of past extinctions. Such events can never be replayed or observed directly, and so cannot be tested or falsified; moreover, evidence from the past is inevitably sketchy and difficult to interpret. Yet, despite these inherent limitations, we can still assess how our available data stacks up against alternative ideas, and arrive at a probabilistic judgement on what is more or less likely. The extinction of the dinosaurs is the most famous example, but there are many others. In this week’s issue of Science, I have a co-authored paper “And then there were none?” with Bert Roberts on the extinction of Australia’s megafauna and the probable role of early modern people. There is a write-up of the story in The Australian, here.
In short, we argue that improved dating methods show that humans and megafauna only co-existed for a relatively short span of time after people invaded Australia, adding weight to the argument that hunting led to the extinction of many large-bodied species. In particular, new methods for direct dating of teeth and bones at a site long been claimed to provide evidence that humans and megafauna lived side-by-side for 20,000 years, has revealed that this site actually shows nothing of the sort — the bones and stone tools are not of the same age and were probably redeposited together due to erosion and floods. Although this latest finding doesn’t ‘prove’ that humans hunted megafauna to extinction, it does withdraw an important piece of supporting evidence for the alternative climate (drought) hypothesis. So, incrementally, science advances by narrowing uncertainty. Email me if you want a PDF of the article.
More broadly, many aspects of climate science can be looked at in a similar light — especially with respect to our efforts to understand the relative importance of past climate forcing effects, and our projections of future change. We cannot ever know what will happen in the future, for a whole variety of reasons (imperfect knowledge, limitations on our models and our ability to parameterise them, uncertainty about human decisions); likewise, we cannot ever be sure just how important greenhouse gases, ice albedo, dust, volcanoes and the sun were in perturbing past climates, nor how abruptly and markedly they did this. Yet we can still assess, based on multiple lines of evidence, what is more or less likely, and make decisions on energy pathways and other globally significant human activities on this basis, under a risk management framework.
In this context, Nature has just produced a great news feature called “The real holes in climate science” (available free, here). To quote:
Like any other field, research on climate change has some fundamental gaps, although not the ones typically claimed by sceptics. Quirin Schiermeier takes a hard look at some of the biggest problem areas.
The four areas of greatest uncertainty they discuss are regional climate predictions and downscaling (an area of particular interest in my research on ecological impacts that are relevant at the population and individual scale), precipitation (where will it get wetter, drier, snowier or more barren and how quickly will these shifts occur?), aerosols (how much warming sulphates and dust suppressing, and what is the heating effect of black carbon?) and palaeoclimatic reconstruction (the ‘tree ring controversy’ and related uncertainties). I would also add feedbacks, abrupt change and slow/fast climate sensitivity to that list…
Also, be sure to read the associated editorial, Climate of suspicion:
With climate-change sceptics waiting to pounce on any scientific uncertainties, researchers need a sophisticated strategy for communication.
This provides sound advice for all scientists wishing to engage on the topic of climate change in the public arena. Which definitely includes me, now that I’ve agreed to ‘debate’ Christopher Monckton and Ian Plimer, with the help of Graham Readfearn (of News.com’s Green Blog), at The Brisbane Institute on 29 January. As many of my readers may suspect, I plan to focus on the underlying motivations for this argument over ‘holes in the science’, rather than getting entangled in the scientific details or background of the antagonists, and will propose some ideas for cutting the Gordian knot (which, in my humble opinion, ultimately boils down to a question of energy economics). This is only fitting, given my evolution of thought on the matter of climate and energy over the last year.