Has the intergovernmental panel on climate change got its numbers wrong? That's what a recent paper in Nature seems to be saying, to the delight of climate change sceptics. Whereas the IPCC forecasts a rise in global mean temperature of around 0.2-0.3 oC per decade, researchers in Germany using computer modelling found that temperatures are likely to remain flat until around 2015, as indeed they have done since about 1998.
Sceptics will argue this shows scientists don't have a clue about climate and that the dire forecasts from models count for nothing. But this would be like saying that because we took a wrong turn on the road from London to Edinburgh, we have no idea where Edinburgh is.
There is actually nothing in the new result that conflicts with the IPCC's position, which has always acknowledged that the poorly understood natural variability of the climate system will superimpose its imprint on the global warming trend. The new findings are an attempt to forecast short-term, decade-scale temperature changes, rather than the longer-term changes usually considered by climate modellers. Over a decade or two, temperatures are much more susceptible to natural variations (which boosted warming in the late 1990s). The current cooling influence is the result of a weakening of heat-bearing ocean currents such as the Gulf Stream. This may persist for about a decade, but then the warming will resume, and by 2030 it should reconnect with IPCC predictions.
So there's no reason to throw out the climate models. And no climate scientist seems particularly surprised by the findings, which simply flesh out the short-term picture. Sceptics see this as dissembling and back-pedalling. But it's a curious logic that uses climate modelling to discredit climate models.
Yet how can the models can be vali- dated when they appear to predict one thing one moment, and the opposite the next? We must simply accept that natural variability compromises short-term predictions—a fact of life that demands great care in framing the right questions and drawing conclusions. We should remain wary of claims that a few hot summers, or a few more hurricanes, prove that catastrophe is imminent, just as we should of suggestions that a few relatively cool years rubbish the IPCC's forecasts
Biofuels and the food crisis
We must be wary too of attempts to invoke global warming as an explanation for every environmental trend. Droughts and storms amplified by climate change may be playing a small part in the global food crisis, but a far bigger source of the problem is attempts to mitigate global warming with biofuels. In 2006, a fifth of US maize was grown to make ethanol, not food. With the US providing 70 per cent of global maize exports, grain prices worldwide were sure to feel the effect.
The rush towards an ill-considered biofuels market is a depressing reminder that the vicissitudes of climate science are nothing compared to its bungled economics. The passion for biofuels in the Bush administration is driven more by a wish for energy independence than by climate concerns, while farmers embrace them largely for reasons of profit. But science has played a part in condoning this shaky vision. It's a little late now for some scientists to be explaining that the benefits will only be felt with next-generation biofuels, which will make much more efficient use of plant matter.
Biofuels aren't the only reason for soaring food prices. Population rise is playing its ever baleful part, as is the increase in oil prices, which makes food costlier to produce and transport. And growing crops for energy introduces a new economic coupling between oil and food: escalating oil prices make it advantageous for farmers to switch to energy crops. The consequences of this new link between two vast sectors of the economy have yet to be carefully evaluated.
Predicting the next earthquake
Back in the 1970s, when Chinese geoscientists claimed that they could predict earthquakes, they were quickly contradicted by the 1976 Tangshan quake that killed several hundred thousand people. The death toll of the recent, also unexpected, quake in Sichuan province may ultimately approach a tenth of that.
It may, in fact, be impossible to predict earthquakes by monitoring geological faults, because the size and timing of slippage is inherently unpredictable from information available at the source. But researchers based at Edinburgh now think that information should be taken from a far wider area around the fault zone, in the pattern and evolution of stress in surrounding rock. They propose using artificially induced seismic waves to map out these stresses, and claim this could enable the time, size and maybe location of earthquakes to be forecast days or even months in advance.
The team says a stress-monitoring site could have forecast the Sichuan earthquake from Beijing, 1,000 kilometres away. A monitoring station's likely price tag of several million dollars dwindles before the cost of damage inflicted by earthquakes this severe. Despite the dismal record of earthquake prediction, this one looks worth a shot.
