That Met Office media controversy in context
The Met Office has today published an explanation of how it forecasts decadal temperature trends following news that it has revised downwards its decadal prediction of likely global temperature rise up to 2017. These measurements focus on short-term fluctuations in the climate system, not long-term climate trends. So why has a scientifically significant – though, in the scheme of things, relatively minor – adjustment to an experimental forecast ended up being heralded in the Daily Mail as evidence the Met Office has admitted climate change isn’t happening?
In the Daily Mail, skeptic blogger James Delingpole claims that the Met Office “quietly readjusted its temperature projections” on Christmas Eve, conceding that “‘global warming’ isn’t happening” after it produced a new decadal forecast. The Met Office has been responding to claims that warming has stopped all week, explaining that the forecast in question only relates to short term fluctuations in the climate – and certainly doesn’t mean warming has stopped. We take a look at what decadal forecasts are, and what they really mean.
Forecasting fluctuations
In an article today, Met Office chief scientist Professor Julia Slingo points out that the organisation has published a decadal forecast – which predicts climate trends for the next five to 10 years – every year since 2007. The forecasts are created for government departments, but it also publishes them online.
The aim of decadal forecasts is to predict fluctuations in the climate system by taking the current state of the climate into account and allowing for how changes in ocean systems can influence the climate.
So why has the latest forecast been revised? Dr Doug Smith, who leads the Met Office’s research and development into decadal climate prediction, explains:
“We have been issuing decadal forecasts for a few years, but this year we upgraded to the very latest version of our climate model, which benefits from many years of improvements in climate modelling. We thoroughly tested this upgrade and demonstrated that the new model provides more skill [i.e. it better matches weather and climate observations] than our older model.
The new model indicates that the average temperature is likely to rise by 0.43 degrees Celsius above the average for the period 1971 to 2000 by 2017, rather than 0.54 degrees, as an earlier forecast had suggested. For more on the new forecast, read our blog here.
Smith explains that fluctuations in global temperature in the next few years are expected due to natural variability, but that they have no sustained impact on long term warming. What’s more, Smith explains that the slightly lower prediction in the new forecast is still within the bounds of the previous prediction. He says:
“The latest experimental decadal prediction provided by the Met Office issued in December 2012 suggests that global temperatures over the next five years are likely to be a little lower than prediction from the previous prediction issued in December 2011, but still near record levels. While they are different, the range of temperatures in the 2012 prediction overlap with the range from the previous prediction provided in December 2011.”
Smith also points out that both this and previous decadal predictions from the Met Office are consistent in that they both predict near record global temperatures in the next few years. He says:
“2000 to 2009 was warmest decade in the instrumental record and our decadal predictions indicate that it is likely that temperatures up to 2017 will continue to be near this record level.”
Why do decadal forecasting?
But if these forecasts don’t reflect the long term trends, why does the Met Office produce them? First, according to Slingo’s article, they are a useful test for new climate models, showing how well they simulate natural variability and how successfully they represent long term human caused warming.
Such forecasts could also help policymakers to adapt to a warming world, the Met Office argues. It says there is emerging evidence that the climate is becoming more volatile and that extreme events, like droughts in Texas, extreme temperatures in Australia and floods in the UK, are becoming more frequent.
Because of this, there’s a greater need for advice on what climate volatility will be like over the next decade in different regions, so that societies can plan, says Slingo. She adds:
“Decadal forecasting has, potentially, an important role to play here in assessing the probability of [extreme weather] in the next few years, because they combine both natural variability and climate change.”
Developing science
So how do such forecasts relate to the bigger picture of understanding of climate change? Overall, scientific understanding of likely climate change is based on a great deal of supporting scientific work. In the case of decadal forecasts, however, the Met Office is careful to emphasise that decadal forecasts, are still in the early stages. For example, the forecasts use data from observations below the ocean surface – which were until recently pretty sparse.
Dr Ed Hawkins, a senior research scientist at Reading University, told Carbon Brief that decadal models are still in their infancy. He says:
“They are all are very much experimental. There are considerable technical challenges in producing such forecasts and there is no clear best strategy yet.”
The new experiments aren’t just confined to the UK. Hawkins explains that groups all over the world are working on similar forecasts. Hawkins was one of the authors on a paper, out last December in the journal Climate Dynamics, which summarises this work. But although this is still a very new area of science, Hawkins says “[a]ll the groups agree on predicting continued warming for the next decade.”
Bad timing
The Met Office released its revised forecast on Christmas Eve last year – leading some commenters to suggest it was trying to cover up the new data. It was unfortunate timing – given that Christmas Eve does look like quite a good time to bury some bad news.
But the timing doesn’t change the fact that the implications some newspapers are drawing from the new data aren’t well founded. In response, the Met Office says that while it publishes its decadal forecasts in “the spirit of openness and transparency”, it doesn’t publicise them because they’re “still at the cutting edge of research”.
Hawkins says the amount of media attention the revision received is “disproportionate”, seeing as “the revision is well within the uncertainty presented in previous forecasts.” He adds:
“Previous forecasts received very little media attention, if at all, because the Met Office have been very clear that they were, and still are, experimental. [But] it is right that the forecasts are openly available to allow proper evaluation afterwards.”