Most of the kerfuffle about climate change focuses on global average temperature change – despite this being a metric that no-one can personally experience by definition. It seems that 2014, if not the hottest-ever, will likely be a statistical tie for the hottest year on record worldwide. But what about other long-term data that might give a clue as to the direction of climate change, in ways that matter more to us all on a daily basis?
One area is weather records. It seems like hardly a month passes with some record-breaking weather event or other making the news. Now, if climate change were being felt, you would naturally expect more hot records than cold records, and you would expect a clustering of hot records in recent years in a way that would be unlikely through mere statistical chance. Also, given that a hotter atmosphere holds more water vapour, and climate change means an intensification of the hydrological cycle, you might expect more wet or dry records to be being set, depending on the location. But is this actually the case? Or are we simply bamboozled by news reports so that we remember recent records and forget about heatwaves a century ago?
Certainly we seem to have experienced record-breaking rainfall in recent years – as a resident of Oxford I have seen the River Thames burst its banks with increasing frequency it seems. But didn’t we also experience a record-breaking cold spell in December 2010? And I remember a few climate sceptic chuckles around the time of a heavy snowfall event a couple of winters ago. (“Where’s your global warming now?” etc) The country with the longest dataset on weather is actually the UK, so this is a good place to examine record-setting trends. In a fascinating paper just published in Weather magazine (the journal of the Royal Meteorological Society), Mike Kendon of the UK Met Office does just that.
There are some methodological issues here: many records have duplicates, such as two heavy rainfall events that are statistically very close, within a margin of error. There also needs to be a hierarchy of records – a nationwide record is more important than a regional one, a record-breaking year more significant than a month, and so on. Even given the caveats, some trends are quickly apparent, however. Here is the graph for temperature – the top chart shows actual records, the bottom chart includes records as scored in the hierarchy:
This result is striking, and accords pretty much with what we would expect with global warming (although the UK is of course not representative of the entire globe). There is a large jump in hot weather records after the 1990s, and a striking overall reduction in cold records despite variation between decades. (Note that the decade after 2010 only includes data up to 2014, so fewer records are set there by definition.) As Kendon writes:
Since 2000, there have been 10 times as many hot records (204) as cold records at all (20). No cold records at all were set during the period 2000–2009, or during 1930–1939, although there have been several since 2010, most prominently the exceptionally cold December of 2010. By contrast, a large number of hot records were set – for example in 2011 (April, spring, autumn, November and the year overall), and in 2006 (July, summer, September, autumn and the year overall). The period since 2000 accounts for two-thirds of all hot-record scores, but only 3% of the cold-record scores.
This tendency towards hot rather than cold records also applies to seasons as well as individual months:
When considering seasonal records only, the period since 2000 accounts for over 70% of hot-record scores but only 2% of cold-record scores. In a similar way, it accounts for 98% of annual hot-record scores (all but one record) but no annual cold-record scores.
The UK’s longest-running temperature data series, the Central England Temperature (CET) series, goes all the way back to 1659 – the longest in the world by far. So where are the heat records in this multi-century view?
Since 2000, the CET series has seen the warmest April (2011), July (2006), September (2006), October (2001), Spring (2011) Autumn (2006) and year (2006) – 7 out of a possible 17 records within the last 15 years of a 355-year series – with only December 2010 coming close to setting a cold record.
So once again, very much what you’d expect if modern UK temperatures were averaging at highs unprecedented in several centuries at least. But what about the precipitation data? Flooding trends are complicated by changes in built-up areas in river catchments and so on, but rainfall data should be worth a look surely? Here’s the same graphical treatment for wet/dry records:
Once again, it is pretty clear that the UK has been setting more rainfall records – so our recent news reports have not been biased by recent memory. Kendon writes:
Were we to assume no trends, we might expect a roughly steady number of records by decade – both wet and dry, but this is far from the case. Since 2000, there have been 10 times as many wet records (106) as dry records (11). Examples of recent periods where many wet records were set include 2012 (April, June, summer and the year overall) and winter 2013/2014, whereas the only prominent dry period based on this record scoring system was spring 2011. The period since 2000 accounts for 45% of all wet-record scores, with the highest record score for the 2010s – which is only half a complete decade. The 1910s also saw a clustering of wet records. By contrast, the dry-record scores remain fairly steady throughout until a pronounced reduction from 2000 onwards, the period since 2000 accounting for only 2% of dry record scores.
This is interesting because there is no increase in drought, and a big increase in rainfall records – exactly what anecdotally people think they have been seeing recently in the UK. This is also pretty much what climate models predict for the UK’s part of the mid-latitudes, where more rainfall, and more intense rainfall, is expected overall.
This also accords with the IPCC view, which is that there has been no increase in many forms of extreme weather – don’t look for a climate change signal in tornados or hurricanes, for example. But heatwaves do seem to have been getting stronger and more frequent, and there is also a signal for more intense precipitation. The UK’s dataset about record-setting gives further weight to this IPCC picture with the longest weather records in the world. Something to consider next time climate sceptics accuse scientists of overstating their case.