UK 2007 flooding: Caused by Climate Change?

The book Climate Change: Biological & Human Aspects (Cambridge U. Press) covers many aspects of 'global warming' including flooding. Though the past has seen its share of flooding that suggest that the present flooding may not be climate related, computer simulations suggest that the future may well see such flooding. In the UK this can occur in future summers as well as the winter, though the risks of such floods in the winter may well be more likely. This suggests that Britain's major flood incidents of 2000, 2003, 2004, 2005 and 2007 may well be connected with climate change.

The below is a single extract (p364-8) from Climate Change: Biological & Human Aspects: just one of the book's number of references to rain and flooding.

Note: This was written long before, and the book published at exactly the same time as, the British floods of July 2007!

Floods have always taken place throughout human history. Noah's flood and similar stories that might date from pre-history appear to relate to the flooding of the Black Sea, which itself arose due to sea-level rise following the last glacial maximum and the onset of the current Holocene interglacial. Furthermore, there has always been a tension between the human desire to exploit the fertility of the floodplains of the world's greatest rivers and the fact that floodplains (by definition) flood.

      In the late twentieth century the floods of the upper Mississippi in the summer of 1993, the winter floods in December 1993 and January 1995 of the Rhine and Bangladesh's chronic flood problem all served to illustrate the material costs and costs to human well-being. The 1993 Mississippi flood had an estimated cost of US$15 billion but only a small death toll, of 48. Following that flood there was concern expressed over the way the river watershed was managed and the flood-control measures that were in place, but irrespectively that flood was due to an extreme weather event following a season of exceptional snow in Iowa and the 1993 precipitation record was the greatest since records began in 1895. Meanwhile, both the 1993 and 1995 floods of the Rhine were also due to extreme weather events and parts of the region in 1993 had three times the average rainfall in December. The cost of the flood in Germany alone was around US$850 million in the money of the day. However, in 1993 lessons were learned so when in 1995 there were floods again, the damage costs were halved. Even so, there was disruption and in the Netherlands some 200,000 people were evacuated from their homes for safety. What is also clear is that in north-west Europe, as illustrated by England and Wales, instrumental measurements have shown that between 1765 and 1995 there has been a slight decrease in the intensity of winter drought years accompanied by a similar increase in the intensity of years with high winter rainfall, with average rainfall also increasing by about 17% (Osborn et al., 2000). This trend is what is expected in that region with climatic warming. However, in the case of Bangladesh there is no historic meteorological evidence of increased rainfall (although future projections tell a different story) but deforestation in the Himalayas and isolated tropical storms appear to have caused much of the problem. Bring climate-change factors into the picture and in the future flood problems will undeniably be aggravated. If the recent climate events in Bangladesh are part of a longer-term future trend, then arguably this future may have already begun. Indeed, factor in the dimension of sea-level rise and matters will get worse still. As stated, the region is low-lying and has previously seen damage from sea flooding. In the case of the 1970 tropical storm some 300,000 perished when a tidal wave swept across the Ganges and Meghina deltas (Burroughs, 1997). The implications are clear.

      With regards to UK climate trends, the autumn 2000 was then the wettest since records began in 1766. The heaviest rainfall that season was across England and Wales, with a total of 489 mm falling between September and November; the most extreme rainfall was in October, which resulted in extensive flooding. UK insurance claims arising from that season's floods came to around £1 billion. Many of the same areas of southern UK flooded again in early 2003. In the summer of 2004 a flash flood in Boscastle, Cornwall, was caused by around 20 cm of rain falling in just 4 hours. Two rivers burst their banks and a 3-m wall of water passed through the town's high street. More than 150 people had to be airlifted to safety and despite about 50-60 cars being washed away nobody died, but around £50 million of damage was caused. Another extreme event took place in January 2005, when 100 mm of rain fell around Carlisle in Cumbria. So intense was the rain that surface-water drainage could not cope and flooding began, quickly before river monitoring (which the UK Environmental Agency then relied on) generated a flood alert. The flood caused local power cuts and interrupted both the terrestrial and mobile phone systems; consequently, some people became trapped in their homes.

      The same year, 2005, saw a similar event on the North Yorkshire moors. After a weekend heatwave in which temperatures reached 33.1 °C, the night of Sunday 19 June saw rainbursts over the north east of England that triggered regional flooding. North Yorkshire as a whole had the best part of a month's worth of rain in 3 hours. One fast-flowing flood that went through the villages of Thirlby, Helmsley and Hawnby was caused by 2.7 cm of rain that fell in just 15 min. The flood was in places 2 m deep and it carried cars along with it. Meanwhile, in river courses torrents washed bridges away. The highest rainfall in the UK on that day took place in Hawarden, Flintshire, in Wales, with 4.3 cm of rain. Two days after the event some 2,500 homes were still without electricity. Indicative of the 'clumping' of rainfall patterns Friday 24 June 2005 saw some 2.8 cm fall on RAF Lyneham in Wiltshire. The usual rainfall for June in that area is 3-6 cm.

      On continental Europe the most disastrous floods in recent times took place in August 2002 along the Elbe and Danube rivers. In the Czech Republic some 200,000 had to leave their homes whereas in Germany 3,600 patients had to be moved from hospitals threatened by waters. Rough cost estimates for the Elbe flood are US$3 billion in the Czech Republic and over $9 billion in Germany. Flood damage (which is not the same as flooding) of this magnitude had never happened in Europe before. There have been 10 extreme floods in Dresden, Germany, since the thirteenth century, with water levels peaking between 8.2 and 8.8 m, but the 2002 flood peaked at 9.4 m. However, the previous record-holder in 1845 had a higher flow rate than the flood in 2002. The reason why the 2002 flood was higher despite a lower flow rate is due to greater river containment as a result of floodplain development. This again exemplifies how non-climate factors can compound extreme weather events. Having said this, the rain that caused the 2002 flood exceeded most previously measured rainfall amounts and intensities, and in the core area of precipitation rainfall levels exceeded the 24-h record observed in Germany since records began and were close to the PMP (Becker and Grunewald, 2003).

      The future Western European picture that is just beginning to emerge is therefore a complex mix of changes in seasonal precipitation (down in summer and up in winter), extreme events (similar frequency but slightly greater intensity) and watershed development (increased floodplain development). This is exemplified by historical analysis of rivers about which there is substantial historical documentation. Among these are some European rivers. In 2003 German researchers led by Manfred Mudelsee looked at records for the middle Elbe and middle Olde rivers, primarily concentrating on events in the past 500 years, for which the records are more reliable. Discerning long-term trends in monthly run off was difficult. They found that there was no substantial trend emerging above the natural variability in events. So, the historical analysis did not provide evidence for a climate-change-related cause for increased European flooding.

      How can this conclusion sit alongside model predictions? Leaving aside model reliability (which is continually improving), historical records look back in time while models look forward. The past has seen just a degree's worth of global warming since the Industrial Revolution, whereas the coming decades are expected to see a Business-as-Usual rise above 1990 temperatures of 2.0-4.5 °C (best-estimate all-model scenarios) with overall low and high estimates of 1.4 and 5.8 °C respectively. In short, retrospective reviews and prospective forecasts are not the same. So, the former may show hardly any change while the latter might suggest a definite future change. What it means is that the early twenty-first century sees us on the cusp. Even so, the actual changes in the future could be quite small, with just marginal excesses over previous instances. Here the problem is that such floods are catastrophic events. Small changes can have big consequences. Matters are fine until river banks are breached (such as with a peak rise of just a few millimetres), then there is a quick transition from no damage being incurred to considerable damage.

      Finally, increased major floods could well happen in the summer despite European summers becoming drier. Seemingly paradoxically, computer models predict an increase in intense summer rainfall with global warming. (Note: This underlining is added to this web page only for emphasis and does not appear in the book.) Instead of the less rain being spread across summer months, there will a tendency for this precipitation to clump into extreme weather events (Christensen and Christensen, 2002). This is because in a warmer world there is more water vapour in the air. Of course, some air is warmer than other air. So when water-laden air (more than historically anticipated at a given latitude) meets cool air the resulting rainfall will be greater. Additionally, matters can even be worse. Consider two bodies of water-laden, warm air colliding. There is nowhere for the air to go but up. At a higher altitude it cools, again releasing water but even more so than in the previous scenario. The consequence of this for flooding is significant. If rainfall is spaced out over weeks then the water has the opportunity to sink into the ground and then into the subsurface geology. If a few weeks' or even a month's worth of rain falls at once, due to an atmospheric collision, then there is no time for the resulting precipitation to drain away and so it remains on the surface. Problems are compounded if the surface is corrugated with hills and valleys, as this volume of rain will run off the hills into the valleys as fast-flowing floods. Of course, there are parts of the Earth's surface that historically are warmer than temperate latitudes and which also have high rainfalls. However, these regions have seen waterways being gouged out and if they are inhabited they have the appropriate drainage systems in place. With climate change, a place that previously only had moderate, well-spaced rain with time becomes exposed to more extreme rainfall regimens; there will not be appropriate drainage systems in place and so flooding results, with property damage and possible loss of life.

Climate Change: Biological & Human Aspects is available from Cambridge University Press, Cambridge UK and its offices overseas including in New York (US), Melbourne (Australia), Madrid (Spain), Cape Town (South Africa) and elsewhere. ISBN 978-0-521-87399-4 (hardback) and ISBN 978-0-521-6919-7 (paperback). See also details at CUP.   It is illustrated with around 70 diagrams and a score or so of tables. It is fully referenced and has a number of explanatory appendices. Aimed at those with differing expertise, it is an introductory text but, at around 500 pages, comprehensively covers a wide range of climate-related issues.

Climate Change: Biological & Human Aspects 2nd edition (2013).