UK Flooding Events and Fake Science
Blöschl et al (2017, ref 1) published a paper in Science that purports to show flooding in S England occurs every year but only ever in January and that flooding is disconnected in time from extreme rainfall events via water storage in soils. The changing pattern with time is ascribed to man-made climate change and implications for insurance were highlighted by the Financial Times.
A simple analysis of actual UK flood records from the UK Met Office shows that floods actually occurred throughout the year and that these are always directly associated with extreme heavy rainfall of either convective or cyclonic origin. According to Blöschl et al, floods in Southern England occur only in January (Fig 1), but the Met Office records show no flooding in Southern England in January at all.
On Monday last week I had a post titled European Floods and Fake Science and in it I said I would return to the subject with a more detailed look at the pattern of UK floods in recent decades as reported by the UK Met Office and to compare this real world with the imaginary world of climate science. There is no semblance of similarity between the two when it comes to UK floods and I dare say this may apply across the board.
The two data sets I use here are the S England flood data published by Blöschl et al in the once-esteemed journal Science [ref 1] and a qualitative diary of extreme UK weather events in the UK published on-line by the UK Met Office [ref 2]. The latter includes journalistic entries on uncommon hot, cold, dry, wet and windy conditions. I have extracted the entries on uncommon wet events that led to flooding and reproduced these in Appendix 1 and Figure 2.
UK Flooding According to Blöschl et al
Blöschl et al’s view of flooding in S England for their study period 1960 to 2010 is shown in Figure 1.
Figure 1 This shows detail of Blöschl et al Figure 2 panel D. Green = Flood, Purple = Precipitation and Light Blue = Soil Moisture. In my earlier post I explain how “Flood” actually equals mean maximum river flow which does not equate to flooding. I don’t know how many S England measuring stations are involved, but it should be clear that averaging many stations will obscure the exceptional event in a small area that leads to actual flooding. The exceptional events are further obscured by running a 10 year average through the data. If you want to identify flood events taking the average of an average is the last thing you should do.
The green line is flooding and the purple line precipitation. Their methodology is to use the date of peak mean precipitation for each year and the date of peak mean river flow each year (big averaging in itself) and to then run a 10-year moving average through the data. A big average of a big average is bound to obscure the extreme weather events that lead to the rare instances of flooding in the UK.
This methodology leads to a picture where the flooding is separated from precipitation by 2 to 4 months and the structure of the flooding dates with time do not match the structure of the changing precipitation with time. This led Blöschl et al to say:
In southern England (Fig. 2D) [shown here as Figure 1], subsurface water storage capacity tends to be much larger than in coastal Norway. The maximum rainfall, which occurs in autumn, therefore tends to get stored, and soil moisture and groundwater tables continuously increase until they reach a maximum in winter. Sustained winter rainfall on saturated soils then produces the largest floods in winter. As a result, the flood timing in southern England is more closely associated with the timing of maximum soil moisture than with the timing of extreme precipitation.
Actual UK Flooding Events
The best source I could find on recent UK flooding records was this Met Office diary of past weather events that includes unusual cold, warm, wet, dry and windy conditions [2]. The unusual wet events that led to flooding are summarised in the Appendix and in Figure 2. The Met Office series begins in 1998 and there is only a 13 year period of overlap with Blöschl et al whose series stops in 2010. While Blöschl et al show 13 consecutive years of floods in January alone, the Met Office Diary has floods in Jan, Feb, April, May, Jun, Jul, Aug, Sep, Oct and Nov (2*). In 1999, 2002, 2003 and 2006 no flooding is recorded by the Met Office. January is NOT particularly prone to flooding at all. The flooding of Jan 2005 (Figure 2) took place in Carlisle in NW England.
Figure 2 Summary of recent UK flooding transcribed from the UK Met Office diary (see the Appendix). Blöschl et al show floods every year, only ever in January, for the S of England. It needs to be noted that the Blöschl et al data are for the S of England while the Met Office data are for the whole of the UK; but this does not affect the total discordance between the Blöschl et al trend and reality – see text for details. Prior to 2010, the UK did not experience floods every year and the floods when they occurred happened throughout the year. Post-2010, there is indeed a significant change in the flooding pattern, missed by Blöschl et al because they mysteriously chose not to include post-2010 data in their analysis. Oval 1 = a concentration of floods in November and December. Ovals 2 and 3 = multiple events in a single year.
Of the 11 flooding events recorded by the Met Office 1998-2010, six occurred in Southern England as follows:
- July 2007 S Midlands
- Oct 2008 Devon
- Nov 2010 Cornwall
- May 2000 S England
- Feb 2001 Kent
- August 2004 SW England
There is no record of flooding in S England in January at all. Reading the Met Office narrative, flooding in England is always associated with extreme heavy rainfall. The exact opposite of what Blöschl et al conclude.
Discussion Part 1 – The Cause and Form of UK Flooding
Reading through the Met Office diary it is clear that there are two principal types of weather event that leads to flooding in the UK. The first is very localised but extreme high rainfall linked to convective thunderstorms that lead to “flash floods” with localised extreme consequences such as Boscastle in August 2004. The second is regional high rainfall linked to deep Atlantic cyclones that on occasions move very slowly across the UK resulting in widespread and sometimes longer term flooding such as happened on the Somerset Levels and Thames Valley in December 2013.
The flood in Boscastle in August 2004, caused by a very heavy and localised convective storm, was catastrophic but affected only a very small area.
Figure 3 The flood in Boscastle caused catastrophic damage to a very small area (image credit The Daily Mail). The flooding in Somerset and the Thames Valley during December 2013 linked to cyclonic storms affected a much larger area for a much longer period (see inset picture up top).
The winter storms of December 2013 covered a much wider area and affected several rivers, regions and considerably more homes and people. And yet, these storms and associated flooding of 2004 and 2013 are both recorded as single events in Figure 2, which is clearly an imperfect approach. It is way beyond the scope of this blog post to try and refine this methodology.
Discussion Part 2 – The Changing Pattern of UK Flood Events
Somewhat ironically, there does indeed appear to be a significant shift in the pattern of UK flooding missed by Blöschl et al because of their flawed methodology and truncated time series. Inspection of Figure 2 shows a clear change in the temporal distribution and frequency of UK flooding since around 2010. There appears to be a shift from single or no flooding events spread throughout the year in the interval 1998-2010 to one where flooding occurs somewhere every year. There are sometimes multiple events throughout the year (ovals 2 and 3) and there is a distinct bias towards the winter months November and December (oval 1).
The change in pattern is not gradual but quite abrupt and this left me wondering what the cause might be? One possibility is that this reflects a change in the recording standard at the Met Office, and that may be part of it, but I don’t think it is the main cause. I personally experienced the storms of December 2015 where NE Scotland was hit by one Atlantic Cyclone after another until eventually at the end of the month, with the ground saturated, rivers already full, Storm Frank caused several rivers to flood with widespread damage to property. The underlying cause appears to be a deeply meandering and “frozen” jet stream that has the ability to throw one cyclone after another at the same chunk of land. It is the repetition that does the damage.
At this juncture I wish to point out that the changed face of UK flooding began with this:
Figure 4 The UK from space on 7th January 2010. Image from the Met Office.
The winter of 2009 and 2010 was one of the most severe experienced by the UK for many decades and it too was caused by a change in the jet stream with circulation going into reverse sending freezing Siberian air across the UK, in a way I imagine happened during the Little Ice Age.
Sensitive to the fact that global warming was supposed to banish snow from UK shores for ever, the Met Office published a paper that sought to explain the aberrant behaviour by recently discovered and greater than expected spectral shifts in solar output than previously known [3]. The change in flooding pattern began with a very quiet Sun (zero sunspots in 2009) and the Sun is but one year away from this level of low activity again following one of the weakest and shortest solar cycles (cycle 24) in 100 years (Figure 5).
If the pattern of winter snowfall and precipitation in the UK has changed in recent years the more likely cause would appear to be low solar geomagnetic activity and not trivial changes in atmospheric CO2.
Figure 5 The geomagnetic activity of the Sun, as measured by sunspot number, was hyperactive during cycles 21, 22 and 23 (1976 – 2009) but has suddenly gone very quiet in the most recent cycle 24. The sudden change in the pattern of UK flooding (Figure 2) is coincident with this change to the Sun. Chart from SIDC.
Discussion Part 3 – More Focussed Rain but not More Total Rain
I want to conclude the discussion by pointing out that more focussed rainfall does not necessarily mean more rainfall (Figure 6). Recent, large, widespread flooding events are mainly down to cyclones riding a “frozen” jet stream hitting the same area repeatedly. This often means less rainfall elsewhere in the UK. Hence recent prolonged flooding events are not connected to a warmer atmosphere holding more water vapour which is the default argument made by the climate science fraternity.
Read more at euanmearns.com
Trackback from your site.