The first speaker of the morning session was Terry Marsh who is leader of the National Hydrology Monitoring Program. He stressed the importance of the many competing factors that affect riverflow, how some floods are caused by excess groundwater while others are caused by overland flow. In 2000 there were floods in Yorkshire, followed by severe droughts in 2003; then came the Boscastle flood in 2004. July 2007 was very wet and in August 2008 floods took place in Northern Ireland and then the flood at Cockermouth in November 2009. Looking back there were very bad floods on the Severn at Shrewsbury in 1946 and in the following year the worst floods in England as a whole occurred. The period 1949-1955 witnessed many extreme events: drought in 1949, bad flooding in 1951 and the Lynmouth disaster of 1952 followed by equally bad coastal flooding in 1953 and river floods in 1954 ending with the highest non official rainfall record of 355mm at Martinstown in 1955. While many individual events can be easily quoted, trends are harder to show. The period 1998-2002 has been the wettest five-year period on record, while the Thames at Teddington has no clear trend in runoff even though river levels have dropped because of channel improvements. Low flows on some rivers have fallen because of greater abstraction for drinking water and industrial use. The river Roding at Redbridge has lower flows because of the diversion of sewage effluent. Marsh suggested that the 1975-76 has not been fully documented but said that the lowest seven-day rainfall ever was recorded by 50% of the monitoring network. From 1951-2001 there was no trend in rainfall, although there may be wetter winters and drier summers. Winter rainfall is more important for water supply, especially if there is a run of dry winters that led to the supply restrictions during 2004-2006.

Ian Strangeways gave a thought provoking talk called ‘Are instrument networks capable of identifying climate change?’ Much of the existing record has been produced by the manual recording of temperature. However, the glass bulb of the mercury thermometer shrinks with time and may, as a result, give an increase in recorded temperature 0.1 degree Celsius per decade. Therefore thermometers must be recalibrated every few years. During bright sunshine and low wind speeds the temperature in the Stevenson screen can be 1 degree Celsius higher than the outside air. Strangeways then showed slides of screens covered with snow, overshadowed by trees that had grown up over time and the effects of a small change in readings as a result of moving the instruments at a site in Wallingford. Ocean temperatures as measured on ships using a bucket to get a sample of water will produce results that also depend on the size and design of the bucket. Buoys are now used more often and these give more stable results. It is important to have the metadata as well as the data itself when using the data. Homogeneity has to be achieved, but just because one reading may appear quite different to those around it does not make the reading invalid. On climate modellers grid squares it is hard to get sensible average temperatures where there is both land and sea in the same square. In 1998 the warm temperatures were cause by an El Ninio event. Since then global temperatures have been falling and the so-called hockey stick graph of temperature change so often quoted is simply not correct. The Little Ice Age was a global event so temperatures have recovered since the end of this event in about 1890. Ian concluded by saying that natural processes in our climate are more important than changes in carbon dioxide.

‘Climate variability or climate change? Implications for design floods’ was the title of the talk given by David Macdonald who is a practising hydrologist for Black and Veatch. Macdonald stressed the importance of the many different causes of floods, for example building on floodplains such as at Banbury in Oxfordshire. On the South Downs houses had been built and nearby grassland had been ploughed up and winter wheat grown. During wintertime the ground was bare, there was more runoff, deep gullies produced and the houses flooded. Beverley Brook in London had become urbanised and local flooding had been caused by surcharging of drains. So far as rainfall was concerned there had been very little change at Oxford since 1770. In the past 100 years there were ups and downs in the record and then Macdonald quoted from the recent Defra report on rainfall and dam safety that climate change was not well understood. On many sites he had investigated there were no real trends in floods. The current advice from Defra is to consider the possible addition of 20% flood flows by 2080 when designing schemes. This is not a requirement but could be used in certain circumstances. At Beddington Park the river Wandle floods. The two-year flood is estimated as 4m3/sec and with a growth factor of 2.5-3.0 gives the 100 year flood as 10-12m3/sec. If an allowance for climate change is made this gives a value of 14m3/sec. Upstream of the site under study on the Wandle a bridge had been built to carry 14m3/sec so as a matter of engineering judgment a design value of 15m3/sec was chosen. On the Cherwell at Banbury there was serious flooding in April 1998 but the flood record did not include events in 1932 and March 1947. At Derby the measured flood record on the Derwent begins in 1935. But a search through newspaper records showed that serious floods took place in 1921, 1922, 1931, and 1932. When estimates of these are added to the measured record the estimates of the 100 year flood increase from 360m3/sec to 450m3/sec. If the gauged record based estimate of Q100 had been increased by 20% for climate change it would still be lower than the historic + gauged estimate.

During the afternoon Colin Clark spoke on the basic physics of climate change, climate forcing and recent changes in temperature. He stressed that the climate sensitivity to a forcing of 1 Watt per square meter was 0.1 deg. C as compared with a GCM derived value of 0.6˚C. The effect of carbon dioxide on earth temperatures was estimated from the data gathered on the Viking and Pioneer space missions to Mars and Venus respectively which have atmospheres mainly composed of Co2. When these and other data are plotted against the partial pressure of Co2 it is found that a doubling of Co2 levels from 300-600 ppmv on earth will cause the temperature to rise by 0.4˚C, which is much lower than all GCM forecasts. Global temperatures have flattened off since 1997 and much of the previous warming was at night while extreme daytime temperatures in CET and Charldon Hill Research Station have declined. Overall the daily temperature range has fallen during the last 50 years. The cause of this may have been shown during late 2001. During the three-day period from 11-14 September 2001 all flights in USA were grounded following the Twin Towers disaster. This caused the daily temperature range to increase by over 1˚C as compared to the three days before and after this time period. In the US as a whole both the frequency of hail and violent tornadoes have decreased during the last century. The same is true for destructive hailstorms in England and Wales.

Geoff Jenkins of the UKMO then presented the latest forecasts based on an ensemble of many GCMS. The frequency distribution of the results, that is the results are classified into groups and then plotted as a bar graph suggest that by 2080 winter rain will increase and summer rainfall drop. There is much uncertainty in these forecasts. For example temperatures may rise by 5˚C at the end of 2100 with high emissions or by only 2˚C with lower emissions. There is natural variability with winter rainfall and local variations can take place. In the future there will be forecasts of future wind speeds and snowfall.

Roger Martin then spoke on the need to address the problem of population which is at the heart of any future policy to improve the quality of life in many countries that are becoming overcrowded. Today England is the most densely populated country in Europe. For many people population control is a taboo subject. Perhaps more important than climate is the crunch of water and oil supply, both of which are crucial for basic living standards of today. Martin questioned whether we should consider using less resources in the future. The quality of the soil has also been neglected in the past. He then showed the changes in World population in the world reminding the meeting that since the meeting started in the morning, world population had increased by 40000. There had been a massive increase in world population during the lifetime of the Queen from 4-9 billion. Can this carry on without causing massive problems even without climate change?

The main discussion period, which lasted almost 45 minutes, was chaired by Andrew Black, president of BHS. Several people asked about the validity of the ensemble forecasts of the UKMO. How could they be judged correct simply by agreement as compared with absolute accuracy? The measured record for the future was questioned in view of measurement problems raised in Ian Strangeways talk, while Roger Martin was asked how population could be controlled. Finally Paul Hardaker gave a thought-provoking summary of the issues as a whole and showed a video of interviews with people in Liverpool on their perception, understanding, and actions in relation to climate change. This was the first Saturday meeting partly organised by BHS and many people felt it was a good day of the week to hold such an event. The presentations will appear on the Royal Meteorological Society website.

Report by Colin Clark