Indicators of Climate Change in the UK

Editors: M.G.R. Cannell, J.P. Palutikof and T. H. Sparks

Produced by the Climatic Research Unit and the Centre for Ecology and Hydrology

Publication year: 1999

This report sets out the results of a search for appropriate indicators of climate change in the UK, following the criteria outlined in the Introduction. The first 4 indicators are measures of our climate. The other 30 indicators (5-34) are various environmental and economic factors, which are affected by climate. The latter indicators cover many aspects of life in Britain.

The accompanying tables summarize, for each of the environmental and economic indicators, whether year-to year variation in the indicator has been related to a climate variable, whether there has been a trend over time in recent years or decades, and how we may expect the indictor to change in future in response to climate change.

It is impossible to generalize from the results. However, we can identify characteristics possessed by different sets of indicators. This is done below, dealing first with the 4 climate indicators and then with the environmental and economic indicators.

As a starting point, we included four indicators based on measures of our climate. Air temperature over central England since 1900 was selected because, perhaps simplistically, we would expect it to show an upward trend in response to global warming. In fact, we find that there is an upward trend in the annual series until around 1950 and again from the mid-1960s to the present but that, somewhat surprisingly, the 1950s and early 1960s were years of relatively lower temperatures.

The other three climate series are included mainly in order to capture possible changes in the atmospheric circulation over the UK, which might occur in response to global warming. Two involve rainfall, looking first at the geographical distribution, by comparing year-to-year changes in the amount of rainfall in Scotland and in SE England, and second at the seasonal distribution. Rainfall amounts evenly distributed throughout the year indicate a maritime climate, whereas a strong contrast between a wet winter and a dry summer suggests a more continental influence. The final climate indicator looks for changes in the atmospheric circulation, by examining an index of the winter-time strength of westerly wind flow across the Atlantic. When the flow is strong, the British Isles experience warmer, more maritime conditions. When the flow is weak, continental (cold, dry) conditions prevail. This index gradually strengthened from around 1970 until the winter of 1994/5, which had the highest index value on record. This was followed by a very steep decline – the 1995/6 winter had the lowest value on record.

Climate models give us clues as to the likely behaviour of some of these indicators in response to global warming:

• Central England temperatures should increase
• Geographical contrasts in rainfall should become more marked, with Scotland becoming wetter and SE England becoming drier

• Rainfall should become more seasonal, i.e., wetter winters and drier summers

However, the most probable change in the behaviour of the westerly wind flow in response to global warming is not known.

Thus, for three of the climate indicators we have some idea of what to expect as a result of global warming. Are there signs that the expected changes are already occurring? Central England temperatures have gradually increased through this century, by about 0.5^oC. However, there is no evidence of a persistent trend in the indicator of geographical contrasts in rainfall, nor of increasing rainfall seasonality.

For 21 of the indictors, it was possible to find a climate variable that was closely related with year-to-year fluctuations. The closeness of the relationship is expressed in the Tables by the correlation coefficient r, where r² indicates the percentage of the variation in the indictor accounted for by the climate variable.

Most relationships are intuitively obvious. Thus, hot, dry summers are associated with high rural ozone levels, many insurance claims for subsidence, more domestic tourism, more outdoor fires, a greater incidence of Lyme disease, greater use of irrigation water, low yields of unirrigated potatoes, low summer hay yields, poor health of beech trees, greater insect abundance and late migration of salmon upstream. Mild winters and springs are associated with poor skiing, lower human mortality in winter, early emergence of leaves, flowers, insects and breeding of birds and larger populations of some insects and birds.

There are, however, two reasons for being cautious in assuming that the fluctuations in the climate variables are wholly responsible for the fluctuations in the indicators. First, in most cases, the correlated climate variable is the ‘best’ one we found. Many climate variables were examined and there is always the possibility that the ‘best’ correlation does not actually reflect a causal relationship.

Second, there is the risk that we are observing fluctuations in indicators caused by some other forcing mechanism. For example, heat island effects from urban growth can change the local temperature regime, with physiological impacts on vegetation, e.g., earlier leafing dates. Removal of habitats due to human activity may be more important for determining the size of bird and insect populations than climatic factors. For the economic indicators, changes in fashion and/or disposable income may control the behaviour of an indicator more strongly than climate.

The Tables show which indicators display an increasing or decreasing trend with time. Indicators which have been increasing in recent years are: sea level, frequency of Thames barrier closures, insurance claims for subsidence, number of outdoor fires, incidence of Lyme disease, use of irrigation water, proportion of potato crop that is irrigated, areas of grapes and maize, dates of leafing of trees, emergence of some insects, arrival of swallows and egg-laying by birds. Other indicators show no clear trend with time, although some suggest greater variability in recent years, such as river flows, groundwater levels in chalk and perhaps weather-related insurance claims.

Some of these indicators were chosen partly because they showed trends which seemed to be weather or climate related. Many of the insect and bird indicators are of this type. Generally, the climate variable we have selected to compare them with also shows an upward trend with time. The behaviour of these indicators in future years bears careful watching. Will they maintain current trends, thus demonstrating, first, the soundness of their selection and, second, that the impacts of climate change are a reality in the UK? Or will they revert to earlier conditions, suggesting that any trends were simply random behaviour or natural variability?

Other indicators contain a trend which we suspect is unrelated to any weather or climate effect, but is rather related to changes in human activity. The number of outdoor fires is such a case – we do not believe that there is a climate-related upward trend with time, rather that the link is with generally higher rates of vandalism. Other examples include the supply of domestic gas and domestic holiday tourism. For these indicators, the trend is irrelevant and, where it is particularly strong as in the case of domestic tourism, has been removed from the series prior to presentation here.

It may be noted that 2(?) of the environmental and economic indicators have been processed to reflect the seasonality of events rather than absolute changes in quantities. This is because seasonality was often found to be more closely related to climate. Thus, the proportion of gas used in winter was chosen because it was more closely related to climate than annual gas consumption (and also this has the effect of removing the long-term trend which is considered to be unrelated to climate). Similar reasoning led to the choice of indicators which measured the proportion of human deaths in winter.

A small set of indicators has been deliberately selected to represent a response due to people’s perception of a shift in the climate. These are the area of vineyards under cultivation, the area of forage maize, and the proportion of the potato crop which is irrigated. All three are increasing although the current climate does not seem to justify it - growers are either anticipating or perceiving change which has not actually yet occurred, or they have other reasons for altering their choices, such as consumer demand.

Some indicators which superficially appear to demonstrate a physical link between climate and the environment, may in fact represent a change in human behaviour. Notable amongst such indicators are the number of outdoor fires and the incidence of Lyme disease. Both rise in hot, dry summers, mainly because people are out and about more in such conditions. However, given the capacity of humans to modify their behaviour, these relationships may break down in the future if, for example, the number of outdoor fires is deemed unacceptably high, or if Lyme disease becomes perceived as a real risk.

The selection of indicators took into account their utility over time – they should not only be a good indicator of sensitivities to climate change and variability for the purposes of this report, but also in the longer term, so that continuous monitoring can take place.

This requires, first, that the data should continue to be available in the same form. For example, we use as one of our indicators the arrival date of swallows, based on observations at four coastal locations. If the observations from one of these locations ceases, this would introduce a bias into the indicator that would probably render it useless for our purposes. Obviously continuity cannot be guaranteed in all cases, but by using well-established long-term records we hope to ensure that a high proportion of the indicators selected here will continue unchanged in the future.

Second, the indicators must continue to be responsive to climate change and variability. A very good example of a possible failure in this respect is the indicator of number of days with ice on Lake Windermere has stood at zero for the last eight years. Until the late 1980s it was responsive to temperature fluctuations, but this seems to be no longer the case – it has ‘bottomed out’ and its future utility as an indicator must be in question.

This report presents a present-day view of the response of the UK environment, economy and society to climate change and variability. The indicators were selected after inspection of their past response to climate. The true success of the project will be measured by the performance of the selected indicators in the future. Will they continue to respond to climate variability in the same way as they have in the past? Will they act as sensitive indicators of changes to our climate due to global warming?