Taking Stock of a Warming World

Transcription

1 1 Taking Stock of a Warming World This chapter summarises observed changes to date in the climate system, and the observable effects that those climate changes have had on ecosystems, agriculture, forestry, human health, and so on. In this chapter, we do not discuss whether human activities lie behind the observed changes or whether they could be natural variations, or what the future may hold. These matters are discussed in chapters 2 4. Chapter 1 simply describes the changes in climate that have been observed, and the extent to which those changes have demonstrably affected natural and human systems. The structure and material in this chapter is based almost entirely on the Fourth Assessment Report (AR4) of the Intergovernmental Panel on Climate Change (IPCC), in particular the Working Group I and II Reports (WGI chapters 3 6 and WGII chapter 1) and the Synthesis Report (SYR Topic 1). 4 Since the AR4, significant new research has been published on recent changes in ice sheets and our understanding of what drives sea-level rise. I provide a few selected references to some of the most recent papers on this subject, but this does not aim to represent a comprehensive update. The literature on observed effects of climate change also continues to grow enormously, but given the breadth of publications, I have not attempted to provide any systematic updates. Selected illustrations of the effects of recent climate change are drawn from individual papers assessed in WGII chapter 1 and more recent publications (references to those specific papers are given in the text). Contents 1.1 Introduction definitions of climate and climate change Changes in the climate system Global and regional changes in temperature...13 Trends in global average temperature...14 Trends in different regions of the globe and atmosphere Trends in snow and ice cover...17 Glaciers, snow cover, and permafrost...17 Arctic and Antarctic sea ice...17 Polar ice sheets Global and regional sea-level changes Changes in precipitation Changes in other climatic extremes Comparing the 20th century climate with the climate in earlier times Twentieth century climate in the context of the past 2,000 years Ice ages and warm periods over the past several hundred thousand years AR4 comprises four volumes: the Working Group I, II, and III Reports (IPCC, 2007a (WGI), 2007b (WGII), 2007c (WGIII)) and the Synthesis Report (IPCC, 2007d (SYR)). Each report has a Summary for Policymakers (SPM) and each Working Group report has a Technical Summary (TS). 11

2 Climate Change 101 An Educational Resource 1.4 Impacts of observed changes in the climate system Effects of changes in snow and ice on polar and mountain regions Effects on water resources and freshwater ecosystems Effects of temperature changes on terrestrial and marine ecosystems Changes in the coastal zone and their relation to sea-level rise Emerging effects on the human environment...30 Effects on human health...31 Effects on agriculture and forestry Consistency and global coverage of observed effects of climate change Direct effects of increasing carbon dioxide concentrations Ocean acidification Carbon fertilisation...34 Boxes Box 1.1: Limitations of proxies for deriving pre-20th century climate information...23 Box 1.2: Effect of changes in sea-ice cover on Antarctic penguins...27 Box 1.3: Effect of algal bloom events on freshwater ecosystems...28 Box 1.4: Impact of higher temperatures on terrestrial species and ecosystems...29 Box 1.5: Coral bleaching role of increasing temperatures and other pressures...30 Box 1.6: The difficulty of identifying economic impacts of climate change Introduction definitions of climate and climate change The increasing reference to climate-related events in news bulletins seems to suggest storms and droughts are happening more frequently. Pictures of disintegrating ice shelves are seen as a signal that the polar ice caps are melting, caused by rising global temperatures. Yet some people remind us that the climate is and has always been changing. So is the world really getting warmer? And if it is, is the warming unusual or part of the normal fluctuations of a complex system? And has any of the observed warming actually made a difference to ecosystems, water resources, and human societies? This chapter addresses these questions based on the available evidence from observations and analysis of data around the world. But before we look at the details, we first need to define what we mean by climate change and what climate is. Some commentators use different definitions of climate change, and a statement that is correct under one definition could be wrong under another definition. Climate is often referred to as the average weather. A single rainy day does not make a wet climate. But if observations over many years have shown that, say, September is usually a wet period, then one may say that the climate during September is wet. Climate parameters, such as temperature, precipitation (rainfall or snowfall), and wind, always have to be defined by averages over a given period. There is no absolute minimum period over which such averaging has to occur, but the convention is somewhere between 20 and 30 years (though even longer averaging could be necessary when data are particularly sparse or noisy ). 12

3 Taking Stock of a Warming World Another important part of climate is its natural range of variability. In some countries the temperature on any given summer day does not vary by more than 2 3 degrees from its average, whereas in other countries the temperature on one day may be 10 degrees colder or warmer than in the following week. The extent to which the weather varies around its average is an important part of the average weather. Statements about the climate essentially tell us not only what sort of weather to expect, but also by how much we can expect the actual weather on any given day or short-term period to deviate from this expectation. Because climate describes statistical properties of weather, a single warm winter day does not mean the climate has changed. It is just a warm day. Not even a single warm year means the climate has changed. A change in the average weather has to persist over an extended period (decades or longer), before we can say that there has been a change in climate. Climate changes can occur not only in the average weather but also in its range of variability. For example, if a region normally receives reliable rain every year over a period of decades, and then starts to receive a lot of rain in some years but little rain in other years, we would say that the climate in this region has changed, even though the long-term average rainfall may have remained the same. It is important to note that, consistent with the IPCC, the expression climate change here simply refers to a detectable change in climate. It makes no judgement regarding the cause of the change. This is an important difference to usage in political negotiations, and often everyday conversations, where the phrase climate change often implies that human emissions of greenhouse gases are responsible for the changes. However, it is one task to identify a statistically significant change, and a quite separate task to identify the cause(s) of the observed change. Chapter 1 looks at changes in climate and their effects without any assumption about underlying causes. Chapter 2 then investigates whether we understand the causes behind these changes, in particular whether human activities are responsible. 1.2 Changes in the climate system The answer to whether the climate system is getting warmer is an unequivocal yes. This is a very strong statement, and scientists do not make such a statement lightly. It is based on the fact that many multiple strands of evidence all point in the same direction of a warming world, including rising global average ocean and air temperatures, widespread melting of snow and ice, and rising global average sea level (WGI 3.2, 4.8, 5.2, 5.5). We examine each of those strands of evidence in turn Global and regional changes in temperature The most direct measure of Earth s climate is the global average temperature of the air at the surface of the planet. It is an important measure of the climate not least because it tells us the average temperature where you and I live, and where the plants we use as food grow. Thousands of meteorological observing stations are collecting local temperature data about 2 m above ground. Large databases compile those 5 Note that when we say that global average temperature or sea level is rising, this does not mean that every part of the world is necessarily experiencing the same change. Some parts experience more change than average while others experience less; a few local changes can even go in the opposite direction to the global trend. 13

4 Climate Change 101 An Educational Resource measurements into a global average and into long-term global trends. The record for global average temperature goes back to about 1850, when reliable direct measurements were first carried out at a sufficient number of stations around the world. Some meteorological stations go back further in time, but there are not enough measurements for an adequate approximation to a global average temperature. The global average temperature includes measurements from not only stations on land but also ships at sea, for the ocean surface covers 70% of the planet. Trends in global average temperature Figure 1.1 shows the global average temperature from 1850 to Several important features leap out from this graph. The first is the clear overall upward trend. The highest temperatures were reached during the 1990s and after 2000: 11 of the past 12 years ( ) rank among the 12 warmest years in the entire instrumental record. Overall, the world is now about 0.76 C warmer than it was in the late 19th century. 6 (WGI 3.2) If we fit a linear trend to the temperatures over the past 150 years, we find that the global average temperature has increased on average by 0.045(±0.012) C per decade. But it is clear from the graph that a linear fit is not a good approximation for how the world has actually warmed it misses the much stronger rate of warming over the past 100, 50, and 25 years (see Figure 1.1). Since 1980, the rate of warming has been 0.177(±0.052) C, which is almost four times greater than the rate of warming averaged over the entire period. (WGI 3.2) It is worth noting that global average temperatures obviously vary from year, and that, for a few years at a time, temperatures can go in the opposite direction of the long-term trend. For example, if in 1986 one had looked back about eight years ( ), one would have seen falling average temperatures (see Figure 1.1). But if one had concluded from this in 1986 that the world was generally in the process of getting colder, nothing would have been further from the truth. A similar argument applies when looking at the most recent global average temperatures. The year 1998 may have been the hottest year on record to date (although another data set suggests that 2005 may have been equally as hot as, if not hotter than, 1998). However, claims that the world has been cooling since 1998, which can be found in the popular press, clearly do not make sense if we recall that climate trends can be established only over decades or more). Average temperatures from the year 2000 up to 2008 inclusive are about 0.17 C higher than the average temperatures of the 1990s, consistent with a continued long-term warming of the atmosphere. Conclusions drawn from a few data points or just a single year are almost always meaningless in relation to a global average long-term trend a signal indicative of climate change is robust only if it persists over an extended period. 6 To be precise: now refers here to , and the late 19th century refers to

5 Taking Stock of a Warming World Figure 1.1: Trend and inter-annual variability in global average surface temperature, Note: Main panel: The left-hand axis shows the difference relative to the mean temperature, and the right-hand axis shows the estimated actual temperature. The dots are global annual average temperatures, and the blue wriggly line is a smoothed line indicating decadal changes. The lightly blue shaded band represents a 90% confidence interval (ie, there is a 90% chance that the real temperature will lie within this band). The straight lines are linear fits to the data for the past 150, 100, 50, and 25 years. Inset panel: The inset panel shows the detailed annual average temperatures for , along with fitted trends for seven-year periods. The seven-year fitted trends are highly variable, including several periods with negative (cooling) trends despite the clear long-term warming during , as evident in the longer-term temperature data in the main graph. Source: Based on WGI FAQ 3.1 Figure 1 and data from United Kingdom Met Office global temperature data set HadCRUT3 (Brohan et al, 2006). Trends in different regions of the globe and atmosphere The warming of the Earth s surface has been widespread over the globe (Figure 1.2). 7 Land regions have generally warmed faster than the atmosphere above the surface of the oceans. The warming is most pronounced at higher northern latitudes, many of which are far away from human habitation (including many sites 7 This does not mean that every single point of the globe has warmed. The south-east United States and the Atlantic ocean south of Greenland have cooled. These are the only two regions with significant cooling. In other regions, such as central South America and central Africa, data are too sparse to establish robust trends. 15

6 Climate Change 101 An Educational Resource in the Arctic region). This confirms that warming is not simply an artefact resulting from the build-up of cities (WGI 3.2). 8 Figure 1.2: Linear trend of annual average temperatures in different parts of the globe, Note: Areas in grey have insufficient data to derive a reliable trend. Source: Based on WGI Figure 3.9. The warming of the atmosphere is mirrored by a warming of oceans: average ocean temperatures have increased down to at least 3,000 m, with strongest warming in the layer from the ocean surface down to 700 m. (WGI 5.2) The warming of the atmosphere also extends vertically upwards: analysis of satellite data and weather balloons since about 1960 shows that the region called the troposphere, which extends from the ground to about km above the Earth s surface, has warmed at a rate similar to the rate of warming near the surface. The rate of temperature change at different altitudes had been a point of discussion a few years ago, where satellite measurements seemed to suggest very little warming in the free troposphere, contrary to what one would expect from model simulations. However, more recent analyses have largely resolved this discrepancy: there were significant errors in the way data from different satellites had been analysed and 8 The question of whether increasing economic activity and urbanisation could distort the surface temperature record has been extensively analysed. It is well known that the large congregation of concrete in urban areas can produce a so-called urban heat island effect, which leads to urban night-time temperatures being higher than in surrounding rural areas. Even though data from individual urban meteorological stations can show clear signs of such an effect, a comparison of large sets of data from rural and urban meteorological stations shows very little difference across larger regions. Altogether, the possible effect of urban heat islands on the global average temperature trend (which of course includes sea surface temperatures, where the urban heat island effect by definition is not present) are estimated to be less than 0.02 C over the 20th century. This possible error, even though it is comparatively small, has been included in the error bars of the global average temperature record shown in Figure 1.1. (WGI 3.2) 16

7 Taking Stock of a Warming World combined. Based on the most recent analyses, tropospheric warming trends at all latitudes now appear to be consistent with models and the trend observed at the surface (WGI 3.4.1; Karl et al, 2006; Santer et al, 2008). The widespread warming of the atmosphere and ocean implies that very large amounts of heat energy have been added to the climate system. Since about 1960, about 80% of the total additional heat energy resulting from the warming has been absorbed by the ocean, with the remainder of the additional heat contained in the warmer atmosphere, continents, and ice. 9 (WGI 5.2) Trends in snow and ice cover Consistent with global warming trends, ice and snow have retreated around the world, most notably through reductions in average snow cover, glaciers, and Arctic sea ice (WGI 4.8). Glaciers, snow cover, and permafrost Glaciers are controlled both by temperature and the amount of snow they are blanketed by, and the great majority of glaciers in both hemispheres have seen an overall decline in size over recent decades (WGI 4.5). Some small glaciers in the Alps, Andes, and Himalaya Ranges are in the process of disappearing. Only a few glaciers have increased, often related to a pronounced increase in snowfall. Average spring snow cover in the northern hemisphere is now estimated to be on average about 5% less than it was in the decades before 1970 (although there are large regional year-to-year variations). The areas where snow cover retreated most strongly are those where warming was most pronounced especially during the spring season (WGI 4.2). Correlated with the warming of the atmosphere and ocean, the temperature at the top of the permafrost layer has increased by about 3 C in the Arctic and in other high altitude plateaus and alpine regions. This has led to thawing and ground instability (subsidence and dislocation of rocks). For example, a significant increase in the number of rock avalanches was observed during the extremely warm summer of 2003 in Europe (WGI 4.7; WGII 1.3). Arctic and Antarctic sea ice Arctic sea ice has declined dramatically over the past few decades. Satellite measurements from 1978 to 2005 have demonstrated a decline of sea-ice extent of more than 7% per decade during summer, and almost 3% in the annual average. Concurrent measurements of sea-ice thickness, based on submarine data, show a reduction of up to 1 m in sea-ice thickness since the late 1980s compared with the typical thickness of about m (WGI 4.4). 9 The total amount of additional heat energy that has already accumulated in the climate system is staggering: the energy added to the oceans alone from 1961 to 2003 is estimated at 8.11(±0.74) Joules (WGI 5.2). By comparison, the total amount of energy that humans currently produce across the globe from the burning of fossil fuels is 200 times less than this amount, only about 4(±0.4) Joules (data for 2006 from IEA, 2008b). This shows that the observed warming of the Earth cannot be explained by the direct addition of energy from human activities, but that a much more powerful mechanism must be at play, which I discuss in chapter 2. 17

8 Climate Change 101 An Educational Resource These trends continued and even accelerated in the past two years: 2007 saw the lowest summer sea-ice coverage on record (Comiso et al, 2008; Stroeve et al, 2008b), and 2008 had the second lowest coverage on record (see also Figure 1.3). There is increasing evidence that record-low ice extents make the sea ice more prone to melt again in the following year because first-season ice is thinner and more liable to break up (Maslanik et al, 2007). In contrast to these Arctic trends, Antarctic sea ice has shown large regional variations (eg, around the Antarctic Peninsula) but no significant overall trend when averaged around the continent. (WGI 4.4) Figure 1.3: Arctic sea ice extent in 1982, 2005, and 2007, and median ice extent over Note: The area of sea-ice cover lost in 2007 is roughly 10 times the size of the United Kingdom or New Zealand. Source: Maps and graphs are from United Nations Environment Programme GRID-Arendal, Maps and Graphics Library ( updated with data for 2007 and 2008 from the National Snow and Ice Data Center ( 18

9 Taking Stock of a Warming World Polar ice sheets The amount of snow and ice stored in glaciers and snow fields is dwarfed by the amount that is contained in the large polar ice sheets 10 that cover Greenland and Antarctica. The two ice sheets together have been shrinking at least since the early 1990s, but uncertainties about the amount of change for the individual ice sheets are large, particularly for Antarctica. Greenland thickened in its high altitude central part due to increased snowfall, but this increase in ice mass at its centre has been more than offset by ice mass loss from melting near the coast. The East Antarctic Ice Sheet has most likely been growing a little due to increased snowfall, whereas the West Antarctic Ice Sheet is probably losing mass due to increased flow of its glaciers. In both Antarctica and Greenland, the break-up of coastal ice shelves (sheets of permanent floating ice) and floating glacier tongues (glacier ice that flows beyond the coast for some distance) has led to increased flow of some glaciers that drain ice from the inner parts of the ice sheets. Such increases in glacier flow have contributed to the observed overall loss of ice mass from the polar ice sheets (WGI 4.6, 4.8). There have also been suggestions that meltwater that drains down deep crevasses can lubricate the flow of glaciers over the bedrock and thus accelerates the flow of ice even when only a moderate amount of melting takes place at the surface. Recent studies in Greenland confirmed that such an acceleration did indeed take place on short timescales of days and weeks, but changes in meltwater flow did not appear to have a significant effect on overall annual or longer-term ice mass loss (Das et al, 2008; Joughin et al, 2008; van de Wal et al, 2008). Another recent study identified that the significant acceleration of a large Greenland glacier was triggered by the advection of warm subsurface ocean waters that melted its ice shelf, pointing to a complex interplay of atmospheric, ocean, and ice sheet mechanisms in the flow of glaciers and overall loss of ice from the ice sheets (Holland et al, 2008). Unfortunately, we do not yet have models that successfully and on a large scale reproduce all these dynamic processes and that would allow us to predict the impact of higher air and ocean temperatures on the rate at which the polar ice sheets could lose ice into the ocean. This limits the extent to which we can model future changes in ice sheets and resulting sea-level rise as the world continues to warm (Alley et al, 2008). There is currently no consensus among scientists on whether and how much the loss of polar ice through their bounding glaciers could speed up further with rising temperatures (more on this issue in chapter 3) Global and regional sea-level changes The warming of the atmosphere and ocean, and the melting and decline of snow, glaciers, and ice sheets have led to a rise in sea levels over the past century (WGI 5.5). Sea levels rise in a warming world for two main reasons. The first reason is that water expands when it warms up. Since the world s oceans cannot expand sideways, they can only go up. The second reason is that snow and ice that have hitherto been stored on land turn into water as they melt and eventually reach the 10 An ice sheet is a large mass of ice that is so thick it covers most or all of the bedrock on which it sits, and its shape is largely determined by melting and snowfall processes. In the current climate, there are only two ice sheets, covering Greenland and Antarctica (which is split into West and East Antarctica, separated by the Transantarctic Mountains). In contrast, glaciers are masses of ice that flow downhill under the force of gravity, and whose shape and flow speed is dominated by the slope and shape of the underlying terrain. 19

10 Climate Change 101 An Educational Resource ocean, thereby adding to the total amount of water. The melting of sea ice and breakup of floating ice shelves do not contribute to sea-level rise since this ice is already floating in water (just as melting ice cubes in a glass do not make the glass overflow but throwing additional ice cubes into an already full glass certainly does). Figure 1.4 shows the change in global average sea level since the late 19th century. Sea level rose by an average of 17(±5) cm over the 20th century (or 1.7 mm per year), but recent satellite measurements from 1993 to 2003 have shown a much higher rate of 3.1(±7) mm per year. It would be premature though to claim on the basis of these more recent data that sea-level rise is accelerating, since limited periods of higher rates of sea-level rise have also been observed during earlier periods (eg, during the early 1980s). The variability in the rate of sea-level rise is most likely due to intermittent changes in the amount of heat absorption by the oceans, and hence the changing rate of thermal expansion of the ocean water. Longterm data indicate that the average rate of sea-level rise accelerated from the mid- 19th to the mid-20th century (WGI 5.5, 5.6). If we put all the contributions to sea-level rise together, it is estimated that for , 57% of the sea-level rise comes from thermal expansion of the warming ocean, 28% from the shrinkage of glaciers and small ice caps, and the remainder (about 15%) from the observed reduction in the ice sheets of Greenland and Antarctica. For the period before the 1990s, data about individual contributions to sea-level rise are much less certain, especially regarding changes in ice sheets. 11 (WGI 5.5, 5.6) Satellites since the 1990s have allowed the precise measurement of global sea level and have provided a richer picture of sea-level changes around the world. They confirm that the global average sea level is rising significantly, but this rise is not uniform: some regions have witnessed more than average increases, while other regions have seen less rise or even a fall in sea levels. Such regional variations in sea-level rise can result from changes in ocean circulation associated with wind and regional climate patterns. In addition, relative local sea-level changes can, of course, also be influenced by changes in land masses, such as geological uplift or land subsidence from water extraction or settlements, rather than changes in absolute sea level (WGI 5.5). This regional variation reinforces the need to look at many different data points and measurement techniques when determining global long-term trends: a single tide gauge measurement from one particular location cannot prove or disprove whether the global average sea level, or even absolute local sea level, has been rising over the 20th century. 11 The AR4 noted that the individual contributions to sea-level rise (thermal expansion, melting of glaciers and ice caps, and changes in ice sheets) match the total observed sea-level rise for , but do not add up for the earlier period from A recent study suggests corrections to the way that thermal expansion is calculated, which would improve the agreement between observations and individual components of sea-level rise for (Domingues et al, 2008). 20

11 Taking Stock of a Warming World Figure 1.4: Global average change in sea levels, relative to the average sea level over Note: Dots are annual average values, the dark blue line is a decadal smoothed average, and the light-coloured band represents a 90% confidence interval (ie, there is a 90% chance that the real temperature will lie within this band). The red line presents results from satellite data since Source: WGI Figure SPM Changes in precipitation So far we have discussed evidence for widespread warming across the world, which is evident in temperature trends, but has also fuelled the reduction of ice and snow and rising sea level. However, the climate system has also changed in other important aspects, in particular precipitation (ie, rain and snowfall) and the occurrence of climatic extremes (such as heavy rainfall, droughts, heat waves, and storms). Large-scale patterns of change have been observed over the course of the 20th century. Average precipitation increased in several high-latitude regions such as northern Europe and northern and central Asia and in eastern parts of North and South America. Significant decreases in precipitation were observed in the Sahel, the Mediterranean, southern Africa, and parts of southern Asia, that is, mostly dry mid-latitude and subtropical regions. In other regions, rainfall may have varied on shorter timescales but data are insufficient to establish clear longer-term trends (WGI 3.3, 3.9). Changes in annual average rainfall tell only part of the story though. Rain and snow can fall in relatively steady amounts, or they can fall in heavy downpours with more extended dry periods in between. Observations suggest that extreme precipitation events have increased, including areas where average rainfall decreased, but concurrently, the incidence of drought has also increased globally since 1970, mostly in the tropics and subtropics (WGI 3.3, 3.8). It may seem odd that precipitation extremes increase at either end of the spectrum, but it is fully consistent with a warming world. The widespread increase in heavy downpours is consistent with the fact that a warmer atmosphere can hold more moisture, and hence more moisture is available to fall as rain in short and intense bursts (WGI 3.8, 3.9). The increase in droughts is in part the consequence of reduced rainfall in many regions, but higher temperatures may also lead to a faster evaporation of moisture from soils: some drought incidents in Australia and Europe 21

12 Climate Change 101 An Educational Resource have been specifically linked to extreme temperatures and heat waves. In addition, changes in ocean surface temperatures and atmospheric circulation patterns as well as reductions in snow pack have also contributed to extended dry periods in some regions (WGI 3.3, 3.9, 4.2) Changes in other climatic extremes Apart from precipitation, other climatic extremes, such as heat waves and tropical cyclones, have also changed in their statistical frequency and/or intensity. Extreme events are generally much harder to monitor because by definition they do not happen very often. A change in the occurrence of something that happens only every few years or even decades is much harder to define and demonstrate scientifically than a change in a continuously measurable quantity. However, we notice extreme events more because they tend to do immediate and noticeable damage to lives and property. Temperature extremes and heat waves have most clearly changed over the past 50 years. The number of extremely hot days has increased, while the number of extremely cold days and frosts has decreased, both being consistent with a warming climate. There is also robust, though not quite as comprehensive, evidence that heat waves, that is, periods of high temperature lasting several days or weeks, have become more common over most land areas (WGI 3.8, 3.9). Tropical cyclones (typhoons and hurricanes) symbolise for many people the potentially destructive consequences of climate change, but the link between cyclones and climate change is not simple. Satellites and other observations show that the activity of the most intense hurricanes has indeed increased in the North Atlantic since about This increase has been correlated with the increase in ocean surface temperatures in this region, which provide the energy that fuels hurricane formation. Some data suggest an increase in intense tropical cyclone activity also in other regions, but there are concerns over data quality especially before 1970, when observing systems relied only on individual observing stations rather than satellites. These limitations make it difficult to reach robust conclusions on long-term global changes in cyclone intensity or frequency. It is also important to point out that the overall number of tropical cyclones seems not to have changed globally; the only clear and significant change is that the proportion of storms in the most intense category has increased in some regions, but with fewer weaker storms (WGI 3.8). Observations also suggest an increase in extreme high water at a broad range of sites since the mid-1970s. In some places, rising average sea level was the main driver behind increased high water extremes: higher average sea level means that spring tides, especially when combined with storms, also reach higher and, therefore, more easily overtop high water marks and existing stop banks. In other regions, changes in storminess were more important in determining trends. Data are insufficient to establish trends further back in time (WGI 5.5). 1.3 Comparing the 20th century climate with the climate in earlier times The observed long-term climate trends, regional climate patterns, and changes in extremes together paint a compelling and unequivocal picture of a warming world and many changes in the climate system that are related to this warming trend. And yet, periods of years are just blips on geological timescales, and some people argue that what looks like a significant trend could simply be an example of natural long- 22

13 Taking Stock of a Warming World term variability of the Earth s climate. So how does the warming trend over the 20th century compare with warming and cooling episodes that may have happened before? Twentieth century climate in the context of the past 2,000 years Getting a sense of the Earth s climate before the late 19th century is difficult because very few direct and reliable, let alone comprehensive, measurements of global average temperature exist. Temperatures, therefore, need to be derived from socalled proxies, that is, natural processes that are affected by the temperature of the surrounding atmosphere and leave a measurable trace (WGI 6.2). One example of a widely used proxy is the thickness of tree-rings trees go through an annual growth cycle and the width of tree rings in some places can be used as an indicator of the ambient temperature in which the tree grew. Very roughly, thin tree rings indicate cold years, while thick tree rings indicate warmer years. Taking into account the limitations of any individual proxy method (see Box 1.1), the best way to estimate previous global annual average temperatures is to take many different proxy studies and combine them or show them side by side to get a sense of the differences between the studies. Figure 1.5 does exactly this. The figure shows temperatures estimated for the northern hemisphere based on a large number of proxies, including tree-rings, pollen sediments, boreholes where the temperature in successive layers of the ground is measured, and ice cores where tiny trapped bubbles of air are used to analyse the composition and infer air temperature in the past. Too few records are available from the southern hemisphere to warrant their inclusion in Figure 1.5 (WGI 6.6). Box 1.1: Limitations of proxies for deriving pre-20th century climate information Proxies have important limitations. First, they need to be calibrated one species of trees will respond very differently to temperature than another one, or even one in another location where there is different soil and different amounts of sunshine and wind. Secondly, temperature is often not the only thing that influences the proxy. In the case of trees, rainfall, wind, and nutrients also may have been changing and their influence can be difficult to separate from the temperature s influence. Thirdly, trees in temperate latitudes grow mostly during the warm summer months, so that tree rings properly measure only summer climate, not annual average temperature. Finally, all proxies derive their data from specific study sites, so that the resulting temperatures can usually give only local, or at best regional, pictures of previous climates (WGI 6.2, 6.6, 6.7). Figure 1.5 reveals three important messages. The first is that different proxy records show substantial differences, most likely due to any combination of the limitations discussed above. The second is that within individual proxy records, temperature variations during the past 1,300 years appear to have been limited to less than about 1 C. The band that shows the greatest overlap between proxies (which might be indicative of hemispheric rather than local temperature variations) shows variations of less than 0.5 C over this time. The third important message is that the late 20th century appears indeed unusually warm in comparison to this long-term climate record (WGI 6.6). 23

14 Climate Change 101 An Educational Resource Figure 1.5: Changes in average northern hemisphere temperatures over the past 1,300 years Note: Changes are given relative to the average temperature in The individual coloured lines represent estimates from selected individual studies that used different sets of proxy data. The shaded bands indicate the amount of overlap between all assessed studies, with darkest shading indicating the greatest overlap. Source: Based on WGI Figure 6.10 Panels b and c; see the original figure for details on the individual studies. A more detailed uncertainty analysis suggests at least a 90% probability that the 50 years were warmer globally than any other 50-year period in the past 500 years, and at least a 66% probability that they were warmer than any other 50-year period in the past 1,300 years (WGI 6.6). This conclusion should not gloss over the fact that, regionally, temperatures may have varied more. For example, there are well-documented records of Vikings settling Greenland and farmers in England producing wine during the Middle Ages. However, based on the large number of proxy studies around the world, there is no evidence for any extended particularly warm periods at a global scale in the more distant past (WGI 6.6). The late 20th century does stand out as unusually warm. Studies going further back in time have increasing uncertainties and more limited coverage. Collectively, the available data provide no convincing evidence that the world as a whole was ever warmer than at present in the 10,000 years since the world came out of its last ice age (WGI ) Ice ages and warm periods over the past several hundred thousand years Before we conclude the discussion of how significant the recent warming is compared with earlier times, it is worth casting a brief look back into even earlier history, namely the succession of ice ages and warm periods over the past several hundred thousand years. Proxy data for these times are much more sparse and tend to be less accurate in terms of both timing (ie, which year, or even decade, they represent in the past) and the climate parameter they measure (such as temperature, sea level, or storminess) than is the proxy data for the past 1,300 years (WGI 6.3, 6.7). Nonetheless, these past climate changes allow us to draw several fundamental conclusions about the Earth s climate system. The first key conclusion is that differences in global average temperature of some 4 7 C changed the planet from one where, in the northern hemisphere at least, land was covered down to mid-latitudes in permanent ice and snow, to a planet with much 24

15 Taking Stock of a Warming World less ice than we see today. Changes of a few degrees Celsius in global average temperature might seem small compared with the temperature differences that we experience in our daily lives, but daily or yearly variations in temperature usually only mean that heat energy is shuffled from one place to another. In contrast, when such temperature changes occurred globally and were sustained over centuries to millennia, they were sufficient to alter completely the face of the Earth (WGI 6.3, 6.4). The second conclusion is that ice ages were not arbitrary and unpredictable changes of the Earth s climate system. They were caused by changes in the way the Earth orbits around the sun, and hence in the amount of solar energy the planet receives. Ice ages are a dramatic illustration of the fact that changing the energy balance of the Earth changes the global average temperature, and along with it many other aspects of the climate (WGI 6.4). We discuss in chapter 2 the various ways in which the energy balance of the Earth changed over the 20th century, and what we can learn from this about the causes for the recent observed climate changes and their effects. A third conclusion is that global-scale changes between ice ages and interglacial periods typically occurred over timescales of millennia rather than centuries. There is, however, evidence that, regionally, large and rather abrupt climatic shifts of more than 10 C occurred within decades in the North Atlantic region between Greenland and northern Europe. Such changes are likely to have been related to shifts in ocean circulation that can occur over the space of only a few years. They indicate that even if the external forcing of the climate system is gradual (such as a steady change in the Earth s orbit around the sun), the climate system can overreact regionally, and could well continue to do so in future in response to future changes in the Earth s overall energy balance. However, the detailed mechanisms behind these overreactions are still only poorly understood (WGI 6.4). The final key conclusion from the study of ice ages and interglacial periods is that the last time the Earth was a few degrees warmer than today (which was about 125,000 years ago), the global sea level was 4 6 m higher, mainly from the retreat of polar ice. Most of the sea level is assumed to have come from the Greenland ice sheet, which was much smaller during this period than it is today. 12 (WGI 6.4) A recent study indicates that sea level rose by up to 1.6 m per century during this most recent previous warm period (Rohling et al, 2008). 1.4 Impacts of observed changes in the climate system Now that we have established that the climate is unequivocally changing, and that the change appears unusual at least compared with the past several hundred, if not several thousand, years, the question arises of course: has anybody noticed? Scientists have clearly noticed and documented these changes, but can we see this change in the way that ecosystems respond? Can we see this change in human society? In other words, does the warming we have seen so far actually matter? 12 The reason we know that the Greenland ice sheet was much smaller during this earlier period comes from ice cores. Every year, a new layer of snow is laid on top of the ice sheet and gradually gets compressed into ice. We can drill into the ice and thus work our way backwards in time, almost like in tree rings. The more ice layers we can go through, the further back in time we travel; in some places in Antarctica and Greenland, ice cores provide information going back almost 1 million years. In several places in Greenland though, the ice core drill hits bedrock when ice layers correspond to about 125,000 years no ice is present from earlier times. This means that snow must have fallen on bare ground about 125,000 years ago. 25

16 Climate Change 101 An Educational Resource The answer to these questions is also a fairly clear yes, though not as unequivocal as for the change in climate itself. There is strong evidence that regional climate changes over the past 30 years, particularly temperature increases, have affected many natural systems on all continents and in most oceans. Evidence for effects on human activities is also emerging, but such effects are generally harder to detect because changes in climate alone are hardly ever the sole reason for changes in human behaviour (WGII 1.3). The next few sections go through some of the key observed and scientifically documented effects step by step, with illustrative examples drawn from the scientific literature Effects of changes in snow and ice on polar and mountain regions The observed rapid retreat of glaciers and reduction in snow cover has already affected ecosystems and human activities in several ways. One set of impacts relates to physical changes in the landscape. When glaciers melt, lakes often form on their surface. As these lakes grow with accumulating meltwater, their walls can burst and send sudden floods downstream, which represents a serious threat to villages located in glacier valleys. Such glacial lakes have formed in increasing numbers and size in many mountain regions around the world, and governments in the Himalaya Ranges and Andes have devoted significant resources to dam lake walls or drain lakes to safe levels. The melting of frozen ground is also destabilising glacier moraine walls in many regions. In addition, anecdotal evidence suggests weakening of permafrost in the Arctic region is destabilising buildings, roads, airfields, and pipelines, but we do not have sufficient scientific publications to identify robust trends (WGII 1.3). The retreat of sea ice is having not only physical impacts but is also changing the habitat of species that depend on it. For example, polar bears use sea ice as a platform from which to hunt seals, and a reduction in sea ice means the bears access to their primary food source is more difficult. The overall effects of the reduction in sea ice on polar bears is still limited, since the gradual retreat of ice means the southernmost regions become unsuitable while others can temporarily become more attractive. In the future, as sea-ice cover is expected to continue to shrink across much of the north polar region (see chapter 3), the habitat of polar bears is likely to be increasingly compromised (WGII 1.3, 4.4.6). For an example from Antarctica, see Box 1.2. Many human activities are already noticeably influenced by the reductions in Arctic snow and sea ice. New shipping routes are opening up but traditional hunting, which requires travel over sea ice, is becoming less reliable. On land, travel in the Arctic tundra often occurs on frozen ground, but the period when such travel is possible decreases as permafrost melts. 13 The reduction of snow and ice cover has also affected commercial and recreational alpine activities such as skiing and climbing. Snow cover has become less reliable in low altitudes, so more commercial ski fields have to use artificial snow. An increasing number of high-altitude mountaineering routes in the Alps, Andes, and Himalaya Ranges are becoming less accessible due to the loss of consistent ice and snow deposition, which increases crevasses and the risk of rock and ice falls (WGII 1.3). 13 For some roads, the period of safe travel on permafrost has almost halved since 1970 from 220 to 130 days per year (WGII 1.3). 26

17 Taking Stock of a Warming World Box 1.2: Effect of changes in sea-ice cover on Antarctic penguins In Antarctica, penguin populations have been affected by local changes in sea ice, but the effect on them is more complex to determine than the effect on polar bears in the Arctic. This is, in part, because Antarctic sea ice has decreased in some places but increased in others, so there is no evidence yet of an overall consistent impact on penguins. Penguin responses to changes in sea ice may also be more complex because different penguin species have different levels of adaptation to ice and ice-free conditions (Forcada et al, 2006). Shrinking sea ice generally reduces the abundance of krill, the penguin s main food source, but the position of floating ice bergs can also alter the distribution of food sources and the penguins ability to access their nesting sites (Barbraud and Weimerskirch, 2001; Forcada et al, 2006) Effects on water resources and freshwater ecosystems The enhanced melting of glaciers and snow cover is leading to higher peak river flows earlier in the season. The resulting shift in peak river flows towards spring and away from the summer months can be problematic because summer is usually the period with highest water demand. The increased spring flows, of course, can be sustained only while the glaciers are melting; under continued warming, this increased flow is expected to eventually reverse and decline as the ice mass of individual glaciers shrinks. Winter flows have increased in some rivers in high northern latitudes where precipitation now falls increasingly in the form of rain rather than snow (WGII 1.3, 3.4). Where snow and ice are not the dominant sources of water, river flows have changed in many places consistent with changes in rainfall patterns: run-off has tended to decrease in many rivers in the subtropics and some dry mid-latitudes, whereas run-off has increased in many high-latitude regions. Although there is evidence of increased heavy precipitation in most regions, there is no evidence yet of a related global increase in river floods, because human modifications of river catchments also influence flood risk. As already noted, the areas affected by drought have increased since the 1970s, but data are insufficient to detect and attribute longterm changes in water availability to this increase in drought risk because many other changes (such as changes in water demand, river management, and land-cover in river catchments) are also important at the local and regional scale (WGII 1.3.2, 3.4). Increasing air temperatures are also noticeably affecting the temperatures and thermal structure of lakes and rivers. Surface temperatures of lakes and rivers have increased C in Europe, North America, and Asia since the 1960s. Higher summer surface temperatures reduce mixing of water from different depth levels in lakes, which in turn affects nutrient concentrations, salinity, and oxygen concentrations. These changes have led to earlier and more intense spring algal blooms in high-altitude and high-latitude lakes by up to 20 days. These changing temperatures and algal bloom events have also affected dependent higher species (WGII 1.3; Box 1.3). 27

18 Climate Change 101 An Educational Resource Box 1.3: Effect of algal bloom events on freshwater ecosystems Some zooplankton species that depend on algae for food have not managed to adapt their breeding cycle to the changes in algae brought on by increased lake and river temperatures, so have become less common (Winder and Schindler 2004). Similarly, cold-water fish species are gradually being replaced by warmth-loving species in some major rivers, which is reflected in changes in fish catch and the wider ecosystem (Daufresne et al, 2003). The main migration of some fish, such as salmon, now occurs up to several weeks earlier in the year, which can affect fish mortality rates because their main food sources along the way may not be available when the salmon most rely on them (Cooke et al, 2004) Effects of temperature changes on terrestrial and marine ecosystems Over the past decade, there has been an explosion of studies into the responses of ecosystems and individual species to regional warming, both on land and in the oceans. Changes in phenology (the timing of seasonal activities of animals and plants) are among the most prominent effects of the observed warming. Spring events such as leaf unfolding or flowering in mid- to higher northern latitudes now occurs some 3 10 days earlier than it did in the early 1970s. Similarly, the timing of birds egg-laying closely follows regional spring temperatures, and the hibernation period is ending earlier for many small mammals. Many migratory birds have changed the date of their migrations, with short-range birds typically responding more strongly to temperature signals than long-range travellers. Some birds have stopped migrating in response to the warmer winters (WGII 1.3). Such changes in the breeding and migration timing of different species may appear to be obvious and successful responses to higher temperatures, but they can create challenges and threaten to unravel ecosystems if different parts of a food web do not change in the same way. Examples from the literature offer powerful demonstrations of this problem (see Box 1.4). Rising temperatures have generally led to an overall greening of vegetation as spring occurs earlier and plants colonise higher latitudes and altitudes in response to warming. However, a range of studies shows that such greening and temperaturedriven migration may come at a cost in species diversity. In many cases, generalists in the world of plants and animals have found it easier to move with the changing climate than have specialists, which may rely on more than warm temperatures for their survival (eg, they may rely on special food sources that have not kept pace with the shift in climate). (WGII 1.3) Similar large-scale changes have been observed not only on land but also in the oceans, where shifts in species ranges and abundances are associated with rising temperatures as well as related changes in ice cover, salinity, oxygen levels, and ocean circulation. These changes are particularly noticeable in the North Sea, where warmth-loving species of plankton have expanded and moved northward by up to 10 degrees of latitude (more than 1,000 km) over recent decades, while cold-loving plankton species declined or moved into polar waters. Perhaps more importantly, the timing of algal blooms and key production of different plankton species that feed on each other also changed, but at different rates (WGII 1.3). 28

19 Taking Stock of a Warming World Box 1.4: Impact of higher temperatures on terrestrial species and ecosystems The end of the hibernation time of the yellow-bellied marmot (a resident of mountain ranges in the United States and Canada) appears to be dictated by spring temperatures. These temperatures have increased, but snow cover has not decreased, possibly due to increased winter snowfalls. As a result, marmots wake earlier, but then have to struggle through a longer period until edible vegetation emerges from the snow. During this period, they are also more vulnerable to predators. Some migrating birds are similarly mis-timed: American robins now arrive in the Colorado mountains on average 14 days earlier than they did in the early 1980s, but their main food source is still buried under snow when they arrive (Inouye et al, 2000). Similar slow disintegrations of food webs and species communities because of warming have been observed in the Yosemite National Park (Moritz et al, 2008). Some longer-distance migrating birds in Europe, such as flycatchers, face opposite challenges to local hibernation: they still arrive at about the same time in Europe from their wintering area in west Africa, but spring is happening up to two weeks earlier in Europe. To adapt to these changes, some of the birds now lay their eggs earlier, which leaves them less time to recover from their migration. Even so, their breeding cycle has not kept up with the change of seasons, because some of the insects that represent their main food source are already declining by the time the eggs hatch. As a consequence, populations of flycatchers have decreased in several study sites in Europe (Both and Visser, 2001; Both et al, 2004). This mismatch in the food web has had significant consequences: for example, the larvae of cod, a commercially harvested fish, depend on plankton during their early growth phase, and several studies suggest the climate-driven change in plankton has added to existing pressures on cod stocks from over-fishing (Beaugrand et al, 2003; Edwards and Richardson 2004). The bleaching of coral reefs is often held up as another prime example of observable impacts of recent climate change. However, both pollution and increasing temperatures influence bleaching episodes. An unambiguous attribution of recent bleaching events to observed temperature increases is therefore difficult at present (see Box 1.5) Changes in the coastal zone and their relation to sea-level rise Many places around the world are experiencing coastal erosion and losses in coastal wetlands. Detailed observations suggest that sea-level rise is one factor, but land subsidence, loss of sea-ice cover and the melting of permafrost along coastlines, 14 and human modifications of the shoreline also contribute to the observed changes. The most important direct human impacts on the coastal zone include extensive shoreline development through buildings and roads, the mining of sand and corals, the clearance of mangrove areas, and the reduction of sediment input from major rivers through the construction of large upstream dams and channels (WGII 1.3). 14 Sea ice reduces the magnitude of waves that hit shorelines. Where sea ice disappears, the increase in wave activity can dramatically increase coastal erosion. This effect can be further increased by thawing of permafrost along the shore line (Forbes et al, 2004; Forbes, 2005). 29

20 Climate Change 101 An Educational Resource Even though coastal erosion and loss of wetlands are significant problems in some regions, the evidence suggests they are at present caused largely by such human modifications of the shoreline and the building of fixed structures in a highly dynamic environment, and that rising global average sea level contributes only to part of the problem. However, further sea-level rise over the next decades and centuries is expected to make existing problems worse in many regions and could create additional challenges (see chapters 3 and 4; WGII 1.3, 6.4, 10.4, 16.4). Box 1.5: Coral bleaching role of increasing temperatures and other pressures Coral reefs have been identified as vulnerable to the effects of climate change, particularly high sea surface temperatures. When temperatures exceed the average seasonal maximum by more than about 1 C, corals expel the symbiotic algae that live in the substrate and provide the coral with nutrients. These algae are mainly responsible for the colour of corals, so corals that have expelled the algae appear bleached hence the phrase coral bleaching. If corals survive a bleaching episode, it usually takes them weeks to months to recover their original symbiotic algae density; they could also be recolonised with other algal species. If bleaching is prolonged, or if ocean surface temperatures exceed the average seasonal maximum by more than about 2 C, corals die (although different species have different tolerance thresholds). Major bleaching episodes have been observed during recent decades, usually during El Niño conditions that lead to particularly high ocean surface temperatures. However, ocean temperatures are not the only factor in bleaching: corals that suffer from high sediment loads, pollution by agricultural nutrients, and over-fishing that can damage coral structures are more susceptible to bleaching and take longer to recover. There is increasing evidence of temperature-related bleaching events, but it is not possible to separate rising ocean temperatures contribution to bleaching from the role of non-climate stresses in this bleaching (WGII 1.3, 4.4.9, 6.4, 6.6) Emerging effects on the human environment Our survey has focused mainly on the effects of recent climate changes on physical systems, such as glacier lakes and river flows, and various ecosystems. Some of these effects necessarily flow through to human society because humans depend on ecosystems such as fisheries, on water from rivers and glacier melt, and on snow and ice cover in the Arctic and in alpine areas, and because some human settlements are affected by changing natural hazards such as increasing glacier lakes. In some cases, the relevance of climate change to date is difficult to separate from concurrent non-climate changes: for example, fish stocks in the North Sea have been affected by severe over-fishing as well as climate changes; water stress in some regions may be caused not only by reductions in water supply but also by changes in land use and water demand; and changes in the coastal zone are often the result of a mixture of direct human activities, changes in the climate, and other environmental changes. In addition, humans, to some extent, adapt to the changes they experience in their environment, which makes it harder to clearly detect signals of a changing climate in human society (WGII 1.3). This difficulty applies particularly to detecting economic impacts of recent climate changes. Numerous studies attribute costs to 30

21 Taking Stock of a Warming World specific climatic extreme events (such as hurricanes or floods), but it is exceedingly difficult to attribute a long-term trend in such costs to changes in the climate due to concurrent changes in human society (see Box 1.6). Some climate-related impacts on the human environment are emerging though, despite the presence of non-climate changes and adaptation, particularly in temperate developed countries. This geographical focus is unlikely to reflect the real distribution of the effect of recent climate changes; it is more likely to be indicative of the generally greater availability of robust long-term data and scientific studies that allow clear conclusions about the role of climate within a constantly changing human environment (WGII SPM). Effects on human health Human health has been demonstrably affected by recent climate changes. The most prominent example is the impact of the 2003 summer heat wave, which led to the deaths of tens of thousands of people in Europe (WGII 1.3, 8.2). The occurrence of heat waves (defined as a number of consecutive days that exceed a certain temperature) is generally more likely where the average seasonal temperature has increased. At the same time, winter mortality in many temperate countries is reducing, but this appears to be primarily the result of better adaptation to the cold rather than warmer winters per se (WGII 1.3, 8.2). Apart from the direct effects of heat and cold, rising spring temperatures are leading to the earlier, and in some places more intense, production of pollen that can trigger allergenic reactions. Similarly, milder winters have led to the spread of ticks that are carriers for tick-borne encephalitis. Evidence is insufficient to determine whether this wider spread of ticks has also led to an increase in the actual incidence of the diseases they carry. The evidence is less clear for changes in the distribution of malaria. Some studies suggest temperature increases in African highlands have contributed to the spread of malaria in these regions, but changes in mosquito control, resistance to drugs, the incidence of other diseases, and other environmental factors make it difficult to determine the relative importance of climate in the observed changes (WGII 1.3, 8.2). Effects on agriculture and forestry Agriculture and forestry also show clear effects of the recent warming, particularly in Europe and North America. Flowering dates for many commercial crops have advanced during spring, along with a reduced incidence of frosts and a longer growing season. Farmers have responded to these changes in some regions by sowing some crops earlier, although the change in planting dates has been less than what one might expect based on the concurrent change in flowering dates for wild plants and fruit trees (WGII 1.3). Forest productivity has generally increased resulting from a combination of lengthening of the growing season, increased carbon dioxide (CO 2 ) concentrations (which can act as food for plant growth see section 1.5), nitrogen deposition, and changed management practices. In some already dry regions, such as the Mediterranean and parts of North America, forest production declined as a result of warmer and drier summers, while in high northern latitudes the productivity of tundra increased. Studies have found that outbreaks of damaging insects and forest fires are correlated with temperature increases, particularly during drought periods, but longerterm trends in forest fires and their causes are still controversial (WGII 1.3, 5.2). 31

22 Climate Change 101 An Educational Resource Box 1.6: The difficulty of identifying economic impacts of climate change Global economic losses related to weather-related natural hazards show rapidly rising costs since the 1970s. It is generally acknowledged that much of this increase is due to the increasing value that is represented by human settlements and activities. There is no agreement in the scientific literature about the extent to which climate trends may have contributed to the rising cost of natural disasters to date, because many other human changes (including disaster management and the way costs are reported around the world) make it difficult to unambiguously detect a climate signal in these changes (WGII 1.3, 7.2) Consistency and global coverage of observed effects of climate change The preceding survey indicates that the effects of recent climate changes have been pervasive. For the AR4, the IPCC surveyed more than 29,000 data sets of observed effects in biological and physical systems and their relationship with recent warming. More than 89% of these data sets 15 showed that physical and biological systems had changed in a way that was consistent with the direction of change that would be expected if recent warming and related climate changes were the dominant cause (WGII 1.4). This analysis was recently extended further, including even more data sets, which further confirmed the significant effect of rising temperatures on human and natural systems (Rosenzweig et al, 2008). While it is generally hard to prove in any individual case that climate was indeed an important reason for a specific change, the broad consistency of changes across the world gives a compelling picture. In some cases, the effects of recent climate changes are still small, but the fact we can discern so many effects as a consequence of less than half a degree global average warming over the past few decades shows how important climate is in determining living conditions on our planet. Most studies concerning the recent effects of climate change come from Europe followed by North America, but effects of recent climate change have been reported from all continents and most oceans (Figure 1.6), and, in all continents, most observed effects are in the direction expected under increasing temperatures (WGII 1.3, 1.4). The under-representation of studies from developing countries and from the southern hemisphere in general do not necessarily imply that effects of climate change have been less in these regions; rather they indicate the absence of long-term data and scientific studies that would allow robust conclusions about the effects of recent climate changes to be drawn (WGII 1.4). Indigenous communities in particular may well hold a wealth of knowledge about recent climate changes and their effects, but the IPCC assessment, and hence this chapter, is limited to those studies that appear in the scientific literature and have undergone independent scrutiny. I would like to stress that none of the discussion in this chapter assumes we know the causes of the observed climate changes. This chapter is concerned only with the 15 Note that effects on human systems, such as agriculture, forestry, and health, have been excluded from this global statistic because the relative influence of climate and non-climate drivers is less clear in many instances. 32

23 Taking Stock of a Warming World question of whether the climate is changing (to which we answered unequivocally yes ) and whether those changes have affected physical, biological, and human systems (to which the answer is also a fairly clear yes ). It is an entirely different question whether we understand the causes of the recent climate changes, and to what extent we can attribute the observed warming and its effects to human activities. These questions are addressed in chapter 2. Figure 1.6: Map of observed changes in physical and biological systems correlated with temperature changes, Note: The size of each point on the map relates to the number of studies that are represented. White areas on the map do not have sufficient climate data for temperature trends in that region to be reliably determined for Source: Based on WGII Figure SPM Direct effects of increasing carbon dioxide concentrations Along with global temperatures, greenhouse gas concentrations, particularly of CO 2, have increased markedly over the past century. As we see in chapter 2, the increase in greenhouse gas concentrations has been identified as the main cause of the observed warming. However, CO 2 can also affect ecosystems directly, independent of its role in the climate system. Our survey of evidence of a warming world would not be complete without a brief look at the evidence of the direct non-climate effects of CO Ocean acidification As the concentration of CO 2 in the atmosphere increases, the oceans absorb an increasing amount of this CO 2. Over the past few decades, the ocean has absorbed about 20 30% of the total human emissions of CO 2 each year. When CO 2 dissolves in ocean water, it forms a weak acid, whereas the ocean naturally is weakly alkaline. 33