Introduction
By 2050, global average temperature under A1b emissions would be between 1.4 and 2.9 °C above the 1961-1990 mean, with an average increase across climate models of around 1.9 °C. Global average sea level would be 12 to 32 cm higher than over the period 1961-1990, with an average increase of 18 cm (note that changes in temperature under A1b are between changes under RCP6.0 and RCP8.5: IPCC 2013). However, the spatial patterns of changes in temperature, precipitation, sea level and other relevant climatic variables vary between climate models, so the projected potential impacts also vary. This section first describes the potential impacts across the world and across sectors under one example plausible climate story, and then assesses the uncertainty in impacts by region and sector.
A coherent story: Impacts under one plausible climate future
Figure 1 and Table 3 show the impacts in 2050 under one illustrative climate model (HadCM3); this particular model has an increase in global mean temperature of 2.2 °C (relative to 1961-1990) in 2050 under A1b emissions, and a global mean sea level rise of 16 cm.
Table 3 Regional and global impacts in 2050, under the A1b emissions and socio-economic scenario, under one plausible climate story. The socio-economic impacts are relative to the situation in 2050 in the absence of climate change (Table 2)
Under this plausible story, approximately 1 billion people are exposed to increased water resources stress due to climate change, relative to the situation in 2050 with no climate change, and almost 450 million people are exposed to a doubling of flood frequency. In contrast, around 1.9 billion water-stressed people see an increase in runoff, and around 75 million flood-prone people are exposed to flooding half as frequently as in the absence of climate change. Approximately 1.3 million additional people are flooded in coastal floods each year. Around a half of all cropland sees a decline in suitability, but about 15 % sees an improvement. Global residential heating energy demands are reduced by 30 % (bringing them back to approximately the 2000 level) but cooling demands rise by over 70 %. The net effect is a reduction in total heating and cooling energy demands of around 15 %. There are, however, considerable regional variations in impact.
Under this story, increases in water scarcity are most apparent in the Middle East, north Africa and western Europe, whilst increases in exposure to river flooding is largest in south and east Asia. The suitability of land for cropping declines in most regions, but increases at the northern boundary of cropland and along some margins in east Asia. Spring wheat yields show a mixed pattern of change, maize yields decline everywhere except in parts of north America and eastern China, and soybean yields tend to increase in parts of south and east Asia, north America and small parts of south America, but decrease elsewhere. Increases in coastal flood risk are concentrated in Asia and eastern Africa, whilst wetland losses focus around the Mediterranean and north America. Cooling energy demands increase particularly in regions where there is currently little demand for cooling, but increase only slightly in some warm regions—because the relative change in requirements is smaller. Heating energy demands decrease most in the warmest regions.
Many regions are exposed to multiple overlaying impacts. For example, under this plausible climate story river flood risk increases across much eastern Asia, coastal flood risk increases substantially, and cooling energy demands increase by more than 70 %. At the same time, the productivity of the three example crops increases in parts of eastern Asia, but decreases across much of northern China. The suitability for agriculture appears to increase in northern and western China, although soil organic carbon contents decline (in this case because conversion of forest to arable land reduces the inputs of carbon from vegetation).
In southern Asia, crop suitability declines, productivity of maize declines but soybean productivity increases (in some parts). River flood risk increases and some coastal megacities see increased flood risk. Cooling energy demands rise by around 30–40 %, but there is little change in heating demands. Water scarcity reduces under this story across many water-scarce parts of southern Asia.
The suitability of cropland for crop cultivation declines across much of sub-Saharan Africa, primarily due to reductions in available moisture; more than 90 % of cropland in southern Africa would see a reduction in suitability for crop production. Maize yields reduce by 20–40 %. River flood risk increases substantially in parts of western Africa, and coastal flood risk increases in particular for many east African coastal cities. Across the Middle East and North Africa crop suitability declines and large populations are exposed to increased water scarcity and increased cooling energy demands; NPP also reduces in many parts of the region.
Within western and central Europe, river flood risk is little affected under this story, but around 200 million people are exposed to increased water resources stress. Crop suitability increases in the north of the region but declines elsewhere, and spring wheat productivity declines across much of central and eastern Europe. Cooling energy demands are increased very significantly—from close to zero in northern Europe—but heating energy demands fall by at least 40 %.
Under this story, the main potential impacts in North America appear to be reductions in crop suitability across much of western and central North America, but increases at the northern margins of agriculture, and mixtures of increases and decreases in crop yields. Cooling energy demands increase very significantly in the eastern parts of North America, where heating energy demands fall. Coastal wetland loss is particularly large along the west coast.
Across South America, maize and soybean yields fall and NPP decreases substantially across the Amazon basin; the suitability for cropping declines in the drier parts of eastern south America, but increases along parts of the west coast.
The impacts plotted in Fig. 1 and tabulated in Table 3 would arise under one particular plausible climate future. In principle it is possible to produce similar stories under other climate models. Table 4 shows the global aggregated impacts for each indicator under another six climate models (and they should be compared with the global row in Table 3). Supplementary Figs. 2-7 show the distribution of impacts under six more climate model patterns, and Supplementary Table 3 presents regional impacts under all 21 climate model patterns used.
Table 4 Global impacts in 2050, under the A1b emissions and socio-economic scenario, under six plausible stories. Compare with the global impacts in Table 3
Uncertainty in projected regional impacts
The uncertainties in regional impacts, by sector, are given in Table 5, which shows the range in estimated impacts across the climate models used (which range from 21 for most indicators to 7 for SOC). Fig. 2 summarises the regional uncertainty in impacts.
Table 5 The range in regional and global impacts in 2050, under the A1b emissions and socio-economic scenario, across all climate models
For most impact sectors, the projected ranges are very large. In some cases—specifically the water and river flooding sectors—this is because of very large uncertainty in projected changes in regional rainfall (in south and east Asia, for example). In some other cases, the large uncertainty is because the sector in a region is particularly sensitive to change (for example where the baseline values in the absence of climate change are small—see forest and NPP change in west and central Asia). In other cases, the uncertainty range is dominated by individual anomalous regional changes. For example, the large range in estimated additional people exposed to coastal flooding is due to one particular climate model producing very considerably higher sea level rises in some regions than the others. There is least uncertainty in projected reductions in heating energy requirements and, for most regions, increases in cooling energy requirements; the greatest uncertainty here is in those regions where requirements are currently low—Europe and Canada—but the percentage changes are sensitive to small changes in temperature.
The considerable uncertainty in each region and sector needs to be interpreted carefully. It is not correct simply to add up the extremes of each range across regions and use this to characterise the global range; the global range will be smaller than the sum of the extremes because no one climate model produces the most extreme response in every region. Similarly, it is not appropriate to define the maximum impact across all sectors in a region as the sum of the maximum impacts for each sector, because again no one single climate model produces the maximum impact in all sectors. Indeed, there are some associations between impacts in different sectors between climate models. For example, models which produce the greatest increase in exposure to water resources stress tend to be those which produce the smallest increase in exposure to river flooding, and the greatest area of cropland with a decline in suitability (see Supplementary Fig. 8 for an example).