The consequences of climate change: how bad could it get?

By: Nigel Arnell

The United Nations Climate Action Summit held in New York on 23rd September was meant to be the occasion where countries and industry organisations made stronger commitments to reduce the emissions of the greenhouse gases that are causing global warming. Whilst 65 countries pledged to achieve net-zero emissions by 2050, the overall commitment – according to some commentators – fell “woefully short” of expectations. Most news headlines after the event focused on Greta Thunberg’s speech where she passionately challenged world leaders to do more: “the eyes of all future generations are upon you”.  Meanwhile, campaign groups such as Extinction Rebellion warn of ‘unprecedented global emergency’, ‘climate breakdown’ and ‘mass extinction’. The best-selling author David Wallace-Wells writes of ‘The Uninhabitable Earth’.

So what are the consequences of climate change, and how bad could it get? The impacts of climate change in the future depend not only on how climate – temperature, precipitation and so on – changes, but also on how societies and economies change. Our estimates of changes in climate depend on two things. First is a projection of how emissions will change. This depends on how economies and energy use change, and we cannot predict this: it will depend on policy choices, such as the actions to reduce emissions announced in New York last week. We, therefore, use ‘scenarios’ to describe plausible changes in emissions, but these should not be seen as predictions. The second part is where climate science fits in. Projections of the effect of an emissions scenario on changes in local weather are made using climate models. Whilst all models produce broadly similar changes in climate (temperatures increase in high latitudes more than in tropical regions, wet areas tend to get wetter and dry areas tend to get drier), the detailed projections can vary considerably between climate models. We can estimate the consequences of these changes in local weather for local climate hazards and resources – such as floods and droughts – using separate ‘impacts’ models.

Figure 1: Change in global heatwave, drought and flood hazard through the 21st century under three plausible emissions scenarios, and numbers of people affected in 2100. The five estimates of people affected for each emissions scenario are from five different socio-economic scenarios. See Arnell et al. (2019) for specific definitions of the indicators.

In practice, policymakers and others want to know the human impacts of climate change. We can estimate the direct impacts – such as the number of people affected by flooding, droughts or heatwaves – by combining our estimates of the physical changes in climate with socio-economic scenarios describing plausible changes in population and the economy. Figure 1 shows the global direct impacts of climate change on exposure to heatwaves, river flooding and drought (from Arnell et al., 2019). The left panels show changes in the physical hazard through the 21st century, under different emissions. The range in estimates for each emissions scenario shows the uncertainty due to different model projections of local temperature and precipitation. By 2100, the chance of experiencing a major heatwave has risen from 8% to 100%, the average proportion of time in drought has gone up from 6 to 30%, and the chance of a flood has increased from 2% to 7% (in some regions the changes are greater). The right-hand panels show the human impacts in 2100, under three of the emissions scenarios and for five socio-economic scenarios. The biggest impacts – in terms of numbers of people affected – are obviously with the highest emissions, but the differences between the socio-economic scenarios can be very large. When presenting such results we often focus on the central estimates (the solid lines in the figures), but we could instead look at the ‘worst-case’ impacts at the top end of the distribution. These can be a lot higher than the central estimates.

These direct impacts have knock-on consequences: changes in the frequency of droughts, for example, could plausibly lead to loss of livelihoods, food insecurity, political instability and displacement of people. It is these potential ‘systemic’ risks that are of greatest significance for policymakers and indeed are behind many of the most extreme warnings about climate emergencies and crises. However, these knock-on consequences depend not only on the physical changes in climate and future socio-economic scenarios that we can model but also – and largely – on how societies and governments react and behave. In order to estimate the likelihood of the real ‘worst-case scenarios’ that are ringing the loudest alarm bells we, therefore, need to link the work of climate scientists, impact modellers and experts in institutions, governance and human behaviour.


Arnell, N.W. et al. (2019) The global and regional impacts of climate change under Representative Concentration Pathway forcings and Shared Socioeconomic Pathway socioeconomic scenarios. Environmental Research Letters 14 084046.  doi:10.1088/1748-9326/ab35a6

Wallace-Wells, D. (2019) The Uninhabitable Earth: a story of the future. Penguin: London

This entry was posted in Climate, Climate change, Climate modelling. Bookmark the permalink.