By: Tristan Quaife
The Paris Agreement, which is signed by 193 countries belonging to the United Nations Framework Convention on Climate Change, aims to limit the rise in global mean temperature to 2°C, and ideally 1.5°C, above pre-industrial levels. To achieve either of these goals will require a significant slowdown in the rate of increase of the concentration of greenhouse gases in the atmosphere. In particular, this means reducing carbon dioxide (CO2) emissions caused by burning fossil fuels, the production of cement and land use changes such as forest clearance. There are also methods proposed for removing CO2 from the atmosphere, such as increasing tree planting, but it is highly unlikely that we can meet the Paris targets without also making significant reductions in our emissions. This blog post explains some of the context.
The current atmospheric carbon dioxide concentration is around 420 parts per million (ppm), compared to a preindustrial level of around 280ppm. Since the late 1950’s there have been regular, high-quality measurements made of atmospheric carbon dioxide concentration at the Mauna Loa Observatory in Hawaii. A plot of these data is shown in Figure 1. By comparison with other observations taken around the world we know that the Mauna Loa record is representative of the global concentration. Note that at the start of the time series the data are not far off the pre-industrial level, hence the majority of the increase takes place in the last 60 years of human history.
Scientists also have a good understanding of the sinks and sources of this atmospheric carbon. There are reliable estimates of emissions from fossil fuel burning and cement production which are acquired from a variety of sources, including economic data. Emissions from land use change are far less certain, but represent a smaller, albeit still significant contribution. We also have a high level of confidence from different observational data sets and modelling studies, of how much carbon is being taken up by the oceans and land.
Figure 1: The Keeling curve; atmospheric CO2 concentration measured at Muana Loa, Hawaii. Image produced by the Scripps CO2 programme
Climate models are a key tool for helping us interpret what the likely impact increased CO2 levels will have on global temperatures, and what the onward impacts of that will be (such as, for example, the associated increase in sea level rise due to melting of ice sheets). In a recent study, led by Ranjini Swaminathan from the University of Reading and the UK National Centre for Earth Observation, we examined the data from a large number of climate models, run under different socio-economic scenarios (the “Shared Socio-economic Pathways” or SSPs for short) to see in which years different global temperature thresholds would be crossed by the models. The results are shown in Figure 2. It’s a complex plot but contains lots of useful information. The temperature thresholds are on the y-axis, and the different SSPs are indicated by the different coloured bars. Each dot represents the year at which an individual climate model passes the corresponding temperature threshold for a given SSP. The white boxes show the median year of threshold exceedance based on all the models running that scenario. The column on the right explains how many models (if any) do not exceed the threshold.
Figure 2: Year of threshold exceedance for CMIP6 models (individual dots), for different Shared Socioeconomic Pathways (indicated by colours). The number of models that stay below the given threshold is shown on the right-hand side. Taken from Swaninathan et al. (2022).
Focusing on the 1.5°C and 2°C thresholds, we can see some interesting outcomes of this exercise:
- Under all SSP scenarios we tested there are some models that exceed 2°C.
- The majority of models predict we will exceed 1.5°C even under SSP1-1.9.
- It is only SSP1-1.9 where the majority of models don’t exceed 2°C.
- Other than for the two SSP1 scenarios all models exceed 2°C.
The significance of this is that the SSP1 scenarios, which carry the subtitle “Sustainability – Taking the Green Road”, all represent significant reductions in fossil fuel emissions, a transition toward greater use of renewables and rapid increases in technology for mitigation and adaptation. SSP1-1.9 is the most optimistic of all scenarios and sees the world moving to net-zero carbon emissions by mid-century. There is an excellent explainer of the different SSPs here if you want to read more here.
An important question to ask, then, is where our knowledge of the actual CO2 concentration from the Mauna Loa observations, and other data, puts us in relation to the different SSP scenarios. This is illustrated in Figure 3, which shows the CO2 concentrations corresponding to the different SSPs with the Mauna Loa CO2 observations overlayed. The observation data are the same as in Figure 1, but with the seasonal cycle removed to make the comparison with the SSPs clearer. In one sense, there is not much to see here – the SSPs start in 2015, and so they haven’t diverged greatly yet and the CO2 observations sit nicely on top. But this means we are still in control of whether or not we take the route implied by SSP1-1.9, and have a fighting chance of avoiding a 2°C rise (and maybe even a 1.5°C rise). But if we want to do this so we must reduce the rate of emissions quickly over the next 30 years.
Figure 3: Atmospheric CO2 concentrations for the different SSPs and from the Muana Loa observations. SSP data from Meinshausen et al. (2020) and observational data from here.