This is the fourth post in a series of posts in which I record and reflect on the University of Exeter’s massive open online course (MOOC). This week’s focus is on climate modelling.
How climate models work
Professor Peter Cox of Exeter, formerly of the UK Met Office, explains that the same models that are used to make the weather forecast are also used to predict the effects of climate change. Except for climate change, slower and less obtrusive processes (performance of carbon sinks, for example) have to be taken into account.
Each step of a climate simulation takes about 20 minutes to run on a super computer. Running a simulation that covers, say 250 years, takes about three months.
Are climate models accurate?
Scientists have tested the models against past climate, and they have been shown to be accurate. However, these models, when run forward into the future, failed to predict observed warming – in fact they predicted a gradual cooling. This was because human interference with the climate had not been factored in – mainly burning of fossil fuels and deforestation.
Says Prof Cox: “So in the case of climate models, we try to simulate how the globe has been known to warm by about 0.8 degrees Celsius since the mid-19th century, and we do that by including the factors we think might have affected climate. These include natural things, like how the sun has varied in its output, and also when volcanoes go off because they tend to cool down the climate system. But we also include human factors, particularly the increasing carbon dioxide that’s been associated with burning fossil fuels and deforestation. When we put in the natural factors, we find that the models can reproduce aspects of the climate, but only until about 1970. From then onwards, the actual warming of the climate and what the models simulate diverge. And in fact, the models tend to have the climate cooling when it should be warming.
“But when we put in the human factors, particularly carbon dioxide increase, we find that models reproduce the observed warming extremely well.”
In looking into the further future, there are so many variables, scientists use “scenarios” – for example, if the world we to cut emissions drastically, that might limit global warming to 2C, but under other scenarios, warming is at 6C over the next 100 years – the same temperature difference between the Ice Age and now, only happening 10 time faster.
Different climate scenarios
The IPCC uses different emissions scenarios to predict future global warming and climate change impacts. These scenarios are called RCPs – Representative Concetration Pathways. For example, under RCP 2.6, under which huge radical measures are taken to cut emissions and emissions gradually tail off to zero, temperatures gradually return to pre-industrial levels. However, under RCP 8.5, emissions continue on their upward trend (a business as usual scenario), the temperature by 2100 will be 2.6C to 4.8C above 1986-2005 levels (which in turn are 0.6C warmer that 1850-1900).
The course directs students to an article by NASA’s Earth Observatory which draws on data from the IPCC’s AR4 of 2007, suggesting that “Based on a range of plausible emission scenarios, average surface temperatures could rise between 2°C and 6°C by the end of the 21st century.”. However, in IPCC AR5 (2014) the evidence of warming hardened and the scenarios became more dramatic.
Climate feedbacks
The climate models try to take account of various unpredictable climate feedbacks. Chief among these are snow and ice feedback (less ice means less reflectivity and more warming), water vapour feedback (warming leads to more vapour, which leads to more warming), cloud feedback (clouds have cooling effect, but this could change if warming continues, and they could cause warming, which would cause more clouds etc), the carbon cycle (at the moment, oceans and land are carbon sinks, but they could become sources of CO2).
Geoengineerging
There are several ways – all based on sound scientific principles – that we might try to stop global warming by technological means. Some involve increasing the reflectivity of the earth’s surface by making clouds brighter or injecting aerosols into the upper atmosphere. Others involve various means of carbon removal, capture and storage. We could geoengineer the ocean so that it absorbs more CO2; we could plant more trees; we could turn power stations away from coal and towards biomass; we could pump CO2 into the empty Yorkshire coal mines (yes, this is being seriously considered by Drax power company in the UK).
However, there are problems. First, geoengineering can stop emissions, but cannot reverse the damage already done. Second, it might be able to alleviate climate change impacts in broad regions, but would not be able to prevent country-to-country variations – there would be winners and losers, with predictable geopolitical repercussions. Also, if the geoengineered solutions failed, there would be a sudden increase in CO2 which most living organisms would not be able to survive. Other than that, sounds like a winner.
Two Degrees 
Pingback: Climate change basics 8: the anthropocene – Two Degrees