Q: Climate modelling is full of uncertainties — how do these affect the science community’s ability to nail its projections to the mast?
A: The temperature projections made back in the early ’90s were done with somewhat rudimentary models. Now it seems those early projections were more extreme because they didn’t include all the necessary processes. That means projections dating to the early ’90s — such as a likely 2°C increase for the planet by 2050 — are now more likely to materialise towards the latter half of this century, if no action is taken to reduce emissions. There’s a lot of empirical support for what these models do and they’ve been quite extensively reviewed, including by the United Nations Intergovernmental Panel on Climate Change (IPCC).
More recently even, the latest set of models from the IPCC have included new work on improving estimates of climate sensitivity. Early results suggest that they’re projecting a warmer future sooner than the previous round five or six years ago.
Q: So what’s making the latest batch of models more climate-sensitive, and therefore, more useful?
A: As I understand it, the new models incorporate energy feedback from clouds more efficiently and more realistically. The net effect is that a more responsive warming to rising CO2 is being projected than before. In fact, we have a good idea from hundreds of thousands of years of ice core data that there’s a good correlation between CO2 concentration and atmospheric temperature.
So, that’s not very good news for us, but it’s a moving target and climate science is evolving. For instance, I think it’s fair to say that we don’t fully understand yet how responsive the atmosphere is to CO2, but we do know with virtual certainty that rising CO2 is causing warming.
Q: But surely all this is grist to the denier’s mill — extreme projections in the early ’90s, cooler projections around 2013/14 and now we’re looking at a warmer future again. And yet, scientific credibility is the key to whether or not people acknowledge that nature is dangerously close to ejecting Homo sapiens from the mothership.
A: I don’t think any modeller would claim their model is 100% spot-on. Models try to simplify and synthesise complex processes so that they can come up with useful predictions. Models are either useful or not useful, but they’re never perfect — the complexity of nature is beyond what models can capture in all details.
In the physics of climate, it’s understood that we cannot precisely predict the future because climate has a characteristic called “deterministic chaos”. To get a perfect prediction, you must measure everything with absolute precision and, even if you’re out by, say, 0.000001% in the measurement of so-called “initial conditions”, the model will deviate from reality after a certain amount of time.
Therefore, it’s not possible to model the weather precisely over long periods — modelling hurricane tracks is a lovely, real-world example. That’s why we need to consider the issues very carefully and take lessons from the results. Models are getting better, though. We’ve all experienced more accurate predictions of weather over the timescale of days and even weeks — and that’s part of research to reduce the uncertainties.
Q: How well are the models serving us on the long-term outlooks?
A: Pretty well. Take the simple ’80s models of, say, the former director of Nasa’s space studies institute Jim Hansen [today Columbia University climate programme director]. He came up with three scenarios — high, middle and low — and here’s an example of how iniquitous the denialist fraternity are, because they always highlight and cite his high scenario and say, “Well, he completely overestimated climate warming.”
But his mid-level scenario has been really useful. Hansen even included the assumption of an El Chichon-sized volcanic eruption in the mid-’90s. He wasn’t predicting a volcanic eruption, he was simulating an uncertainty: What if there was a significant eruption? How would it affect the projections? How much time might it buy us? After all, big volcanic eruptions can cause hemispherical or even global cooling for a few years, because volcanoes emit sulphate aerosols, which create a “sunscreen effect” that reflects radiation back into the atmosphere.
In fact, we did have an eruption — Mount Pinatubo in the Philippines in June 1991. You don’t often get volcanoes of a size big enough to cause such a substantial sunscreen effect. They only occur every few decades at best. So it was quite an extraordinary projection by Hansen. Certainly the newer models have done a very good job of predicting the warming we would have expected back in the ’90s, so that denialist talking point doesn’t hold any water.
Q: So if denialism were my side-hustle — which seems to be the case with many of the deniers — how could I go about poking holes in the trends?
A: Fluctuations caused by the energy fluxes between atmosphere and ocean are popularly used by the denialist fraternity to cherry-pick and cast doubt. When we get an El Niño event, which changes the circulation of water in the Pacific, it releases a lot of that energy back into the atmosphere. During El Niño years you tend to get above-average warm temperatures; during La Niña years the ocean tends to absorb energy out of the atmosphere.
In 1998 we had a strong El Niño, which made it a very significant El Niño year — and that was the hottest year on record at that time. Even without factoring in El Niño, we’ve repeated the temperature record many times since then. But that’s where this idea of a “warming hiatus” between the end of the ’90s and 2012 came from. To even have a snowball’s hope of showing that, you have to start your record in extremely hot 1998 and trace it through to 2012, where you had a bit of a La Niña. Temperatures didn’t cool, they simply plateaued for a while, but this “hiatus” was still the warmest period since records began.
Q: Right, but that’s just one period. Why fixate on that?
A: Looking at the long-term warming trends by drawing a line through a 40- or 50-year period, and considering the trends since 2012 where we still have most of the hottest years on record since 1998, you can see the denialist-supporting wiggle in all this, and that’s not entirely surprising. Earth-system science is such a complex and multi-faceted area and there are all sorts of tricks you can use to confuse people — fossil fuel emission uncertainties and clouds and ocean circulation and model sensitivity.
But none of the sceptics will take the record from 1991 to 1998 and say, “Ooh, global warming’s accelerated.” They’d rather take the period from 1998 to 2012 and say, “Ah! Global warming has gone away!” We call that “cherry-picking”. If somebody makes that argument, either the person genuinely doesn’t understand the science, or they do understand the science and they’re trying to fool you. For both of those eventualities that person is not worth listening to as an expert. Discount it. It’s just junk. Mendacious. Completely false. Puerile.
Q: Climate projections seem to do fairly well when considering large swathes of the planet, but struggle on a granular level. Why the gap?
A: On a broader scale it’s certainly more feasible to come up with general conclusions about climate change risks. For example, in the northern hemisphere the preponderance of evidence shows that wild species have been shifting their geographic ranges — the areas where they naturally occur — towards the pole and upwards in elevation when that’s available to them, because they’re tracking their preferred “climate space”; and that the quantum of those range shifts matches the predictions from the climate models.
There are people who say the temperature records have been fiddled with to show global warming and to argue that point you have to argue that the wild species in the northern hemisphere have also been duped by that fraud because they’re all responding in the way they should [laughs].
Q: What happens when you get more specific and local?
A: Then the uncertainties pile up so much that it becomes hard to say anything specific. To predict impacts at a particular point climatically starts to make little sense because there are all sorts of things that could push that prediction out of whack.
For instance, you could say that the species Protea Magnifica occurs with a 100% likelihood in the fynbos biome, but it becomes impossible to pinpoint the certainty of its occurrence within a 10m2 range. With respect to rainfall projections, however, one of the few exceptions at the regional scale is the Western Cape, where the models predict the entire westerly frontal systems are likely to start moving south, thus creating conditions of a greater frequency of drought in this part of the world.
This also holds for other Mediterranean-climate areas of the world like Western Australia, South America, California and the Mediterranean itself. This is one of the most consistent and well agreed weather-system shifts that occurs. The circulation from the poles to the tropics is strengthened and it pushes the fronts more towards the south.
Q: To what extent do these extreme individual events create honest signals of larger climate breakdown?
A: There’s no question that the trends we’re seeing in rainfall indicate much more intense events, but, interestingly, also longer dry spells between them. So, even where you don’t get a change in rainfall, you’re seeing a concentration of rain into shorter, sharper events with flooding — and then the dry spells in between can seal off the soils, so you get this increased run-off effect.
We’ve seen flooding in Joburg, more broadly in Gauteng, and KwaZulu-Natal, for example. This terrible recent sequence of cyclones that hit Mozambique has also brought home the reality of these risks. And we’re only in 2019. Let’s say this doesn’t bode well. If extreme weather does occur over the next few decades, we’re going to have to be ready.
Look at the drought in Makhanda. Cape Town nearly ran out of water. The likelihood of a Cape Town drought was something we identified in 2000 when we were doing some work for local government and looked at rainfall records going back to the 1920s. We saw this very significant natural drought in the 1920s and 1930s and it was obvious that, with warmer temperatures and more people, Cape Town could run into problems.
We’re now seeing these extremes starting to bite. Elsewhere in the world, we’ve seen hectic heatwaves in Australia leading, for example, to mass mortality of fruit bats. Europe’s heatwaves have been extraordinary. The melting of Greenland ice is unprecedented. You’re looking beyond South Africa’s shores. And there are all sorts of signs that the system is starting to show all the signs of this shift into a much warmer world. DM
Professor Midgley is also co-author of leading South African climate change studies dating back to the year 2000, and was awarded Germany’s Humboldt Research Award in June. According to Stellenbosch University, “the award is granted in recognition of a researcher’s entire achievements to date and to academics whose fundamental discoveries, new theories or insights have had a significant impact on their own discipline and who are expected to continue producing cutting-edge achievements in the future.”
"Thou almost make me waver in my faith to hold opinion with Pythagoras" ~ Shakespeare, Merchant of Venice