Uploaded January 2017 | Updated November 2026, 17 hours ago
This talk was held at the Cambridge Conference on Catastrophic Risk on December 12-14, 2016.
In a recent talk at the Royal Society, climate sceptic Matt Ridley proposed a novel trichotomous characterisation of the climate-change problem, as outlined in the title above. He described himself as a “lukewarmer”. However, by virtue of the complex nonlinear nature of climate, such a categorical delineation of the problem is inappropriate. Rather, to be consistent with the nature of key uncertainties in estimating how increasing levels of atmospheric carbon dioxide interact with the natural climate system, notably with the hydrological cycle, estimates of future climate should instead be framed in terms of probability and risk: What, for example, is the risk of catastrophic climate change (the sort that would pose an existential threat to significant sections of humanity) in the coming century or so? It is simply inconsistent with our present state of knowledge to conclude that this risk is negligible. Indeed, according to both observations and modelling, this risk lies in the 10s of per cent, and is therefore substantial by any relevant measure.
The biggest uncertainty preventing a more categorical estimate of future climate lies in the problem of cloud feedback: How will low- and high-level cloud respond to increasing levels of carbon dioxide? Current computers are not powerful enough to allow clouds to be resolved and they must therefore be represented by quasi-empirical parametrisation formulae. Here I call on the climate sceptic community (including, perhaps, the incoming US Administration) to acknowledge this risk-based approach to the assessment of future climate and to broadly accept IPCC estimates of risk (whilst recognising that these estimates can and must be improved). I call on economists and climate scientists to work closer together to provide the most reliable estimates possible of expected climate impacts. To aid such work, I call on European countries to come together to fund a dedicated (exascale) flagship climate-computing centre – a sort of Climate CERN – which would allow the resolution of global climate models to be driven down to scales where key cloud systems can be represented using the laws of physics. Such a centre is essential if we are to understand and quantify better the nature of climate extremes.
This talk was held at the Cambridge Conference on Catastrophic Risk on December 12-14, 2016.
In a recent talk at the Royal Society, climate sceptic Matt Ridley proposed a novel trichotomous characterisation of the climate-change problem, as outlined in the title above. He described himself as a “lukewarmer”. However, by virtue of the complex nonlinear nature of climate, such a categorical delineation of the problem is inappropriate. Rather, to be consistent with the nature of key uncertainties in estimating how increasing levels of atmospheric carbon dioxide interact with the natural climate system, notably with the hydrological cycle, estimates of future climate should instead be framed in terms of probability and risk: What, for example, is the risk of catastrophic climate change (the sort that would pose an existential threat to significant sections of humanity) in the coming century or so? It is simply inconsistent with our present state of knowledge to conclude that this risk is negligible. Indeed, according to both observations and modelling, this risk lies in the 10s of per cent, and is therefore substantial by any relevant measure.
The biggest uncertainty preventing a more categorical estimate of future climate lies in the problem of cloud feedback: How will low- and high-level cloud respond to increasing levels of carbon dioxide? Current computers are not powerful enough to allow clouds to be resolved and they must therefore be represented by quasi-empirical parametrisation formulae. Here I call on the climate sceptic community (including, perhaps, the incoming US Administration) to acknowledge this risk-based approach to the assessment of future climate and to broadly accept IPCC estimates of risk (whilst recognising that these estimates can and must be improved). I call on economists and climate scientists to work closer together to provide the most reliable estimates possible of expected climate impacts. To aid such work, I call on European countries to come together to fund a dedicated (exascale) flagship climate-computing centre – a sort of Climate CERN – which would allow the resolution of global climate models to be driven down to scales where key cloud systems can be represented using the laws of physics. Such a centre is essential if we are to understand and quantify better the nature of climate extremes.










