Tackling the Arctic Cloud Problem

The Arctic surface is warming faster than anywhere else on the planet. There is an urgent need within the climate science community to understand why this is happening and to predict how this warming will continue in the future. However, the global computer models we use to make these predictions cannot even reproduce the temperature increase that we have already witnessed.

These models are particularly bad at reproducing Arctic clouds. Arctic low-level clouds are long-lived, yet models fail to reproduce their long lifetime properly. Clouds play a key role in our climate by reflecting solar energy and trapping heat from the surface, and these persistent Arctic clouds are likely contributing to the surface warming we have seen in the region. If we cannot reproduce these clouds properly in our models, we will not be able to confidently predict how much the Arctic surface will continue to warm in the future. For this reason, we need to develop new ways to model Arctic clouds which are appropriate for the unique polar environment. Model shortcomings are largely because of the way we represent internal processes within the clouds on very small scales; scales much smaller than a global climate model can show. To represent such small-scale processes in large-scale models, we use equations built from measurements of clouds taken primarily at the mid-latitudes. These mid-latitude equations are not suitable for polar clouds, and we need new Arctic-relevant versions to capture these clouds correctly in our models. A significant modelling challenge is the lifetime of these Arctic clouds. In our models, we find that Arctic clouds disappear too quickly, in a matter of hours, when they should survive for up to several days. To improve on this failing, we need to study these clouds on small scales and understand how they interact with their environment. Such scales cannot be achieved with low-resolution global models, so we need to use different high-resolution tools for this task. However, there are a number of reasons why these high-resolution models cannot be used to model clouds across large scales, like the entire Arctic. This is a common challenge in cloud science, as the divide between investigating important small-scale interactions at high resolution and studying the climate impact of clouds across an entire region presents a significant problem in our effort to improve predictions of their role in future climate change. During this Fellowship, I will address this problem with a focus on clouds in the Arctic. I will improve the equations representing small-scale cloud processes by using two linked computer models which use the same methods to represent clouds but work on very different scales. One operates at high resolution and can capture small-scale dynamics in detail, while the other can be used to model the atmosphere around the globe in lower resolution. The link between these models will allow me to study and adapt small-scale processes in the first model then directly test how these changes affect clouds modelled across the entire Arctic with the second model. My experience working with both models will enable me to develop new, more realistic, equations to model polar cloud processes, using measurements recently obtained in the Arctic to indicate how these processes operate in reality. Ultimately, we cannot confidently predict future Arctic warming using global computer models without tackling this Arctic cloud problem.

Grant reference
Natural Environment Research Council
Total awarded
£555,528 GBP
Start date
1 Jan 2022
4 years 11 months 30 days
End date
31 Dec 2026