Effects of artificial heat added by sea ice perturbation methods

A common approach to assessing how polar amplification affects lower latitude climate is to perform coupled ocean-atmosphere experiments in which sea ice is perturbed to a future state. England et al. criticize this approach (England et al., 2022, to appear in J. Climate), using a simple 1-dimensional energy balance model to show that sea ice perturbation experiments do not isolate the true effect of sea ice loss on the climate, because they add an artificial heat flux required to melt sea ice and keep it in the future state. They thus show that the temperature response in sea ice perturbation experiments is greater than the true temperature response to sea ice loss alone, which can be directly diagnosed in the energy balance model. This suggests a similar overestimation of the response to sea ice loss in perturbation experiments in comprehensive models. We extend these results. First, we simplify the additional heat argument to its essential elements, making it clear why sea ice perturbation experiments must include artificial warming. Second, we evaluate the effect of additional heat in an energy balance model that includes moist energy transport, a key process in modulating polar climate change that is missing in the England et al. study. Third, we make initial strides in assessing spurious effects in a comprehensive model sea ice perturbation experiment. We show that the artificially large Arctic warming in sea ice perturbation experiments implies an artificially large meridional gradient in the surface temperature response to sea ice loss. Similarly, we show that the sea ice perturbation experiment overestimates the midlatitude jet response to sea ice loss. We finish by discussing how comprehensive model experiments could be designed to more thoroughly quantify the effect of additional heat.