Observations of Weak Liquid Water Path Response to Aerosols in Ship Tracks.

Oral Presentation

The assessment of aerosol-cloud interactions remains a major source of uncertainty in understanding climate change. Ship emissions can artificially brighten clouds through the injection of aerosols, which act as cloud condensation nuclei. These “shiptracks” can be used to evaluate the impacts of aerosols on clouds, whilst simultaneously disentangling any effects of the surrounding meteorology. Previous research has relied on analysis of shiptracks visible in satellite imagery, whereas recent work in predicting shiptrack locations by advecting ship positions in ERA5 winds has allowed for analysis of a greater number of shiptracks. This work has suggested that there are many shiptracks in which the liquid water path (LWP) is increased. 

However, analysis of a counterfactual scenario in which shiptrack locations from the wrong year are used suggests that there is a bias in the way enhancements are calculated, resulting in false positive signals in the response of cloud liquid water content. In assuming that background gradients in the cloud properties are linear, when they are in fact non-linear, the comparison of “outside” and “inside” track regions can yield false positive enhancements even in counterfactual situations where no shiptracks are present. This bias is intricately linked to the correlation between the winds used to predict track locations and cloud properties. Our research focuses on unravelling and correcting this bias, aiming to reveal the true time evolution of a cloud’s response to an aerosol perturbation. 

By correcting for this bias, we present a more accurate depiction of the shiptrack response, showcasing that, on average, the LWP in shiptracks is largely unchanged from that of the surrounding clouds. In some cases, however, a LWP response can be recovered. This work emphasises the need for careful consideration of counterfactuals in shiptrack analyses and reveals a bias that is otherwise difficult to detect.