Evaluating simulations of ship tracks in a km-scale model

[Geoengineering News]

https://acp.copernicus.org/articles/26/4251/2026/acp-26-4251-2026.html

Authors: Anna Tippett, Paul R. Field, and Edward Gryspeerdt

26 March 2026

Abstract

Clouds, and in particular their adjustments following an aerosol perturbation, remain a major source of uncertainty in climate projections, due to the wide range of scales over which cloud processes act on. This uncertainty limits our capability to simulate potential solar radiation management strategies, such as marine cloud brightening (MCB). A “natural”, or “opportunistic”, experiment for investigating MCB is analysis of ship tracks, as they mimic the intended effect and allow us to investigate time evolving cloud adjustments. In this study, we model a real case of ship tracks and evaluate model performance in representing the lifetime of cloud adjustments through comparisons with satellite observations. Requiring accuracy in individual cases creates a particularly challenging test for simulations of aerosol-cloud interactions, but it is necessary to ascertain whether this model is suitable for simulating MCB accurately.

Our findings highlight a key deficiency in activation parameterisations when simulating high aerosol number concentrations – such as those expected in MCB scenarios. While the model can replicate the mean cloud properties within ship tracks, it struggles to capture the temporal evolution of the adjustments. Specifically, in precipitating clouds, both the enhancement in droplet number concentration (Nd) and subsequent adjustment to liquid water path (LWP) are overestimated and persist for too long. This discrepancy between model and observations is linked to excessive model sensitivity to Nd perturbations in precipitating conditions, leading to unrealistically strong suppression of drizzle, and ultimately resulting in simulated ship tracks which overestimate the cooling effect in these cases. We identify scenarios in which current formulations of parameterisations are not suitable for use in simulating high-concentration aerosol perturbations, such as MCB, and scenarios in which models are more capable.

Source: EGU