Invisible Ship Tracks Produce Mean-State Cloud Microphysics Perturbation in Tropical Trade Cumulus

[Geoengineering News]

https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2025GL121160

Authors: Chris J. Wright, Jihu Liu, Yang Cao, Yannian Zhu, Daniel Rosenfeld, Robert Wood, Joel A. Thornton

First published: 29 July 2026

https://doi.org/10.1029/2025GL121160

Abstract

Aerosol effects in cumulus-topped boundary layers remain relatively unconstrained and drive uncertainty in climate modeling. We assess the impacts of shipping aerosol on trade cumulus, combining satellite retrievals and large eddy simulations (LES). Over the Bay of Bengal shipping lane, sparse cumulus cloud fields rarely produce visually identifiable plume features embedded in clouds. That is, the ship tracks are generally “invisible.” Nonetheless, we show using MODIS Aqua that this shipping lane produces a shift in the probability distribution of cloud droplet number (Nd), constituting a strong mean-state perturbation in trade cumulus composed entirely of invisible ship tracks. Liquid water path and cloud fraction adjustments are undetectable. An LES ensemble of the shipping lane shows similar sensitivity of Nd and lack of consistent domain-average adjustments. We also discuss the limitations of using aerosol mass in ACI parameterizations, and evaluate the radiative response to ship emissions and fuel regulations.

Plain Language Summary

Emissions of particulate from human activity brighten clouds, which has partially masked the effect of global warming. The magnitude of this masking is a large source of uncertainty in climate models. Here, we use an “opportunistic” experiment of a highly polluted shipping lane and the regulation of that pollution to determine the effect of pollution on an understudied type of cloud: trade cumulus. We find that the number of cloud droplets, which is strongly related to cloud brightness, is strongly affected by the ships prior to the regulation of sulfur emissions. Other changes to the cloud, such as cloud coverage and thickness, are not affected. High resolution simulations show good agreement with observations.

Source: AGU