A Case Study on Ship Tracks in the Bay of Biscay

The impacts of aerosol perturbations on cloud microphysical structures and the cloud radiative response (aerosol indirect effects) have been studied intensively since the mid 20th century and yet remain one of the largest uncertainties of climate sensitivity.

Ship tracks, which are characterised by recurring streak-line patterns of increased cloud albedo under certain environmental conditions are often used as an illustrative example for such aerosol-cloud effects. They are predominantly observed in stratocumulus regions with relatively shallow boundary layers along the worlds major shipping corridors. Although the radiative effect of ship emissions on a global scale is still under debate, they indeed provide an ideal test bed for studying aerosol-cloud effects of warm clouds. The ships exhaust particles are injected into pristine regions of very low aerosol concentrations, providing a large perturbation to the system where aerosol effects are distinguishable.

This work will focus on a case study from the 26th – 28th of January 2003 in and around the Bay of Biscay. During this time period ship tracks were observed in a region influenced by high pressure subsidence, followed by a cold front propagating through the domain from the Atlantic towards the European coast. MODIS observations of cloud optical thickness are available over this region (see Figure), which are used for model evaluation.

For these studies we use the limited-area model COSMO, maintained and developed by the German weather service. This model is suited for idealised case studies, operational weather forecasting as well as climate projections. In the current study, aerosol microphysics following Vignati et al (2004) and a two-moment cloud microphysics by Seifert and Beheng (2006) are included. Hence we use prognostic relations for aerosol evolution, cloud condensation nuclii (CCN) production and cloud microphysical interactions, which allows us to describe the transformation and cloud interaction of the ship emissions in a bulk sense. Results will be shown for simulations with a 2-km horizontal resolution, which were obtained using a one-way nesting approach: 2-km simulation nested within a 12-km simulation, which in turn is driven by ERA-interim at the lateral boundaries.

In this framework the meso-scale aerosol-cloud effects due to shipping emissions are studied, which are evaluated using simulated and observed cloud optical thickness. Furthermore changes in the horizontal cloud distribution, vertical extent, drizzle rates and cloud microphysical properties, as well as boundary layer (BL) profiles of moisture, wind and temperature were compared between clean BL conditions, without shipping emissions and pristine aerosol concentrations as low as 20-50/cm³, and polluted BL conditions with shipping emissions. Model evaluation shows a satisfactory representation of the synoptic-scale environment including the simulation of very sharp inversions (Possner et al. 2013). Due to the large uncertainties with respect to the emission mass flux itself and the emission size distribution, differently aged plumes with differently scaled mass fluxes are emitted to estimate the sensitivity of the cloud response with respect to emission uncertainties.

Further analysis is split here into the discussion of aerosol effects within two different dynamical regimes and geographical regions. During the studied time period, the ship plume aerosols are injected into the BL under stable (subsidence regime) and convective conditions as the cold front passes over the marine domain (frontal regime). During the subsidence regime we have defined two regions (see Figure) in which different types of aerosol-cloud interactions are investigated. Region I is characterised by multiple clearly distinguished ship tracks criss-crossing the stratocumulus deck on both, the 26th and 27th of January 2003.

In this region one can study the localised response of the cloud to shipping emissions attributable to individual ships passing through this domain. Here one expects strong gradients in CCN concentration between clean and ship-plume conditions from 50 – 1000/cm³ associated with a very localised microphysical and radiative response of the cloud. For the remainder of this abstract this region will be referred to as the emission zone.

Some of the aerosol from this emission zone are advected downwind, leading to a more horizontally homogenous, but also smaller, CCN increase in region II, as the shipping emissions experience some mixing and dilution. The second region of detailed investigation (region II) includes a cloud band stretching across the Bay of Biscay on the 26th of January, which is characterised by a large optical thickness ranging between 24 to 32. This cloud structure is absent in the clean simulation. However, in preliminary simulations with initial cloud seeding by advected ship emission aerosol, parts of this cloud structure begin to form. The chain of proposed dynamical and microphysical processes leading to the observed cloud formation in region II is investigated.

In the frontal regime, we investigate the effect of the cold front on BL aerosol distribution and concentration, and determine if an aerosol effect on the cold sector cumulus cloud field is present.

Vignati, E., J. Wilson, and P. Stier, 2004: M7: An efficient size-resolved aerosol microphysics module for large-scale aerosol transport models. JGR, 109, D22202.

Seifert A., and K. D. Beheng, 2006: A two-moment cloud microphysics parameterization for mixed-phase clouds. Part I: Model description. Meteorol. Atmos. Phys., 92, 45-66.

Possner A. , E. Zubler, O. Fuhrer, U. Lohmann, and C. Schär, 2013 (submitted): A case-study on modelling low-lying inversions and stratocumulus cloud cover in the Bay of Biscay. Weather and forecasting.