Real-case simulations of aerosol–cloud interactions in ship tracks over the Bay of Biscay

Abstract

Abstract. Ship tracks provide an ideal test bed for studying aerosol-cloud interactions (ACIs) and for evaluating their representation in model parameterisations. Regional modelling can be of particular use for this task, as this approach provides sufficient resolution to resolve the structure of the produced track including their meteorological environment whilst relying on the same formulations of parameterisations as many general circulation models. In this work we simulate a particular case of ship tracks embedded in an optically thin stratus cloud sheet which was observed by a polar orbiting satellite at 12:00 UTC on 26 January 2003 around the Bay of Biscay.The simulations, which include moving ship emissions, show that the model is indeed able to capture the structure of the track at a horizontal grid spacing of 2 km and to qualitatively capture the observed cloud response in all simulations performed. At least a doubling of the cloud optical thickness was simulated in all simulations together with an increase in cloud droplet number concentration by about 40 cm −3 (300 %) and decrease in effective radius by about 5 µm (40 %). Furthermore, the ship emissions lead to an increase in liquid water path in at least 25 % of the track regions.We are confident in the model's ability to capture key processes of ship track formation. However, it was found that realistic ship emissions lead to unrealistic aerosol perturbations near the source regions within the simulated tracks due to grid-scale dilution and homogeneity.Combining the regional-modelling approach with comprehensive field studies could likely improve our understanding of the sensitivities and biases in ACI parameterisations, and could therefore help to constrain global ACI estimates, which strongly rely on these parameterisations.

The resolution dependence of cloud effects and ship‐induced aerosol‐cloud interactions in marine stratocumulus

“…Most of the background Aitken mode particles that are sufficiently large to activate consist of sulphate advected into the simulation domains from the driving ECHAM5‐HAM climatology. These particles are then mixed downward from the free troposphere into the PBL, as was seen in previous simulations of this case [see Possner et al , , Figure 5] and in other studies of stratocumulus‐topped PBL [ Mechem et al , ; Berner et al , ; Kleinman et al , ; George et al , ]. A small additional source of accumulation mode particles in the PBL consists of local sea‐salt emissions.…”

“…In this study the mass flux is parameterized based on in situ exhaust measurements [ Hobbs et al , ] of five cargo vessels (NYK Sunrise, Tai He, Star Livorno, Fremo Scorpius, and Sea Pearl) running on marine fuel oil, which were obtained within the marine boundary layer. However, the literature‐based PM emission fluxes of 20.84 kg s −1 per ship were insufficient to generate an in‐plume aerosol perturbation comparable to size distribution measurements of the exact same exhaust plume [ Hobbs et al , ; Possner et al , ]. Consequently, a significant underestimation of the cloud response in comparison with the MODIS retrieval obtained on 26 January for this particular marine boundary layer was simulated at 2 km resolution [ Possner et al , ].…”

“…However, the literature‐based PM emission fluxes of 20.84 kg s −1 per ship were insufficient to generate an in‐plume aerosol perturbation comparable to size distribution measurements of the exact same exhaust plume [ Hobbs et al , ; Possner et al , ]. Consequently, a significant underestimation of the cloud response in comparison with the MODIS retrieval obtained on 26 January for this particular marine boundary layer was simulated at 2 km resolution [ Possner et al , ]. To account for this deficit, a tenfold scaling of the PM emission flux to 208.4 kg s −1 , which is equivalent to assuming about a doubling in size of the particles, was applied.…”

“…To account for this deficit, a tenfold scaling of the PM emission flux to 208.4 kg s −1 , which is equivalent to assuming about a doubling in size of the particles, was applied. Given this modification,the size distributions measured in the vicinity of the stack [ Hobbs et al , ] were matched in ship_ 02 [ Possner et al , ].…”

“…This work continues from a previous study [ Possner et al , ], which evaluated the simulated cloud response to the ship exhaust at the 2 km scale qualitatively and quantitatively using the Moderate Resolution Imaging Spectroradiometer (MODIS) cloud optical depth retrieval and observations of various field measurements of ship tracks [ Ferek et al , ; Hobbs et al , ; Hudson et al , ; Durkee et al , ; McComiskey et al , ]. In Possner et al [] the structure of the tracks forming behind a moving emission source was successfully captured, while the magnitude of the change in τ within the track was found to agree with the MODIS retrieval obtained at 12 UTC on 26 January 2003. Almost all processes, except for the large‐scale subsidence, involved in the formation of a strong inversion and of the track and its radiative effect, are parameterized.…”