[Paper Review] Stratocumulus over SouthEast Pacific: Idealized 2D simulations with the Lagrangian Cloud Model

Mirosław Andrejczuk, Alan Gadian+1 more|arXiv (Cornell University)|Nov 1, 2012

Climate variability and models23 references3 citations

TL;DR

This study uses a Lagrangian Cloud Model (LCM) with a 2D idealized setup to simulate stratocumulus clouds over the Southeast Pacific, based on VOCALS observations. The model successfully reproduces key cloud properties like liquid water mixing ratio and droplet number concentration, and captures the observed aerosol–cloud droplet relationship, though it underestimates large droplet concentrations and shows sensitivity to microphysical parameterization choices.

ABSTRACT

In this paper a LES model with Lagrangian representation of microphysics is used to simulate stratucumulus clouds in idealized 2D set-up based on the VOCALS observations. The general features of the cloud simulated by the model, such as cloud water mixing ratio and cloud droplet number profile agree well with the observations. The model can capture observed relation between aerosol distribution and concentration measured below the cloud and cloud droplet number. Averaged over the whole cloud droplet spectrum from the numerical model and observed droplet spectrum are similar, with the observations showing a higher concentration of droplets bigger than 25 μm. Much bigger differences are present when comparing modelled and observed droplet spectrum on specific model level. Despite the fact that microphysics is formulated in a Lagrangian framework the standard deviation of the cloud droplet distribution is larger than 1 μm. There is no significant narrowing of the cloud droplet distribution in the up-drafts, but the distribution in the up-drafts is narrower than in the down-drafts. Modelled and observed standard deviation profiles agree well with observations for moderate/high cloud droplet numbers, with much narrower than observed droplet spectrum for low droplet number. Model results show that a significant percentage of droplets containing aerosol bigger than 0.3 μm didn't reach activation radius, yet exceeding 1 μm, what is typically measured as a cloud droplets. Also, the relationship between aerosol sizes and cloud droplet sizes is complex; there is a broad range of possible cloud droplet sizes for a given aerosol size.

Motivation & Objective

• To validate a Lagrangian Cloud Model (LCM) against real-world VOCALS observations of stratocumulus clouds.
• To assess the model's ability to reproduce observed relationships between aerosol size, cloud droplet number, and droplet spectrum.
• To investigate the sensitivity of microphysical and dynamical cloud properties to numerical parameters in the LCM framework.
• To explore the limitations and strengths of the Lagrangian approach in representing droplet spectrum broadening and activation processes.

Proposed method

• The LCM tracks millions of Lagrangian parcels representing aerosol and cloud droplets, each with identical chemical and physical properties.
• Microphysical processes—condensational growth, activation, deactivation, and collision-coalescence—are simulated in a Lagrangian framework and coupled with Eulerian dynamics and thermodynamics.
• The model uses a bin-based representation of droplet and aerosol size distributions, with collision events mapped to a grid for efficiency.
• Sensitivity tests vary the number of parcels, bin resolution, collision frequency, and threshold parameters to assess numerical robustness.
• Model output is compared with in-situ VOCALS observations of cloud water mixing ratio, droplet number concentration, and droplet spectra.

Experimental results

Research questions

• RQ1 Can the Lagrangian Cloud Model accurately simulate observed stratocumulus cloud properties such as liquid water mixing ratio and droplet number concentration?
• RQ2 How well does the model reproduce the observed relationship between aerosol concentration and cloud droplet number below the cloud layer?
• RQ3 To what extent does the model capture the observed droplet spectrum, particularly the concentration of large droplets (>25 µm)?
• RQ4 How sensitive are the model results to numerical parameters such as the number of parcels, bin resolution, and collision frequency?
• RQ5 What is the role of Lagrangian microphysics in shaping droplet spectrum broadening, especially in updrafts and downdrafts?

Key findings

• The model reproduces observed cloud water mixing ratio and cloud droplet number profiles well, with good agreement in bulk cloud properties.
• Observed droplet spectra show higher concentrations of droplets larger than 25 µm compared to the model, indicating underestimation of large droplet formation.
• The standard deviation of the droplet spectrum is larger than 1 µm in the model, and there is no significant narrowing in updrafts, though updrafts show narrower spectra than downdrafts.
• For low droplet number conditions, the model produces a narrower droplet spectrum than observed, suggesting limitations in representing broadened spectra under low-CCN conditions.
• A significant fraction of aerosol particles larger than 0.3 µm fail to activate but still grow beyond 1 µm, leading to misclassification as cloud droplets in observations.
• The model shows a broad range of possible cloud droplet sizes for a given aerosol size, highlighting the complex, non-linear relationship between aerosol and droplet size.