This study aims to explore possible causes in the decreasing trend of liquid water path (LWP) with increasing cloud droplet number concentrations (CDNC) obtained by satellite-based studies. Datasets comprise cloud properties derived from UCLALES-SALSA simulations of a typical marine stratocumulus cloud setup (nocturnal drizzling stratocumulus cloud DYCOMS-II RF02 of the Second Dynamics and Chemistry of Marine Stratocumulus field campaign (Ackerman et al., 2009. Mon. Wea. Rev., 137, 1083–1110). Model-based properties were used to calculate LWP and CDNC with two different equations commonly used in satellite data analysis. Deviations between model and satellite-retrievals are included as relative percentage error.

Cloud top properties included in these datasets were obtained with a sampling methodology that mimics this so-called penetration depth bias (Grosvenor et al. 2018, AMT, 11(7), 4273–4289). Datasets include properties in a cloud upper region comprising as many model layers as needed to reduce by three units the cloud optical thickness in the infrared region (2.38 um -4.00 um to get a surrogate of MODIS retrievals at the 2.13 um um channels) below its maximum value. All properties are calculated as extinction coefficient weighted average values that consider all model layers in the selected cloud upper region. Cloud optical thickness was calculated for the visible wavelength band (0.25 um - 0.65 um) as a surrogate of MODIS retrievals at 0.66 um.

Simulations were performed in a horizontal model domain size of 51 Km by 51 Km and a vertical domain extending up to 1.4 Km. Horizontal and vertical resolutions were set to be 75 m and 20 m, respectively. Equations were resolved with a maximum time step of 1 s using an hourly period for spinup without nudging.

To determine the effect of aerosol on cloud properties, we ran a set of 3 simulations with initial accumulation model aerosol concentrations of 65, 150 and 300 #/cm3 (corresponding to the total number concentrations of 78, 180 and 360 #/cm3). In our datasets, we have included cloud properties sampled at 2 h, 6h and 10h from the start of the run. Each file was labeled using this information, for example the file CloudTop_prop_acc150_6h.nc contains snapshots of cloud top properties at 6.2 h, 6.3 h and 6.5 h for the simulation scenario initialized with an aerosol concentration of 150 #/cm3 in the accumulation mode.

Datasets have been divided in two containers according to the spatial resolution.

Container 1 : 2003163_DYCOMS_II_RF02_cloud_prop_75 m_75m including 9 ncfiles, three sampling times and three simulation scenarios initialized with different aerosol loadings. Each file shows cloud property fields in the original horizontal resolution of simulations.

Container 2 : 2003163_DYCOMS_II_RF02_cloud_prop_1425m_1425m including 9 ncfiles, three sampling times and three simulation scenarios initialized with different aerosol loadings. Original horizontal resolution in simulations. Each file shows cloud property fields spatially aggregated in subdomains of 1.425 Km by 1.425 Km (19 model vertical columns each one with 75 m by 75 m). Spatially-aggregated properties are obtained using cloud optical thickness in cloudy columns as a weighting factor during horizontal averaging operations along subdomains.