Spatial variability of liquid water path in marine low cloud: Part I. Probability distributions and mesoscale cellular scales

Aspects of the liquid water path (LWP) spatial probability distribution function (pdf) and power spectra in marine low cloud over the subtropical NE and SE Pacific Ocean are examined using daytime retrievals for two months of data from the Moderate Resolution Imaging Spectroradiometer (MODIS) on the NASA Terra satellite. It is shown that the pdf can be described to a good degree of accuracy using a modified Gamma distribution in accordance with earlier work. The pdf is used to validate and constrain a simple Gaussian physical model of the mesoscale variability that links cloud LWP variability with the underlying boundary layer thermodynamical structure. Importantly, this model couples variability of the cloud LWP with the cloud fraction, and so may be used in large-scale numerical models to account for unresolved variability. Optical depth spatial variability is shown to be well modeled using an adiabatic assumption with constant droplet concentration. A simple formulation for albedo biases is presented that is implemented as a simple reduction in the cloud fraction. Power spectral analysis is used to determine the scale dependence of the LWP pdf. It is found that for the majority of power spectra, characteristic scales exist that are related to mesoscale cellular convection (MCC). However, despite the existence of characteristic scales in many cases, it is shown that in the ensemble, the liquid water path exhibits a surprisingly universal scaling with an exponent close to -5/3 over scales up to at least several hundred kilometers.