Airborne measurements of the water droplet spectra n(r) (r is radius) (0.23 ≤ r ≤ 150 μm) were made at about 14 levels in stratus layers during May and August 1981, 130 km SW of San Diego. All May days were found to be associated with air masses approximating marine air; all August days, with air masses approximating continental air. One May day and one August day were selected to represent these approximate marine and continental air masses, respectively. Four parameters were calculated from 15 n (r) observed along each level extending 6.44 km and from the average n (r) over the entire run. These parameters are the number of droplets, the mean droplet radius, the cross-sectional area and the liquid water content.
Profiles of the four parameters were constructed as a function of the distance from the cloud base, defined to be at the level where the liquid water content is 0.02 g m−3. The droplet spectra on the marine and continental air mass days differed primarily because n(r) on the continental days contained many more small droplets. The profile of the liquid water content was closely examined for the region from 250 below to 250 m above the cloud base. The profiles of the cross-sectional area and the liquid water content were closely approximated by a model below cloud base for the selected days.
The average horizontal scale size of the four parameters was estimated to be 2.9 km in and near the clouds. The horizontal variability of a parameter was defined to be the ratio of the standard deviation to its average at a level. This variability was a maximum near the cloud base and top and was a minimum near mid-cloud, suggesting enhancement of horizontal variation by vertical mixing across the saturation levels. The correlation coefficients between pairs of parameters along each level showed no consistent vertical pattern except that the number of droplets and the liquid water content are positively correlated at all levels. These correlations do not support the presence of inhomogeneous mixing on the selected days.
The apparent horizontal variability created by an inadequate sampling volume by the spectrometers was examined. The horizontal sampling distances required to obtain a reliable average of each parameter were also estimated.