a. Measurements and methods

The Small Cumulus Microphysics Study (SCMS) was conducted in July–August of 1995 along the east-central coast of Florida (i.e., in the vicinity of Cape Canaveral). The objective of the SCMS was to investigate the initial development and early evolution of warm cumulus clouds, specifically addressing 1) the onset of precipitation, 2) the evolution of droplet and raindrop size distributions, and 3) the processes of entrainment and mixing. This investigation used aircraft measurements from 12 flights of the National Center for Atmospheric Research (NCAR) C-130 during the SCMS and a total of 411 cumulus cloud traverses to examine the characteristics of humidity halos. Typically one to three aircraft traverses were made through each cloud sampled. The NCAR C-130 attempted to traverse through the center of clouds; therefore cloud diameter is generally equivalent to in-cloud traverse length for this investigation.

The small cumulus sampled on different days during the SCMS developed in relatively similar ambient thermodynamic environments that were systematically documented using aircraft and rawinsonde measurements. Aircraft measurements collected during take-off and landing on each flight provided vertical profiles of humidity, stability, and vertical wind shear in the undisturbed ambient environment. These profiles, collected in cloud-free areas typically located within 5–20 km of the aircraft cloud study region, were used as a point of comparison with aircraft humidity measurements near cloud edge regions to determine the presence and extent of humidity halos.

The humidity approaches the value of the undisturbed environment as the distance from the cloud increases within the cloud-free detrainment region surrounding a cloud. For purposes of this investigation, the cloud edge is defined using total particle concentration measurements from the forward scattering spectrometer probe (FSSP). These data had a collection rate of 10 Hz, approximately equivalent to an aircraft flight distance of 11 m. The cloud boundary was defined as the location where the cloud droplet number concentration decreased to, and remained below, 10 cm⁻³. This threshold value was chosen to avoid clear-air background aerosol and particle concentrations that existed during SCMS (e.g., Laird et al. 2000, 2001) and have been previously observed in maritime environments (e.g., Meszaros and Vissy 1974; Hoppel et al. 1989). Absolute humidity measurements were examined to determine the extent of the humidity halo and were collected using the fast-response Lyman-alpha hygrometer with a sample rate of 25 Hz (∼4.5 m distance) and response time of 2 ms. Humidity measurements collected at 25 Hz were interpolated to 10 Hz to determine the humidity halo existence and lengths away from cloud boundaries.

Similar to PH96 and LU03, the horizontal extent of the halo was represented by the e-folding distance prior to and following each cloud traverse. The e-folding distance was determined through a comparison of the average in-cloud absolute humidity with the absolute humidity of the undisturbed environment at the height of each aircraft cloud traverse. The e-folding distance was defined as the distance from the cloud boundary where the measured absolute humidity decreased to, and remained below, the e-folding absolute humidity (ρ_υe); ρ_υe is defined by

View Expanded

where _υc and _υenv are the average absolute humidities of the cloud and the ambient environment at the height of the aircraft traverse (z), respectively. Figure 1 presents data from a cloud traverse showing an extension of the cloud humidity halo in the downshear direction with no evidence of a halo in the upshear direction. A “No Halo” region is defined as the coincident location of the e-folding distance, as determined by absolute humidity measurements, with the location of cloud boundary, as determined by FSSP measurements. Measurements from cloud traverses where the aircraft did not reach the maximum extent of the humidity halo (i.e., e-folding distance) prior to a subsequent cloud traverse were not included in the analyses.

b. Clouds, shear environment, and humidity halos

The winds within the ambient environment at levels ±200 m relative to each cloud traverse height were used to determine the vertical wind shear. Values of shear were generally less than ±2 m s⁻¹ km⁻¹, with an average absolute vertical wind shear of 1.0 m s⁻¹ km⁻¹. The mean diameter of the clouds included in the current study was 1.1 km. The conceptual model developed by PH96 and their results describing humidity halos were derived from measurements in a stronger vertical wind shear environment compared to those sampled during the SCMS (Fig. 2) and for clouds with a larger mean diameter (Fig. 3). The average vertical wind shear and cloud diameter for isolated cumulus traversed by PH96 were 6.1 m s⁻¹ km⁻¹ and 2.3 km, respectively. Figures 2 and 3 show noticeable differences in the vertical wind shear and cloud diameter distributions for the 411 cloud traverses of the current study and the 31 clouds sampled by PH96.

Figure 4 shows that approximately 70% of halos observed during SCMS had lengths of less than one cloud radius and about 98% of observed humidity halos had a radial length less than four cloud radii. Similarly, PH96 found humidity halos were generally less than four cloud radii; however, halos in a few instances had a larger radial length. For larger trade wind cumulus clouds with an average diameter of about 3.6 km, LU03 found all halos were less than two cloud radii in horizontal extent. The average humidity halo lengths, relative to the cloud radius, reported by PH96 were 0.3, 0.7, and 1.3 for upshear, cross-shear, and downshear regions, respectively. The mean halo lengths, relative to cloud radius, for SCMS clouds were found to be 0.6, 0.7, and 0.8 for upshear, cross-shear, and downshear regions, respectively. These translate into halo lengths of approximately 660, 770, and 880 m for a mean SCMS cloud of 1.1-km diameter.

The SCMS measurements show a larger frequency of observed humidity halos than earlier studies. On average, approximately 86% of cloud boundary traverses had a humidity halo. PH96 found halos were present with frequencies of 74%, 40%, and 36% on the downshear, cross-shear, and upshear regions of clouds. LU03 identified the occurrence of a humidity halo approximately 44% and 55% of the time for upshear and downshear regions, respectively. Figure 4 shows that, regardless of the radial direction, humidity halos were observed with a frequency between 77% and 90%. Similar to previous studies, halos were most frequently observed in the downshear region (87%) and the frequency did not vary significantly in the cross-shear (86%) or upshear (83%) regions. The difference in frequency of halo occurrences between upshear and downshear regions was much smaller than previously reported observations. The observed high frequency and small differences in the frequency of halos in all radial directions was likely caused by the absence of a strong vertical wind shear environment during the SCMS.

Cloud age was shown by PH96 to have a significant effect on the radial lengths of the humidity halo. The appearance of penetrated SCMS clouds was examined using forward-looking flight videos to estimate the age of each cloud. Cloud age was determined using similar criteria employed by PH96. The visual clarity of cloud top and edges were used to distinguish younger clouds with well-defined, sharp boundaries from older clouds with ragged, nebulous boundaries. Not all clouds could be adequately classified using the forward-looking flight videos. The cloud age was estimated for 341 of the 411 SCMS cloud traverses and each was placed in one of two cloud age categories (young, old). The young cloud category represented cloud lifetimes up to and including the mature stage.

Figure 5 shows cloud age had a considerable effect on the existence of humidity halos, their average length, and the degree of their variability in length. Humidity halos were more prevalent for older clouds with an approximate frequency of occurrence of 90%. Alternatively, younger clouds had humidity halos with a frequency of 57%–85%. Although an increase of halo length was not always observed from upshear to downshear regions for clouds of similar age, a considerable increase in the average and variability of the halo length for each region occurred with an increase in cloud age. The smallest differences were in the cross-shear regions where old clouds had an average halo length 1.5 times larger than young clouds. The upshear and downshear regions showed an increase in average halo length of nearly one cloud radius from young to old clouds.