Drop shapes derived from a previously conducted artificial rain experiment using a two-dimensional video disdrometer (2DVD) are presented. The experiment involved drops falling over a distance of 80 m to achieve their terminal velocities as well as steady-state oscillations. The previous study analyzed the measured axis ratios (i.e., ratio of maximum vertical to maximum horizontal chord) as a function of equivolumetric spherical drop diameter (D_eq) for over 115 000 drops ranging from 1.5 to 9 mm. In this paper, the actual contoured shapes of the drops are reported, taking into account the finite quantization limits of the instrument. The shapes were derived from the fast line-scanning cameras of the 2DVD. The drops were categorized into D_eq intervals of 0.25-mm width and the smoothed contours for each drop category were superimposed on each other to obtain their most probable shapes and their variations due to drop oscillations. The most probable shapes show deviation from oblate spheroids for D_eq > 4 mm, the larger drops having a more flattened base, in good agreement with the equilibrium (nonoblate) shape model of Beard and Chuang. Deviations were noted from the Beard and Chuang model shapes for diameters larger than 6 mm. However, the 2DVD measurements of the most probable contour shapes are the first to validate the Beard and Chuang model shapes for large drops, and further to demonstrate the differences from the equivalent oblate shapes. The purpose of this paper is to document the differences in radar polarization parameters and the range of error incurred when using the equivalent oblate shapes versus the most probable contoured shapes measured with the 2DVD especially for drop size distributions (DSDs) with large median volume diameters (>2 mm). The measured contours for D_eq > 1.5 mm were fitted to a modified conical equation, and scattering calculations were performed to derive the complex scattering amplitudes for forward and backscatter for H and V polarizations primarily at 5.34 GHz (C band) but also at 3 GHz (S band) and 9 GHz (X band). Calculations were also made to derive the relevant dual-polarization radar parameters for measured as well as model-based drop size distributions. When comparing calculations using the contoured shapes against the equivalent oblate spheroid shapes, good agreement was obtained for cases with median volume diameter (D₀) less than around 2 mm. Small systematic differences in the differential reflectivity (Z_dr) values of up to 0.3 dB were seen for the larger D₀ values when using the oblate shapes, which can be primarily attributed to the shape differences in the resonance region, which occurs in the 5.5–7-mm-diameter range at C band. Lesser systematic differences were present in the resonance region at X band (3–4 mm). At S band, the impact of shape differences in the polarimetric parameters were relatively minor for D₀ up to 2.5 mm. Unusual DSDs with very large D₀ values (>3 mm) (e.g., as can occur along the leading edge of severe convective storms or aloft due localized “big drop” zones) can accentuate the Z_dr difference between the contoured shape and the oblate spheroid equivalent, especially at C band. For attenuation-correction schemes based on differential propagation phase, it appears that the equivalent oblate shape approximation is sufficient using a fit to the axis ratios from the 80-m fall experiment given in this paper. For high accuracy in developing algorithms for predicting D₀ from Z_dr, it is recommended that the fit to the most probable contoured shapes as given in this paper be used especially at C band.

This note builds on prior technique development related to the classification of rain types utilizing C-band polarimetric (CPOL) radar measurements. While the prior work was preliminary and limited in scope, the authors elaborate here on the basis of the drop size distribution (DSD)-based indexing technique for rain-type classification (convective/stratiform/mixed), and place it on firmer footing by testing the methodology against texture- and disdrometer-based methods as applied to Darwin datasets. A microphysical-based methodology is attractive as it links more directly to the underlying rainfall physical processes. Statistics of the DSD parameters, namely, histograms of log10(Nw) and D0, for convective and stratiform rain types across the premonsoon buildup and monsoon regimes were derived and further separated for over land and over ocean regions. The maximum value for mean D0 (1.64 mm) and the largest histogram standard deviation (0.32 mm) occurred for convective rain over land during the buildup regime. The largest differences in D0 and NW histograms were found to be for convective rain between the buildup and monsoon regimes (independent of land or ocean areas). Stratiform rain histograms were found to be very similar during the buildup regime with little land–ocean differences. However, somewhat larger land–ocean differences were found for the monsoon stratiform rain. The main histogram characteristics of the “mixed” or “uncertain” rain type were closer to the convective rain type than to stratiform, across both regimes and land–ocean areas. Additionally, the Nw versus D0 cluster of points (mean ±1σ) for convective rain agrees very well with the previously published range of values for maritime convective (equilibrium-like) DSDs.

Analysis of drop size distributions (DSD) measured by collocated Meteorological Particle Spectrometer (MPS) and a third-generation, low-profile, 2D-video disdrometer (2DVD) are presented. Two events from two different regions (Greeley, Colorado, and Huntsville, Alabama) are analyzed. While the MPS, with its 50-μm resolution, enabled measurements of small drops, typically for drop diameters below about 1.1 mm, the 2DVD provided accurate measurements for drop diameters above 0.7 mm. Drop concentrations in the 0.7–1.1-mm overlap region were found to be in excellent agreement between the two instruments. Examination of the combined spectra clearly reveals a drizzle mode and a precipitation mode. The combined spectra were analyzed in terms of the DSD parameters, namely, the normalized intercept parameter N_W, the mass-weighted mean diameter D_m, and the standard deviation of mass spectrum σ_M. The inclusion of small drops significantly affected the N_W and the ratio σ_M/D_m toward higher values relative to using the 2DVD-based spectra alone. For each of the two events, polarimetric radar data were used to characterize the variation of radar-measured reflectivity Z_h and differential reflectivity Z_dr with D_m from the combined spectra. In the Greeley event, this variation at S band was well captured for small values of D_m (<0.5 mm) where measured Z_dr tended to 0 dB but Z_h showed a noticeable decrease with decreasing D_m. For the Huntsville event, an overpass of the Global Precipitation Measurement mission Core Observatory satellite enabled comparison of satellite-based dual-frequency radar retrievals of D_m with ground-based DSD measurements. Small differences were found between the satellite-based radar retrievals and disdrometers.

Comparisons of drop shapes between measurements made using 2D video disdrometer (2DVD) and wind-tunnel experiments are presented. Comparisons are made in terms of the mean drop shapes and the axis ratio distributions. Very close agreement of the mean shapes is seen between the two sets of measurements; the same applies to the mean axis ratio versus drop diameter. Also, in both sets of measurements, an increase in the oscillation amplitudes with increasing drop diameter is observed. In the case of the 2DVD, a small increase in the skewness was also detected. Given that the two sets of measurements were conducted in very different conditions, the agreement between the two sets of data implies a certain “robustness” in the mean shape of oscillating drops that may be extended to natural raindrop oscillations, at least in steady rainfall and above the surface layer.

We report on measurements of drop size distributions (DSD) using collocated instruments (a Droplet Measurement Technologies, Inc., Meteorological Particle Spectrometer and a 2D-video disdrometer) from two locations with different rainfall climates (Greeley, Colorado, and Huntsville, Alabama, with measurements from the latter that include the outer rainbands of Hurricane Irma). The combination of the two instruments gives what we term as the “full” DSD spectra, the shape of which generally cannot be represented by the standard gamma model, but instead requires the additional flexibility of the generalized gamma model, which includes two shape parameters (μ and c). The double-moment normalization of DSDs using the third and fourth moments is used to arrive at the intrinsic shapes of the DSD with two shape parameters that are shown to capture simultaneously the drizzle mode as well as the precipitation mode, together with a “plateau” region between the two. The estimation of μ and c is done with a global search using nonlinear least squares, and the error residuals are examined to check the sensitivity of the parameters to a preselected, allowed tolerance around the minimum error in the μ, c plane. This leads to a range of plausible fits for a given normalized DSD mainly governed by the c parameter. The stability or invariance of the shape of the normalized DSDs from the two sites is examined, and on average the shapes are similar with some variability at the large normalized diameter end that is explained by the aforementioned range of plausible fits. Heuristic goodness-of-fit methods are described that demonstrate that the generalized gamma model outperforms the standard gamma model with only one shape parameter (μ).

Two rain events are analyzed using two collocated 2D-video disdrometers (2DVD) and a C-band polarimetric radar at 15-km distance. Both events had moderate-to-intense rainfall rates, but the second event had an embedded convective line. For the first event, the fall speed distribution for a given drop diameter interval showed a narrow and symmetric distribution with a mode at the expected value; the second event produced a wider distribution with a significant skewness toward lower fall speeds. The “slower” drops in the second event were detected while the convective line was directly over the 2DVD site. Drop shape information from the two 2DVD instruments showed that, during the passage of the convection line, around 30%–40% of the drops did not have an axis of rotational symmetry, whereas for event 1, it was only 5%. The implications are that for event 1 the dominant mode of drop oscillation is the axisymmetric mode, and that within the convective line of event 2 other fundamental modes were frequent. The radar data for the second event were analyzed in terms of the self-consistency among the radar-measured quantities. The K_dp/Z_h versus Z_dr variations within the line convection were not consistent with the corresponding variation determined from the scattering calculations using the measured 1-min drop size distributions and using the “reference” drop shapes. Also found were low ρ_hv regions within the line convection that were considerably lower than the scattering calculations. These findings are consistent with the asymmetric oscillation modes inferred from the 2DVD measurements for event 2 (probably collision induced) within the convective line.