Experiments at the Mainz vertical wind tunnel were carried out to deepen our present knowledge on the aerodynamics of natural hailstones and to improve current fall velocity parameterizations. Thus, the present study provides parameterizations accounting for shape effects and thus variability when calculating the fall velocity of ice particles growing in the accretional growth mode. For this purpose, two sets of 89 3D-printed replicas of natural hailstones were freely levitated in the Mainz vertical wind tunnel from which information on the fall speed, Reynolds number, and drag coefficient was obtained. Sphericity as a shape descriptor was found to be an important parameter to account for the variability of CD. Thus, by taking their shapes into account, the variations in CD of natural hailstones could be reduced by a factor of 2. The comparison of the present results with previous CD parameterizations for nonspherical particles revealed that they are not adequate to represent CD for natural hailstones. The drag coefficients from the present study were overestimated by 26%- 142% in those studies. Based on sphericity, new parameterizations were derived describing the variations in CD and Re for natural hailstones. Furthermore, a parameterization was developed applicable for a lump graupel growing into a hailstone of about 8 cm in maximum dimension, which is valid for a wide range of Reynolds numbers, i.e., from 1 3 101 to 1 3 105.

Two supercell thunderstorms that occurred in Del Rio and Burkburnett, Texas in the spring of 2020 produced greater than 5 cm diameter hailstones. Fourteen hailstones, eight from Del Rio and six from Burkburnett, were cut in half along the maximum principal axis. Half the hailstone was dissected at 0.5 cm increments, providing 198 subsamples of meltwater (0.25 ml) for stable isotope analysis (delta H-2 and delta O-18) using a Picarro L2130-i water analyzer and half used to create a thin section for identification and analysis of ice crystal morphology. Deuterium serves as a proxy for relative altitude, providing a record of hailstone formation and growth trajectory, while crystallographic analysis differentiates wet and dry growth. Results indicate significant variations in isotopic composition and ice crystal morphology, reflecting distinct growth trajectories in the supercell environment, providing evidence that hailstone formation and growth occurred at varying altitudes and reflect recycling in the updraft. The formation of a hailstone embryo at higher altitudes challenges the paradigm of embryo formation only at the cloud base or lower altitudes compared to the outer growth layers. Dendritic crystal morphologies observed in Del Rio hailstones are consistent with rapid freezing in the presence of elevated supercooled liquid water content. Mass ratio analysis provided that most of the hailstones approached the density of bubble free ice while the sphericity index reflected that most hailstones approached 0.80 sphericity but can more accurately be described as triaxial ellipsoids. The goal of this research was to provide better information on hailstone growth and trajectory through supercell convective thunderstorms.

The method of ice crystal replication in Formvar (polyvinyl formal resin) was introduced by Vincent Schaefer in 1941 [1]. At that time no aircraft based optical instrumentation was available to study the morphology of ice crystals. In spite of the rapid advance of the sophisticated optical particle probes, the formvar replication technique, applied in its very original form, turns out to be a valuable complimentary method for the ice crystal habit characterization [4].In addition to habit and surface morphology, some formvar replicas preserve residual particles that may have acted as ice nucleating particles (INPs). Early attempts to identify these nuclei (e.g. Kumai, 1951 [2] and Koenig, 1960 [3]) were limited by poor instrumental resolution and lack of accurate elemental analysis. Modern scanning electron microscopy and X-Ray spectroscopic techniques allows us to revisit this approach. The ice nucleating particles preserved within the replicas can be characterized and attributed to the ice crystal habits and sampling environmental conditions. Based on several case studies, including ice nucleation experiments conducted in AIDA chamber and analysis of formvar replicas of ice crystals collected from free atmosphere and by airborne probes, we evaluate the potential of the ice replication method combined with SEM analysis for INP identification.References:[1] Schaefer, V. J.: A method for making snowflake replicas. Science, 93 (1941) pp. 239-240.[2] Kumai, M.: Electron-microscope study of snow-crystal nuclei. J. Atmos. Sci., 8 (1951) pp. 151-156.[3] Koenig, L. R.: The chemical identification of silver-iodide ice nuclei: a laboratory and preliminary field study. J. Atmos. Sci., 17 (1960) pp. 426-434.[4] Miloshevich, L. M. and Heymsfield, A. J.: A Balloon-Borne Continuous Cloud Particle Replicator for Measuring Vertical Profiles of Cloud Microphysical Properties: Instrument Design, Performance, and Collection Efficiency Analysis, J. Atmos. and Oceanic Tech., 14 (1997), pp. 753-768.

Microplastics are ubiquitous atmospheric particulate pollutants, yet the inflow wind distribution, transport and accretion on hailstones within convective storm systems remain poorly understood. This study provides comprehensive analysis of microplastics preserved in hailstones supercell storms in Texas: Del Rio (11 April 2020) and Uvalde (28 April 2021). Twenty-one hailstones were chosen, melted, filtered, and retained particles analyzed using FTIR microscopy for polymer and morphological characterization. Microplastics were detected in every hailstone (N = 283 particles), confirming ubiquity in the lower troposphere. Fragments dominant morphology (69.4%), fibers (23.6%), films (6.6%), and spheres (0.4%). Fifty-eight distinct polymer species identified, polycarbonate (n = 87), epoxy (n = 23), and polybutadiene (n = 14) the most abundant chemical classes. Kruskal–Wallis H test (p = 0.0176) revealed significant heterogeneity among hailstones, implying emission sources and transport histories. HYSPLIT back-trajectory and land-use revealed low-level inflow air masses traversed manufacturing zones along the U.S. Mexico border and urban population contributing to microplastic loading within the storm. Findings demonstrate that storms effectively scavenge and archive airborne microplastics, offering an approach for assessing inflow air mass transport and emission patterns. Hailstones represent underutilized archive for evaluating emission sources, inflow air mass transport, and monitoring microplastic pollution.

Abstract Ice replicas of natural hailstones together with earlier measurements for natural hailstones are used to develop a relationship that characterizes the drag coefficient and terminal velocity of hailstones through their stages of development and melting. The ice replicas were produced based on 3D scans of natural hailstones. These particles were then levitated in the wind tunnel, and their terminal velocity, mass, and dimensions were measured as the particles were melting. Constant temperatures are used to characterize the properties of the hailstones as they melt. A primary goal of this study was to find a relationship that includes the apparent increase in drag coefficient found at relatively high Reynolds numbers (Re) and associated terminal velocities found in the wind tunnel for nonmelting hailstones, which has not been considered in earlier representations of hail terminal velocities. For this reason, several drag coefficient models for regularly and irregularly shaped particles using particle sphericity and shape are tested against the observations. Although several of the sphericity/shape models fit the data reasonably well, the increase in the drag coefficient at the higher Re is overpredicted in most of the models. Models and drag coefficients dependent only on the Re are also tested against the observations. A consistency is found between the observations and a relationship between the Reynolds number and maximum or equivalent diameters fit to the data. Significance Statement Earlier studies have largely assumed that hail is spherical with solid ice density and that unmelted and melted hail fall equally fast. This study develops consistent relationships across the range of conditions encountered by unmelted and melting hailstones that can be readily incorporated into models.

On 1 August 2021, a vigorous hailstorm hit Azzano Decimo, in northeastern Italy. The supercell storm produced hailstones up to 10 cm in maximum dimension, which is quite unusual in this area. The storm’s environment registered one of the largest potential instabilities (>3400 J kg−1) ever observed by the local operational Udine radiosonde site. In this paper, we analyze the mesoscale environment supporting the hailstorm. Observations from the nearby operational Fossalon di Grado dual-polarization radar showed the presence of a pronounced Bounded Weak Echo Region and differential reflectivity column, both proxies for intense updrafts; however, Doppler velocities revealed only weaker winds, with the mesocyclone mostly confined to midlevels. Two independent observers in Azzano Decimo collected nine hailstones, including one with a maximum dimension of 9 cm. The physical structure of these hailstones was analyzed in the National Center for Atmospheric Research cold room, including normal and cross-polarized light photographs of thin sections to identify the different growth layers inside each hailstone. Additionally, ice samples were taken from 1 cm ×1 cm ×1 mm pieces from these thin slices. The stable isotopic ratio analyses were performed on these specimens using a Picarro cavity ring-down spectrometer. Isotopic content of the hailstone layers revealed significant variability, including some internal layers that showed signs of kinetic fractionation owing to evaporating liquid being incorporated into the growth layer, likely from evaporation of surface liquid during wet growth or collection of recirculated raindrops that experienced evaporation prior to participating in hail growth. Despite such large isotope variability, the Jouzel model analysis suggested that the major growth happened at high altitudes (between 8 and 10 km), which is also in agreement with a reversible-adiabatic parcel model and radar observations from the event.

A novel knowledge-guided convolutional neural network (KGCNN) is presented in this work for the prediction of various three-dimensional shape parameters, drag coefficient, mass, and density of falling snowflakes. To compensate for the lack of extensive data on real snow, the neural network model is pretrained on images of synthetic snowflakes that are geometrically similar to real snowflakes. The model is then fine-tuned on real snowflake images from available datasets for effective drag coefficient prediction. Existing drag coefficient correlations are integrated into the final layer of the KGCNN to remove the burden of learning the relationship between the Reynolds number and drag coefficient from the model. The shape parameter outputs are also regularized by custom knowledge-guided loss functions to ensure physical interpretability and allow for simultaneous prediction of particle volume and density. Pretraining and custom loss functions were found to reduce normalized root-mean-square error (NRMSE) on mass by 11.8%. Integration of drag coefficient correlations reduced mass NRMSE by 30.1% over similar models directly predicting drag coefficient and outperformed existing physical correlations for snowflake drag coefficient. Predictions of shape parameters, volume, and density by the KGCNN were found to be consistent with experimental values.

The growth trajectory of hailstones within clouds has remained elusive due to the inability to trace them directly, impeding the comprehension of their underlying growth mechanisms. This study investigated hailstone vertical growth trajectories by detecting the stable isotope signatures (2H and 18O compositions) of different shells in 27 hailstones from 9 hailstorms, which allowed us to capture the ambient temperature during hailstone growth. The vertical growth trajectories were obtained by comparing the isotopic compositions of water condensate in clouds, derived from the Adiabatic Model, with those measured in hailstones. Although hailstone growth was primarily observed in the −10°C to −30°C temperature layer, the embryo formation height and subsequent growth trajectories significantly varied among hailstones. Embryos formed over a wide range of temperatures (−8.7°C to −33.4°C); four originated at temperatures above −15°C and 16 originated at temperatures below −20°C, suggesting ice nuclei composed of bioproteins and mineral dust, respectively. Among the 27 measured hailstones, 3 exhibited minimal vertical movement, 16 exhibited a monotonic rise or fall, and the remaining 8 exhibited alternating up-down trajectories; only one experienced “recycling” during up-down drifting. Trajectory analysis revealed that similar-sized hailstones from a single storm tended to form at similar heights, whereas those larger than 25 mm in diameter exhibited at least one period of upward growth. Vertical trajectories derived from isotopic analysis were corroborated by radar hydrometeor observations.

This study analyzes in situ aircraft microphysical measurements in deep snow-producing clouds during the Investigation of Microphysics and Precipitation for Atlantic Coast-Threatening Snowstorms (IMPACTS) field programs in winter 2020, 2022, and 2023 to characterize and compare the microphysics in the updraft and downdraft regions. Measurements were acquired from the NASA P-3 aircraft equipped with a full complement of particle probes and instruments for direct measurements of condensed water content and thermodynamic and 3D wind measurements. We identified the P-3 aircraft flight days collected within snowbands, generating cells, and peripheral regions from 14 flights. These are composited and used for the analysis. Temperatures sampled ranged between -27 degrees and +1 degrees C. The data are partitioned by air vertical velocity, with strong updrafts defined as >0.5 m s(-1), very strong updrafts as >1 m s(-1), and strong downdrafts as <-0.5 m s(-1). This partitioning revealed precipitation mass concentrations that were 2x higher in the strong updrafts and 3x higher in the very strong updrafts than in the downdrafts, a result of particle growth and relative fallout within the updrafts. Total particle concentrations at the concentrations > 1 mm were about the same in each region. However, fallout of the larger aggregates through the updrafts at temperatures > -5 degrees C and into the melting layer results in previously unreported shedding and lofting of the shed particles in the updrafts to subfreezing temperatures. This observation is supported by the overflying NASA ER-2 Doppler radar measurements.

The thermodynamic and kinematic environments favoring the growth of large hailstones (diameter >= 19 mm) may be classified into five distinct types. In this study, we present a simple semi-three-dimensional hailstone growth trajectory model and explore the respective microphysical mechanisms of large-hailstone production (LHP) under these five types of environments. Type 1 environment is characterized by the strongest updraft and adequate cloud water supply in tropical plains. This type has the highest growth rates for embryos and surpasses the greatest melting loss among all five types of environments. Type 2 over tropical hills has a deeper updraft with greater growth at high altitudes. Type 3 over midlatitude plains is characterized by a relatively thick growth zone and the second-highest growth rate. Type 4 over high-latitude plains has a high mass growth during ascent but a medium descent growth rate on average. Over elevated terrains, type 5 has the shallowest melting zone and the lowest melting rate, allowing hailstones to reach the surface with the least mass loss. The responses of hailstone growth to initial hailstone embryo heights are type dependent but are insensitive to initial hailstone radii. A longer ascent (descent) growth duration leads to a greater mass increment and a higher potential for low (high)-level-seeded embryos to grow into large hailstones. Sensitivity tests show that kinematic conditions measured by wind shear are important in all types. Strong wind shear serves as a preferable environment for hailstone growth across five types by extending the growth duration.

Mesoscale bands develop within winter cyclones as concentrated regions of locally enhanced radar reflectivity, often corresponding to intensified precipitation rates lasting several hours. Surface precipitation characteristics are governed by the microphysical properties of the ice-phase particles aloft, yet their unique microphysical evolutionary pathways and ambient environmental dependencies in banded regions remain poorly understood, in part due to a paucity of observations within clouds. Addressing this need, the Investigation of Microphysics and Precipitation for Atlantic Coast-Threatening Snowstorms measured properties of winter cyclones from airborne in situ and remote sensing platforms. Observations collected within a banded region of a decaying-stage northeast United States cyclone revealed a microphysical pathway characterized by precipitation fallout from a weak generating cell layer through an ∼ 2 km deep subsaturated downdraft region. Sublimation was a dominant evolutionary process, resulting in a > 70 % reduction in the initial characteristic ice water content (IWC). This vertical evolution was reproduced by a one-dimensional (1D) particle-based model simulation constrained by observations, conveying accuracy in the process representation. Four sensitivity simulations assessed evolutionary dependencies based on observationally informed perturbations of the ambient relative humidity, RH, and vertical air motion, w. Perturbations of ∼ 2 % RH significantly varied the resultant characteristic IWC loss, by as much as 29 %, whereas comparable perturbations of w had negligible effects. Intrinsic particle evolution during sublimation demonstrated a notable imprint on vertical profiles of radar reflectivity, but the Doppler velocity was more strongly governed by the ambient w profile. These findings contextualize radar-based discrimination of sublimation from other ice-phase processes, including riming and aggregation.

The NASA Investigation of Microphysics and Precipitation for Atlantic Coast-Threatening Snowstorms (IMPACTS) field campaign provides high-quality, high-altitude aircraft lidar (532 nm), radar (W band), and in-cloud microphysical aircraft data taken during wintertime storm events impacting the United States. This study evaluates two mass-dimensional relationships [Brown and Francis (BF95); Heymsfield (H14)] and two lidar-radar microphysical retrieval algorithms [CloudSat and CALIPSO Ice Cloud Property Product (2C-ICE); VarPy (a variational method derived from the satellite lidar-radar data community)] to estimate aircraft-retrieved volume extinction coefficient (a-), ice water content (IWC), and effective radius (re) during the 2020 IMPACTS deployment. BF95 and H14 have a close 1:1 correlation (R2 = 0.98) with in situ observations of a-. However, only BF95 displays a linear, consistent, and almost temperatureindependent low bias for IWC and re, which likely arises from the environmental conditions used to determine each. Unlike the field-campaign-derived BF95 and H14 relationships, VarPy and 2C-ICE directly ingest the aircraft-based lidar and radar data to simulate a-, IWC, and re. For all three microphysical parameters, VarPy and 2C-ICE retrieval errors became notably more pronounced around the dendritic growth zone (from -15 degrees to -10 degrees C) and near freezing (>=-5 degrees C), which suggests that both algorithms experience difficulty addressing riming and aggregation processes and with larger particles (dendrites and plates) due in part to their simplified ice particle assumptions. However, the mean-melt diameter ice-particle assumption did yield more accurate IWC estimates, which led to slightly better overall results for VarPy.

The backscattering properties of randomly oriented complex rosette ice aggregates at the radar frequencies of about 9, 35, and 94 GHz are computed using the boundary element method. A Monte Carlo model is used to generate the rosette aggregates, and 65 aggregates are selected from the statistical runs that are within +/- 30% of a mass-dimension relation that is consistent with the Met Office's cirrus microphysics scheme in its weather and climate models. The area-dimension relationship is shown to be generally consistent with an observed area-size power law. The budding rosettes and rosette aggregates have maximum dimensions between about 10 mu m and 1 cm. To test the budding rosette and rosette aggregate model, data from NASA's IMPACTS campaign are used. The IMPACTS data consists of four frontal snowstorm cases that achieved the best co-incident measurements between the in-situ and remote sensing aircraft. We show that the rosette aggregate model predicts the time series of radar reflectivity data generally well for all four cases.

Remote sensing radars from airborne and spaceborne platforms provide critical observations of clouds to estimate precipitation rates across the globe. The ability of these radars to detect changes in precipitation properties is advanced by Doppler measurements of particle fall speed. Within mixed-phase clouds, precipitation mass and its fall characteristics are especially sensitive to the effects of riming. In this study, we quantified these effects and investigated the distinction of riming from aggregation in Doppler radar vertical profiles using quasi-idealized particle-based model simulations. Observational constraints of a control simulation were determined from airborne in situ and remote sensing measurements collected during the Investigation of Microphysics and Precipitation for Atlantic Coast-Threatening Snowstorms (IMPACTS) for a wintry-mixed precipitation event over the northeastern United States on 4 February 2022. From the upper boundary of a one-dimensional column, particle evolution was simulated through vapor deposition, aggregation, and riming processes, producing realistic Doppler radar profiles. Despite a modest observed amount of supercooled liquid water (0.05 g m-3), riming accounted for 55 % of the ice-phase precipitation mass, cumulatively increasing reflectivity by 44 % and Doppler velocity by 68 %. Independent evaluation of process-based sensitivities showed that, while radar reflectivity is comparably sensitive to either riming- or aggregation-based particle morphology, the Doppler velocity profile is uniquely sensitive to particle density changes during riming. Thus, Doppler velocity profiles advance the diagnosis of riming as a dominant microphysical process in stratiform clouds from single-wavelength radars, which has implications for quantitative constraints of particle properties in remote sensing applications.

Although all ice crystals are unique, many can be grouped together by shape or habit, with members of a habit class sharing similar representations of properties such as fall velocity and growth rate. A decision tree algorithm designed to be adaptable to any particle imaging probe, thus enabling the creation of habit size distributions over a size range larger than that of any probe on its own, is used to classify ice crystals imaged by three airborne cloud probes in mid-latitude winter cyclones during the Investigation of Microphysics and Precipitation for Atlantic Coast-Threatening Snowstorms (IMPACTS) field campaign. Crystals are sorted into seven habit classes based on their morphological properties: sphere, column/needle, plate, graupel, dendrite, aggregate, and irregular. Although adaptability was its primary goal, the algorithm was found to be moderately skillful for identifying idealized habit images. Quantitative tests of the algorithm's adaptability displayed mixed results, as Two-Dimensional Stereo Probe (2DS) classifications showed moderate correlation with Particle Habit Imaging and Polar Scattering Probe (PHIPS) classifications, but only weak correlation with High Volume Precipitation Spectrometer (HVPS) classifications. The algorithm was applied to random sets of images from each probe in a case study of a mesoscale snow band sampled on 7 February 2020. In the case study, qualitative analysis of particle images revealed general agreement on classifications among the probes, supporting the algorithm's applicability to multiple cloud probes. Most classifications appeared correct upon manual inspection, suggesting that in practical use, the algorithm is reasonably able to classify non-idealized images.

During the Multidisciplinary drifting Observatory for the Study of Arctic Climate (MOSAiC) expedition, the Balloon-bornE moduLar Utility for profilinG the lower Atmosphere (BELUGA) was deployed from an ice floe drifting in the Fram Strait from 29 June to 27 July 2020. The BELUGA observations aimed to characterize the cloudy Arctic atmospheric boundary layer above the sea ice using a modular setup of five instrument packages. The in situ measurements included atmospheric thermodynamic and dynamic state parameters (air temperature, humidity, pressure, and three-dimensional wind), broadband solar and terrestrial irradiance, aerosol particle microphysical properties, and cloud particle images. In total, 66 profile observations were collected during 33 balloon flights from the surface to maximum altitudes of 0.3 to 1.5 km. The profiles feature a high vertical resolution of 0.01 m to 1 m, including measurements below, inside, and above frequently occurring low-level clouds. This publication describes the balloon operations, instruments, and the obtained data set. We invite the scientific community for joint analysis and model application of the freely available data on PANGAEA.

The secondary ice process (SIP) is a major microphysical process, which can result in rapid enhancement of ice particle concentration in the presence of preexisting ice. SPICULE was conducted to further investigate the effect of collision-coalescence on the rate of the fragmentation of freezing drop (FFD) SIP mechanism in cumulus congestus clouds. Measurements were conducted over the Great Plains and central United States from two coordinated aircraft, the NSF Gulfstream V (GV) and SPEC Learjet 35A, both equipped with state-of-the-art microphysical instrumentation and vertically pointing W- and Ka-band radars, respectively. The GV primarily targeted measurements of subcloud aerosols with subsequent sampling in warm cloud. Simultaneously, the Learjet performed multiple penetrations of the ascending cumulus congestus (CuCg) cloud top. First primary ice was typically detected at temperatures colder than -10 degrees C, consistent with measured ice nucleating particles. Subsequent production of ice via FFD SIP was strongly related to the concentration of supercooled large drops (SLDs), with diameters from about 0.2 to a few millimeters. The concentration of SLDs is directly linked to the rate of collision-coalescence, which depends primarily on the subcloud aerosol size distribution and cloud-base temperature. SPICULE supports previous observational results showing that FFD SIP efficiency could be deduced from the product of cloud-base temperature and maximum diameter of drops measured similar to 300 m above cloud base. However, new measurements with higher concentrations of aerosol and total cloud-base drop concentrations show an attenuating effect on the rate of coalescence. The SPICULE dataset provides rich material for validation of numerical schemes of collision-coalescence and SIP to improve weather prediction simulations

The stability of ice crystal orientation is studied by modeling the airflow around ice crystals at moderate Reynolds number, where an ice crystal is approximated by a cylinder with three parameters: diameter D, length L, and zenith angle of the axis O. In this paper, the torque acting on ice crystals is simulated at different O first, and then a special O with zero horizontal torque, denoted as Oe, is sought as an equilibrium of ice crystal orientation. The equilibrium is classi-fied into two kinds: stable and unstable. Ice crystals rotate to Oe of stable equilibriums while deviating from Oe of unstable ones once they are released into quiet air. Multiple equilibriums of ice crystal orientation are found via numerical simula-tions. A cylinder with D/L close to one has three equilibriums, two of which are stable (i.e., Oe = 0 & DEG; and 90 & DEG;). A cylinder with D/L away from one has only two equilibriums, one of which is stable (i.e., either Oe = 0 & DEG; or 90 & DEG;). In addition, an asym-metric cylinder has two, three, or five equilibriums, and their Oe is sensitive to the distance between its geometrical center and its center of gravity. The sensitivity of Oe to crystal asymmetry suggests large symmetric ice crystals tend to become asymmetric (or irregular) and subsequently oriented randomly.

Coincident radar data with Doppler radar measurements at X, Ku, Ka, and W bands on the NASA ER-2 aircraft overflying the NASA P-3 aircraft acquiring in situ microphysical measurements are used to characterize the rela-tionship between radar measurements and ice microphysical properties. The data were obtained from the Investigation of Microphysics and Precipitation for Atlantic Coast-Threatening Snowstorms (IMPACTS). Direct measurements of the con-densed water content and coincident Doppler radar measurements were acquired, facilitating improved estimates of ice particle mass, a variable that is an underlying factor for calculating and therefore retrieving the radar reflectivity Ze, median mass diameter Dm, particle terminal velocity, and snowfall rate S. The relationship between the measured ice water content (IWC) and that calculated from the particle size distributions (PSDs) using relationships developed in earlier studies, and between the calculated and measured radar reflectivity at the four radar wavelengths, are quantified. Relationships are derived between the measured IWC and properties of the PSD, Dm, Ze at the four radar wavelengths, and the dual-wavelength ratio. Because IWC and Ze are measured directly, the coefficients in the mass-dimensional relationship that best match both the IWC and Ze are derived. The relationships developed here, and the mass-dimensional relationship that uses both the measured IWC and Ze to find a best match for both variables, can be used in studies that characterize the properties of wintertime snow clouds.

Polarimetric microphysical retrievals reveal a great potential for the evaluation of numerical models and data assimilation. However, the accuracy of ice microphysical retrievals is still poorly explored. To evaluate these retrievals and assess their accuracy, polarimetric radar measurements are spatially and temporally collocated with in situ aircraft measurements obtained during the OLYMPEX campaign (Olympic Mountain Experiment). Retrievals for ice water content (IWC), total number concentration N-t, and mean volume diameter D-m of ice particles are assessed by comparing an in situ dataset obtained by the University of North Dakota (UND) Citation II aircraft with X-band Doppler on Wheels (DOW) measurements. Sector-averaged range height indicator (RHI) scans are used to derive vertical profiles of microphysical retrievals. The comparison of these estimates with in situ data provides insights into strengths, weaknesses, and the accuracy of the different retrievals and quantifies the improvements in polarimetry-informed retrievals compared to conventional, non-polarimetric ones. In particular, the recently introduced hybrid ice water content retrieval exploiting reflectivity Z(H), differential reflectivity Z(DR), and specific differential phase KDP outperforms other retrievals based on either (Z(H), Z(DR)) or (Z(H), K-DP) or non-polarimetric retrievals in terms of correlations with in situ measurements and the root mean square error.