Understanding the microstructure of cloud systems is fundamental for improving our comprehension of precipitation mechanisms and for assessing the potential for weather modification. This analysis leverages data acquired from a sophisticated UAV drone mission to dissect the microphysical properties of a mixed stratiform-convective cloud system. The primary objective is to elucidate the typical cloud characteristics in the region and provide a direct observational basis for evaluating conditions suitable for artificial precipitation enhancement. The deployment of large UAV drones offers a unique advantage in probing complex cloud systems over mountainous terrain, providing rapid-response, in-situ measurements that are critical for real-time decision-making.
The mission was conducted on 26 May 2024 over eastern Guizhou. The synoptic situation was characterized by an upper-level trough moving out of the region, coupled with low-level shear lines, creating favorable conditions for widespread precipitation. The UAV drone platform, a large fixed-wing model, was equipped with a suite of cloud physics instrumentation, enabling detailed measurement of cloud particle spectra and imagery. This configuration allows for the investigation of particle growth processes, phase composition, and water content within the cloud layers that are typically challenging to observe with conventional ground-based or satellite platforms.
Methodology of UAV-Based Cloud Probing
The core of this analysis is the dataset obtained from the onboard cloud and precipitation particle measurement system. The instrumentation comprised three primary probes: a Cloud Droplet Probe (CDP), a Cloud Imaging Probe (CIP), and a Precipitation Imaging Probe (PIP). Their respective measurement ranges allow for a comprehensive view of the particle population, from tiny cloud droplets to large hydrometeors.
| Probe | Measurement Principle | Size Range | Primary Output |
|---|---|---|---|
| Cloud Droplet Probe (CDP) | Forward Scattering | 2 – 50 μm | Spectrum of small cloud particles (NCDP) |
| Cloud Imaging Probe (CIP) | Beam Occlusion & Imaging | 25 – 1550 μm | Spectrum & images of large cloud particles (NCIP) |
| Precipitation Imaging Probe (PIP) | Beam Occlusion & Imaging | 100 – 6200 μm | Spectrum & images of precipitation particles (NPIP) |
Particle number concentration (N, in cm-3) and liquid water content (LWC, in g m-3) are derived from the measured spectra. The effective diameter (Deff) is calculated as the ratio of the third to the second moment of the particle size distribution (PSD).
$$ D_{eff} = \frac{\int D^3 N(D) dD}{\int D^2 N(D) dD} $$
$$ LWC = \frac{\pi \rho_w}{6} \int D^3 N(D) dD $$
where $\rho_w$ is the density of water, D is particle diameter, and N(D) is the particle size distribution function. Data quality control involved removing spurious values, de-spiking, and discarding the first size bin from the imaging probes to minimize instrumental artifacts.
Flight navigation and situational awareness were supported by real-time weather radar data. The flight track of the UAV drone was overlaid on composite reflectivity maps to correlate in-situ microphysical measurements with the macro-scale cloud structure. The mission profile involved horizontal traverses within the upper layers of three distinct cloud entities at altitudes between approximately 5,000 and 5,400 meters, with ambient temperatures ranging from +1.7°C to -3°C.
Horizontal Distribution of Cloud Microphysical Quantities
The UAV drone flight captured significant spatial variability in cloud microstructure. The overarching observation was a clear inverse relationship between particle concentration and size across the different probe ranges. A statistical summary of the entire flight leg highlights this fundamental characteristic.
| Particle Type | Mean Number Concentration (cm-3) | Mean Effective Diameter (μm) | Typical Size-Conc. Relation |
|---|---|---|---|
| Small Cloud Particles (CDP) | 202.02 | 15.69 | High Concentration, Small Size |
| Large Cloud Particles (CIP) | 8.78 | 284.72 | Low Concentration, Large Size |
| Precipitation Particles (PIP) | 0.73 | 599.01 | Very Low Concentration, Very Large Size |
The horizontal transects revealed three primary regions of interest, temporally defined as Period 1 (15:35-15:57), Period 2 (16:13-16:47), and Period 3 (17:00-17:32), corresponding to different cloud clusters. In regions where the UAV drone coincided with radar reflectivity factor (Z) greater than or equal to 25 dBZ, the microphysical signature was particularly pronounced. In these zones, the average small particle concentration (NCDP) increased to 253.5 cm-3 with Deff of 17.5 μm. More strikingly, the large cloud particles exhibited an effective diameter (Deff, CIP) of 460 μm, and precipitation particles reached a Deff, PIP of 1092 μm. This configuration—high concentration of small droplets alongside large, low-concentration ice particles—is indicative of active ice crystal growth via deposition and aggregation, coupled with efficient collection processes. The following expression conceptualizes the growth rate of a collector particle in a population of smaller droplets, relevant for both coalescence and accretion:
$$ \frac{dm}{dt} \approx \frac{\pi}{4} E D^2 V_t w $$
where $dm/dt$ is the mass growth rate, $E$ is the collection efficiency, $D$ is the diameter of the collector particle, $V_t$ is its terminal velocity, and $w$ is the liquid water content of the smaller droplets encountered.
Vertical Profiling via Particle Spectra
By segregating data into specific altitude bins, the particle size distributions (PSDs) provide insight into the dominant microphysical processes at different cloud levels. The PSD is often represented by a modified gamma or exponential function. For analysis, average spectra were computed for three layers: 4970-5020 m, 5310-5369 m, and 5370-5429 m.
The CIP spectra for large cloud particles consistently showed a unimodal distribution, peaking in the 50-100 μm range. However, the spectral width and tail varied with height. In the lowest layer (4970-5020 m), the spectrum decayed rapidly for diameters above 700 μm, characteristic of a warm cloud process zone. The middle layer exhibited a broader spectrum with a slightly more populated tail beyond 1000 μm, suggesting the onset of ice-phase processes and mixed-phase conditions. The highest layer showed the lowest overall concentration but the highest concentration of particles at the maximum detectable size (~1550 μm), likely indicating a region dominated by ice crystal aggregation.
The PIP spectra for precipitation particles displayed a near-exponential decay across all layers: $ N(D) = N_0 e^{-\Lambda D} $. The key difference was in the intercept parameter $N_0$ and the slope $\Lambda$. The middle layer (5310-5369 m) had the highest $N_0$ for particles in the 200-400 μm range, identifying it as a probable active precipitation generation layer. The highest layer, while having moderate concentrations in the 200-400 μm range, was distinguished by the presence of particles exceeding 3000 μm in diameter, strongly suggesting growth dominated by ice-phase mechanisms like riming and aggregation.
Comparative Microphysics of Distinct Cloud Systems
The flight of the UAV drone captured three distinct cloud entities, allowing for a comparative analysis. Period 1 was characterized by the highest ambient temperature (predominantly >0°C). Periods 2 and 3 were within sub-zero temperatures, with Period 3 experiencing the coldest conditions.
Cloud Water Distribution
The frequency of occurrence of different liquid water content (LWC) thresholds, derived from CDP measurements, reveals the availability of supercooled liquid water (SLW), a critical factor for precipitation efficiency and seeding potential.
| LWC Threshold (g m-3) | Period 1 Frequency (%) | Period 2 Frequency (%) | Period 3 Frequency (%) |
|---|---|---|---|
| > 0.001 | 68.94 | 81.57 | 94.36 |
| > 0.01 | 39.78 | 45.21 | 83.47 |
| > 0.1 | 2.18 | 17.59 | 50.20 |
| > 0.5 | 0.00 | 2.91 | 8.20 |
The data clearly shows that SLW was relatively scarce in Period 1 but became increasingly abundant and widespread in Periods 2 and 3, with Period 3 showing the most significant enrichment.
Statistical Microphysical Parameters
The mean microphysical properties for each period, as measured by the UAV drone‘s instrumentation, are summarized below.
| Parameter | Period 1 | Period 2 | Period 3 |
|---|---|---|---|
| NCDP (cm-3) | 254.44 | 157.07 | 220.52 |
| Deff, CDP (μm) | 6.16 | 13.45 | 19.53 |
| NCIP (cm-3) | 7.87 | 9.27 | 7.27 |
| Deff, CIP (μm) | 119.49 | 156.37 | 480.19 |
| NPIP (cm-3) | 0.36 | 1.05 | 0.36 |
| Deff, PIP (μm) | 253.47 | 286.59 | 1042.04 |
The trends are revealing. Period 1 had the highest concentration of small cloud droplets but the smallest effective diameter, indicating a cloud dominated by homogeneous condensation with weak collision-coalescence and minimal ice process involvement. Period 2 showed a notable increase in the concentration of precipitation-sized particles (NPIP) alongside moderate droplet sizes, pointing to a mixed-phase cloud where particle growth was active but somewhat limited. Period 3 was dominated by large particles; despite having a lower concentration of large cloud and precipitation particles, their effective diameters were substantially larger than in other periods. This, combined with the high SLW content, signifies a cloud system where growth via deposition, riming, and aggregation was the dominant and highly efficient microphysical pathway.
Particle Spectra and Growth Mechanisms
The average particle size distributions for each period further illuminate the dominant growth mechanisms. The PSDs can be parameterized. For the small droplet spectrum (CDP), a gamma distribution is often applicable:
$$ N(D) = N_0 D^\mu e^{-\Lambda D} $$
where $\mu$ influences the shape. In Period 1, the spectrum was narrow and steep ($\mu$ low, $\Lambda$ high), indicative of a population growing primarily by diffusion. In Periods 2 and 3, the spectra broadened ($\mu$ increased, $\Lambda$ decreased), suggesting the increasing importance of stochastic collection and ice crystal interactions.
For the larger particles (CIP & PIP), the spectra in Period 1 decayed rapidly, consistent with weak collection. In Period 2, the CIP spectrum showed a more populated tail, and the PIP spectrum had its highest concentration, reflecting active warm-rain and initial ice processes. The spectra in Period 3 were the broadest, with significant mass spread over a wide diameter range, a classic signature of efficient ice-phase growth where the mass doubling time can be expressed in terms of the available SLW and collection kernel (K):
$$ \tau_{growth} \propto \frac{1}{SLW \cdot K(D)} $$
The imagery from the CIP and PIP probes corroborated this analysis. In Period 1, images showed mostly small spherical droplets. In Period 2, images began to show irregular and budding ice crystals. In Period 3, the imagery was dominated by large, complex aggregates of ice crystals and rimed particles, visually confirming the dominance of ice-phase microphysics.
Discussion and Implications
The data collected by the large UAV drone provides an unprecedented, high-resolution view into the microphysical anatomy of mixed cloud systems over eastern Guizhou. The observed inverse relationship between particle concentration and size across different hydrometeor categories is a fundamental feature of precipitation formation, where a few particles efficiently scavenge the available cloud water from a large population of smaller droplets.
The transition observed from Period 1 to Period 3 represents a classic evolution in cloud microphysics driven by thermodynamic environment and dynamic forcing. Period 1 represents a cloud where the “collision-coalescence” warm-rain process was likely the primary, though inefficient, mechanism due to the small droplet sizes. The introduction of ice-phase processes in Periods 2 and 3, particularly in the SLW-rich environment of Period 3, dramatically increased precipitation formation efficiency through the “seeder-feeder” mechanism and the Bergeron-Findeisen process. The presence of large aggregates in Period 3 confirms that aggregation was a key terminal growth process, enhancing fall speeds and precipitation yield.
For artificial precipitation enhancement, these observations are critical. The UAV drone mission successfully identified regions with high SLW content (e.g., Period 3) coexisting with a natural population of ice crystals. Such regions represent prime targets for glaciogenic seeding, where additional ice nuclei could further exploit the SLW reservoir and enhance precipitation via the ice multiplication chain. The ability of the UAV drone to provide real-time identification of these “seedable” zones—characterized by specific thresholds of SLW (>0.1 g m-3), temperature (-1.6°C to -3°C), and the presence of large, growing ice particles—demonstrates its operational utility. The platform’s agility allows it to not only diagnose but also potentially engage these regions with targeted seeding material, closing the loop between observation and intervention.
Conclusion
This analysis, grounded in direct measurements from a large UAV drone, has detailed the microphysical characteristics of a mixed stratiform-convective cloud system. The key findings are: (1) The cloud system exhibited a classic structure with high concentrations of small cloud droplets and lower concentrations of progressively larger hydrometeors. (2) Significant microphysical heterogeneity was observed horizontally and vertically, with particle growth mechanisms shifting from predominantly condensational growth in warmer regions to efficient ice-phase deposition, riming, and aggregation in colder, SLW-rich regions. (3) The UAV drone platform proved exceptionally capable of delineating these distinct microphysical zones in real-time, providing clear signatures for regions with high precipitation potential and susceptibility to weather modification.
The systematic differences in cloud microstructure across the observed cloud regions underscore the importance of in-situ, targeted observations. The deployment of UAV drones for cloud physics research and operational weather modification represents a significant technological advancement, offering a powerful tool to validate remote sensing products, refine numerical model parameterizations, and ultimately execute more precise and effective cloud seeding campaigns. Future missions with similar UAV drone platforms should aim to couple these detailed microphysical observations with high-resolution cloud dynamics measurements to build a more complete picture of precipitation formation in complex mountainous terrain.