-
During the years 2017–19, the Rainfall Enhancement Experiment along the Eastern Side of Taihang Mountain project was conducted in North China, and the data used in this study were obtained during this project. The aim of the project was to comprehensively investigate the fine structure of clouds in North China and establish a concept model for cloud seeding. Several papers have analyzed the data from this project in recent years, including studies on radar retrieval of cloud microphysics, aerosol-cloud interaction (Yang et al., 2019), effects of topography on severe storms with heavy precipitation, improvements of microphysical schemes using in-situ data (Guo et al., 2019; Lei et al., 2020), and cloud seeding effects on convective cloud under variable wind shear conditions (Yang and Lei, 2022). However, most of these studies have primarily focused on shallow convective clouds or precipitation cores within mesoscale weather systems. On 21 June 2017, a Kingair-350 aircraft conducted vertical spiral and horizontal measurements from 0800 to 1100 UTC to examine the microphysical properties of a snowfall event. The aircraft took off from Sijiazhuang airport specifically for this purpose. The goal of the measurements was to determine the microphysical structures of the cloud. -
During this project, cloud properties were collected using the Airborne atmospheric measurements King-Air 350 platform, which was equipped with several microphysical probes. These probes provided measurements of a wide range of hydrometeors spectra as well as bulk ice or liquid parameters, such as LWC (liquid water content) and IWC (ice water content), as described in the papers mentioned in section 2. The key instrumentation used included a Mie-scattering FCDP-100 (Fast Cloud Droplet Probe model 100, SPEC inc), a CIP (cloud imager probe, DMT inc) and HVPS-3 (high volume particles spectrometer. SPEC inc), a CPI (cloud particle imager, SPEC inc), an AIMMS-20 to obtain ambient wind speed, relative humidity, and temperature, and a Nevrozov Total Water Content probe. It is noteworthy that the 2D imager (CIP) used in this study had a sharpened probe front end and the raw data from 2D images were processed using an ITA (Interarrival Time Threshold Algorithm) to remove artifacts due to the shattering of large particles on the probes (Field et al., 2006). The specifications of the airborne instruments used in this study are summarized in Table 1.
Probe Measurement Range Bin interval Fast cloud droplets probe (FCDP) Cloud droplet spectra 2–50 μm 0.2–3 μm Cloud imager probe (CIP) 2D imageries of cloud and precipitation particles, and the size spectra 25–1550 μm 25 μm High volume particles spectrometer (HVPS) 2D imageries of cloud and precipitation particles, and the size spectra 150–19200 μm 150 μm Cloud Particle Imager (CPI) High-resolution imageries of cloud and precipitation particles 10–2000 μm 2.3 μm Nevrozov total water content (TWC) Bulk parameters of ice and liquid phase particles, namely LWC and IWC 0.005–3 g m–3 − AIMMS-200 Flight track of aircraft as well as ambient parameters, including temperature, relative humidity altitude: 0–13.7 kmtemperature: –50°C to 50°C − Table 1. Airborne instrumentations and their specifications.
-
To clarify the fine structure of microphysics and study the evolution of the cloud structure, we employed the WRF model 4.4.1 for all experiments. The model was initialized with 0.25° × 0.25° ERA5 reanalysis data, with a grid spacing of 9 km and 3 km, to better resolve the microphysical processes. The simulation area was divided into two nested domains, with grid points of 186 (lat)× 186 (lon) and 292 (lat) × 283 (lon). The inner domain covered the region where in-situ observations were conducted, and the model was warm-started before conducting the experiments.
Except for cloud microphysics, the two domains have the same parameterizations which include the Rapid Radiative Transfer Model for longwave radiation, the Dudhia scheme for shortwave radiation. The Yonsei University Planetary Boundary scheme, and the Grell-Fraitas cumulus scheme are only applied in the outer domain. In cloud modeling, the parameterization of ice microphysics is essential for the accurate simulation of quantitative precipitation forecasts and the microphysical–dynamic interaction. Therefore, different microphysics schemes can result in variable representations of cloud properties, such as cloud droplet size distribution, ice water content, and precipitation formation. Using various microphysics schemes to study the WCB can help identify the impact of microphysics parameterizations on cloud structure of WCB and determine which scheme is more suitable for simulations. Thus, in this study, we selected four completely different microphysical schemes (Table 2) to compute the vertical and horizontal composites of the WCB, including the conventional two-moment parameterization scheme (Morrison two-moment, MORR), P3 scheme Morrison and Milbrandt (2015), fast bin scheme (HUJI Fast Bin, FB) (Khain et al., 2010), and a newly developed triple-moment six class bulk scheme (Tsai and Chen, 2020) which is provided by NTU.
Scheme Abbreviation Features and description Morrison Two moments MORR Conventional scheme with two-moment Gamma distribution functions to address particle evolution. Cloud activation based on supersaturation. P3 2ICE P3 Updated version of the Morrison scheme, additional categories were used to describe the riming degree of the ice particles HUJI Fast Bin HUJI Fast version of the HUJI bin scheme, solving the stochastic equations for more realistic physical processes NTU NTU Newly developed triple-moment microphysical scheme, considering the terminal velocity and densities of the ice, while cloud droplet activation depends on both updraft and relative humidity. Table 2. Brief description of the four microphysics schemes used in this simulation
The NTU scheme is a bulk parameterization scheme that follows the conventional triple-moment approach. However, it employs a more suitable treatment of ice-phase processes to account for the variations in the shape and density of different hydrometeors. In contrast to other schemes, the NTU scheme redefines snow as snow aggregate to avoid the artificial conversion of cloud ice to snow. For the P3 scheme, although it also predicts changes in ice particle shape during riming, it does not provide a detailed treatment of aggregation, as noted by Tsai and Chen (2020). Meanwhile, the fast bin scheme predicts the evolution of particle spectra explicitly, without making any assumptions. As such, it has demonstrated some superiority over bulk methods in many aspects of cloud simulations (Yin et al., 2017). Table 2 summarizes some of the key specifications and features of each scheme.
Probe | Measurement | Range | Bin interval |
Fast cloud droplets probe (FCDP) | Cloud droplet spectra | 2–50 μm | 0.2–3 μm |
Cloud imager probe (CIP) | 2D imageries of cloud and precipitation particles, and the size spectra | 25–1550 μm | 25 μm |
High volume particles spectrometer (HVPS) | 2D imageries of cloud and precipitation particles, and the size spectra | 150–19200 μm | 150 μm |
Cloud Particle Imager (CPI) | High-resolution imageries of cloud and precipitation particles | 10–2000 μm | 2.3 μm |
Nevrozov total water content (TWC) | Bulk parameters of ice and liquid phase particles, namely LWC and IWC | 0.005–3 g m–3 | − |
AIMMS-200 | Flight track of aircraft as well as ambient parameters, including temperature, relative humidity | altitude: 0–13.7 kmtemperature: –50°C to 50°C | − |