This paper investigates immersion mode ice‐nucleating particle (INP) concentrations measured with a continuous flow diffusion chamber at temperatures of −20°C, −25°C, and −30°C in Beijing from 4 May to 4 June 2018. Aerosol conditions in the experiment period are classified into four types: clean, light PM2.5 pollution, heavy PM2.5 pollution, and dust. The PM mass concentrations and the aerosol size distributions are distinctly different in the four aerosol conditions. INP concentration increases as temperature decreases from −20°C to −30°C for all the four aerosol conditions. At a given temperature, INP concentrations progressively increase as the aerosol condition changes from clean to light PM2.5 pollution and to heavy PM2.5 pollution condition. At −25°C and −30°C, the dust condition has a much higher INP concentration than the other aerosol conditions. The INP concentration is closely related to the number concentration of aerosols with diameters of 1.0–2.5 μm (). Using a similar method as in the previous parameterization of INP concentrations for global and fairly clean aerosol conditions, a new parameterization of the INP concentration based on is provided for the complicated aerosol conditions with high aerosol number concentrations in Beijing.

This study investigates the effects of aerosol vertical distribution on a deep convective cloud system. We intend to elucidate the mechanisms for aerosols entering the cloud from different heights, and how they affect cloud microphysics and precipitation. A thermal bubble is released at 1.5 km initially to run an idealized case using the Weather Research and Forecast (WRF) model. The aerosol layer with high concentration was initially put at different altitudes in the model to study the mechanisms and the number of aerosols entering the cloud. It was found that there are three mechanisms for aerosols from different heights to enter the cloud, depending on their relative height with the thermal bubble. Aerosols from lower altitudes (below 1 km) enter the cloud through pumping, while aerosols from higher altitudes (2–3 km, 3–5 km) enter the cloud through entrainment. Both mechanisms lead to low cloud condensation nuclei (CCN) concentration in the cloud. Only aerosols from intermediate altitudes (1–2 km), which is the same as the initial height of the thermal bubble, enter the cloud mainly by ascending with the bubble and lead to high CCN concentration in the cloud. The differences in activated CCN concentration affect the microphysical processes and precipitation remarkably. For the simulations with an initial aerosol layer at 1–2 km and 0–5 km, aerosols can enter the cloud more efficiently than the other four simulations. More activated CCNs in these two simulations lead to more graupels with smaller sizes at higher altitudes, which delays the precipitation but makes the precipitation last longer. However, the accumulated precipitation is similar in all six simulations, no matter what aerosol vertical distribution is like. The results in this study indicate that the altitude of aerosol layers determines the mechanisms for aerosols entering clouds, CCN concentration in the cloud, and to what extent the cloud microphysical processes and precipitation are affected.

This study investigates the statistical characteristics of isolated deep convection initiation (DCI) and its relation to topography in the North China area during the summer season of 2017–2019. A total of 2,534 isolated DCI events are identified using the infrared brightness temperature data from Himawari‐8 satellite with a resolution of 0.05° × 0.05°. The spatial distribution of DCI shows clustering over mountains and hills, indicating that the local topography may be critical for DCI. Topography is described with elevation and relief amplitude in this study. Results show that normalized DCI counts increase with elevation and relief amplitude. For the regime characterized by fewer morning clouds (regime one), DCI strongly depends on elevation and relief amplitude, while for the regime with more morning clouds (regime two), topography only moderately impacts DCI. Linear regression analyses suggest the existence of significant linear relationships between DCI and relief amplitude when elevation is relatively high. Topography is also categorized into four types: plain/sea areas, plateaus, hilly areas, and mountains. Normalized DCI counts show that mountains and hilly areas are more favorable for initiating isolated deep convection, especially for regime one. Moreover, DCI counts in regime one show a stronger diurnal variation and a peak occurring 2 hr earlier than that of regime two. We suggest that special attention should be paid to elevation and relief amplitude, as well as morning cloud cover condition when forecasting DCI in the North China area. Further study is necessary to understand this DCI occurrence preference on specific topography.

Ice nucleating particles (INPs) can be important for deep convective storms through affecting microphysical, thermodynamical, and radiative properties and precipitation. Since the mass of a hailstone is predominantly contributed by supercooled cloud droplets, INPs in supercooled cloud droplets during deep convective storms can be indicated from INPs in hailstones. In this study, hailstones are collected from eight deep convective cases in three regions of China, and immersion freezing INPs in hailstones are measured using droplet freezing experiments over a wide temperature range from −27.5°C to −4.0°C. INP concentrations can vary by 2–3 orders of magnitude, which is consistent with the result in a previous study on hailstones for the same eight cases that insoluble aerosol number concentrations in hailstones also have a variability of 2–3 orders of magnitude. Using the fitting method, three INP modes are suggested to be included for most hailstone samples, while one to two INP modes can be good enough for a couple of hailstone samples. Furthermore, the INPs for warmer mode are more heat‐sensitive, indicating possible existence of proteinaceous biological species.

Insoluble particles influence weather and climate by means of heterogeneous freezing process. Current weather and climate models face considerable uncertainties in freezing-process simulation due to limited information regarding species and number concentrations of heterogeneous ice-nucleating particles, particularly insoluble particles. Here, for the first time, the size distribution and species of insoluble particles are analyzed in 30 shells of 12 hailstones collected from China using scanning electron microscopy and energy-dispersive X-ray spectrometry. A total of 289 461 insoluble particles were detected and divided into three species – organics, dust, and bioprotein – utilizing machine learning methods. The size distribution of the insoluble particles of each species varies greatly among the different hailstones but little in their shells. Further, a classic size distribution of organics and dust followed logarithmic normal distributions, which could potentially be adapted in future weather and climate models despite the existence of uncertainties. Our findings highlight the need for atmospheric chemistry to be considered in the simulation of ice-freezing processes.

Drizzle initiation through the collision and coalescence of cloud droplets plays a crucial role in warm cloud precipitation. Recent theoretical studies suggest that the influence of collisional growth on the droplet size distribution can be quantified by a non‐dimensional drizzle number (Dz). Here, large‐eddy simulations with Lagrangian microphysics are employed to evaluate the theory by simulating a tall convection‐cloud chamber under various conditions. Results show that the smaller the Dz, the larger the impact of collisions on the right tail of the droplet size distribution, consistent with the theory. The simulations confirm that the collision rate can be estimated from the droplet size distribution interacting only with cloud droplets of the same size at the mode radius. This suggests that the idealized theory can be a useful tool to design a cloud chamber for drizzle investigation, as well as to represent drizzle formation in models of real atmospheric clouds.