The impacts of enhanced heterogeneous ice nucleation (HET) on the properties of deep convective clouds (DCCs) have been investigated in cloud-resolving simulations with the WRF-CHEM model. The study focuses on a case observed during the Deep Convective Clouds and Chemistry (DC3) field campaign. For the simulated DCCs, which had cold cloud-base temperatures, an inverse relationship exists between ice crystal mass produced through HET and anvil ice crystal number concentrations. This seems to be due to the indirect competition between HET and subsequent homogeneous freezing (HOM) for liquid droplets. Furthermore, our simulations suggest that HET enhancements at warmer temperatures are more efficient in depleting liquid droplets below and hence have larger impacts on anvil properties than HET enhancements at colder temperatures do. This temperature dependence indicates that similar increases in the number of ice nucleating particles (INPs) may potentially have different impacts on DCCs, depending on the INP type and at which temperatures they can nucleate ice crystals. We also found that the reduced anvil ice number concentrations due to the enhanced HET may lead to optically thinner anvil clouds. The reduction in cloud optical depth comes from a decrease in ice crystal mass concentrations, and in some runs also from an increase in ice crystal sizes. These results suggest potentially large impacts of INPs on the properties of DCCs, especially if precipitation is predominantly produced through ice processes in the DCCs. The results underscore the importance of fully understanding the temperature-dependent ability of aerosol particles to nucleate ice crystals.

Abstract. The year of 2015 was an extremely dry year for Southeast Asia where the direct impact of a strong El Niño was in play. As a result of this dryness and the relative lack of rainfall, an extraordinary quantity of aerosol particles from biomass burning remained in the atmosphere over the Maritime Continent during the fire season. This study uses the Weather Research and Forecasting model coupled with Chemistry to understand the impacts of these fire particles on cloud microphysics and radiation during the peak biomass burning season in September. Our simulations, one with fire particles and the other without them, cover the entire Maritime Continent region at a cloud-resolving resolution (4 km) for the entire month of September in 2015. The comparison of the simulations shows a clear sign of precipitation enhancement by fire particles through microphysical effects; smaller cloud droplets remain longer in the atmosphere to later form ice crystals, and/or they are more easily collected by ice-phase hydrometeors in comparison to droplets under no fire influences. As a result, the mass of ice-phase hydrometeors increases in the simulation with fire particles, and so does rainfall. On the other hand, the aerosol radiative effect weakly counteracts the invigoration of convection. Clouds are more reflective in the simulation with fire particles as ice mass increases. Combined with the direct scattering of sunlight by aerosols, the simulation with fire particles shows higher albedo over the simulation domain on average. The simulated response of clouds to fire particles in our simulations clearly differs from what was presented by two previous studies that modeled aerosol–cloud interaction in years with different phases of El Niño–Southern Oscillation (ENSO), suggesting a further need for an investigation on the possible modulation of fire–aerosol–convection interaction by ENSO.

Aerosol emissions from forest fires may impact cloud droplet activation through an increase in particle number concentrations (“the number effect”) and also through a decrease in the hygroscopicity κ of the entire aerosol population (“the hygroscopicity effect”) when fully internal mixing is assumed in models. This study investigated these effects of fire particles on the properties of simulated deep convective clouds (DCCs), using cloud‐resolving simulations with the Weather Research and Forecasting model coupled with Chemistry for a case study in a partly idealized setting. We found that the magnitude of the hygroscopicity effect was in some cases strong enough to entirely offset the number/size effect, in terms of its influence on modeled droplet and ice crystal concentrations. More specifically, in the case studied here, the droplet number concentration was reduced by about 37% or more due solely to the hygroscopicity effect. In the atmosphere, by contrast, fire particles likely have a much weaker impact on the hygroscopicity of the pre‐existing background aerosol, as such a strong impact would occur only if the fire particles mixed immediately and uniformly with the background. We also show that the differences in the number of activated droplets eventually led to differences in the optical thickness of the clouds aloft, though this finding is limited to only a few hours of the initial development stage of the DCCs. These results suggest that accurately and rigorously representing aerosol mixing and κ in models is an important step toward accurately simulating aerosol‐cloud interactions under the influence of fires.