Autoconversion (Au) and accretion (Ac) have been recognized to be the main two factors in determining the complicated aerosol effects on cloud–precipitation process through their interaction mechanism. However, numerical estimates of these aerosol...
Autoconversion (Au) and accretion (Ac) have been recognized to be the main two factors in determining the complicated aerosol effects on cloud–precipitation process through their interaction mechanism. However, numerical estimates of these aerosol–cloud interactions remain highly uncertain and are poorly understood due to the lack of measurements of specific aerosol–cloud variables. While there have been only few monitoring studies to clarify the aerosol–cloud parameterizations and/or explore their uncertainties, the KORUS-AQ campaign has been launched, and provided the opportunity to estimate the influence of aerosol on cloud–precipitation interactions based on comprehensive and detailed observations (i.e., aircraft, ground sites, and satellites).
In this study, we diagnosed, based on such observations, the precipitation susceptibility (So) which expressed as a function of liquid water path (LWP) from the estimation by CLAVR-x satellite data during KORUS-AQ campaign period over Northeast Asia, and modeling studies are performed to enhance our understanding on the mesoscale aerosol–cloud–precipitation processes. As a first step, the aerosol-related parameters measured from DC-8 aircraft data were examined to compare with the results from CLAVR-x satellite data and explored the uncertainties between them. The results showed the cloud droplet number concentration (Nd) used in CLAVR-x satellite data was well corresponding the cloud condensation nuclei (CCN) and condensation nuclei (CN) at lower altitude, but autoconversion and accretion rate in model were not well reproduced in comparison with observation. Therefore, as a next step, WRF-Chem model was employed, and various sensitivity studies on Au and Ac parameterizations were carried out to compare the modeling results against observations. The results of modeling-observation intercomparison study showed that modeling with default parameterization (BASE) were compared with the CLAVR-x products, while several discrepancies between model and observation were sometimes found, which includes the contribution of microphysical process rates according to LWP, Nd versus precipitation rate (R) relation particularly at low LWP condition, and the ratio of accretion to autoconversion (Ac/Au ratio). In the final step, we further carried out various sensitivity tests to reduce these discrepancies, particularly focusing on improving the relationship between Nd and R according to LWP and the Ac/Au ratio through the repeated amendment-and-adjustment process based on both observation and modeling results. The differences of Nd between WRF and WRF-Chem were both examined, and simulated aerosol chemical compositions were evaluated against AMS measurements obtained from DC-8 aircraft, and the activation process was also assessed through sensitivity test of bulk hygroscopicity updated by observation data.
Our repeated amendment-and-adjustment approach suggested five categories from the classification of Au and Ac over Northeast Asia: 1) different accretion rate with autoconverted liquid water, 2) scaled autoconversion rate, 3) susceptibility fitting factor of power law, 4) subgrid variability, and 5) other parameterizations. In each of the 5 categories, So expressed as function of LWP, relationship between Nd and R, and change of the Ac/Au ratio were investigated, and the parameterization candidates matching individual analysis criteria were optimally organized, respectively.
Our results showed that the variation of Au and Ac depending on LWP value has been improved and overestimated So was corrected from the modified contribution of autoconversion and accretion rate. In addition, bulk hygroscopicity methods formulated by combining specific hygroscopicity estimation from direct observation were thus reasonably predicted aerosol concentrations, suggesting the higher possibilities of model performance on both cloud and aerosol chemical species over Northeast Asia. We also pointed out the limitations on the Au and Ac parameterization and highlighted the overestimation of Au related to the initial formation of precipitation at low LWP, as well as the underestimation of Ac in association with the precipitation intensity. Further research, therefore, will need to understand what further processes are associated with the role of aerosol through aerosol–cloud–precipitation physics over Northeast Asia.