Improving Global Climate Model (GCM) Predictability of Mixed-Phase Clouds and Aerosol Interactions at High Latitudes with ARM Observations

Principal Investigator(s):
Xiaohong Liu, University of Wyoming

Zhien Wang, University of Wyoming

The objective of this proposal is to improve the predictability of mixed-phase clouds and aerosol interactions in the Community Atmosphere Model version 6 (CAM6) through comparison with the ARM observations. Mixed-phase clouds composed of a mixture of cloud droplets and ice crystals are prevalent at high latitudes and have been suggested to play an important role in the surface energy budgets, thereby affecting the surface temperature in these regions. While they are important in the climate system, the understanding of mixed-phase clouds at high latitudes is still very limited, owing to complex interactions among the synoptic scale atmospheric state, cloud-top radiative cooling, cloud dynamics and microphysics, boundary layer turbulence, and land/ocean/ice surface properties.

In this proposal, we aim to address the following science questions: (1) Can GCMs simulate reasonably well the mixed-phase cloud properties as well as their differences between northern and southern high latitudes? (2) If not, how do the representations of ice microphysical processes (e.g., ice nucleation, Wegener-Bergeron-Findeisen (WBF) process) and aerosol processes (e.g., convective transport and scavenging) in GCMs affect the simulations of aerosol and mixed-phase cloud properties, through coupling with dynamics and radiation? and (3) What role do seasonally varying aerosols play in the different mixed-phase cloud properties between northern and southern high latitudes? To tackle these science questions, we will test several new treatments/parameterizations developed from the process understanding of mixed-phase clouds, utilizing an integrated modeling approach. This includes running the CAM6 in the weather forecast mode through the DOE Cloud-Associated Parameterizations Testbed (CAPT) and in the single-column mode, to facilitate the comparisons with field campaign data. The new treatments/parameterizations to be tested include: (1) new ice nucleation parameterizations of marine organic aerosols acting as INPs; (2) inclusion of subgrid cloud treatments such as heterogeneous mixing between cloud liquid and ice (pocket structures) in mixed-phase clouds, which can have a critical impact on the WBF process in mixed-phase clouds for the conversion of cloud liquid to cloud ice; (3) a unified treatment of aerosol vertical transport and wet scavenging in convective clouds; and (4) a new higher-order turbulence closure scheme, which unifies the treatments of boundary-layer turbulence, shallow convection, and cloud macrophysics in CAM6.

We will run the CAPT simulations for multiple years to compare with the long-term ARM data at the Arctic sites (North Slope of Alaska Utqiagvik and Oliktok Point), and with the observation data obtained from the ARM West Antarctic Radiation Experiment (AWARE) in 2016 and the Measurements of Aerosols, Radiation, and Clouds over the Southern Ocean (MARCUS) from September 2017 to April 2018. Modeled probability density functions (PDFs) of cloud occurrence, LWP, IWP, and phase partitioning will be compared between the northern and southern high latitudes and also with observations. Seasonal variabilities of aerosol and mixed-phase cloud properties will be analyzed. New model treatments/parameterizations will also be tested in the free-run mode to examine their performance when large-scale feedbacks and land/ocean/ice changes are included.

The observation data used in model evaluations include multi-sensor mixed-phase cloud retrievals from the ARM NSA sites and from the AWARE and MARCUS campaigns. With the multi-sensor measurements at these sites, stratiform mixed-phase cloud properties, including droplet effective radius, droplet concentration and liquid water path for the liquid phase, and ice concentration, ice effective radius and ice water path for the ice phase will be retrieved. MPL, HSRL, and Doppler lidar measurements will be used to develop a dataset of aerosol vertical distributions for aerosol extinctions and aerosol types, especially over the NSA sites.