Improving Simulations of Cirrus Cloud Thinning by Utilizing Satellite Retrievals

The mechanisms of cirrus cloud development are not fully understood and are not accurately implemented in global climate models (GCMs). Consequently, uncertainties arise in the representation of the cirrus cloud thinning (CCT) within GCMs. A key source of this uncertainty stems from the contribution of homogeneous (hom) and heterogeneous (het) ice nucleations in the development of cirrus clouds. Het, commonly regarded as the predominant ice production process, has garnered substantial attention in cirrus cloud studies through both observational field campaigns and GCM simulations. However, recent research, conducted by our group, developed new global-scale satellite retrievals from the Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observation (CALIPSO) and unveiled that the hom process holds far greater significance for cirrus clouds than previously recognized. This observation implies that the effectiveness of CCT might surpass previous estimates, considering that the cooling efficacy of CCT depends on the fraction of cirrus clouds strongly affected by hom along with the optical depth of these hom-affected cirrus clouds.

As a primary step toward harnessing the insights of this novel CALIPSO retrieval and quantifying the net radiative effects (CRE) of cirrus clouds at the top of the atmosphere, at the surface, and within the atmosphere, we employ a 1-dimensional radiative transfer model (RTM) known as libRadtran. The RTM experiments involve utilizing vertical profiles of ice water content (IWC) and effective diameter (D_e) from CALIPSO retrievals for both hom and het cirrus clouds across different geographical regions (polar and midlatitudes) and seasons (winter and summer). The outcomes illustrate that a transition from hom to het cirrus (due to natural processes or CCT) yields substantial cooling effects. This is particularly pronounced in the Arctic during winter, resulting in a CRE decrease of approximately -3 W m^-2 at the surface for cirrus overcast conditions. This cooling arises due to the absence of solar radiation which otherwise partially counteracts the infrared CRE. Sensitivity tests underscore the significant impact of atmospheric thermodynamic profiles and low-level liquid clouds on the magnitude of CRE.

The subsequent phase of our research entails the implementation of these new CALIPSO retrievals in a GCM, known as the Community Atmosphere Model, Version 6 (CAM6), through a simple yet prognostic approach. This encompasses the deactivation of the so-called pre-existing ice treatment, which typically yields unrealistically low ice particle concentrations (due to the use of layer-mean ice mass mixing ratios rather than the much lower ice mixing ratios near cloud top where ice nucleation occurs). “Turning off” the pre-existing ice treatment is also justified by the absence of orographic gravity waves in CAM6 (Lyu et al., 2023, JGR), which would enhance the vertical motions required for reaching the supersaturations needed for hom. Additionally, we estimate D_e from CALIPSO retrievals, accounting for different temperature ranges, latitude bands, seasons, and surface types (land or ocean). This is done for both the het-only and the CALIPSO retrieval constrained (hereafter natural) simulations. By computing terminal ice fall velocity (V_t) from D_e, we modify the simulated IWC as a function of V_t. With this parameterization for het-only conditions, larger D_e is expected which corresponds to faster V_t and consequently, lower IWC, cloud lifetime, and cloud cover. Currently, this is an ongoing procedure that includes running three CAM6 experiments: control (default CAM6), het-only, and natural conditions. The intent is to comprehensively assess the impact of CCT within a complex atmospheric system.