Repository logo

Simulations of Arctic mixed-phase clouds using a new aerosol-linked ice nuclei parameterization in a prognostic ice prediction scheme




Carpenter, James Michael, author
Kreidenweis, Sonia M., advisor
DeMott, Paul J., advisor
Randall, David A., committee member
Eykholt, Richard, committee member

Journal Title

Journal ISSN

Volume Title


Despite the nearly universally-accepted notion that the Arctic is one of the most important areas to fully understand in the face of a changing global climate, observations from the region remain sparse, particularly of clouds and aerosol concentrations and sources. Low-level, mixed-phase clouds in the Arctic are capable of remarkable persistence, lasting for several days when our knowledge of the Wegener-Bergeron-Findeisen (WBF) process suggests that complete conversion to ice, or glaciation, should occur much faster, within a couple of hours. Multiple attempts at simulating these long-lived, mixed-phase clouds have been unable to accurately reproduce all cloud properties observed, with a major consequence being poor representation of radiative transfer, with important consequences for long-term climate simulations. Recent observational campaigns have sought to characterize ice-nucleating particles (IN) not just in the Arctic, but around the planet. A product of these campaigns, the DeMott IN parameterization (DeMott et al., 2010) seeks to provide a means for accurately implementing IN concentration calculations in a global model using minimal, readily-available proxy measurements or estimates of number concentrations of particles having diameters larger than 0.5 microns. In this study, the performance of this parameterization is tested in a cloud-resolving model capable of high resolution simulations of Arctic mixed-phase boundary layer stratus clouds. Three mixed-phase cloud case studies observed during the Indirect and Semi-Direct Aerosol Campaign (ISDAC) and Mixed-Phase Arctic Cloud Experiment (M-PACE) are simulated with varying complexity in their cloud microphysical packages. The goal is to test the new aerosol-linked parameterization as well as the sensitivity of the observed clouds to ice nuclei concentrations. In an effort to increase the realism of the aerosol-cloud interactions represented in the cloud-resolving model, a new, simple prognostic scheme for the activation of ice nuclei is incorporated. The new scheme imposes a finite budget on potential ice nuclei, which are depleted through ice activation and growth, and can potentially be replenished by sublimating ice crystals. Results are contrasted with simulations in which no depletion of IN is assumed. In this study, we found that while the DeMott IN parameterization successfully predicted available IN concentrations within observational error, the model was unable to predict sufficiently high pristine ice concentrations for one of the case studies. There were likely issues with the model or initialization in this case. For two of the case studies, the model performed exceptionally well, predicting accurate ice number concentrations as well as cloud droplet concentrations, leading to reasonable predictions of downwelling longwave radiation at the surface. In all cases, the model failed to predict reasonable cloud ice water contents. In the future, tests of ice crystal habits and growth rates may improve microphysical representation and predicted ice water contents. Replenishment of scavenged ice nuclei via surface fluxes and long-range transport can be included in the simulations to increase realism, but more observations are needed to accurately quantify these effects.


Rights Access




Associated Publications