Characterization of Oceanic Post-Cold Frontal Clouds and their Model Representation

Principal Investigator(s):
Catherine Naud, The Trustees of Columbia University in the City of New York

James Booth, CUNY-City College of New York
Andrew Gettelman, NCAR

Extratropical cyclones are the subject of active research because of their role in transporting heat from the equator to the poles, and also because they can generate extreme weather in some of the most populous regions of the planet. Recent studies have demonstrated the importance of the clouds in these systems, as they control the radiation budget of the midlatitude temperate regions. However, general circulation models (GCMs) have difficulty generating the right amount of clouds in these cyclones, and more specifically in the region of the storm behind the cold front: the post-cold frontal zone. This part of the storm is mainly populated by boundary layer clouds, i.e. clouds occupying the first few kilometers above the surface. However, the clouds’ properties (e.g. liquid and ice water content, altitude, depth, opacity) are not well documented. Recent model experiments with improved boundary layer representations proved insufficient for solving the modeled cloud issue. The reason may stem from other model issues such as the cloud microphysics parameterization which simulates the properties of these clouds or biases in the large scale conditions, such as the strength or temperature of the cyclones.

To help improve the models, a detailed characterization, based on observations, of the post-cold frontal cloud conditions is needed. The ARM Eastern North Atlantic (ENA) site offers the necessary datasets to 1) explore the relationship between large scale conditions and cloud properties in post-cold frontal zones and 2) allow to test each aspect of the models independently.  In addition, the ARM Macquarie Island Cloud and Radiation Experiment (MICRE) campaign will allow for a similar analysis of the southern hemisphere. The same analysis at two very different locations will help ensure the generality of our findings.

Our methodology identifies extratropical cyclones and associated cold fronts using observation-based datasets called reanalysis. Once post-cold frontal conditions have been identified, ARM radiosoundings of temperature profiles will be used to characterize the thermodynamic structure of the lower troposphere. Separately, the measurements from the radar, LIDAR and radiometers will be used to derive the cloud properties (e.g., cloud cover, cloud vertical extent, cloud thermodynamic phase profiles). This analysis will be carried out for both the ARM ENA site and MICRE campaign. For each region, conditional subsetting will be used to examine the relationships between the cloud properties and large scale and surface conditions. This method entails selecting the observations based on local or large scale conditions and comparing the cloud properties between the most extreme of these conditions (e.g. the wettest versus the driest conditions). Additionally, the results for the two observation stations will be contrasted to test the sensitivity of the results to the large-scale conditions. These datasets will be used for a statistical analysis and the detailed analysis of case studies. The same cases will be used to test two models: WRF (a fine resolution, regional model for weather simulation) and CAM (a coarser resolution, global model for climate simulation).

Due to the overlap in parameterizations in CAM and WRF, using the two models provides an opportunity to learn from both climatological and case study experiments. Three configurations of the CAM model will be used to generate climatological post-cold frontal conditions for comparison with the observations and begin to separate the most important parameterizations. Separately, a perturbed physics ensemble (multiple experiments with independent and small changes in each) will be generated using WRF, based on case studies that are determined by the observational analysis. These runs will provide ensemble sensitivity metrics of clouds (for properties such as optical thickness, liquid and ice content), as well as modeled moisture and heating tendencies to parse the roles of different physics in the model generation and maintenance of clouds. Finally, case studies will be generated with CAM, so that we can compare the GCM physics with that of the numerical weather model and observations.

This research will generate three main deliverables related to post-cold frontal clouds: (1) a multi-scaled analysis of the observed cloud properties, (2) the determination of the relative importance of the large-scale circulation and local thermodynamics, and (3) an evaluation the causes of clouds biases, and a recommendation for removing the biases in GCMs. This work will help advance our understanding of cloud formation and removal in extratropical cyclones, and ultimately our ability to better predict whether the cyclones will become more destructive in a warming climate.