George Tselioudis (PI)
Goddard Institute for Space Studies
gtselioudis@giss.nasa.gov

Cloud, Radiation, and Precipitation Changes with Dynamic Regime: An Observational Analysis and Model Evaluation Study

We propose to analyze (1) global satellite observations and regional observational subsets of clouds, radiation, and precipitation and (2) output from the GISS Global Climate Model (GCM) and Single Column Model (SCM) and from the Regional Atmospheric Modeling System (RAMS) in order to address the following scientific questions:    What are the main regimes of variability of midlatitude cloud, radiation, and precipitation fields and what is their relationship to the large-scale atmospheric dynamics?    How does subgrid-scale (defined with respect to a GCM) variability of atmospheric dynamics, in particular vertical motion, affect cloud, radiation, and precipitation properties in midlatitude clouds?    Besides subgrid variability in vertical motion, are there additional important contributions to subgrid cloud, radiation, and precipitation variability from other variables, e.g., saturation mixing ratio (temperature) and horizontal advection (both of air masses and hydrometeors)?    What is the relationship between the subgrid dynamical and thermodynamical processes that control these subgrid cloud, radiation, and precipitation distributions and the resolved-scale variables on the GCM grid?  Compositing and clustering techniques will be applied to satellite and ground-based observations to define the full spectrum of variability of the midlatitude cloud, radiation, and precipitation fields and to examine the dependence of those fields on atmospheric dynamics. The analysis will include both global satellite and reanalysis datasets as well as regional satellite and field study subsets from the Data Integration for Model Evaluation (DIME) web archive of the Global Energy and Water Cycle (GEWEX) Cloud System Study (GCSS). The results of this analysis will be provided to the DIME database so that it can be available to the larger community.  These compositing and clustering techniques will also be applied to output from model runs with the GISS GCM and SCM. For the GCM, multi-year, current-climate runs will be analyzed and statistical composites and clusters derived for both the global midlatitude area and for smaller regions characteristic of the different midlatitude dynamic regimes. The runs with the SCM will span periods of a month to a season for the same smaller regions to obtain statistically meaningful composites and clusters. Direct comparisons of the observational and modeling composites and clusters will allow us to identify deficiencies in the cloud, radiation, and precipitation fields of the these models and provide some insights into the physical processes responsible for the identified model errors. As with the observational analysis, output and statistics from these simulations (and also from the RAMS runs discussed below) will also be made available via the DIME web site.  Finally, we will run RAMS for the individual regions of interest, again for timescales of a month to a season, to test the hypothesis that it is the unresolved dynamical variability in the coarse-resolution GCM and SCM that is responsible for differences in the modeled and observed cloud, radiation, and precipitation statistics. Specifically, we will run RAMS across a range of resolutions, from coarse resolutions similar to those of a GCM, through mesoscales, and up to “cloud resolving” scales. To all of these simulations at these different resolutions, we will apply the same compositing and clustering techniques in order to determine how changing resolution alters the model statistics and the agreement between model and observations. In other words, the use of a regional model like RAMS with flexible resolution will allow us to do three things: (1) quantify the sensitivity of the cloud, radiation, and precipitation variability to subgrid-scale (to a GCM) dynamical variability; (2) explicitly characterize this subgrid variability in key dynamical and thermodynamical variables; and (3) derive quantitative relationships between this subgrid variability and the variables at the scale of the GCM grid.

+ Back to Participant Listing