Summary of achievements

The connection between clouds and climate is central to the problem of understanding Earth's changing climate. The clouds that are included in climate models are not realistic, especially low-level clouds associated with the turbulent atmospheric boundary layer. This deficiency of climate models is regarded as the leading source of uncertainty in projections of climate change. Brian's research focuses on clouds and their representation in climate models from several angles. In one approach, the Community Atmosphere Model is run in a short-term forecast mode with high-frequency output saved over regions of interest, following the Cloud-Associated Parameterization Testbed (CAPT) framework. This allows us to evaluate the simulations using spatially and temporally limited observations from, for example, field campaigns. It also allows us to investigate biases in the the model physics that are manifest as the model diverges from realistic initial conditions. Another approach is to use idealized climate model simulations to better understand the roles of the boundary layer and cloud processes in climate and climate change. Particularly interesting has been the use of a water-covered Earth configuration called the aquaplanet. These experiments reproduce the Earth-configuration's tropical cloud response to a change in climate, and provide a simplified view of this response that emphasizes the role of shallow cumulus convection in the disagreement among models. Brian's other projects also focus on the atmospheric boundary layer and clouds, including an analysis of the statistics of low clouds from satellite-based lidar measurements and investigating the structure of the arctic boundary layer across climate models.

Publications

Zhang, Y.Y., B. Stevens, B. Medeiros, and M. Ghil. 2009: Low-Cloud Fraction, Lower-Tropospheric Stability, and Large-Scale Divergence. Journal of Climate, 22, 4827-4844, doi:10.1175/2009JCLI2891.1.

Figure 1: High resolution figure

Abstract: This paper explores the capability of the mixed-layer model (MLM) to represent the observed relationship between low-cloud fraction and lower-tropospheric stability; it also investigates the influence of large-scale meteorological fields and their variability on this relationship. The MLM's local equilibrium solutions are examined subject to realistic boundary forcings that are derived from data of the 40-yr European Centre for Medium-Range Weather Forecasts Re-Analysis (ERA-40). The MLM is successful in reproducing the positive correlation between low-cloud fraction and lower-tropospheric stability. The most accurate relationship emerges when the forcings capture synoptic variability, in particular, the daily varying large-scale divergence is a leading factor in improving the regression slope.

The feature of the results is mainly attributed to the model cloud fraction's intrinsic nonlinear response to the divergence field. Given this nonlinearity, the full range of divergence must be accounted for since a broad distribution of divergences will give a better cloud fraction overall, although model biases might still affect individual MLM results. The model cloud fraction responds rather linearly to lower-tropospheric stability, and the distribution of the latter is less sensitive to sampling at different time scales than divergence. The strongest relationship between cloud fraction and stability emerges in the range of intermediate stability values. This conditional dependence is evident in both model results and observations. The observed correlation between cloud fraction and stability may thus depend on the underlying distribution of weather noise, and hence may not be appropriate in situations where such statistics can be expected to change.

Figure caption: Large-scale divergence over ocean inferred from ERA-40. (top) The seasonal mean of JJA in 2001, and (bottom) the daily mean on 16 Jul 2001.