Diagnosis and Sensitivity of the 200 hPa Circulation in NCAR Community Climate Models

The purpose of this report is to determine the nature of systematic errors in several different seasonal cycle versions of the National Center for Atmospheric Research (NCAR) Community Climate Models (CCMs) and to assess the dynamical impact of such errors on climate perturbation studies. Our diagnosis focuses on the seasonal mean December- January-February (DJF) 200 hPa circulation and evaluates the sensitivity of that flow to anomalous tropical forcing. Other features of the CCM climatologies for January and July, including the mass fields and hydrologic cycle, are evaluated in Hurrell et al. (1993). A linear barotropic stationary wave model is used to examine the sensitivity of observed and CCM climatological flows to idealized tropical forcing. The forcing consists of a mass source/sink dipole over the equatorial Pacific that mimics the horizontal structure of anomalous rainfall during El Niño. The climatological flows on which the forcing is imposed differ among the various CCMs, and these in turn differ from observations. Differences between the responses to tropical forcing imposed on each base state relative to observations then indicate the dynamical significance of the CCM climate errors. The northern wintertime teleconnections during El Niño are particularly strong with the Northern Hemisphere (NH) flow, and our assessment of the significance of the CCM climate biases focuses on how such errors impact the NH stationary wave response. Of course, the full CCM behavior during El Niño is considerably more complicated than just the steady barotropic response to tropical forcing. As such, the approach taken herein should be viewed as only one component of a more thorough model evaluation that ultimately requires diagnosis of CCM integrations with interannual sea surface temperature (SST) anomalies. Relationships between climate errors in CCM divergent and rotational flow components are examined in the context of the linear model. These experiments are designed to determine the link between various features of the CCM 200 hPa circulation biases, and to provide guidance for, model development. Our principal results common to all the CCM versions studied herein are: * The zonally averaged zonal wind at 200 hPa has a westerly bias in the tropics. The error is not zonally uniform but is strong locally with a 20 m s-1 maximum over the equatorial central Pacific. This is associated with an erroneous westward shift of the CCMs' subtropical Pacific stationary wave troughs in both hemispheres.