Some Diagnostics of the Northern Wintertime Climate Simulated by the ECHAM3 Model

Abstract

The control climate of an atmospheric general circulation model, ECHAM3, is validated with emphasis on the wintertime planetary- and cyclone-scale wave propagation and storm tracks in the Northern Hemisphere. By applying a set of time filters, it is shown that ECHAM3 has too much variance in the period range of 6?10 days. Geographically, the North Pacific and North Atlantic storm tracks are too strong and extend too far into the western parts of the continents. This overestimated variance compensates the deficient low-frequency variability in ECHAM3, so that the total variance is well simulated by the model. These problems seem to have a relatively small effect on the simulation of cyclone-scale eddies. The cyclone climatology of ECHAM3 is studied using a simple method in which a cyclone is defined as a local minimum in the 1000-hPa height field. The climatology is found to be reasonably close to that observed. The propagation of planetary- and cyclone-scale waves in ECHAM3 is studied using various approaches based on lag-correlation maps of geopotential height. Zonal wavelength and group velocity calculated from the unfiltered data are close to that observed, but the easterly phase speed is too large in the model. Synoptic-scale eddies in the model have their phase speed, horizontal wavelength, and vertical tilt close to those in the observations, but the lifetime of the eddies is too long. The essential characteristics of the vertical wave propagation from the troposphere to the stratosphere are captured by the model. It is speculated that some of the problems in the model simulation might be a consequence of inadequate orographic forcing in ECHAM3. The mountain pressure torque in the model is of the right magnitude, but this is partly due to the erroneous sea level pressure field. However, identifying the possible role of orography in the quality of the model simulation would need further experiments.