Diagnosing CH₄ models using the equivalent length in the stratosphere

Abstract. The equivalent length, a measure for mixing strength in the atmosphere, in meridional direction in the current application, is used here to investigate the causes of the atmospheric chemistry model (TM3, TM5-4DVAR and LMDz-PYVAR) biases in the stratosphere. Compared to measurements, we find that the modeled surf zone (a strongly stirred region caused by planetary wave breaking in mid-latitude stratosphere in the winter hemisphere), especially in the southern hemisphere, is not strong enough. We assume that this is due to an underestimation of the planetary wave breaking magnitude in the models. Consequently, the region with meridional uniform stratospheric CH₄ concentrations has smaller latitudinal coverages in the models than the measurements, especially in the southern hemisphere between June and September. During the southern winter, a region with both vertically and horizontally well mixed CH₄ concentrations occur between 450 and 850 K (~ 18 and 30 km) in surf zone latitudes. Such a region is absent in the models, and underestimations of CH₄ concentrations within it are visible in comparisons with measured CH₄ profiles. The modeled polar vortex breaks too fast and during the vortex period CH₄ concentration differences across its barrier are underestimated compared to the measurements.