What influences the middle atmosphere circulation in the IFS?

TitleWhat influences the middle atmosphere circulation in the IFS?
Publication TypeMiscellaneous
Year of Publication2017
AuthorsPolichtchouk, I, Hogan, RJ, Shepherd, TG, Bechtold, P, Stockdale, T, Malardel, S, Lock, S-J, Magnusson, L
Secondary TitleTechnical Memorandum
Number809
Abstract

The presence of large (~20 K) middle atmosphere temperature biases motivates an investigation into the representation of the middle atmosphere circulation in the European Centre for Medium-Range Weather Forecasts (ECMWF) Integrated Forecast System (IFS). The focus of this report is on the Brewer-Dobson circulation (BDC) and the zonal-mean temperature and zonal wind structure, and their dependence on: 1) the gravity wave drag parameterizations; 2) the cubic octahedral (TCo) grid and the stochastic physics (SPPT) scheme; 3) the “sponge” in the upper part of the model, designed to prevent wave reflection; and 4) the inclusion of an approximate diurnal cycle of ozone and the reduction of solar UV output. Long free-running model runs are used to assess the performance and sensitivity of the modelled climate. Nudged runs, where the resolved wave fluxes entering the stratosphere are constrained by the observations, are used to assess the model’s performance and sensitivity for specific Northern Hemisphere (NH) and Southern Hemisphere (SH) winter conditions. All the results apply to the T255L137 resolution, where a substantial part of the inertia-gravity wave activity is unresolved.

Overall, the parametrized non-orographic gravity wave drag (NOGWD) has the largest impact on the middle atmosphere circulation in all runs. The polar cap downwelling responds to perturbations in NOGWD flux, but there are complex interactions between NOGWD and resolved wave drag in both polar regions in the time-mean and during the seasonal cycle. NOGWD also affects the quasibiennial oscillation (QBO) and the mesospheric semi-annual oscillation (SAO): The QBO frequency and the QBO and the SAO westerly phase amplitude all increase in response to an increase in the NOGWD, and the SAO easterly phase amplitude decreases in response to a decrease in the NOGWD. The nudging framework allows us to isolate the impact of parametrized orographic gravity wave drag (OGWD) on the polar cap downwelling: NOGWD completely compensates for the removal of OGWD in the NH, but the compensation has an impact on the resolved wave drag.

The TCo grid coupled with the SPPT scheme increases the strength of the lower-stratospheric upwelling and the frequency and westerly phase amplitude of the QBO compared to the control run at the same spectral truncation with linear grid and without SPPT. In the mesosphere, removing the sponge increases the westerly phase of the SAO by 15 times as the resolved waves are allowed to freely propagate into the mesosphere. However, the sponge has no impact on the QBO and little impact on the BDC. The inclusion of the diurnal cycle of ozone and the reduction in solar UV have little impact on the stratospheric circulation despite greatly improving the global warm mesospheric temperature bias.