Impacts of Introducing Convective Gravity Wave Parameterization upon Tropical Variability in the Met Office Unified Model

A convective gravity-wave (CGW) parameterization is implemented in the Met Office Unified Model (UM) and its impacts on tropical stratospheric variability are investigated using a climate configuration of the UM. A pair of AMIP-II experiments is performed with and without the CGW parameterization (SC and CTL experiments, respectively) for 12.5 years from September 1978. Two major impacts in the tropical stratosphere appear by including the CGW parameterization, upon (i) the quasi-biennial oscillation (QBO) and (ii) the annual cycle in the upper stratosphere. The period of the simulated QBO becomes longer (29 months) when the CGW parameterization is included, and this longer period is the same as in the reanalysis in 1979-1990. Also, phase-stalling in the easterly QBO is simulated in the SC experiment in the same year as in the reanalysis (1989), although the stalling is weaker than in the reanalysis. The differences in the QBO period and amplitude between the two experiments are likely related to the differences in the spectral shape and temporal variability between the CGWs that take into account spatiotemporal variation of convective sources and the background gravity waves (BGW) that launch at every model gridpoint constantly. The CGWs have a much broader spectrum with peaks at higher phase speeds than the BGWs. The other major impact of the CGW parameterization is strengthening of the annual cycle of zonal wind in the upper stratosphere, which closely matches the reanalyses in both amplitude and phase. This is because the tropical momentum flux into the stratosphere by the CGWs has a strong annual cycle, due to the seasonality of tropospheric convection, with a maximum momentum flux near 10° in the summer hemisphere by eastward-propagating waves.