Stratospheric measurements of water vapor (H2O) and methane (CH4) by the Halogen Occultation Experiment (HALOE) on board the Upper Atmosphere Research Satellite (UARS) during 1991-2000 are used to examine seasonal variation in an ascent rate of the annually varying signature of [H2O]+2[CH4] (^H) in the tropical lower stratosphere. By calculating a vertical lag-correlation between the ^H profiles at a certain and the next time step, the ascent rate is estimated for every latitude, altitude and time. Seasonal variation of the ascent rate is not the same as that of the actual vertical velocity, but a good indicator of the actual one.

The ascent rate of ^H exhibits a large seasonal variation with a maximum of 0.2-0.4 mm/s (increasing with height from 60 to 15 hPa) during December-January, and a minimum of 0.15-0.2 mm/s for June-July within 15 degrees of the equator. On the other hand, the ascent rate of ^H anomalies from the time mean profile at each altitude shows no significant seasonal cycle. This difference between the seasonal cycles in ascending motions of ^H and its time anomaly suggests that we should use raw data of ^H to examine seasonal variation of vertical velocity, since a seasonal cycle in vertical velocity can produce vertical inhomogeneity of the time average of ^H. The observed amplitude ranges from 0.05 to 0.1 mm/s increasing with height between 15 and 60 hPa, and is larger in the Southern Hemisphere than in the Northern Hemisphere. The hemispheric difference of the seasonal amplitude is interpreted as the sum of symmetric and asymmetric modes of seasonal cycle with respect to the equator: the former is associated with a larger extratropical pump by planetary waves in the Northern winter season; and the former is with seasonality in each hemisphere. We also found a double peak structure in the ascent rate for December-February between 20 and 40 hPa.

Analysis of temperature fields from Singapore rawinsonde data, the National Centers for Environmental Prediction (NCEP) reanalysis data, and the U.K. Meteorological Office (UKMO) assimilation data, confirms that the ascent rate variation is closely anti-correlated with temperature variation above 40 hPa, so that ascent anomalies lift up the stratified isentropic surfaces relatively, resulting in negative temperature anomalies. In the 40-60 hPa level, however, the ascent anomalies precede the negative anomalies of temperature variations by 1-2 months. Assuming that a radiative heating is expressed using the Newtonian cooling form, the radiative relaxation coefficient obtained from a regression analysis between variations of ascent rate and temperature is shown to be 1/(10-20 day) for the 20-40 hPa layer, and 1/(20-50 day) for the 40-60 hPa layer. The long relaxation timescale observed between 40 and 60 hPa is twice as much as the timescales used generally (about 10-20 days), but consistency between the phase lag and the regression analysis confirms that these values could be valid for the seasonal variation.