Water vapour is a key climate variable since it accounts for about 60% of the natural greenhouse effect for clear skies (Kiehl and Trenberth, 1997) and provides the largest positive feedback in model projections of climate change (Held and Soden, 2000). During the 1980s and 1990s, stratospheric water vapour (H2O) was observed from balloon measurements over Boulder to increase with a trend of 1-1.5 %/yr (Oltmans et al., 2000). While this trend seemed to be confirmed by HALOE observations between 1992 and 1996 (Randel et al., 2004), the Boulder and HALOE observations diverged from 1997 to 2001. Since tropical tropopause temperatures slightly decreased from 1980 to 2000 (Zhou et al., 2001) decadal-scale changes in H2O and temperature seem uncorrelated. The small-scale processes ruling the H2O transport through the tropical troposphere and the related dehydration mechanisms are currently under debate (e.g., Sherwood and Dessler, 2000; Holton and Gettelman, 2001; Schoeberl et al., 2006).
Both HALOE and Boulder data show a sudden drop in stratospheric H2O in 2001 and persistent low values since then (Randel et al., 2006), correlated with a cold anomaly in tropopause temperatures during this period which is consistent with the observed H2O decrease. Randel et al. (2006) showed that the reduced TTL cold point temperature is related to an increase in the mean tropical upwelling associated with an enhanced Brewer-Dobson circulation (BDC) since 2001. This is consistent with observations of a sudden increase in extra-tropical tropospheric wave forcing from both hemispheres that drives the BDC (Dhomse et al., 2008). However, there is also some evidence that other processes may have a larger impact on stratospheric H2O in some years (Fu et al., 2006; Hanisco et al., 2007; Dhomse et al., 2008). It is unclear to date if the observed changes in stratospheric H2O and global circulation are part of a low-frequency natural variability or a monotonic trend which may be related to climate change.
Besides import of tropospheric air, oxidation of methane is another source of stratospheric H2O. The methane increase during 1980-2000 can explain about 1/3 of the stratospheric H2O increase. The efficiency of methane oxidation is controlled by the stratospheric composition (Saueressig et al., 2001; Röckmann et al., 2004). An open question is how the methane oxidation efficiency will evolve in future due to changing atmospheric composition and to which degree this will be compensated by a further increase of methane, for which various scenarios are under debate.