127SOLA, 2016, Vol. 12, 127−134, doi:10.2151/sola.2016-028

AbstractThe radiative and chemical interactions of stratospheric water

vapour (SWV) mean that SWV has a significant influence on the climate. Based on the Stratospheric Water and OzOne Satellite Homogenized (SWOOSH) SWV data (100−1 hPa, 60°S to 60°N), variations in the middle and upper SWV are mainly investigated over the past two decades. Water vapour variations below 10 hPa are found to result from upward transport of the lower SWV. Using empirical orthogonal function analysis and regression analysis of the anomalous upper SWV variability, it is found that upper SWV variations have an 11-year period controlled by the solar cycle as well as a 2-year cycle regulated by the quasi-bien-nial oscillation (QBO). Our results also show that state-of-the-art chemistry–climate models perform poorly in simulating the upper SWV variations. It is possibly because the effect of lower–middle SWV changes on the upper SWV variations is too strong in the simulations.

(Citation: Xie, F., J. Li, W. Tian, X. Zhou, and X. Ma, 2016: The variations in middle and upper stratospheric water vapour over the past two decades. SOLA, 12, 127−134, doi:10.2151/sola. 2016-028.)

1. Introduction

Due to its radiative and chemical significance, variations in stratospheric water vapour (SWV) can influence climate change (Dessler et al. 2008, 2013; Shindell 2001; Forster and Shine 2002; Solomon et al. 2010); consequently, a better understanding of these variations would help to improve our predictions of future climate change (Oninas et al. 2001; Stenke and Grewe 2005; Tian et al. 2009). Understanding the observed SWV variations and determining the factors that control such variations have become subjects of great interest in recent years (Holton and Gettelman 2001; Rosenlof et al. 2001; Fueglistaler et al. 2005; Randel et al. 2006; Dhomse et al. 2008; Wang et al. 2014, 2015). Observations have shown that SWV increased between 1981 and 2000, began to decrease after 2000, before increasing again after 2005 (Oltmans et al. 2000; Hurst et al. 2011). Although these changes in SWV levels have been widely studied (Sherwood and Dessler 2000; Rosenlof 2003; Scaife et al. 2003; Fueglistaler and Haynes 2005; Jiang et al. 2007, 2015; Fu 2013; Hegglin et al. 2014; Xie et al. 2014a), the characteristics of the SWV variations and their con-trolling factors remain debated. Because of the vertical transport of water vapour (known as the SWV tape recorder signal; Mote et al. 1996), the changes in SWV are considered to result from changes in tropospheric water vapour entering the stratosphere.

Consequently, the SWV is thought to be controlled by the tropical cold-point tropopause temperature (Randel and Jensen 2013), because of the freeze-drying that occurs when tropospheric air enters the stratosphere in the tropics (Brewer 1949). Although water vapour would freeze at the cold-point tropopause, it is thought that strong and deep convection might transport ice into the stratosphere (Rosenlof 2003). A simulation study showed that variations in transport explain 30% of the lower SWV changes over the past several decades (Tian and Chipperfield 2006). In particular, the upward transport in the Asian summer monsoon is thought to be capable of transporting a large amount of water vapour into the stratosphere (Gettelman et al. 2004; Fu et al. 2006; Su et al. 2011; Bian et al. 2011, 2012). As volcanic eruptions can change the tropopause temperature and bring tropospheric water vapour into the stratosphere, SWV can increase significantly after such eruptions (Considine et al. 2001; Joshi and Jones 2009). Garfinkel et al. (2013a) suggested that the decrease in SWV since 2000 was caused by an increase in sea surface temperature over the western Pacific warm pool, which has cooled the tropopause. In addition, Dessler et al. (2013) emphasized the role of variations in the quasi-biennial oscillation (QBO) as a primary control on the SWV changes.

As much of the attention has focused on understanding changes in the lower SWV, we have limited knowledge of the vari-ations in the middle and upper SWV. Because the tape recorder signal can be found at about 10 hPa, it is believed that variations in the middle and upper SWV are also controlled mainly by the upward transport of water vapour (Mote et al. 1996). Using a cou-pled chemistry–climate model, the water vapour trend in the upper stratosphere has been shown to be controlled mainly by methane oxidation, and the contribution from the chemical decomposition of methane in the middle and upper SWV has also been estimated (Rosenlof et al. 2001; Tian and Chipperfield 2006). The influence of the 11-year cycle in solar irradiation on the middle–upper SWV has been investigated using simulations (Woods et al. 2000; Gabriel et al. 2011). The limited observations indicate that the semi-annual oscillation (SAO) and the QBO signals consist in the upper SWV (Jackson et al. 1998; Dunkerton 2001). Tao et al. (2015a, b) used a chemical Lagrangian model to investigate the middle and upper SWV variations caused by the QBO combined with a sudden stratospheric warming event. These studies ana-lyzed the different effects of various factors on the middle–upper SWV; however, the nature of the variations in the middle and upper SWV over recent decades, and the factors that have con-trolled these variations, have not been studied in detail.

2. Data and methods

The stratospheric water vapor (SWV, 100−1 hPa, 60°S to 60°N) is divided into lower SWV (LSWV, 100−70 hPa, 60°S to 60°N), lower–middle SWV (LMSWV, 70−30 hPa, 60°S to 60°N), middle–upper SWV (MUSWV, 30−10 hPa, 60°S to 60°N), and upper SWV (USWV, 10−1 hPa, 60°S to 60°N) in this study. The

The Variations in Middle and Upper Stratospheric Water Vapour over the Past Two Decades

Fei Xie1, 2, Jianping Li1, 2, Wenshou Tian3, Xin Zhou4, 5 and Xuan Ma1, 2

1College of Global Change and Earth System Science, Beijing Normal University, Beijing, China2Joint Center for Global Change Studies, Beijing, China

3Key Laboratory for Semi-Arid Climate Change of the Ministry of Education, College of Atmospheric Sciences, Lanzhou University, Lanzhou, China

4State Key Laboratory of Numerical Modeling for Atmospheric Sciences and Geophysical Fluid Dynamics, Institute of Atmospheric Physics, Chinese Academy of Sciences, Beijing, China

5University of Chinese Academy of Sciences, Beijing, China

Corresponding author: Jianping Li, College of Global Change and Earth System Science (GCESS), Beijing Normal University, Xueyuan South Road, HaiDian, Beijing 100875 China. E-mail: ljp@bnu.edu.cn. ©2016, the Meteorological Society of Japan.

128 Xie et al., Variations in Middle and Upper Stratospheric Water Vapour

MUSWV and USWV from 1992 to 2013 based on the SWOOSH data. The data used in the study are described in Section 2. The LMSWV and MUSWV increase from 1992 to 2000, but decrease after 2000, which is essentially in agreement with the variations in LSWV over the past two decades (Figs. 1a and 1e). A large volca-nic eruption in the early 1990s and a series of smaller eruptions in the tropics in the 2000s affecting tropical tropopause temperature may be the cause of the lower and middle SWV variations after 1992 and 2000, respectively (Glaze et al. 1997; Textor et al. 2003; Joshi and Shine 2003; Joshi and Jones 2009; Chiodo et al. 2014; Schieferdecker et al. 2015). The variations in LSWV, LMSWV, and MUSWV from 1992 to 2013 all show a linear decreasing trend. Figures 1b and 1d show the lead–lag correlations among the low-pass filtered LSWV, LMSWV, and MUSWV (Low-pass filter is in order to get rid of the impact of QBO and high-fre-quency signal on the correlation coefficient between water vapor variations at different layers). Significant maximum negative correlations and are observed at a lag of ~6 months (R = ~0.85) when the LSWV leads LMSWV and ~12 months (R = ~0.81) when the LMWV leads MUSWV. It is found that the average speed of vertical velocity of BD circulation between 100 hPa and 30 hPa (interval 70 hPa) is −0.0005 Pa/s and between 70 hPa and 10 hPa (interval ~60 hPa) is −0.00025 Pa/s. Thus, the water vapor from 100 hPa to 30 hPa needs ~6 months and from 70 hPa to 10 hPa needs ~12 months. The upward propagation of water vapor signal is well consistent with the speed of BD circulation. It illustrates that the variations in water vapour below 10 hPa result mainly from the LSWV variations because of the LSWV upward transport (Figs. 1a, 1b, 1c, and 1d).

However, the USWV variations are not in agreement with the MUSWV variations (Fig. 1e). The linear trend of USWV is increasing whereas that of the MUSWV is decreasing. In addition, the lead–lag correlation between the MUSWV and USWV varia-tions is not significant when the MUSWV leads USWV (Fig. 1f ). This implies that the variations in USWV are not solely related to middle and lower SWV variations. It should be pointed out that the USWV also shows an increase over the period 1992−2000 and a decrease after 2000, which is a nature of the variations in the middle and lower SWV caused by large volcanic eruption. How-ever, the variations in the USWV may not be related to volcanic eruptions because the effects of major volcanic eruptions on SWV are considered to be limited basically below the 10 hPa level (Löffler et al. 2015).

To further analyze the variations in the USWV, an EOF analysis was performed on the USWV variations from 1992 to 2013. Figure 2a shows that the leading principal component (PC1, which accounts for ~30% of the variance) of USWV variations is strongly correlated with the variations in USWV, indicating that the leading EOF mode may well represent the temporal variability in USWV. Figure 2b shows the EOF mode (EOF1). The maximum loadings of the EOF1 spatial pattern are located in the tropical upper stratosphere, suggesting that the variations in the leading mode of the USWV result from a process in the upper stratosphere. However, the upward transport of lower and middle SWV anomalies, thereby influencing the USWV, deems to not a decisive factor on controlling the USWV variations.

To identify the process that controls PC1, Fig. 3a shows the results of a spectral analysis of PC1, which highlights a decadal variation (11-year period) and an interannual variation (2-year period). This indicates that the variations in PC1 are caused by both the solar cycle and QBO, which affect stratospheric water in different phases (Woods et al. 2000; Gabriel et al. 2011). To analyze the relationship between PC1 and the solar cycle, a low-pass filter was applied to PC1. However, the simultaneous correla-tion between the low-pass-filtered PC1 and the solar cycle is not significant (Fig. 3b). In fact, a lead correlation revealed that the low-pass-filtered PC1 is significantly anti-correlated with the solar cycle when the solar cycle exceeds PC1 variations for ~3 years (maximum correlation coefficient of −0.8; Fig. 3c). Solar activity can influence the upper stratospheric chemical compounds via chemical reactions (Texier et al. 1988; Matthes et al. 2004, 2011; Egorova et al. 2004; Austin et al. 2008; Gray et al. 2010); i.e.,

SWV variations were calculated by removing the seasonal cycle from the original time series and then dividing by its standard deviation (normalized).

The Stratospheric Water and OzOne Satellite Homogenized (SWOOSH) dataset is a merged record of stratospheric ozone and water vapour measurements taken by a number of limb sounding and solar occultation satellites over the past 30 years, and gives similar results to the Global Ozone Chemistry and Related trace gas Data Records for the Stratosphere (GOZCARDS) project (Froidevaux et al. 2015). The primary SWOOSH product is a monthly-mean zonal-mean gridded data set containing ozone and water vapor data from the SAGE-II/III, UARS HALOE, UARS MLS, and Aura MLS satellite instruments, spanning 1984 to present. For both ozone and water, corrections are applied to the SAGE, HALOE, and UARS MLS data to force agreement with the Aura MLS measurements. The corrections are additive offsets that vary with latitude and height, but not with time, and are determined from coincident observations closely matched in space and time during the instrument overlap time periods. Tropopause variables are calculated from the Modern-ERA Retrospective Analysis for Research and Applications (MERRA).

As there are too many missing values in the data before 1992 (after 1991 the HALOE data are available; however, the data col-lected during 1991 were contaminated by volcanic activity), this study uses data covering the period 1992−2013. This dataset has many kinds of horizontal and vertical grids. In this study, the hor-izontal resolution is 2.5° zonal mean (latitude: 90°S to 90°N) and vertical pressure range has 31 levels (316−1 hPa). The measure-ments are homogenized by applying corrections calculated from data collected during periods of instrument overlap. For further details, see http://www.esrl.noaa.gov/csd/groups/csd8/swoosh/usersguide.pdf.

The solar cycle index is based on monthly values of the F10.7 index (10.7 cm solar radio flux) from NOAA’s Space Environment Center (www.sec.noaa.gov) for the period 1992−2013. The 10-hPa zonal wind fields from the ERA-interim data are used to define the QBO index.

The statistical significance of the correlation between two autocorrelated time series was determined via a two-tailed Student’s t-test using the effective number (N eff ) of degrees of freedom (DOF). We determined the N eff of DOF (Bretherton et al. 1999) using the following approximation (Xie et al. 2014b; Sun et al. 2014):

1 1 21N N NN jN

j jeff XXj

N

YY≈ +−

=∑ r r( ) ( ) ,

where N is the sample size, and rXX and rYY are the autocorrela-tions of the two sampled time series, X and Y, respectively, at time lag j.

The formulae to calculate the BD circulation in a pressure coordinate system are given by Edmon et al. (1980),

ω ω φ φ θ θ φ* ( cos ) [cos (( ) / )]= + ′ ′−a v p

1

where q is the potential temperature, a is the radius of the earth, v is mean meridional wind, w is mean vertical velocity in pressure coordinates. Subscripts p and f denote derivatives with pressure p and latitude f, respectively. The overbar denotes the zonal mean and the accent denotes the deviations from the zonal mean value.

The water vapor data from 16 chemistry-climate models is also used in the study (Morgenstern et al. 2010). These 16 models attended the Chemistry-Climate Model Validation Activity (CCMVal2, which is to improve understanding of Chemistry- Climate Models and their underlying General Circulation Models through process-oriented evaluation, along with discussion and coordinated analysis of science results).

3. Results

Figure 1 shows the variations in the LSWV, LMSWV,

129SOLA, 2016, Vol. 12, 127−134, doi:10.2151/sola.2016-028

an increase in solar radiation increases atomic oxygen in the first excited electronic state in the upper stratosphere related to the photo dissociation of ozone, and the atomic oxygen can react with water vapour. Thus, the PC1 of USWV is nearly anti-correlated with solar radiation. The QBO impacts the stratospheric water vapor by two processes: QBO influences the dynamical transport of chemical compounds and results in a further feedback of distribution anomalies of chemical compounds (Wallace et al. 1993; Randel and Wu 1996); In the equatorial upper stratosphere, the negative/positive QBO (easterly/westerly phase) is associated with positive/negative temperature anomaly, which negatively/positively influence the ozone anomalies through photochemical reactions (Hauchecorne et al. 2010) that could photo-chemically bring about water vapor anomalies. To analyze the relationship between PC1 and the QBO, high-pass filtering was applied to PC1 (Fig. 3d), revealing that the high-frequency oscillation of PC1 is indeed related to the QBO. It must be noted that the semi-annual oscillation (SAO) signal can also be found in the high-frequency variations of PC1. Figure 3 shows that the solar cycle and the QBO, the main controlling factors on the stratosphere, might have a joint impact on the changes in USWV.

Figures 4a and 4b show the meridional cross sections of zonal mean USWV variations regressed onto the solar cycle and QBO indices, respectively. Both solar cycle and QBO seem to could cause maximal ~0.15 ppmv water vapor anomalies; however, the affected region by solar cycle is evidently larger than that by QBO. The USWV regressed onto solar cycle/QBO explains ~19%/12% variance. Figure 4c shows the joint impact of solar cycle and QBO on USWV. The spatial pattern of the joint impact is similar to the EOF1 pattern of USWV variations (Fig. 2b). EOF1 is considered to relate to solar cycle and QBO.

To investigate in more detail the variations in SWV at dif-

Fig. 1. (a) LSWV (100−70 hPa, 60°S−60°N averaged) and LMSWV (70−30 hPa, 60°S−60°N averaged) variations over the past two decades based on SWOOSH data. (b) Lead–lag correlation between the low-pass filtered LSWV and LMSWV variations. Positive lags refer to LSWV leading LMSWV. Low-pass filtering removes signals whose periods are shorter than 60 months from the original time series. The dotted lines denote the 95% confidence level. For explanation of statistical significance, see Section 2. (c) As for (a), but for LMSWV and MUSWV (30−10 hPa, 60°S−60°N averaged). (e) As for (a), but for MUSWV and USWV (30−10 hPa, 60°S−60°N averaged). (d) As for (b), but for LMSWV and MUSWV. (f ) As for (b), but for MUSWV and USWV.

Fig. 2. (a) PC1 and USWV variations. (b) Spatial pattern of EOF1. PC and EOF were calculated from EOF analyses of variations in zonal mean upper SWV over the period 1992−2013 in the SWOOSH data. The square root of the cosine of latitude was used as a weighting function in the EOF analysis (Sun et al. 2013).

130 Xie et al., Variations in Middle and Upper Stratospheric Water Vapour

ferent levels and to understand the chemical processes of solar influencing the USWV (e.g., explaining why the USWV variations lag solar activity), climate–chemistry model is required. Figure 5 shows the upper SWV variations from the simulations performed by models in CCMVAL2 and from SWOOSH. Figure 5 illustrates that the simulated upper SWV variations are not consistent with the observation. This result indicates that state-of-the-art chem-istry–climate models perform poorly in simulating the USWV, which is in agreement with previous studies (Eyring et al. 2010; Forster and Thompson 2011). However, the simulated upper SWV variations are in agreement with the simulated lower–middle SWV with the former lagging the latter (Fig. 6) in a number of models. This result is not corresponding to our result based on observations, i.e., the upper SWV variations are not in agreement with the lower-middle SWV. It implies that the impact of lower–middle SWV changes on the upper SWV caused by transport process may be too strong in the simulations, which covers the effect of chemical processes related to the solar cycle on upper SWV. In addition, the poor model performances likely indicate two problems in the models for water vapor simulation: first, most models have the largest error in simulating the amount of water vapor (lifted by convection) in the upper troposphere (Jiang et al. 2012); Second, the models also have difficulties in simulating the water vapor vertical transport across the tropopause from upper troposphere into the lower stratosphere and above, and even the reanalysis models have water vapor and chemical tracer vertical transport problems (Jiang et al. 2015; Minschwaner et al. 2016). Above analysis may explain the poor performances of chemistry- climate models in simulating the stratospheric water vapor shown in this study.

Fig. 4. (a) Meridional cross section of zonal mean upper stratospheric water vapor variations from 1992 to 2013 regressed onto the solar cycle index. Water vapor regressions are shown as local content (ppmv) vari-ations, with contour intervals of 0.05 ppmv. The solar cycle index leads the water vapor time series by 3.5-year in the regression, and its period is from 1989 to 2010. (b) Same as (a), but with water vapor regressed onto the simultaneous QBO index. (c) is the sum of (a) and (b). The value in the top right corner is the corresponding explained variance. The explained variance is calculated as the square of correlation coefficient between regressed USWV variations and averaged USWV variations.

Fig. 3. (a) Spectral analysis of the PC1 of USWV variations. The black line represents the spectral distribution. The green and red lines denote the 95% confidence level and the red noise spectrum, respectively. (b) Low-pass filtered PC1 (LowPC1) and simultaneous solar cycle index. Solar cycle based on F10.7 index (10.7 cm solar radio flux) from NOAA’s Space Environment Center (www.sec.noaa.gov). The F10.7 index multiplied by -1. Low-pass filtering removes signals whose periods are shorter than 60 months from the original time series. (c) Lead correlation between solar cycle and low-pass filtered PC1. A positive lead refers to the solar cycle leading PC1. The dotted lines denote the 95% confidence level. For explanation of statistical significance, see Section 2. (d) High-pass filtered PC1 (HighPC1) and simultaneous QBO × −1 index. QBO index is represented by the 10-hPa U wind. High-pass filtering removes signals whose periods are longer than 30 months from the original time series.

131SOLA, 2016, Vol. 12, 127−134, doi:10.2151/sola.2016-028

4. Conclusions

This study mainly investigated the variations in middle and upper SWV over the past two decades and attempted to identify the main controlling factors of these variations. We found that the upward transport of the LSWV variations (the tape-recorder phenomenon) controls the variations in water vapour between 100 and 10 hPa. However, the USWV variations are different to those of the middle and lower SWV. Based on EOF and regression analyses of the anomalous upper SWV variability, we found that the upper SWV variations, with an 11-year period controlled by the solar cycle as well as a 2-year cycle regulated by the QBO, dominates the overall variability. Nevertheless, while simulations by state-of-the-art chemistry–climate models of upper SWV vari-ations remain poor, it will be challenging to investigate how the solar cycle influences the upper SWV.

Acknowledgements

This work was jointly supported by the National Natural Sci-ence Foundation of China (41225018, 41575039, and 41305036), 973 Program (2014CB441202). We thanks the datasets from SWOOSH and CCMVal2. We also thank NCAR for providing the WACCM4 model (https://www2.cesm.ucar.edu/ models/current).

Edited by: T. Hirooka

Fig. 5. Comparison of 5-point smoothed upper SWV variations from CCMVAL2 models (the model name indicated at the upper left for each panel) (Eyring et al. 2010) and SWOOSH from 1992−2007. Black line is upper SWV from SWOOSH. Red lines are upper SWV from respective model’s output.

132 Xie et al., Variations in Middle and Upper Stratospheric Water Vapour

References

Austin, J., and co-authors, 2008: Coupled chemistry climate model simulations of the solar cycle in ozone and temperature. J. Geophys. Res., 113, D11306.

Bian, J., R. Yan, H. Chen, D. Lu, and S. T. Massie, 2011: Forma-tion of the summertime ozone valley over the Tibetan Pla-teau: The Asian summer monsoon and air column variations. Adv. Atmos. Sci., 28, 1318−1325.

Bian, J., L. L. Pan, L. Paulik, H. Vömel, H. Chen, and D. Lu, 2012: In situ water vapour and ozone measurements in Lhasa and Kunming during the Asian summer monsoon. Geophys. Res. Lett., 39, L19808.

Brewer, A. W., 1949: Evidence for a world circulation provided by measurements of helium and water vapour distribution in the stratosphere. Quart. J. Roy. Meteor. Soc., 75, 351−363.

Bretherton, C. S., and co-authors, 1999: The effective number of spatial degrees of freedom of a time-varying field. J. Climate, 12, 1990−2009.

Chiodo, G., D. R. Marsh, R. Garcia-Herrera, N. Calvo, and J. A. García, 2014: On the detection of the solar signal in the tropical stratosphere. Atmos. Chem. Phys., 14, 5251−5269.

Considine, D. B., J. E. Rosenfield, and E. L. Fleming, 2001: An interactive model study of the influence of the Mount Pina-tubo aerosol on stratospheric methane and water trends. J. Geophys. Res., 106, 27711−27727.

Dessler, A. E., Z. Zhang, and P. Yang, 2008: Water-vapour climate feedback inferred from climate fluctuations, 2003−2008. Geophys. Res. Lett., 35, L20704.

Dessler, A. E., M. R. Schoeber, T. Wang, S. M. Davis, and K. H. Rosenlof, 2013: Stratospheric water vapour feedback. Proc. Nat. Acad. Sci., 110, 18087−18091.

Dhomse, S., M. Weber, and J. Burrows, 2008: The relationship

Fig. 6. 5-point smoothed lower–middle SWV (100−10 hPa, 60°S to 60°N) vs. upper SWV (10−1 hPa, 60°S to 60°N). Both from CCMVAL2 models from 1970 to 2000. The red line in each panel is the upper SWV. Blue line is the lower–middle SWV.

133SOLA, 2016, Vol. 12, 127−134, doi:10.2151/sola.2016-028

between tropospheric wave forcing and tropical lower strato-sphere water vapour. Atmos. Chem. Phys., 8, 471−480.

Dunkerton, T. J., 2001: Quasi-biennial and subbiennial variations of stratospheric trace constituents derived from HALOE observations. J. Atmos. Sci., 58, 7−25.

Edmon, H. J., B. J. Hoskins, and M. E. McIntyre, 1980: Eliassen- Palm cross-sections for the troposphere. J. Atmos. Sci., 37, 2600−2616.

Egorova, T., E. Rozanov, E. Manzini, M. Haberreiter, W. Schmutz, V. Zubov, and T. Peter, 2004: Chemical and dynamical response to the 11-year variability of the solar irradiance simulated with a chemistry-climate model. Geophys. Res. Lett., 31, L06119.

Eyring, V., T. Shepherd, D. Waugh, et al. (eds.), 2010: SPARC CCMVal Report on the Evaluation of Chemistry-Climate Models. SPARC Report No. 5, WCRP-30/2010, WMO/TD - No. 40.

Forster, P., and K. Shine, 2002: Assessing the climate impact of trends in stratospheric water vapour. Geophys. Res. Lett., 29, doi:10.1029/2001gl013909.

Forster, P. M., D. W. J. Thompson, et al. (eds.), 2011: Strato-spheric changes and climate, Chapter 4 in Scientific Assess-ment of Ozone Depletion: 2010, Global Ozone Research and Monitoring Project–Report. 52516, World Meteorological Organization, Geneva, Switzerland.

Fu, Q., 2013: Ocean–atmosphere interactions: Bottom up in the tropics. Nature Climate Change, 3, 957−958.

Fu, R., and co-authors, 2006: Short circuit of water vapour and polluted air to the global stratosphere by convective transport over the Tibetan Plateau. Proc. Nat. Acad. Sci., 103, 5664− 5669.

Fueglistaler, S., and P. H. Haynes 2005: Control of interannual and longer-term variability of stratospheric water vapour. J. Geophys. Res., 110, D24108.

Fueglistaler, S., M. Bonazzola, P. Haynes, and T. Peter, 2005: Stratospheric water vapour predicted from the Lagrangian temperature history of air entering the stratosphere in the tropics. J. Geophys. Res., 110, D08107.

Froidevaux, L., and co-authors, 2015: Global OZone Chemistry And Related trace gas Data records for the Stratosphere (GOZCARDS): methodology and sample results with a focus on HCl, H2O, and O3. Atmos. Chem. Phys., 15, 10471− 10507.

Garfinkel, C. I., D. W. Waugh, L. D. Oman, L. Wang, and M. M. Hurwitz, 2013: Temperature trends in the tropical upper troposphere and lower stratosphere: Connections with sea surface temperatures and implications for water vapour and ozone. J. Geophys. Res., 118, 9658−9672.

Gabriel, A., H. Schmidt, and D. H. W. Peters, 2011: Effects of the 11 year solar cycle on middle atmospheric stationary wave patterns in temperature, ozone, and water vapour. J. Geo-phys. Res., 116, D23301.

Gettelman, A., and co-authors, 2004: The impact of monsoon circulations on the upper troposphere and lower stratosphere. J. Geophys. Res., 109, D22101.

Glaze, L. S., S. M. Baloga, and L. Wilson, 1997: Transport of atmospheric water vapour by volcanic eruption columns. J. Geophys. Res., 102, 6099−6108.

Gray, L. J., and co-authors, 2010: Solar influence on climate. Rev. Geophys., 48, RG400.

Hauchecorne, A., and co-authors, 2010: Response of tropical stratospheric O3, NO2 and NO3 to the equatorial Quasi-Bien-nial Oscillation and to temperature as seen from GOMOS/ ENVISAT. Atmos. Chem. Phys., 10, 8873−8879.

Hegglin, M. I., and co-authors, 2014: Vertical structure of strato-spheric water vapour trends derived from merged satellite data. Nature Geoscience, 6, 768−776.

Holton, J. R., and A. Gettelman, 2001: Horizontal transport and the dehydration of the stratosphere. Geophys. Res. Lett., 28, 2799−2802.

Hurst, D. F., and co-authors, 2011: Stratospheric water vapour trends over Boulder, Colorado: Analysis of the 30 year Boul-

der record. J. Geophys. Res., 116, D02306.Jackson, D. R., M. D. Burrage, J. E. Harries, L. J. Gray, and J.

M. Russell, 1998: The semi-annual oscillation in upper stratospheric and mesospheric water vapour as observed by HALOE. Quart. J. Roy. Meteor. Soc., 124, 2493−2515.

Jiang, J. H., N. J. Livesey, H. Su, L. Neary, J. C. McConnell, and N. A. D. Richards, 2007: Connecting surface emissions, convective uplifting, and long-range transport of carbon monoxide in the upper troposphere: New observations from the Aura Microwave Limb Sounder. Geophys. Res. Lett., 34, L18812.

Jiang, J. H., and co-authors, 2012: Evaluation of cloud and water vapor simulations in CMIP5 climate models using NASA A-train satellite observations. J. Geophys. Res., 117, D1410.

Jiang, J. H., H. Su, C. Zhai, L. Wu, K. Minschwaner, A. M. Molod, and A. M. Tompkins, 2015: An assessment of upper- troposphere and lower-stratosphere water vapour in MERRA, MERRA2 and ECMWF reanalyses using Aura MLS obser-vations. J. Geophys. Res., 120, 11468−11485.

Joshi, M. M., and G. S. Jones, 2009: The climatic effects of the direct injection of water vapour into the stratosphere by large volcanic eruptions. Atmos. Chem. Phys., 9, 6109−6118.

Joshi, M. M., and K. P. Shine, 2003: A GCM study of volcanic eruptions as a cause of increased stratospheric water vapour. J. Climate, 16, 3525−3534.

Matthes, K., 2011: Atmospheric science: Solar cycle and climate predictions. Nature Geoscience, 4, 735−736.

Matthes, K., U. Langematz, L. L. Gray, K. Kodera, and K. Labitzke, 2004: Improved 11-year solar signal in the Freie Universität Berlin climate middle atmosphere model (FUB-CMAM). J. Geophys. Res., 109, D06101.

Minschwaner, K., H. Su, and J. H. Jiang, 2016: The upward branch of the Brewer-Dobson circulation quantified by tropical stratospheric water vapor and carbon monoxide measurements from the Aura Microwave Limb Sounder. J. Geophys. Res., 121, 2790−2804.

Morgenstern, O., and co-authors, 2010: Review of formulation of present-generation stratospheric chemistry-climate models and associated external forcings. J. Geophys. Res., 115, D00M02.

Mote, P. W., and co-authors, 1996: An atmospheric tape recorder: The imprint of tropical tropopause temperatures on strato-spheric water vapour. J. Geophys. Res., 101, 3989−4006.

Oninas, V., and co-authors, 2001: Radiative cooling by strato-spheric water vapour: big difference in GCM results. Geo-phys. Res. Lett., 28, 2791−2794.

Oltmans, S. J., H. Vomel, D. J. Hofmann, K. H. Rosenlof, and D. Kley, 2000: The increase in stratospheric water vapour from balloonborne, frostpoint hygrometer measurements at Wash-ington, DC, and Boulder, Colorado. Geophys. Res. Lett., 27, 3453−3456.

Randel, W. J., F. Wu, H. Voemel, G. E. Nedoluha, and P. Forster, 2006: Decreases in stratospheric water vapour after 2001: links to changes in the tropicsal tropopause and the Brewer- Dobson circulation. J. Geophys. Res., 111, D12312.

Randel, W., and F. Wu, 1996: Isolation of the ozone QBO in SAGE II data by singular value decomposition. J. Atmos. Sci., 53, 2546−2559.

Randel, W. J., and E. J. Jensen, 2013: Physical processes in the tropical tropopause layer and their role in a changing climate. Nature Geoscience, 6, 169−176.

Rosenlof, K. H., and co-authors, 2001: Stratospheric water vapour increases over the past half- century. Geophys. Res. Lett., 28, 1195−1198.

Rosenlof, K. H., 2003: Atmospheric science—How water enters the stratosphere. Science, 302, 1691−1692.

Schoeberl, M. R., 2004: Extratropical stratosphere-troposphere mass exchange. J. Geophys. Res., 109, D13303.

Scaife, A. A., N. Butchart, D. R. Jackson, and R. Swinbank, 2003: Can changes in ENSO activity help to explain increasing stratospheric water vapour? Geophys. Res. Lett., 30, 1880.

Schieferdecker, T., S. Lossow, G. P. Stiller, and T. Clarmann,

134 Xie et al., Variations in Middle and Upper Stratospheric Water Vapour

2015: Is there a solar signal in lower stratospheric water vapour? Atmos. Chem. Phys., 15, 9851−9863.

Sherwood, S. C., and A. E. Dessler, 2000: On the control of strato-spheric humidity. Geophys. Res. Lett., 27, 2513−2516.

Shindell, D. T., 2001: Climate and ozone response to increased stratosphere water vapour. Geophys. Res. Lett., 28, 1551− 1554.

Solomon, S., and co-authors, 2010: Contributions of stratospheric water vapour to decadal changes in the rate of global warm-ing. Science, 327, 1219−1223.

Stenke, A., and V. Grewe, 2005: Simulation of stratospheric water vapour trends: impact on stratospheric ozone chemistry. Atmos. Chem. Phys., 5, 1257−1272.

Su, H., and co-authors, 2011: Observed increase of TTL tem-perature and Water Vapor in polluted clouds over Asia. J. Climate, 24, 2728−2736.

Sun, C., J. Li, F. Jin, and R. Q. Ding, 2013: Sea surface tempera-ture inter-hemispheric dipole and its relation to tropical precipitation. Environ. Res. Lett., 8, 044006.

Sun, C., J. Li, F. Jin, and F. Xie, 2014: Contrasting meridional structures of stratospheric and tropospheric planetary wave variability in the Northern Hemisphere. Tellus A, 66, 25303.

Tao, M., P. Konopka, F. Ploeger, J.-U. Grooß, R. Müller, C. M. Volk, K. A. Walker, and M. Riese, 2015a: Impact of the 2009 major sudden stratospheric warming on the composition of the stratosphere. Atmos. Chem. Phys., 15, 8695−8715.

Tao, M., P. Konopka, F. Ploeger, M. Riese, R. Müller, and C. M. Volk, 2015b: Impact of stratospheric major warmings and the quasi-biennial oscillation on the variability of strato-spheric water vapour. Geophys. Res. Lett., 42, 4599−4607.

Textor, C., H.-F. Graf, M. Herzog, and J. M. Oberhuber, 2003: Injection of gases into the stratosphere by explosive volcanic eruptions. J. Geophys. Res., 108, 4606.

Texier, H., S. Solomon, and R. R. Garcia, 1988: The role of molecular hydrogen and methane oxidation in the water

vapour budget of the stratosphere. Quart. J. Roy. Meteor. Soc., 114, 281−295.

Tian, W., and M. P. Chipperfield, 2006: Stratospheric water vapour trends in a coupled chemistry-climate model. Geophys Res. Lett., 33, L06819.

Tian, W. S., M. P. Chipperfield, and D. R. Lu, 2009: Impact of increasing stratospheric water vapour on ozone depletion and temperature change. Adv. Atmos. Sci., 26, 423−437.

Wallace, J., R. Panetta, and J. Estberg, 1993: Representation of the equatorial quasi- biennial oscillation in EOF phase space. J. Atmos. Sci., 50, 1751−1762.

Wang, T., A. Dessler, M. Schoeberl, W. Randel, and J. Kim, 2015: The impact of temperature resolution on trajectory modeling of stratospheric water vapour. Atmos. Chem. Phys., 15, 3517− 3526.

Wang, T., W. J. Randel, A. E. Dessler, M. R. Schoeberl, and D. E. Kinnison, 2014: Trajectory model simulations of ozone (O3) and carbon monoxide (CO) in the lower stratosphere. Atmos. Chem. Phys., 14, 7135−7147.

Woods, T. N., W. K. Tobiska, G. J. Rottman, and J. R. Worden, 2000: Improved solar Lyman irradiance modeling from 1947 through 1999 based on UARS observations. J. Geophys. Res., 105, 27195−27215.

Xie, F., J. P. Li, W. S. Tian, Y. J. Li, and J. Feng, 2014a: Indo- Pacific warm pool area expansion, modoki activity, and trop-ical cold-point tropopause temperature variations. Sci. Rep., 4, 4552.

Xie, F., J. P. Li, W. S. Tian, and J. K. Zhang, 2014b: The relative impacts of El Niño Modoki, canonical El Niño, and QBO on tropical ozone changes since the 1980s. Environ. Res. Lett., 9, 064020.

Manuscript received 14 February 2016, accepted 5 May 2016SOLA: https://www. jstage. jst.go. jp/browse/sola/