Satellite Measurements of the Madden–Julian Oscillation in Wintertime Stratospheric Ozone over the Tibetan Plateau and East Asia

ZHANG Yuli,LIU Yi,LIU Chuanxi?,and V.F.SOFIEVA

1Key Laboratory of Middle Atmosphere and Global Environment Observation, Institute of Atmospheric Physics,Chinese Academy of Sciences,Beijing100029

2University of Chinese Academy of Sciences,Beijing100049

3Joint Center for Global Change Studies,Beijing100875

4Finnish Meteorological Institute,Helsinki,Finland

Satellite Measurements of the Madden–Julian Oscillation in Wintertime Stratospheric Ozone over the Tibetan Plateau and East Asia

ZHANG Yuli1,2,LIU Yi1,3,LIU Chuanxi?1,3,and V.F.SOFIEVA4

1Key Laboratory of Middle Atmosphere and Global Environment Observation, Institute of Atmospheric Physics,Chinese Academy of Sciences,Beijing100029

2University of Chinese Academy of Sciences,Beijing100049

3Joint Center for Global Change Studies,Beijing100875

4Finnish Meteorological Institute,Helsinki,Finland

We investigate the Madden–Julian Oscillation(MJO)signal in wintertime stratospheric ozone over the Tibetan Plateau and East Asia using the harmonized dataset of satellite ozone profiles.Two different MJO indices—the all-season Real-Time multivariate MJO index(RMM)and outgoing longwave radiation-based MJO index(OMI)—are used to compare the MJO-related ozone anomalies.The results show that there are pronounced eastward-propagating MJO-related stratospheric ozone anomalies(mainly within 20–200 hPa)over the subtropics.The negative stratospheric ozone anomalies are over the Tibetan Plateau and East Asia in MJO phases 4–7,when MJO-related tropical deep convective anomalies move from the equatorial Indian Ocean towards the western Pacific Ocean.Compared with the results based on RMM,the MJO-related stratospheric column ozone anomalies based on OMI are stronger and one phase ahead.Further analysis suggests that different sampling errors,observation principles and retrieval algorithms may be responsible for the discrepancies among different satellite measurements.The MJO-related stratospheric ozone anomalies can be attributed to the MJO-related circulation anomalies, i.e.,the uplifted tropopause and the northward shifted westerly jet in the upper troposphere.Compared to the result based on RMM,the upper tropospheric westerly jet may play a less important role in generating the stratospheric column ozone anomalies based on OMI.Our study indicates that the circulation-based MJO index(RMM)can better characterize the MJO-related anomalies in tropopause pressure and thus the MJO influence on atmospheric trace gases in the upper troposphere and lower stratosphere,especially over subtropical East Asia.

Madden–Julian Oscillation,stratospheric ozone,tropopause,subtropical jet stream

1.Introduction

The Madden–Julian Oscillation(MJO)is a dominant form of intraseasonal variability(30–60 days)in the tropical troposphere,especially during the extended boreal winter (November–April)when the Indo-Pacific warm pool is centered near the equator.It is characterized by slow(～5 m s?1), eastwardpropagating,large-scaleoscillationsintropicaldeep convection,especially over the equatorial Indian and western Pacific oceans(e.g.,Madden and Julian,1971,1972; Zhang,2005;Lau and Waliser,2012).There are interactions between the MJO and a wide range of high-impact weather and climate processes,e.g.,tropical weather,tropicalcyclones,ENSO,andmonsoonmigration.Recentstudies have suggested that the MJO can disturb the winter stratospheric polar vortex,leading to stratospheric sudden warming events(e.g.,Garfinkel et al.,2012;Liu et al.,2014).A better understanding of the MJO may play an important role in bridging the gap between climate prediction and weather forecasting(e.g.,Waliser,2012;Zhang,2013).

With the advent of the National Aeronautics and Space Administration’s Aqua and Aura satellites,and the European Space Agency’s enviromental satellite(ENVISAT),the footprints of the MJO in atmospheric compositions are gradually being discovered(e.g.,Tian and Waliser,2012).Based on the Nimbus-7 Total Ozone Mapping Spectrometer,Sabutis et al.(1987)first reported evidence for 30–50-day variability in the total column ozone(TCO)over specific locations over the southern Indian and southeast Pacific oceans.Gao and Stanford(1990)went on to identify low-frequency variations(about1–2months)inthe8-yearNimbus-7TOMSTCOdata.Recently,as shown by Tian et al.(2007),the intraseasonal TCO anomalies are mainly obvious in the subtropics over the Eastern Hemisphere and the Pacific,while TCO anomalies are rather small over the equator.MJO convection and related wave dynamics were well proven to have a systematic relationship with subtropical TCO anomalies.For example,the subtropical negative TCO anomalies typically flank or lie to the west of equatorial enhanced convection and are co-located with the subtropical upper-tropospheric anticyclones generated by equatorial anomalous convective forcing.A similar MJO-related ozone signal in the subtropics was also discovered by Liu et al.(2009),who investigated the dynamic formation of a record ozone minimum event (also called ozone“mini-hole”)in mid-December 2003 over the Tibetan Plateau using the satellite measurements of the Michelson Interferometer for Passive Atmospheric Sounding (MIPAS)and Global Ozone Monitoring by Occultation of Stars(GOMOS)instruments and chemical transport model simulations.Liu et al.(2010)further examined the formation of ozone minimum events in December–January–February over the Tibetan Plateau between 1979 and 2001,suggesting that most of these events are contributed by both the anomaly around the tropopause layer and stratospheric transport related to polar vortex disturbances.Recently,Li et al. (2012)investigated the vertical structure of boreal wintertime MJO-related subtropical ozone variations(November 2004–February 2010)using the ozone pro files from the Tropospheric Emission Spectrometer(TES)and Aura Microwave Limb Sounder(MLS)datasets,as well as in-situ measurements in the Southern Hemisphere.This study suggested that the spatiotemporal patterns of the subtropical ozone anomalies in the lower stratosphere(60–100 hPa)are very similar to those of the total column,which are both dynamically driven by the vertical movement of the subtropical tropopause.It indicates that the subtropical TCO anomalies mostly arise from ozone anomalies in the lower stratosphere,supporting the hypothesis of Tian et al.(2007).The robust connection between the MJO and the intraseasonal variations of subtropical stratospheric ozone implies that the stratospheric ozone variations might be predictable with similar lead times over the subtropics.

Given that the tropopause height determines the length of the stratospheric air column,in which most of the total ozone column is contained,the variation in the tropopause height plays a dominant role in initiating the MJO-related ozone variation in the subtropics(e.g.,Tian et al.,2007;Li et al., 2012).In a recent case study,Liu et al.(2009)discovered that the MJO can lead to substantial reduction of the TCO over the Tibetan Plateau by shifting the upper-tropospheric subtropical jet to the north of the plateau.It is therefore interesting to examine the relative contribution of the tropopause height and upper tropospheric subtropical jet to the formation of MJO-related ozone anomalies,especially over the Tibetan Plateau and East Asia.

In this study,we investigate the MJO-related ozone anomalies over the Tibetan Plateau and East Asia using both reanalysis data and satellite measurements.The results based on two different definitions of MJO phases are compared. Section 2 describes the data and methods.Section 3 presents the main results.Conclusions are summarized in section 4.

2.Data and methods

2.1.Reanalysis and satellite measurements

In this study,the MJO-related circulation,tropopause pressure and ozone anomalies are investigated based on the daily mean European Centre for Medium Range Weather Forecasts Reanalysis Interim(ERA-Interim)data between 2005 and 2011(e.g.,Dee et al.,2011).The daily outgoing longwaveradiation(OLR)isderivedfromtheAdvancedVery High Resolution Radiometer instrument on board the National Oceanic and Atmospheric Administration’s(NOAA) polar orbiting spacecraft(Liebman and Smith,1996).

The harmonized dataset of ozone pro files(HARMOZ, 2005–2011)is also used to study the variability of ozone over the Tibetan Plateau and East Asia.These data are from the MIPAS,the Scanning Imaging Absorption Spectrometer for Atmospheric Chartography(SCIAMACHY),and GOMOS instruments.The harmonized dataset has a common pressure grid in netCDF format(network common data form). There are 13 pressure levels(200,170,150,130,115,100, 90,80,70,50,40,30,and 20 hPa)between 200 and 20 hPa, which correspond to a vertical resolution of～1 km below 20 km and 2–3 km above 20 km(So fieva et al.,2013).The altitude ranges of the MIPAS,SCIAMACHY and GOMOS ozone products are 400–0.05 hPa,250–0.05 hPa,and 250–10?4hPa,respectively.

The MIPAS instrument on board ENVISAT is a Fourier transform limb emission spectrometer,measuring the signatures of various trace gases(Fischer et al.,2008).We use a combination of the MIPAS version V5RO3220(2005–April 2011)and V5RO3221(April 2011–2012)ozone products.The vertical resolution of the MIPAS ozone profile is 3–5 km in the stratosphere with an estimated error of 1%–4%.

The SCIAMACHY ozone pro files in HARMOZ are retrieved based on exploiting scattered radiances in the UV and visible ranges.The vertical resolution is 3–5 km with an estimated error of 10%–15%(Rahpoe et al.,2013).SCIAMACHY ozone pro files are usually of poor quality in cloudy conditions.In the harmonized dataset,ozone data at altitudes contaminated by clouds are filtered out to exclude poor quality in cloudy conditions(So fieva et al.,2013).So,the actual error of SCIAMACHY we use in our study is smaller than 10%–15%.

GOMOSonboardENVISATisaself-calibratedmediumresolution stellar occultation spectrometer(Kyro¨la¨ et al., 2004;Bertaux et al.,2010).GOMOS ozone pro files generally cover a vertical range from 15 to 100 km,with a vertical resolution of 2–3 km and an estimated error of 0.5%–1%in the stratosphere(Tamminen et al.,2010).

2.2.Methods

Wheeler and Hendon(2004)developed the widely-usedReal-time Multivariate MJO(RMM)index,based on circulation and(OLR),to calculate the state of the MJO.Recent studies have developed additional MJO indices,such as a velocity potential MJO index,which replaces OLR with the 200 hPa velocity potential(Ventrice et al.,2013),and an MJO index based solely on the outgoing longwave index(OMI)(Kiladis et al.,2014).

The daily RMM index(Wheeler and Hendon,2004), which determines the phases of MJO events,is obtained from the Australian Bureau of Meteorology website(http://cawcr. gov.au/staff/mwheeler/maproom/RMM/).Empirical orthogonal function(EOF)analysis decomposes the combined fields of meridionally averaged satellite-observed OLR and zonal winds at 850 and 200 hPa between 15°S–15°N.The time series of the two leading EOFs(RMM1 and RMM2) vary mostly on intraseasonal timescales(typically 30–60 days).The OMI index(Kiladis et al.,2014)is a straightforward application of an EOF of OLR.The daily values of the OMI PC1( first principal component)and OMI PC2(second principal component)are obtained from NOAA’s Earth System Research Laboratory(http://www.esrl.noaa.gov/psd/ mjo/mjoindex/).

There are eight phases during a lifecycle of the MJO, which indicate the geographic location of the MJO-related convective anomalies.A comparison of MJO phases between RMM and OMI(Fig.1a)in boreal winter 2005–2011 shows a slight delay of MJO phases in RMM compared to OMI.Considering the strength of the MJO,only active MJO phases—amplitudes of MJO indices(RMM12+RMM22)1/2and(PC12+PC22)1/2greater than 1.0—are considered in this study(Fig.1b).

To derive the MJO signal in stratospheric ozone(i.e., MJO-related ozone anomalies),the daily climatology has been firstly removed from the daily mean value.Then,a 20–100-day bandpass filter has been applied to the daily anomalies.The stratospheric column ozone anomalies are defined as the integrated value between 20 and 200 hPa.The bandpass filter substantially reduces the effective sample size.As a result,the regular Student’s t-test is no longer suitable for testing the significance of MJO-related composites(e.g.,Tian et al.,2011;Liu et al.,2014).In the present study,a twotailed Student’s t-test with reduced degrees of freedom is applied to determine the significance of the MJO-related composite.The effective sample number is estimated as

where N′and N are the effective sample number and actual sample number,respectively;and r1and r2are the 2-day lag auto-correlations for RMM1(PC1)and RMM2(PC2),respectively.As a result,N′=0.13N based on the RMM index and N′=0.1N based on the OMI index,which are approximately consistent with previous studies(Tian et al.,2011;Liu et al.,2014).

Given that the MJO is especially active over the equatorial Indian Ocean and western Pacific Ocean,we focus on theinfluence of the MJO on the subtropical stratospheric ozone over the Tibetan Plateau and East Asia.The Tibetan Plateau and East Asia are defined as the regions(25°–40°N,75°–105°E)and(25°–40°N,105°–135°E),respectively.The latitudinal location of the 200 hPa westerly jet over the Tibetan Plateau and East Asia are defined as the mean latitudes of the maximum 200 hPa westerly wind within 75°–105°E and 105°–135°E,respectively.

3.Results

3.1.Evolution of the MJO-related stratospheric column ozone anomalies

To illustrate the influence of the MJO on stratospheric ozone,Fig.2 shows composites of the ERA-Interim MJO-related ozone anomalies(i.e.,a 20–100-day bandpass filter has been applied)of the stratospheric column ozone(SCO) during boreal winter(i.e.,December,January,and February). Limited by small sample sizes,some anomalies are not statistically significant.

In RMM phase 1,there are positive SCO anomalies over the Tibetan Plateau,East Asia,and the western Pacific(Fig. 2a).In RMM phase 2(Fig.2b),the positive SCO anomalies over the Tibetan Plateau and East Asia become stronger and greater than+12 Dobson unit(DU).In addition,there are negative SCO anomalies to the west of the positive SCO anomalies.In RMM phase 3,as the negative OLR anomalies move eastward to the south of Bay of Bengal,the positive SCO anomalies propagate eastward with less strength, and the weak negative SCO anomalies over the west part of the Tibetan Plateau move eastward(Fig.2c).In RMM phase 4,the negative SCO anomalies become stronger and move to the Tibetan Plateau(Fig.2d)when OLR anomalies arrive at the Maritime Continent.As the OLR anomalies enhance over the Maritime Continent in RMM phase 5,the negative SCO anomalies are significantly enhanced(less than?14 DU)with a center that propagates eastward to over East Asia (Fig.2e).In RMM phases 6–8,the negative SCO anomalies quickly dissipate(Figs.2f–h)after the negative OLR anomalies arrive in the western equatorial Pacific.Meanwhile,as the positive OLR anomalies are active over the equatorial Indian Ocean,there are positive SCO anomalies developing over the Tibetan Plateau and East Asia(Figs.2f–h).The evolution of SCO anomalies based on OMI MJO phases is similar to that based on RMM MJO phases(compare right to left panels in Fig.2).One of the major differences is that the negative SCO anomalies are more persistent with a stronger amplitude over East Asia based on OMI than those based on RMM(compare Figs.2m and n to 2e and f),especially in phase 6.This can be attributed to different definitions of the MJO phase based on the two MJO indices.As a result of being ahead of phase based on OMI(Fig.1),some negative SCO anomalies in RMM phase 5 are shown in OMI phase 6,making stronger negative SCO anomalies in OMI phase 6 than in RMM phase 6(compare Fig.2n to 2f).

3.2.Vertical structure of the MJO-related stratospheric ozone anomalies

In this section,we examine the vertical structure of the MJO-related stratospheric ozone anomalies over the Tibetan Plateau and East Asia using reanalysis data and satellite ozone profiles.

The number of ozone profiles from three satellite measurements between 2005 and 2011 over the Tibetan Plateau and East Asia are given in Table 1.SCIAMACHY has almost twice the total profiles of MIPAS,over both the Tibetan Plateau(22 005 cf.12 796)and East Asia(22 053 cf.12 273)regions.The numbers of GOMOS profiles are less than a quarter of those of MIPAS profiles,over both the Tibetan Plateau(1937 cf.12 796)and East Asia(1949 cf.12 273).As a result,the GOMOS dataset is the leastrepresentative dataset among the three sets of satellite measurements used in this study.Figure 3 shows the vertical structure of the MJO-related ozone anomalies based on ERAInterim reanalysis and different satellite measurements from MIPAS,SCIAMACHY,and GOMOS.The ozone anomalies from ERA-Interim reanalysis(Figs.3a,e,i and m)show that there are significant MJO-related ozone anomalies between 20 and 200 hPa,which is generally consistent with previous studies(e.g.,Li et al.,2012).This is also the reason for defining the ozone column between 20 and 200 hPa as the SCO in this study.The significant MJO-related ozone anomalies derived from MIPAS(Figs.3b,f,j and n)andSCIAMACHY(Figs.3c,g,k and o)measurements are generally consistent with those from the ERA-Interim reanalysis. However,the result from GOMOS measurement(Figs.3d,h, l and p)is different.This could be attributable to the scarcity of GOMOS ozone profiles(Table 1).The different observation principles and retrieval algorithms may also contribute to the discrepancies among the three sets of satellite measurements.As shown in Figs.3a–d and 3i–l,both reanalysis data and satellite measurements suggest that there are negative SCO anomalies over the Tibetan Plateau(East Asia)in RMMphases4–6(5–7),whiletherearepositiveSCOanomalies over the Tibetan Plateau(East Asia)in RMM phases 7–8 and 1–3(8 and 1–4).The one–phase delay between the results over the Tibetan Plateau and those over East Asia can be attributed to the eastward propagation of the MJO and its circulation anomalies.The minimal ozone anomalies over East Asia are shown in RMM phase 5 and OMI phase 6,respectively.This one-phase difference between RMM and OMI indices has been explained in section 3.1.It is also noted that there is vertical tilt with altitude in the ozone anomalies between 200 and 20 hPa over the Tibetan Plateau(Figs.3a–h).Compared to ERA-Interim reanalysis,the vertical tilt is clearer in the satellite measurements(compare Figs.3a to 3b–d).The vertical tilt structure has been reported in a previous study(Li et al.,2012).In contrast,no vertical tilt structure can be discerned over East Asia,in either ERA-Interim reanalysis or satellite measurements(Figs.3i–p).

Table 1.Number of ozone profiles from the different instrument measurements(MIPAS,SCIAMACHY and GOMOS)over the Tibetan Plateau and East Asia for each MJO phase(according to the RMM and OMI indices)between 2005 and 2011.

To test if the satellite sampling errors can contribute to the discrepancies among different satellite measurements and reanalysis shown in Fig.3,the MJO-related ozone anomalies from ERA-Interim reanalysis have been interpolated to the geophysical locations of ozone profiles from MIPAS(Figs. 4a,d,g and j),SCIAMACHY(Figs.4b,e,h and k)and GOMOS(Figs.4c,f,i and l)measurements.The differences among the subsampled ERA-Interim at different satellite locations(compare the three columns in Fig.4)indicate that the sampling error is responsible for the differences among satellite measurements shown in Fig.3.Despite the similarity of pressure–phase distribution,the strength of subsampled ERA-Interim ozone anomalies over the Tibetan Plateau at SCIAMACHY locations is slightly larger than those at MIPAS locations.In addition,the smaller anomalies at higher altitudes over the Tibetan Plateau at GOMOS locations based on RMM confirm the lesser representation of GOMOS measurements than the other two measurements because of less GOMOS samples.The correlations between satellite measurements and subsampled ERA-Interim are generally larger than the ones between satellite measurements and fully sampled ERA-Interim(not shown),suggesting that the influence of sampling error could be important.

As shown in Fig.3,the result from GOMOS measurements is quite noisy and somewhat different to that from MIPAS and SCIAMACHY measurements.Compared to the original GOMOS measurements,the result of subsampled ERA-Interim at GOMOS locations,shown in Fig.4,is more similar to the results from MIPAS and SCIAMACHY measurements,shown in Fig.3.The improved result indicates that the observation principles and/or the retrieval algorithms of GOMOS measurements also play an important role in generating the discrepancies among satellite measurements and reanalysis.

It is also noted that the amplitude of MJO-related ozone anomalies between 20 and 50 hPa is greater in ERA-Interim reanalysis than those in satellite measurements(compare the left column to the right three columns in Fig.3).Figure 4 suggests that the amplitudes of ozone anomalies between 20 and 50 hPa are comparable to those in ERA-Interim reanalysis after the interpolation.Therefore,the differences of ozone anomalies in 20–50 hPa between ERA-Interim reanalysis and satellite measurements shown in Fig.3 could be a systematic difference rather than caused by the satellite sampling errors.

3.3.Attribution of the MJO-related ozone anomalies

Previous studies have indicated that the wintertime stratospheric ozone anomalies over the Tibetan Plateau can be attributed to the tropopause height and upper tropospheric circulation pattern(i.e.,200 hPa subtropical westerly jet)(e.g., Liu et al.,2009,2010).To better understand the mechanism responsible for the negative stratospheric ozone anomalies over the Tibetan Plateau and East Asia,Fig.5 shows the MJO-related anomalies in tropopause pressure and 200 hPa horizontal winds during MJO phases 3–6.

In RMM phase 3,when the MJO convective anomalies become active over the equatorial Indian Ocean,there is an anticyclonic anomaly center in the upper troposphere moving towards the Tibetan Plateau(Fig.5a).The anticyclonic anomaly is coupled with an uplifted tropopause(Fig.5a). As a result,the negative stratospheric ozone anomalies move towards the west part of the Tibetan Plateau(Fig.2c).As the MJO-related convective anomalies travel across the Maritime Continent in RMM phases 4–5,the coupled anticyclonic anomaly intensifies and moves eastward(Figs.5b–c), leading to enhanced negative stratospheric ozone anomalies over the Tibetan Plateau and East Asia(Figs.2d–e).After RMM phase 6(Fig.5d),as the MJO convective anomalies move towards the equatorial western Pacific Ocean,the MJO-related circulation anomalies weaken over the Tibetan Plateau and East Asia.

The results based on the OMI index are quite similar (Figs.2i–p and 5e–h).The major difference is that the coupled anticyclonic circulation anomaly and the uplifted tropopause are stronger during MJO phases 5–6 over East Asia based on OMI index than those based on RMM index (compare Figs.5g and h to 5c and d).This difference is consistent with the difference of MJO-related SCO anomalies(compare Figs.2m and n to 2e and f).

It is also noted that,in Fig.5,MJO-related anomalies in tropopause pressure based on RMM show persistent eastward propagation through phases 3–6(Figs.5a–d).In contrast,the anomalies based on OMI show westward propagation from phases 4 to phase 5(Figs.5f–g).This difference is amplified by contrasting the different dates based on RMM and OMI (i.e.,removing the samples that have the same definition of MJO phase based on RMM and OMI)(Fig.6).Therefore,the circulation-based MJO index(i.e.,RMM)can better characterize the eastward propagation of the MJO-related anomalies in tropopause pressure and thus the MJO influence on atmospheric trace gases in the upper troposphere and lower stratosphere over subtropical East Asia.

To explore the relative contributions of the tropopause height and upper-tropospheric subtropical jet to the MJO-related stratospheric ozone anomalies over the Tibetan Plateau and East Asia,the MJO-related anomalies in tropopause height and the latitudinal location of the 200 hPa subtropical jet are compared with the amplitude of the SCO anomalies for each MJO phase.Because of the regional averageusedinFig.7,theamplitudeofMJO-relatedSCOanomalies over the Tibetan Plateau(?8 to 10 DU)and East Asia (?12 to 7 DU)are smaller than those shown in Fig.2.Compared to over the Tibetan Plateau,the amplitudes of negative SCO anomalies over East Asia are generally larger,especially in satellite measurements(compare Figs.7a–h to Figs. 7i–p).Generally,the negative SCO anomalies are observed during RMM phases 4–6 and OMI phases 4–7 over the Tibetan Plateau.As the eastward propagation of the MJO,they are observed during phases 5–7 over East Asia.As a resultof being ahead of phase in OMI,the minimal SCO anomalies are shown in OMI phase 6,which are one phase ahead of minima in RMM(phase 5).However,the timing of the negative ozone anomalies is slightly different based on different satellitemeasurementsduetothesamplingerroranddifferent retrieval algorithms,as discussed in section 3.2.Over both regions,the peak of the negative SCO anomalies coincides well with that of the negative anomalies of tropopause height and that of the positive anomalies of the latitudinal location of the 200 hPa subtropical jet.The result suggests that the anomalies in both the tropopause height and westerly jet play important roles in creating the negative stratospheric ozone anomalies over both regions.However,compared to results based on RMM,the 200 hPa westerly jet anomalies are relatively smaller than the amplitudes of the tropopause anomaly based on OMI(Figs.7e–h and 7m–p).Therefore,the westerly jet in the upper troposphere may play a less important role in generating the negative stratospheric ozone anomalies based on the OMI index due to the fact that the OMI definition does not include the factor of circulation.

4.Conclusions

The MJO-related stratospheric ozone anomalies during boreal winter were analyzed based on satellite-borne ozone profiles from MIPAS,SCIAMACHY,and GOMOS measurements.All the satellite measurements suggest pronounced MJO-related ozone anomalies(greater than±10 DU)between 200 and 20 hPa over the Tibetan Plateau and East Asia. According to the circulation-based MJO index(RMM),there are negative stratospheric ozone anomalies over the Tibetan Plateau in MJO phases 4–6,when the MJO-related convective anomalies are active over the Maritime Continent.In MJO phases 5–7,as the MJO-related convective anomalies move from the Maritime Continent towards the equatorial western Pacific Ocean,there are negative stratospheric ozone anomalies over East Asia.The MJO-related ozone anomalies between 200 and 20 hPa show a vertical tilt structure over the Tibetan Plateau.However,no vertical tilt structure can be discerned over East Asia,in either ERA-Interim reanalysis or satellite measurements.

TheMJO-relatedstratosphericozoneanomaliesarequantitatively different based on different satellite measurements. Further analysis suggests that the discrepancies among different satellite datasets can be mainly attributed to the different sampling errors,observation principles,and retrieval algorithms of the three satellite instruments.

The occurrence of the MJO-related stratospheric ozone anomalies can be attributed to the uplift of the tropopause and the northward shift of the subtropical jet in the upper troposphere.Meteorological analysis suggests that the negative SCO anomalies over both the Tibetan Plateau and East Asia are dynamically associated with the uplifted tropopause and northward shifted subtropical jet.Compared to the results based on the RMM index,the upper tropospheric westerly jet may play a less important role in generating the stratospheric ozone anomalies,as shown by the results based on the OMI index.

Because of the different definitions of the two MJO indices,there are pronounced differences between the results based on the circulation-based(RMM)and convection-based (OMI)indices over East Asia.Compared to the results based on the OMI index,the anomalies in tropopause pressure based on the RMM index propagate eastward with steady velocity,indicating that the circulation-based MJO index (RMM)can better characterize the eastward propagation of MJO-related anomalies in tropopause pressure,and thus the MJO influence on the atmospheric trace gases in the upper troposphere and lower stratosphere,especially over subtropical East Asia.

Acknowledgements.This work was funded by the National Natural Science Foundation of China(Grant No.41105025), the Dragon 3 Programme(ID:10577)and the High Resolution Earth Observation Funds for Young Scientists(Grant No. GFZX04060103).ERA-Interim and NCEP/NCAR reanalysis data have been kindly provided by the European Centre for Medium-Range Weather Forecasts and National Centers for Environmental Prediction.

REFERENCES

Bertaux,J.L.,and Coauthors,2010:Global ozone monitoring by occultation of stars:An overview of GOMOS measurements on ENVISAT.Atmos.Chem.Phys.,10,12 091–12 148,doi: 10.5194/acp-10–12091-2010.

Dee,D.P.,and Coauthors,2011:The ERA-Interim reanalysis: configuration and performance of the data assimilation system.Quart.J.Roy.Meteor.Soc.,137,553–597,doi:10.1002/ qj.828.

Fischer,H.,and Coauthors,2008:MIPAS:An instrument for atmospheric and climate research.Atmos.Chem.Phys.,8, 2151–2188,doi:10.5194/acp-8–2151-2008.

Garfinkel,C.I.,S.B.Feldstein,D.W.Waugh,C.Yoo,and S. Lee,2012:Observed connection between stratospheric sudden warmings and the Madden–Julian Oscillation.Geophys. Res.Lett.,39,L18807,doi:10.1029/2012GL053144.

Gao,X.H.,and J.L.Stanford,1990:Low-frequency oscillations in total ozone measurements.J.Geophys.Res.,95,13 797–13 806.

Kiladis,G.N.,J.Dias,K.H.Straub,M.C.Wheeler,S.N.Tulich, K.Kikuchi,K.M.Weickmann,and M.J.Ventrice,2014:A comparisonofOLRandcirculation-basedindicesfortracking the MJO.Mon.Wea.Rev.,142,1697–1715.doi:http://dx.doi. org/10.1175/MWR-D-13-00301.1.

Kyr¨ol¨a,E.,and Coauthors,2004:GOMOS on Envisat:an overview.Advances in Space Research,33,1020–1028.

Lau,W.K.-M.,and D.E.Waliser,2012:Intraseasonal Variability in the Atmosphere-ocean Climate System.2nd ed.Springer, Heidelberg,Germany,581 pp.

Li,K.-F.,B.Tian,D.E.Waliser,M.J.Schwartz,J.L.Neu,J.R. Worden,and Y.L.Yung,2012:Vertical structure of MJO-related subtropical ozone variations from MLS,TES,and SHADOZ data.Atmos.Chem.Phys.,12,425–436.

Liebman,B.,and C.A.Smith,1996:Description of a com-plete(interpolated)outgoinglongwaveradiationdataset.Bull. Amer.Meteor.Soc.,77,1275–1277.

Liu,C.X.,Y.Liu,Z.N.Cai,S.T.Gao,D.R.L¨u,and E.Kyr¨ol¨a, 2009:A Madden-Julian Oscillation-triggered record ozone minimum over the Tibetan Plateau in December 2003 and its association with stratospheric“l(fā)ow-ozone pockets”.Geophys. Res.Lett.,36,L15830,doi:10.1029/2009GL039025.

Liu,C.X.,Y.Liu,Z.N.Cai,S.T.Gao,J.C.Bian,X.Liu, and K.Chance,2010:Dynamic formation of extreme ozone minimum events over the Tibetan Plateau during northern winters 1987–2001.J.Geophys.Res.,115,D18311,doi: 10.1029/2009JD013130.

Liu,C.X.,B.J.Tian,K.-F.Li,G.L.Manney,N.J.Liversey, Y.L.Yung,and D.E.Waliser,2014:Northern Hemisphere mid-wintervortex-displacementandvortex-splitstratospheric sudden warmings:Influence of the Madden-Julian Oscillation and Quasi-Biennial Oscillation.J.Geophys.Res.,119, 12 599–12 620,doi:10.1002/2014JD021876.

Madden,R.A.,and P.R.Julian,1971:Detection of a 40–50 day oscillation in the zonal wind in the tropical Pacific.J.Atmos. Sci.,28,702–708.

Madden,R.A.,and P.R.Julian,1972:Description of global-scale circulation cells in the tropics with a 40–50 day period.J.Atmos.Sci.,29,1109–1123.

Rahpoe,N.,C.von Savigny,M.Weber,A.V.Rozanov,H.Bovensmann,and J.P.Burrows,2013:Error budget analysis of SCIAMACHY limb ozone profile retrievals using the SCIATRAN model.Atmospheric Measurement Techniques,6, 2825–2837,doi:10.5194/amt-6-2825-2013.

Sabutis,J.L.,J.L.Stanford,and K.P.Bowman,1987:Evidence for 35-50 day low frequency oscillations in total ozone mapping spectrometer data.Geophys.Res.Lett.,14,945–947.

Sofieva,V.F.,and Coauthors,2013:Harmonized dataset of ozone profiles from satellite limb and occultation measurements.Earth System Science Data,5,349–363,doi:10.5194/essd-5-349-2013.

Tamminen,J.,and Coauthors,2010:GOMOS data characterisation and error estimation.Atmos.Chem.Phys.,10,9505–9519,doi:10.5194/acp-10–9505-2010.

Tian,B.J.,Y.L.Yung,D.E.Waliser,T.Tyranowski,L.Kuai, E.J.Fetzer,and F.W.Irion,2007:Intraseasonal variations of the tropical total ozone and their connection to the Madden-Julian Oscillation.Geophys.Res.Lett.,34,L08704, doi:10.1029/2007GL029451.

Tian,B.J.,D.E.Waliser,R.A.Kahn,and S.Wong,2011:Modulation of Atlantic aerosols by the Madden-Julian Oscillation.J.Geophys.Res.,116,D15108,doi:10.1029/2010JD015201.

Tian,B.,and D.E.Waliser,2012:Chemical and biological impacts.Intraseasonal Variability in the Atmosphere-Ocean Climate System,2nd ed.,W.K.M.Lau and D.E.Waliser,Eds., Springer-Verlag,Berlin,Heidelberg,569–585.

Ventrice,M.J.,M.C.Wheeler,H.H.Hendon,C.J.Schreck III,C. D.Thorncroft,andG.N.Kiladis,2013:Amodifiedmultivariate Madden-Julian Oscillation index using velocity potential.Mon.Wea.Rev.,141,4197–4210,doi:10.1175/MWR-D-12–00327.1.

Waliser,D.E.,2012:Predictability and forecasting.Intraseasonal Variability in the Atmosphere-Ocean Climate System. 2nd ed.,W.K.M.Lau and D.E.Waliser,Eds.,Springer-Verlag,Berlin,Heidelberg,433–476.

Wheeler,M.C.,and H.H.Hendon,2004:An all-season real-time multivariate MJO index:Development of an index for monitoring and prediction.Mon.Wea.Rev.,132,1917–1932.

Zhang,C.D.,2005:Madden-JulianOscillation.Rev.Geophys.,43, RG2003,doi:10.1029/2004RG000158.

Zhang,C.D.,2013:Madden-Julian Oscillation:Bridging weather and climate.Bull.Amer.Meteor.Soc.,94,1849–1870.

:Zhang,Y.L.,Y.Liu,C.X.Liu,and V.F.Sofieva,2015:Satellite measurements of the Madden–Julian Oscillation in wintertime stratospheric ozone over the Tibetan Plateau and East Asia.Adv.Atmos.Sci.,32(11),1481–1492,

10.1007/s00376-015-5005-y.

8 January 2015;revised 18 May 2015;accepted 27 May 2015)

?Corresponding author:LIU Chuanxi

Email:lcx@mail.iap.ac.cn