El Niño-Southern Oscillation (ENSO) and local sea surface temperature (SST) have been regarded as the important factors influencing the precipitation, evaporation and circulation over the Asian-Australian monsoon (AAM) region. The moisture source is determined directly by precipitation and evaporation. This present paper studies the impacts of ENSO and local SST on moisture source in the AAM region during boreal summer. Relative roles of ENSO and local SST are also discussed by using the singular value decomposition (SVD) and conditional SVD (CSVD) methods. The authors identify one major coupled mode between the interannual variations of apparent moisture sink (<Q2>) and SST for the period 1979–2008. Spatial structure of the major mode exhibits two key regions of moisture source, one over the western-central southern Indian Ocean (SIO) where <Q2> is negative, and the other over the northwest and north side of Australia, where <Q2> is positive. In the corresponding map of SST, negative values are also seen in the former region, but are positive in the latter region. The interannual variation of moisture source in the AAM region has an outstanding positive correlation with the local SST. Furthermore, ENSO also has a remarkable correlation with the principle component (PC) of the first empirical orthogonal function (EOF) mode of <Q2>. ENSO and local SST work cooperatively to determine the variations of moisture source in the AAM region.

Introduction

The study on moisture budgets over the Asian-Australian monsoon (AAM) region is an important part of the Climate Variability and Predictability (CLIVAR) project and the Global Energy and Water Cycle Experiment (GEWEX) project. This is of great importance to the economics of dozens of countries and the livelihoods of more than 60% of the world's population. A number of studies focus on the moisture budgets over the AAM region. In boreal summer, the moisture source region lies in the southern Indian Ocean and the Australian monsoon region, while the moisture sink region lies in the Asian monsoon region. There exist two cross-equatorial water vapor transport channels in the AAM region, through which a lot of water vapor is transported from the moisture source region to the moisture sink region (Qiao et al., 2002; Qiao and Lin, 2006). The variation of moisture source with its profound linkage to the Asian summer monsoon plays a primary role in the moisture budgets over the AAM region.

It is well known that surface evaporation is the main source of water vapor in the atmosphere, while rainfall is the main sink. The net moisture source of the air column from the surface to the top of the atmosphere equals evaporation minus rainfall per unit area (Yanai et al., 1973). The evaporation and rainfall play equivalent roles in the variation of moisture source. As mentioned in the papers Qiao et al., 2002 and Qiao and Lin, 2006, the moisture source in the AAM region lies in the southern Indian Ocean (SIO). It is obvious that the variation of oceanic evaporation and rainfall has a close relationship with the local sea surface temperature (SST). Thus, the variability of moisture source is also related tightly with the local SST over the SIO.

There is much research paying attention to the important role of local SST over the Indian Ocean in the variations of rainfall and atmospheric circulation. During the positive SIO dipole phase, the SST anomalies of extra-tropical Indian Ocean will enhance the Mascarene high by interacting with regional atmosphere, which enhances the cross-equatorial flow off Somali, and finally leads to an increase in the Indian summer monsoon rainfall (Yang and Ding, 2006). Variability of SST over the North Indian Ocean has remarkable influence on regional precipitation variability over the Indian Subcontinent (Roxy and Taninoto, 2007). Dry (wet) years in southwest Western Australia (SWWA) are associated with cold (warm) SST anomalies in the eastern Indian Ocean off the northwest shelf of Australia (England et al., 2006). Local positive SST anomalies induce a reorganization of the large-scale atmospheric circulation across the Indian Ocean basin, which increases convective precipitation over southwest Western Australia (Ummenhofer et al., 2008). In the atmospheric models, it has also been indicated that the Indian Ocean SST anomalies can drive realistic rainfall anomalies (Rocha and Simmonds, 1997; Reason and Mulenga, 1999; Reason, 2001; Goddard and Graham, 1999). On the other hand, the El Niño-Southern Oscillation (ENSO) has been considered as a major factor influencing the interannual variability of the AAM (Webster et al., 1998). SST anomalies over the tropical Pacific Ocean also have notable impact on the variation of rainfall and atmospheric circulation in the AAM region. The positive SST anomalies in summer in the tropical Indian Ocean or Pacific can not only cause the response of atmospheric circulation and precipitation in the tropical region, but also lead to the weakening of East Asia and South Asia summer monsoons (Zou and Liu, 2002). The relative roles of remote and local SST forcing in shaping AAM anomalies have been investigated. In different regions, the local SST and the ENSO play different roles. In some regions, the local SST is most important, while in some other regions, the ENSO is essential (Li et al., 2005).

Furthermore, the variability of SST also has great impact on oceanic evaporation. Yu (2007) gave the relation to compute the rate of evaporation (Evp):

formula
where qs is the saturation specific humidity at the SST, qa is the near-surface atmospheric specific humidity, U is the near-surface wind, and ce is a turbulent exchange coefficient and is determined by the atmospheric stability, the air-sea temperature differences, and the wind speed (Fairall et al., 2003). Here, ce, U, and qa are all influenced by SST (Cayan, 1992; Zhang and McPhaden, 1995). So evaporation is strongly influenced by SST.

The above studies revealed the importance of SST anomalies in the variability of evaporation, rainfall and atmospheric circulation and indicated that SST anomalies in the SIO and eastern Pacific must have great impact on the variation of moisture source in the AAM region. However, there has been little work done in this aspect. Therefore, it is necessary to further investigate the impact of SST on the moisture source in the AAM region. The reanalysis data, the relation to compute the vertically integrated apparent moisture sink <Q2> and the methodologies are described. The characteristics of interannual variation of moisture source in the AAM region are described. The coupling modes of SST and moisture source will be investigated and the relative roles of local SST and ENSO will be compared.

Data and methodology

The NCEP-DOE Reanalysis II daily data for the period 1979–2008 is used, which is an improved version of the NCEP Reanalysis I model that fixed errors and updated parameterizations of physical processes. The zonal wind (u), meridional wind (v), the vertical p-velocity (ω), relative humidity (r) and temperature (T) at multiple pressure levels for 1000–10 hPa and the surface pressure fields are used in this study. These variables are available on regular grids with a horizontal resolution of 2.5° × 2.5°. The sea surface temperature (SST) is derived from the NOAA extended reconstructed sea surface temperature version 3b and for the 1979–2008 period with a horizontal resolution of 2° × 2°. The Oceanic Niño Index (ONI) is derived from the Climate Prediction Center of NOAA for the period 1979–2008. The daily apparent moisture sink Q2 is computed by the formulas (Yanai et al., 1973)

formula
where q is the mixing ratio of water vapor, the horizontal wind, ω the vertical p-velocity, and L the latent heat of condensation.

Integrating (1) from pT to ps we have:

formula
where ps is the surface pressure, and pT = 10 hPa.

In this paper, to focus on the interannual variation, the oscillation with periods of 2–8 years was extracted from the monthly data of both <Q2> and SST. Furthermore, June-July-August mean indicates the seasonal mean in boreal summer. The singular value decomposition (SVD) and conditional SVD (CSVD) (An, 2003) methods are used to analyze the coupled characteristics of <Q2> and SST over the moisture source region of the AAM domain. Here, the <Q2>(x, y, t) and SST(x, y, t) are functions of space (x, y) and time (t), and the ONI(t) is a function of time only. For the CSVD method, we define new variables <Q2>*(x, y, t) and SST*(x, y, t), of which signals are obtained by removing the signal covariant with the simultaneous ONI(t) from <Q2>(x, y, t) and SST(x, y, t), respectively. Thus,

formula
where cov and var indicate the temporal covariance between two variables and the variance, respectively. From the definition of Eq. (3), the <Q2>* and SST* are completely uncorrelated with the simultaneous ONI.

Characteristics of interannual variation of moisture source in the AAM region in boreal summer

Climatic characteristics of moisture budgets over the AAM region

The AAM domain spans from about 40°E to 160°E and from 30°S to 40°N (Wang, 2006). Seasonal mean <Q2> over the AAM region in boreal summer for 30-year averages from 1979 to 2008 is shown in Figure 1. The mean <Q2> pattern exhibits maximum moisture sink in the Asian summer monsoon region, whereas a large region of moisture source in the southern Indian Ocean (SIO). A distinct latitudinal distribution is found in the moisture source region along 0°–30°S, where the contours distribute almost follow the latitude except when they encounter the continent. Based on this feature, the region (40°–160°E, 30°S–0°) is selected as the moisture source region of the AAM domain for the remainder of this paper.

Time-space structure of the interannual variation of <Q2> over moisture source region

To focus on the interannual variation, the oscillation with periods of 2–8 years was extracted from the raw data. The percentage variance explained by the 2–8 year oscillation (Figure not shown) exceeds 70% in most areas of the selected region (40°–160°E, 30°S–0°), especially in the central SIO region, which indicates that the interannual variation is outstanding in the moisture source region of the AAM domain. The EOF method is used to reveal the temporal and spatial characteristics of the interannual variation of moisture source. The percentage variance of the first EOF mode is 16.14%, which passes the North's significance test (North et al., 1982) for α = 0.05.

The spatial feature of EOF1 (Figure 2a) is a distribution along the northwest-southeast direction in most of the SIO. There is a large area of positive values in the western-central SIO and a large area of negative values in the northwest and north side of Australia, indicating that the two above regions are the key areas of the interannual variations of moisture source in the AAM region. Compared with Figure 1, the negative centers also locate in the two regions in the 30-year mean distribution. So, we need to pay special attention to the two regions. Figure 2b presents the time series of the first principal component (PC1) and the simultaneous ONI. Note that in the EOF analysis, no SST information was involved. However, PC1 has good accordance with the simultaneous ONI, especially in the peaks. The correlation coefficient between them is 0.72, passing the significance test at α=0.05 level, indicating that ENSO has close relationship with the interannual variation of moisture source in the AAM region. The other modes of EOF can't pass the significance test and therefore are not discussed here.

Coupled characteristics of <Q2> and SST in moisture source region of the AAM domain

The interannual variation of <Q2> is determined by both surface evaporation and precipitation that are influenced by the atmospheric circulation. SST plays an important role in the variation of all the three factors. Many researches have emphasized the importance of local SST on the variation of atmospheric circulation and compared the different influences of ENSO and local SST on atmospheric circulation over the AAM region. Besides, the features of the East Asian monsoon circulation anomalies and summer rainfall anomalies and a teleconnection between the central Pacific and East Asia also have been studied in the different phases of ENSO cycles (Wang et al., 2000; Huang et al., 2004). Therefore, we investigate the impacts of ENSO and local SST on the variation of <Q2> over moisture source region of the AAM domain. The coupled characteristics of <Q2> and SST are studied by using the SVD and conditional SVD (CSVD) (An, 2003) analysis methods. The relative roles of ENSO and local SST will be discussed.

Figure 3 is the heterogeneous coherence coefficient of the first two SVD modes. The percentage variance of the first two SVD modes are 48.97% and 21.29%. The mode correlation coefficient for the first two SVD modes are both 0.90. The spatial structure of <Q2> in SVD1 mode is almost inversely similar to the EOF1 pattern with a spatial correlation coefficient of −0.93. The coefficient is negative in the region of 50°–110°E and 30°–15°S, and the center is at about 75°E and 25°S. A positive center is in the region of 80°–160°E and 15°S–0°. The corresponding coefficient for SST is also negative in the region of 50°–110°E and 30°–15°S, but positive in the region of 95°–160°E and 15°S–0°. Comparing Figure 3b with 3a, we can find negative values in both <Q2> and SST fields over the western-central SIO and positive values in both <Q2> and SST fields in the northwest and north side of Australia. The <Q2> anomaly has a good positive correlation with the local SSTA in both regions. When the SST is higher (lower) over the western-central SIO, moisture source there is weaker (stronger). When the SST is lower (higher) over the northwest and north side of Australia, moisture source there is stronger (weaker). In Figure 3c, there is no systematic center in the strong moisture source areas. Comparing with Figure 3d, the <Q2> has no obvious correlation with local SST.

The climate system in the tropical Indian Ocean is influenced by ENSO through an atmospheric bridge (Klein et al., 1999). The coupled pattern of <Q2> and SST in the SVD analysis is influenced by both the local SST variation in the SIO and ENSO. In order to get the respective impact of local SST and ENSO on the variations of moisture source in the AAM region, we use the CSVD method to isolate the coupled pattern excluding the influence of ENSO. To do so, the signal of simultaneous ONI (ENSO index) has been removed from <Q2> and SST anomaly to represent the influence of local SST. The ratio of the variance after removing the ENSO influence to the total variance defines the variance fraction. Figure 4 shows the fraction of variance of <Q2> and SST over the moisture source region of the AAM domain. In most of the region, the fractional variance in Figure 4a is higher than 90%, but in the key areas of moisture source, they are minimum centers with the value of about 50%. So, ENSO has little impact on the variation of <Q2> in most of the moisture source region. In the key regions, however, it is relatively important, accounting for about 50% of the interannual variation of moisture source. Similar to Figure 4a, the fractional variance is higher than 90% in most area in Figure 4b. The minimum value of about 60% is also exhibited in the key regions of SST, which indicates that SST is determined mainly by the local air-sea interaction in most of the moisture source region, but in the key regions, local air-sea interaction and ENSO play almost equivalent roles on the interannual variations of SST.

Figure 5 is the same as Figure 3, but for CSVD. The percentage variances of the first two CSVD modes are 35.95% and 21.61%. The mode correlation coefficient for the first two CSVD modes are 0.86 and 0.93. In Figure 5a, the positive and negative values are distributed in turns. In Figure 5b, the coefficient is almost positive in all of the moisture source region. So, after the ENSO influence is excluded, the spatial structure of <Q2> in the CSVD1 is not silimar to the EOF1 pattern at all. The CSVD1 can't represent the major characterists of interannual variation of moisture source. However, the spatial structure of CSVD2 is homologous with SVD1 of <Q2> and inversely similar to the EOF1 pattern. The PC1 of SVD and PC2 of CSVD for <Q2> (Figure 6a) have good accordance, with a correlation coefficient of 0.47. In Figure 5c, a large area of negative values is seen in the western-central SIO, while a large area of positive values is in the northwest and north side of Australia. The spatial structure of SVD1 and CSVD2 for SST is also very similar to each other. In Figure 5d, there is also a large area of negative values over the western-central SIO, but positive values over the northwest and north side of Australia. The correlation coefficient between PC1 of SVD and PC2 of CSVD for SST (Figure 6b) is 0.45. In the two key areas, the positive correlation between <Q2> and local SST is also outstanding in CSVD2.

Based on the above analysis, we can make a simple conclusion. In the western-central SIO and the northwest and north side of Australia, the variation of SST is dominated not only by the local air-sea interaction but also by ENSO. Both local SST and ENSO play important role on variation of moisture source in these regions. The positive (negative) <Q2> is corresponding with positive (negative) local SST over both key regions. When the local SST is higher, the moisture source is weaker in the western-central SIO. When the local SST is lower, the moisture source is stronger in the northwest and north side of Australia. ENSO enhanced the anomaly of moisture source caused by the local SST. ENSO and local SST work cooperatively to influence the moisture source variation in the AAM region.

Conclusions and Discussion

The variation of the AAM is influenced by both ENSO and local SST. This work studied the relative impacts of ENSO and local SST on the interannual variation of moisture source in the AAM region in boreal summer. In the map of 30-year mean of apparent moisture sink (<Q2>), the Asian summer monsoon region is opposite to the SIO and the Australian monsoon region. There is a huge region of moisture source spanning from 40° to 160°E and from 30°S to 0°. The EOF1 pattern shows there are two key regions of the interannual variation of moisture source, one over the western-central SIO, and the other over the northwest and north side of Australia. The variations of moisture source in the two regions are opposite. The PC1 of EOF analysis is accordant with ONI in a very high degree, indicating that the interannual variation has a close relationship with ENSO. The coupled characteristics of <Q2> and SST over the moisture source region of the AAM domain are analyzed by using the SVD and CSVD method. The SVD1 of <Q2> almost inversely duplicates the EOF1, which indicates that the local SST over the SIO is a primary factor for the interannual variability of moisture source in the AAM region. In both key regions, there exists obvious positive correlation between <Q2> and local SST. For the CSVD method, the ENSO signal was excluded from both <Q2> and local SST data over the moisture source region of the AAM domain. Then, we find that the CSVD1 of <Q2> can't duplicate the EOF1 anymore, but the CSVD2 is very similar to SVD1 in both <Q2> and SST distribution. The fact that the CSVD1 and SVD1 are unlike, but the CSVD2 and SVD1 are alike, shows obvious evidence that both ENSO and local SST have great impacts on the variations of <Q2> over the moisture source region. If either ENSO or local SST is removed, the interannual variation of <Q2> will be not the same as the actual situation. ENSO and local SST work cooperatively to influence the interannual variations of moisture source in the AAM region. For the CSVD analysis, one caveat must be added. In an El Niño event, positive SST anomalies usually appear in the Indian Ocean about 3 months after SST anomalies peak in the tropical Pacific (Klein et al., 1999), which suggests a more complicated interaction between ENSO and the AAM that has not been addressed here. In this study, the simultaneous ONI is removed from the <Q2> and SST data for the CSVD analysis based on the fact that PC1 has good accordance with the simultaneous ONI (Figure 2a). Finally, this work just showed some statistical results to prove the important roles of ENSO and local SST over the SIO in the interannual variations of moisture source in the AAM region. The mechanism involving how the ENSO and local SST influence the interannual variations of moisture source in the AAM region will be discussed in further work.

Acknowledgements

This work was supported by the National Natural Science Foundation of China under Grant No. 40605028, the Natural Science Foundation of Guangdong, China under Grant No. 10151027501000101, and the Fundamental Research Funds for the Central Universities of China. The authors appreciate the constructive and useful comments of the reviewers for improving the presentation of this paper.

The text of this article is only available as a PDF.

References

An, S.-I.
2003
.
Conditional maximum covariance analysis and its application to the tropical Indian Ocean SST and surface wind stress anomalies
.
J. Climate
,
16
:
2932
2938
.
Cayan, D. R.
1992
.
Latent and sensible heat flux anomalies over the northern oceans: The connection to monthly atmospheric circulation
.
J. Climate
,
5
:
354
369
.
England, M. H., Ummenhofer, C. C. and Santoso, A.
2006
.
Interannual rainfall extremes over southwest Western Australia linked to Indian Ocean climate variability
.
J. Climate
,
19
:
1948
1969
.
Fairall, C. W., Bradley, E. F., Rogers, D. P., Hare, J. E., Grachev, A. A. and Edson, J. B.
2003
.
Bulk parameterization of air–sea fluxes: Updates and verification for the COARE algorithm
.
J. Climate
,
16
:
571
591
.
Goddard, L. and Graham, N. E.
1999
.
Importance of the Indian Ocean for simulating rainfall anomalies over eastern and southern Africa
.
J. Geophys. Res.
,
104
:
19099
19116
.
Huang, R. H., Chen, W., Yan, B. L. and Zhang, R. H.
2004
.
Recent advances in studies of the interaction between the East Asian winter and summer monsoon and ENSO cycle
.
Adv. Atmos. Sci.
,
21
:
407
424
.
Klein, S. A., Sode, B. J. and Lau, N.-C.
1999
.
Remote sea surface temperature variations during ENSO: Evidence for a tropical atmospheric bridge
.
J. Climate
,
2
:
917
932
.
Li, T., Tung, Y-C. and Hwu, J.-W.
2005
.
Remote and local SST forcing in shaping Asian-Australian monsoon anomalies
.
J. Meteor. Soc. Japan
,
83
(
2
):
153
167
.
North, G. R., Bell, T. L., Cahalan, R. F. and Moeng, F. J.
1982
.
Sampling errors in the estimation of empirical orthogonal functions
.
Mon. Wea. Rev.
,
110
:
699
706
.
Qiao, Y. T. and Lin, M. J.
2006
.
The characteristics of moisture transport over Asian-Australian monsoon region
.
Acta Scientiarum Naturalium Universitatis Sunyatseni
,
45
(
6
):
102
105
.
(In Chinese with English abstract)
Qiao, Y. T., Luo, H. L. and Jian, M. Q.
2002
.
The temporal and spatial characteristics of moisture budgets over Asian and Australian monsoon regions
.
J. Tropical Meteorology
,
8
(
2
):
113
120
.
Reason, C. J. C.
2001
.
Subtropical Indian Ocean SST dipole events and southern African rainfall
.
Geophys. Res. Lett.
,
28
:
215
218
.
Reason, C. J. C. and Mulenga, H.
1999
.
Relationships between South African rainfall and SST anomalies in the southwest Indian Ocean
.
Int. J. Climatol.
,
19
:
1651
1673
.
Rocha, A. and Simmonds, I.
1997
.
Interannual variability of south–eastern African summer rainfall. Part 2: Modeling the impact of sea surface temperatures in rainfall and circulation
.
Int. J. Climatol.
,
17
:
267
290
.
Roxy, M. and Taninoto, Y.
2007
.
Role of SST over the Indian Ocean in influencing the intraseasonal variability of the Indian summer monsoon
.
J. Meteor. Soc. Japan
,
85
(
3
):
349
358
.
Ummenhofer, C. C., Gupta, A. S., Pook, M. J. and England, M. H.
2008
.
Anomalous rainfall over southwest western Australia forced by Indian Ocean sea surface temperatures
.
J. Climate
,
21
:
5113
5134
.
Wang, B.
2006
.
The Asian Monsoon
,
New York
:
Springer/Praxis Publishing
.
Wang, B., Wu, R. G. and Fu, X. H.
2000
.
Pacific-East Asian teleconnection: How does ENSO affect East Asian climate?
.
J. Climate
,
13
:
1517
1536
.
Webster, P. J., Magana, V. O., Palmer, T. N., Shukla, J., Tomas, R. A., Yanai, M. and Yasunari, T.
1998
.
Monsoons: Processes, predictability, and the prospects for prediction
.
J. Geophys. Res.
,
103
:
14451
14510
.
Yanai, M., Esbensen, S. and Chu, J. H.
1973
.
Determination of bulk properties of tropical cloud clusters from large-scale heat and moisture budgets
.
J. Atmos. Sci.
,
30
:
611
627
.
Yang, M. Z. and Ding, Y. H.
2006
.
Diagnostic study of the variation of Indian Ocean sea surface temperature and its impact on Indian summer monsoon rainfalls
.
Acta Oceanologica Sinica
,
28
(
4
):
9
16
.
(In Chinese with English abstract)
Yu, L.
2007
.
Global variations in oceanic evaporation (1958–2005): The role of the changing wind speed
.
J. Climate
,
20
:
5376
5390
.
Zhang, G.-J. and McPhaden, M. J.
1995
.
The relationship between sea surface temperature and latent heat flux in the equatorial Pacific
.
J. Climate
,
8
:
589
605
.
Zou, L. Y. and Liu, X. F.
2002
.
Numerical study on the effect of SSTA in the tropical Indian Ocean and Pacific on Asian summer monsoon
.
J. of Nanjing Institute of Meteorology
,
25
:
166
172
.