This study aims to document the seasonal cycle of chlorophyll a interannual variability and to determine the dynamic framework for understanding the mechanisms that produce it. We investigated the variation in sea surface chlorophyll a concentration from autumn 1997 to summer 2010 in the South China Sea using the Sea-Viewing Wide Field-of-View Sensor products. Using the Season-Reliant Empirical Orthogonal Function method, we revealed the interannual and seasonal variations and their dependency on climate variability. The time coefficient of the first mode shows a sudden drop in 2007. Corresponding season-reliant empirical orthogonal function (spatial pattern) is the largest in the southern centre in the autumn, which can be explained by a large increase in chlorophyll a concentration during which Typhoon Hagibis passed through. The second mode reflected an anomalous event of low phytoplankton biomass in the western center of the sea during the summer of 1998 and 2010 due to an El Niño event. Time coefficient is negatively related with the Multivariate El Niño–Southern Oscillation Index. The third mode revealed an interannual variability of chlorophyll a in the region to the northwest of Luzon Islands, which appeared from autumn to the next spring. To better understand the chlorophyll a variation off the South Vietnam coast, we analyzed the summer chlorophyll a variability in this local region solely using the Empirical Orthogonal Function method.

Introduction

The South China Sea (SCS) is the largest marginal sea in the tropics, located between the western Pacific and the eastern Indian Ocean. It covers 3.5 million km2 from the equator to 23°N and from 99°E to 121°E, with the bathymetry characterized by vast shelf regions in the west and a deep basin in the east (Figure 1). The SCS is connected with the East China Sea to the northeast by Taiwan Strait, the Pacific Ocean to the east by Luzon strait, and the Java Sea and the Indian Ocean to the southwest by Karimata Strait. The SCS is surrounded by south China to the north, Vietnam to the west and by large and small islands including Taiwan Islands, Philippines Islands, Palawan Islands, Karimata Islands, Sumatra islands and Malay Peninsula.

SCS is influenced deeply by the Southeast Asian monsoon system (Wyrtki, 1961). The general circulation pattern of the SCS is a basin-wide cyclonic gyre driven by northeast monsoon in winter and an anticyclonic gyre driven by the southwest monsoon in summer, with additional influence by Kuroshio in the northern part (Qu, 2000; Su, 2004; Liu et al., 2008; Wang and Tian, 2010). The SCS plays an important role in the East Asian monsoon due to its high water temperature and large heat content in the upper ocean. A previous study indicated the physical characteristics, including sea surface temperature (SST), sea surface height, wind and sea surface heat flux in the SCS showed multi-scale variability (Xie et al., 1998; Qu, 2001; Gao and Zhou, 2002; Xie et al., 2007; Zeng and Wang, 2009; Zhuang et al., 2010). The interannual variability of physical parameters in the SCS is closely related to El Niño–Southern Oscillation (ENSO), which is a quasiperiodic variation in climate which arises from a complex interaction between the tropical Pacific Ocean and the atmosphere, and the possible mechanisms have been widely discussed (Zhou et al., 1998; Fang et al., 2006; Wang et al., 2006; Rong et al., 2007; Yan et al., 2010). Previous studies on multi-scale variability of phytoplankton on the large scale of SCS focused on the seasonal phytoplankton variability (Shen et al., 2008) or on the chlorophyll a concentration anomaly during 1997–1998 period (Xie et al., 2003; Zhao and Tang, 2007), as well as the decadal variability (Liu et al., 2012). Recently, two reports on the understanding of the seasonal and interannual changes in phytoplankton abundance in the South China Sea have been published (Qiu et al., 2011; Tang et al., 2011). The first report revealed a significant interannual variability of chlorophyll a linked with climate in the SCS. The second report illustrated the seasonal and non-seasonal variability of the central western SCS. They indicated high chlorophyll variability off the southeast Vietnam coast (Qiu et al., 2011). However, these findings did not show us the chlorophyll a concentration interannual variability in different seasons in the SCS.

In this article, we will reanalyze the chlorophyll a in the SCS with the Season-Reliant Empirical Orthogonal Function (S-EOF) method (Figure 2). After an analysis of the entire SCS, we will focus on the summer chlorophyll a variation of Southeast Vietnam Offshore, which traditionally was considered as one of the most dynamically active areas in the SCS.

Data and methods

The data used in the present study include the chlorophyll a concentration, SST, wind stress curl, Multivariate ENSO Index (MEI). The SCS subset of sea surface chlorophyll a concentration is extracted from the Ocean Biology Processing Group (OBPG) of the Goddard Space Flight Center (GSFC), National Aeronautics and Space Administration (NASA). We use the latest monthly and seasonal Sea-viewing Wide Field-of-view Sensor (SeaWiFS) standard ocean color products, which were produced and distributed via the ocean color web (http://oceancolor.gsfc.nasa.gov/ftp.html), with the reprocessing versions SeaWiFS (R2010.0). We chose the mapped seasonal products with 9 km × 9 km spatial resolution from autumn 1997 to summer 2010, all together 13 years data (52 images). The seasonal product was averaged from 91 or 92 days of the satellite daily products by OBPG, with spring defined from the 80th or 81st to the 171st or the 172nd, summer defined from 172nd or 173rd to 263rd or 264th, autumn defined from 264th or 265th to 354th or 355th, and winter defined from 355th or 356th to 79th or 80th of the next year.

To reduce the size of the data set for the subsequent calculation, the original 9 km × 9 km data are re-binned to a 1/3° × 1/3° grid using the geometric mean. The missing data in these re-binned data are then estimated using the Kriging interpolation with a Gaussian semivariogram model and a fixed search radius of 3 points (Stein, 1999). Except for summer 2008, our averaging and smoothing process results in an almost gap-filled time series in the study area. The summer 2008 is still missing about 25% of the data. For consistent analysis, the missing data are replaced with the mean value of summer 2007 and summer 2009. As satellite data of chlorophyll a can be approximated with a lognormal distribution, we log-transformed the data for analysis. After that, every point was weighted using the cosine of latitude (for equal area weighting) to get a data matrix Xm×n, where m is the number of spatial points, and n is the time length. For EOF analysis, the climatologically seasonal means from spring to winter during autumn 1997 to summer 2010 are calculated and removed from the seasonal values to obtain the seasonal anomalies. For analysis the seasonal characteristics interannual variability, four seasons of data are considered as one spatial variable.

The SST data used in this article are the extracted SCS subset of the Tropical Rain Measuring Mission (TRMM) from the Asian-Pacific data Research Center of the University of Hawaii. The data is nearly free of cloud over the SCS (Wentz et al., 2000). The data have a resolution of 0.25 × 0.25°. The data cover the period from December 1997 to December 2009.

Wind data used here are the Quikscat Scatterometer monthly mean field products produced by the Center for Satellite Exploitation and Research (CERSAT) at the French Institute of Research for the Exploitation of the Sea (Ifremer) (http://cersat.ifremer.fr/data/discovery/by_parameter/ocean_wind/mwf_quikscat). The data have a resolution of 0.5 × 0.5°. The data are available from August 1999 to October 2009.

MEI data, which used to scale the strength of the ENSO, are acquired from the Climate Diagnostics Center, NOAA (http://www.cdc.noaa.gov/Correlation/mei.data).

The EOF analysis is a decomposition of a signal or data set in terms of orthogonal basis functions which are determined from the data and it has been used routinely in physical oceanography and atmospheric studies (Nagar and Singh, 1991), as well as in biological oceanography recently (Wilson and Adamec, 2001; Yoder and Kennelly, 2003; Venegas et al., 2008). In EOF analysis, a temporally and spatially chlorophyll data set, Chl(x,t), with location x = 1, M at time t = 1, N is decomposed according to , where ai(t) are the temporal components of the spatial components φi(x). The temporal and spatial components are calculated from the eigenvectors and eigenfunctions of the N × N covariance matrix of the data set. Although the traditional EOF analysis demonstrates the dominant modes of the spatial and temporal variability of the chlorophyll a variability, it cannot obtain coherent seasonal characteristics. Compared with traditional EOF analysis, the four seasonal anomalous chlorophyll a fields are treated as an integral block to construct covariance matrix in S-EOF analysis. Consequently, the principal component (PC) then will be divided into four consecutive seasonal fields after the analysis (Wang and An, 2005). The S-EOF analysis is able to identify regions of energetic variability as can conventional EOFs. Consequently, herein we do not a priori limit our attention to a specific season or region in the study area, but allow the chlorophyll a data to determine these by themselves (Minobe, 2002).

Results and discussion

The analysis of the entire SCS

In different areas and different times in the SCS, impact factors on chlorophyll are generally different. The first three S-EOF modes of the chlorophyll a concentration account for 53% of the total variance, which can explain most of the useful variability in the SCS (Figure 2).

The first S-EOF mode, which accounts for 24.6% of the total variance, is shown in Figure 3. In this mode, the variability concentrated on autumn and the next summer. Combined with pc 1, it is clear that the maximum variability focused on autumn 2007 to summer 2008 of the first S-EOF mode. According to its spatial mode, almost all the SCS shows increase of chlorophyll a in autumn 2007. A sudden strong enhancement of chlorophyll a appeared in center of SCS, near Vietnam coast and a slight enhancement of chlorophyll a in the southwest of Luzon Strait in the autumn. The SST and wind stress curl show great differences on November 2007. The cooling down with the stronger wind stress curl can explain the increase of chlorophyll a in most of the SCS. In the southwest of Luzon Strait, the wind direction changes in late autumn, and the northeast monsoon prevails. The northeast monsoon strengthens in November. During the northeast monsoon, strong wind stress induces Ekman pumping, which transports subsurface nutrients stimulating the growth of the phytoplankton (Wang et al., 2010). A strong upwelling and cooling down in the center of SCS near Vietnam coast (Figure 4), had been induced by Hagibis (Sun et al., 2010). A storm warning was issued on 18 November, and on 21 November, the Joint Typhoon Warning Center upgraded Hagibis to typhoon status. Hagibis then wobbled off the coast of Vietnam beginning on 22 November where it gradually weakened. However, on 24 November, the Hagibis met Typhoon Mitag and interacted with it. Then, it turned back eastward to the SCS. It weakened to a depression east of the Philippines on 28 November. These increases of chlorophyll in the east of Luzon strait were caused by a slight cooling down in this area (Figure 3). Strong Ekman pumping and mesoscale cold eddies induced by the long forcing of typhoon tend to bring deep-layer cold, nutrient-rich water to the euphotic zone and trigger phytoplankton bloom. The typhoon induces large wind stress and low SST in the Western SCS. In the Southwestern Luzon Strait, A seasonal upwelling occurs and leads negative SST anomaly. In the summer of 2008, there is a low chlorophyll a concentration in the west coast of Vietnam, which also can be explained by a warming up in that year during this time (not shown in this article).

The second S-EOF mode, which can explain 18.6% of the total variance, exhibits the largest amplitudes in summer, especially in the center-west SCS, near Vietnam coast. From pc 2, it shows in 1998 and 2010, the chlorophyll a decreased in the three seasons: winter (Figure 5b), spring (Figure 5c) and summer (Figure 5d). The decreasing magnitude in the summer is the most clear. The chlorophyll a in the west-center of SCS has been reported as a jet-shaped high chlorophyll a concentration (Kuo et al., 2000). An anomalous event in all of the SCS was reported in the summer of 1998 and compared with chlorophyll reported from other summers, the difference was especially evident in the western SCS (Zhao and Tang, 2007). In 2009–2010, although it is not as a strong an ENSO year as 1997–1998, the chlorophyll a shows similar magnitude of down trending in the SCS. The pc 2 shows high negative correlation ship with spring MEI (with R2 = 0.72), which can explain that in the ENSO year, the chlorophyll a decreases in the summer. During and after an El Niño, the SCS is anomalously warm, and the SST anomalies are large in the east of Vietnam in summer (Liu et al., 2004). The warmer SST may limit vertical mixing because the water column is stabilized by thermal stratification. Hence, the growth of phytoplankton may be restricted by the rapidly exhausted nutrients in the euphotic zone where no deep-layer nutrient supplies.

The third S-EOF mode of the chlorophyll concentration anomalies explains 10.1% of the total variance. The corresponding spatial pattern and time coefficients are shown in Figure 6. The third S-EOF mode shows the seasonal characteristics of chlorophyll a variability induced by Luzon cold eddy. The Luzon cold eddy is one of the most important circulation features in the northern SCS (Soong et al., 1995; Qu, 2000). The eddy will enhance the productivity in the area (Chen et al., 2007; Shen et al., 2008). The Luzon cold eddy, begins from October, matures the next January and ends in May. The spatial pattern (Figures 6a–c) reveals the chlorophyll interannual variability off the northwest coast of Luzon Island. The La Niña events have been regarded as the modulation of the west Pacific water entering the northern SCS which will contribute to the weakening of the Luzon eddy (Zheng et al., 2007). However, we do not find significant relationships between the chlorophyll a with La Niña events. The maximum correlation coefficient is the summer MEI with pc 3, with the value is 0.3894. From pc 3, the chlorophyll a in the Luzon cold eddy area shows the lowest from autumn 1997 to spring 1998 and highest in autumn 2001 to spring 2002 and autumn 2009 to spring 2010. In the summer, the chlorophyll shows symmetrical characteristic variability in the northern and the southern Vietnam Coast, which we will analyze in the next section.

The analysis of the local areas in the SCS

From the analysis of these results, it was concluded that in summer, offshore South Vietnam revealed significant chlorophyll a variability in the study area, which was traditionally recognized as one of the most obvious dynamic regions in the SCS. Therefore, we chose the summer chlorophyll a concentration of this area separately to analyze using EOF method.

The area was located in the southwest SCS (8°N–16°N, 108°E–115°E). The first S-EOF mode occupies 53% of the variance contribution, which represents the strong chlorophyll a concentration induced by the cold filament in the area (Figure 7). The time coefficient series are very similar with that of Figure 5, which exhibited that there were 2 extreme lower chlorophyll a events in the area, one is in 1998 and the other is in 2010. The coefficients are similar may be because the main variability of Figure 5 is concentrated on this area and this season. The second EOF mode, which occupies 12% of the variance contribution, reveals a dipole structure of the chlorophyll a concentration. The circulation in this area splits into a weakened cyclonic gyre north of 12 N and a strong anticyclonic gyre in the south (Wang et al., 2006). The dipole structure of chlorophyll a concentration variability may induced by an anticyclonic eddy in the south and a cyclonic eddy in the north during the summer in the study area.

Conclusions

The S-EOF methods, which deals with a season-dependent time series of data, using four season data as its spatial state vector, is a useful tool to get the seasonal characteristics of the interannual variability. Compared with the conventional EOF analysis, the seasonally combined EOF analysis gets more detailed seasonal information of the chlorophyll a interannual variability in the SCS, which can account for the SCS interannual variability being strongly modulated by the seasonal forcing.

The strong enhancement of chlorophyll a concentration in autumn 2007 in the center of SCS was not seen in the previous time series of chlorophyll a analysis studies (Qiu et al., 2011; Tang et al., 2011), although both of them included data from 2007. This may be because the smoothing of the four seasonal data reduced the significance of abnormal signals, unlike the S-EOF methods used in the article. The interannual chlorophyll a variability induced by the summer upwelling off the South Vietnam coast and the offshore spread of cold water revealed similar trends as the previous studies. However, a dipole structure of summer chlorophyll a variability in this area is another new finding in this research.

Acknowledgements

Many thanks to the SeaWiFS Project within NASA GSFC, for providing the SeaWiFS products, CERSAT at the Ifremer for providing the QuikSCAT products, Asian-Pacific Data Research Center of the University of Hawaii for providing the TRMM SST products.

Funding

The research work was supported by the Public Science and Technology Research Funds of Ocean (No. 201205040), the Sun Yat-sen University Cultivation Fund for Young Teachers (Grant No. 42000-3321400) and the Innovation Group Program of State Key Laboratory of Tropical Oceanography, South China Sea Institute of Oceanology, Chinese Academy of Sciences (No. LTOZZ1201).

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

References

Chen, Y. L. L., Chen, H. Y., Lin, L. L., Lee, M. A., Chang, J.,
2007
.
Effects of cold eddy on Phytoplankton production and assemblages in Luzon Strait bordering the South China Sea
.
J. Oceanogr.
63
(
4
),
671
683
.
Fang, G. H., Chen, H. Y., Wei, Z. X., Wang, Y. G., Wang, X. Y., Li, C. Y.,
2006
.
Trends and interannual variability of the South China Sea surface winds, surface height, and surface temperature in the recent decade
.
J. Geophys. Res.
111
(
C11
),
C11S16
.
Gao, T. Z., Zhou, F. X.,
2002
.
Monsoonal characteristics revealed by intraseasonal variability of Sea Surface Temperature (SST) in the South China Sea (SCS)
.
Geophys. Res. Lett.
29
(
8
),
1222
.
Kuo, N., Zheng, Q., Ho, C.,
2000
.
Satellite observation of upwelling along the western coast of the South China Sea
.
Remote Sens. Environ.
74
,
463
470
.
Liu, F., Chen, C., Zhan, H.,
2012
.
Decadal variability of chlorophyll a in the South China Sea: a possible mechanism
.
Chin. J. Oceanology and Limnology
30
(
6
),
1054
1062
.
Liu, Q., Jiang, X., Xie, S., Liu, W.,
2004
.
A gap in the Indo-Pacific warm pool over the South China Sea in boreal winter: Seasonal development and interannual variability
.
J. Geophys. Res.
109
,
C07012
.
Liu, Q. Y., Kaneko, A., Su, J. L.,
2008
.
Recent progress in studies of the South China Sea circulation
.
J. Oceanogr.
64
(
5
),
753
762
.
Minobe, S.,
2002
.
Interannual to interdecadal changes in the Bering Sea and concurrent 1998/99 changes over the North Pacific
.
Prog. Oceanogr.
55
, (
1–2
),
45
64
.
Nagar, S. G., Singh, S. V.,
1991
.
Variations of surface pressure over India during the southwest monsoon
.
Theor. Appl. Climatol.
44
(
2
),
95
112
.
Qiu, F. W., Fang, W. D., Fang, G. H.,
2011
.
Seasonal-to-interannual variability of chlorophyll in central western South China Sea extracted from SeaWiFS
.
Chin. J. Oceanol. Lim.
29
(
1
),
18
25
.
Qu, T. D.,
2000
.
Upper-layer circulation in the South China Sea
.
J. Phys. Oceanogr.
30
(
6
),
1450
1460
.
Qu, T. D.,
2001
.
Role of ocean dynamics in determining the mean seasonal cycle of the South China Sea surface temperature
.
J. Geophys. Res.
106
(
C4
),
6943
6955
.
Rong, Z. R., Liu, Y. G., Zong, H. B., Cheng, C.,
2007
.
Interannual sea level variability in the South China Sea and its response to ENSO
.
Glob. Planet Change
55
(
4
),
257
272
.
Shen, S. H., Leptoukh, G. G., Acker, J. G., Yu, Z. J., Kempler, S. J.,
2008
.
Seasonal variations of chlorophyll a concentration in the northern South China Sea.IEEE Geosci
.
Remote Sens. Lett.
5
(
2
),
315
319
.
Soong, Y. S., Hu, J. H., Ho, C. R., Niiler, P. P.,
1995
.
Cold-core eddy detected in South China Sea
.
Eos Trans.
, AGU
,
76
(
35
),
345
.
Stein, M. L.,
1999
.
Interpolation of Spatial Data: Some Theory for Kriging.
Springer-Verlag
,
New York
.
Su, J. L.,
2004
.
Overview of the South China Sea circulation and its influence on the coastal physical oceanography outside the Pearl River Estuary
.
Cont. Shelf Res
.
24
(
16
),
1745
1760
.
Sun, L., Yang, Y. J., Xian, T., Lu, Z. M., Fu, Y. F.,
2010
.
Strong enhancement of chlorophyll a concentration by a weak typhoon
.
Mar. Ecol. Prog. Ser.
404
,
39
50
.
Tang, S. L., Dong, Q., Liu, F. F.,
2011
.
Climate-driven chlorophyll-a concentration interannual variability in the South China Sea
.
Theor. Appl. Climatol.
103
(
1–2
),
229
237
.
Venegas, R. M., Strub, P. T., Beier, E., Letelier, R., Thomas, A. C., Cowles, T., James, C., Soto-Mardones, L., Cabrera, C.,
2008
.
Satellite-derived variability in chlorophyll, wind stress, sea surface height, and temperature in the northern California Current System
.
J. Geophys. Res.
113
(
C3
),
C03015
.
Wang, B., An, S. I.,
2005
.
A method for detecting season-dependent modes of climate variability: S-EOF analysis
.
Geophys. Res. Lett.
,
32
,
L15710
.
Wang, C. Z., Wang, W. Q., Wang, D. X., Wang, Q.,
2006
.
Interannual variability of the South China Sea associated with El Nino
.
J. Geophys. Res
.
111
(
C3
).
Wang, G. H., Chen, D. K., Su, J. L.,
2006
.
Generation and life cycle of the dipole in the South China Sea summer circulation
.
J. Geophys. Res.
111
(
C6
),
C06002
.
Wang, J., Tang, D., Sui, Y.,
2010
.
Winter phytoplankton bloom induced by subsurface upwelling and mixed layer entrainment southwest of Luzon Strait
.
J. Marine Syst.
83
,
141
149
.
Wang, Q., Tian, Y.,
2010
.
Definition of the South China Sea summer monsoon onset
.
Chin. J. Oceanol. Lim.
28
(
6
),
1281
1289
.
Wentz, F. J., Gentemann, C., Smith, D., Chelton, D.,
2000
.
Satellite measurements of sea surface temperature through clouds
.
Science
288
(
5467
),
847
850
.
Wilson, C., Adamec, D.,
2001
.
Correlations between surface chlorophyll and sea surface height in the tropical pacific during the 1997–1999 El Nino-Southern Oscillation event
.
J. Geophys. Res.
106
(
C12
),
31175
31188
.
Wyrtki, K.,
1961
. Physical oceanography of the Southeast Asian waters: scientific results of marine investigation of the South China Sea and the Gulf of Thailand. Scripps Inst. of Oceanogr., La Jolla, Califonia.
Xie, A., Chung, Y. S., Liu, X., Ye, Q.,
1998
.
The interannual variations of the summer monsoon onset over the South China Sea
.
Theor. Appl. Climatol.
59
(
3–4
),
201
213
.
Xie, S. P., Xie, Q., Wang, D. X., Liu, W. T.,
2003
.
Summer upwelling in the South China Sea and its role in regional climate variations
.
J. Geophys. Res.
108
(
C8
),
3261
.
Xie, S. P., Chang, C. H., Xie, Q., Wang, D. X.,
2007
.
Intraseasonal variability in the summer South China Sea: Wind jet, cold filament, and recirculations
.
J. Geophys. Res.
112
(
C10
),
C10008
.
Yan, Y. F., Qi, Y. Q., Zhou, W.,
2010
.
Interannual heat content variability in the South China Sea and its response to ENSO
.
Dyn. Atmos. Oceans
50
(
3
),
400
414
.
Yoder, J. A., Kennelly, M. A.,
2003
.
Seasonal and ENSO variability in global ocean phytoplankton chlorophyll derived from 4 years of SeaWiFS measurements
.
Glob. Biogeochem. Cy.
17
(
4
),
1112
.
Zeng, L. L., Wang, D. X.,
2009
.
Intraseasonal variability of latent-heat flux in the South China Sea
.
Theor. Appl. Climatol.
97
(
1–2
),
53
64
.
Zhao, H., Tang, D. L.,
2007
.
Effect of 1998 El Nino on the distribution of phytoplankton in the South China Sea
.
J. Geophys. Res.
112
(
C2
),
C02017
.
Zheng, Z. W., Ho, C. R., Kuo, N. J.,
2007
.
Mechanism of weakening of west Luzon eddy during La Nina years
.
Geophys. Res. Lett.
34
(
11
),
L11604
.
Zhou, F. X., Yu, S. Y., Liu, Q.,
1998
. The interannual oscillation of meridional sea level wind and its impact on sea surface temperature in the South China Sea. p. 523–533. Proceedings of the 3rd International Conference on East Asia and Western Pacific Meteorology and Climate. 1996 May 16–18, Chungli, Taiwan.
Zhuang, W., Xie, S. P., Wang, D. X., Taguchi, B., Aiki, H., Sasaki, H.,
2010
.
Intraseasonal variability in sea surface height over the South China Sea
.
J. Geophys. Res.
115
,
C04010