The present study aims to investigate the temporal variability of chromophoric dissolved organic matter through remote sensing and to determine its influence factors in the Pearl River Estuary. A medium resolution imaging spectrometer, chromophoric dissolved organic matter product was evaluated with in situ absorption coefficient (R2 = 0.9605, RMS = 0.1672 and MRE = 0.3930). L2 daily products were then averaged into monthly data to analyze variability in the specified area from January 2003 to December 2009. Accumulated anomaly analysis and the wavelet analysis showed that the variability had a significant period of one year from 2003 to 2009, accompanied with 3 and 6 month periods in 2004 and 2008 separately. Cross wavelet transform was used to analyze the relationships between chromophoric dissolved organic matter and its influence factors including rainfall and phytoplankton in time frequency space. Relationships with salinity and light were also analyzed. It was found that photobleaching induces lower dissolved organic matter in summer even though there is greater discharge due to large rainfall intensity during this season. However, the anomalous strong precipitation may cause significantly higher findings in the study area. An example was a heavy flood caused by strong precipitation in 2008 induced higher findings for the period for less than half a year.

Introduction

Chromophoric dissolved organic matter (CDOM, also called Gelbstoff) represents the light absorbing fraction of dissolved organic carbon (DOC). It absorbs in the ultraviolet and blue range of the visible spectrum. Therefore, CDOM could alter the optical field of water columns in coastal regions and estuaries (Keith et al., 2002). Moreover, CDOM could release inorganic carbon through photochemical processes and affect carbon cycling in the coastal area.

The Pearl River is the largest river in South China with a mean annual discharge volume of 3.26 × 1011 m3. Eighty percent of the annual discharge volume occurs during the wet season, from April to September (Yin et al., 2004). Terrestrial original materials discharge into the South China Sea through a delta with eight river outlets and then mix with the sea water (Chen et al., 2003a, 2003b). CDOM in the Pearl River is considered to be of terrestrial origin since the CDOM is larger nearshore and lower offshore (Hong et al., 2005). The geographical extent of terrestrially dominated regions varies seasonally, depending on the magnitude of river discharge.

Many studies on CDOM in the Pearl River Estuary (PRE) emphasize on the absorbance and fluorescence of CDOM, algorithms for CDOM, and distribution of CDOM (Callahan et al., 2004;Chen et al., 2003b, 2004b; Hong et al., 2005). There are sparse reports on continuous temporal variability analysis of CDOM and its influence factors based on the satellite data in coastal areas and estuaries. The CDOM low frequency (seasonal and inter-annual) variability in coastal areas and its influence factors are therefore not clear. Since CDOM absorption coefficient can be detected based on its peculiar spectral characteristics using remote sensing technique, it is especially convenient to analyze the variability of CDOM in optically complex water. The Medium Resolution Imaging Spectrometer (MERIS) CDOM product is retrieved using “Neural Network” algorithm which encompasses Cass 1 as well as Cass 2 water with the maximum possible accuracy (Bricaud and Morel, 1999).

The purpose of this study is to analyze the temporal variability of CDOM in the PRE and to identify its influence factors. First, the MERIS products are evaluated using in situ CDOM absorption coefficient data. Furthermore, remote-sensed CDOM variability trends and periods are analyzed using the accumulated anomaly analysis and wavelet analysis. In addition, the influences of rainfall, chlorophyll-a, salinity and light on CDOM are discussed.

The factors that influence CDOM absorption coefficient variability are complicated and studied by many researchers (D’sa and Dimarco, 2009; Retamal et al., 2007). Physical, chemical and biological processes all affect the temporal variability and optical properties of CDOM (Kowalczuk et al., 2005). The CDOM variability essentially lies on the balance between sources and sinks. The CDOM is derived primarily from terrestrial and anthropogenic organic matter and local phytoplankton in the semi-enclosed river estuary and coastal waters. CDOM is removed from water generally in three ways, coagulation and precipitation of the high molecular weight fraction, microbial degradation and photobleaching (Kowalczuk et al., 2005).

Data and methods

Eight cruises were carried out in the Pearl River Estuary, PRE20030311, PRE20030419, PRE20090815, PRE20091022, PRE20091122, PRE20091213, PRE20100319 and PRE20100704. The numbers after the PRE are field investigation dates. Figure 1 shows the location of the sampling stations during the field measurements and the specified study area investigated with use of ocean color imagery. The specified study area (region C) is located in the plume front, and the remote sensing data in the region C is less influenced by the atmospheric correction error due to deeper water depth compared with high-latitude shallower water region of the PRE (Wong et al., 2004). During these cruises, the downward irradiance just above water (Ed, W m−2)was obtained, followed by the protocol described by Mobley (1999).

Synchronous surface water samples were collected and filtered promptly through Whatman glass microfibre GF/F filter with 25 mm of diameter produced by Whatman International Ltd Maidstone England. The CDOM absorbance was measured using Gelbstoff Optical Analysis Laboratory System (GOALS) at wavelength of 440 nm with a sampling interval of 0.38 nm for all cruises (D’sa et al., 1999). The Gelbstoff Optical Analysis Laboratory System includes the World Precision Instruments Liquid Waveguide Capillary Cell, light source and spectrometer. The World Precision Instruments Liquid Waveguide Capillary Cell contains a rigid quartz capillary tubing with the length of 5 m coated by an amorphous polymer optical cladding. When the water sample was injected into one end of the quartz capillary tubing, a light source was introduced into the same end of the waveguide via an optical fiber. The light attenuated by water sample was recorded by the spectrometer at the other end. The spectra were then processed following the method proposed by D’Sa et al. (1999). Before the measurement of water samples, ultrapure water was measured in the same way as the reference. The chlorophyll a in the cruises of PRE20090815, PRE20091022, PRE20091122 and PRE20100704 were extracted with 90% acetone solution and analyzed with Turner Designs 10 fluorometer. Salinity was measured using YSI 6600 instrument in the cruises of PRE20090815, PRE20091022, PRE20091122 and PRE20100704.

The evaluation of MERIS CDOM product is based on the MERIS Level 2 Reduced Resolution (1.1 km at nadir) data containing remote sensing reflectance, water color constituents, other water properties and atmospheric information. MERIS default CDOM absorption coefficient at 440 nm is retrieved using the neural network algorithm. In order to reduce the effects of temporal variability on the in situ data and reach the greatest possibility of a match, only the stations that were undertaken within ±3 h windows around the satellite overpass were chosen to evaluate the CDOM absorption coefficient from satellite (Bailey and Werdell, 2006).There are six stations available, indicated by grey pentagrams (Figure 1). A box of 3×3 pixels is defined, centered on the location of in situ measurement. The median value of the box is applied to compare with the in situ data.

MERIS Level 2 Reduced Resolution daily CDOM absorption coefficient products from January 2003 to December 2009 were collected and calculated to monthly CDOM data. Three hundred and thirteen satellite images were used in total and an average of four images were used to derive the monthly composite images. The monthly average CDOM absorption coefficient in area C was calculated for temporal variability analysis. MERIS daily chlorophyll a products were also calculated to monthly data and chlorophyll a was averaged in area C. MERIS includes two chlorophyll a products (Algal Pigment Index 1 and Algal Pigment Index 2), which are derived by the blue/green ratio and the neural network, respectively. The Algal Pigment Index 2 is applied in this study since the signal of blue/green ratio is interfered with by the CDOM and suspended sediment in the complex coastal and estuary water. The TRMM and others combined monthly surface precipitation 0.25°× 0.25° data were applied to calculate the monthly rainfall rate in Pearl River watershed.

An accumulated anomaly was used to reveal the abrupt variability trend of CDOM absorption coefficients in area C (Wei, 2007). For a time series , the accumulated anomaly at time t can be expressed as

formula

Continuous wavelet transform studies localize variability of power within a time series. It can determine both the dominant modes of variability and how those modes vary in time by decomposing a time series into time-frequency space. Therefore in this study, the continuous wavelet transform is used to find the CDOM dominant modes and their variability in time. The Morlet wavelet is utilized as the mother function. A Morlet wavelet is defined as

formula

where ω0 is the dimensionless frequency and η is the dimensionless time. The ω0 is set to 6 since it provides a good balance between time and frequency localization (Baliunas et al., 1997; Grinsted et al., 2004; Kumar and Foufoula-Georgiou, 1997; Torrence and Compo, 1997).

Results and discussion

The in situ CDOM absorption coefficient at 440 nm ranged from 0.03 to 1.05 m−1, decreasing from the river estuary toward the coastal ocean. Chlorophyll a concentrations varied from 0.79 to 33.13 mg m−3 with the average level of 6.95 mg m−3. The salinity ranged from 0.1 to 11.23. Figure A1 (available in the Appendix in the online supplementary files) shows the comparison result between MERIS CDOM products and in situ CDOM. The root mean squared error is 0.17 and the mean relative error was less than 40%. It was indicated that the MERIS CDOM product possessed the capability and potential to detect variability of CDOM in the PRE.

Figure 2 shows the monthly CDOM variability from January 2003 to December 2009. The CDOM absorption coefficients at 440 nm varied from approximately 0.1 to around 1.5 m−1. The lowest CDOM absorption coefficient occurred in April 2005 and the highest value occurred in December 2009. The CDOM absorption coefficient had a significant negative correlation with rainfall rate (R = −0.28, p = 0.016), whereas no significant correlation existed between CDOM and chlorophyll a (R = 0.01, p > 0.05). A significant positive correlation was found between chlorophyll a and rainfall rate (R = 0.67, p < 0.01).

Generally, greater CDOM absorption coefficients occur in winter and lower levels of CDOM are present in summer. Some similar variability features occurred in other estuaries and marginal areas. Kahru and Mitchell (2001) have obtained the variability of CDOM absorption in California Current from SeaWiFS data, and the minima of CDOM occurred during late summer. Vecchio et al. (2004) have studied spatial and seasonal distribution of CDOM in the Middle Atlantic Bight, and they found that surface water exhibits lower CDOM absorption in August and September. Morel et al. (2009) demonstrated that the seasonal CDOM cycle in the Northern Hemisphere steadily exhibits a minimum in summer. They found that the lowest level of CDOM in the Mediterranean Sea occurs in August.

There is an anomalous high CDOM level in the summer of 2008. Generally, the chlorophyll a presents an inverse variability with the CDOM. However, in the summer of 2008, the chlorophyll a also presents anomalous high levels. The precipitation shows similar variability with the chlorophyll a.

There are some missing data because of cloudy coverage and gaps in satellite coverage. The missing data occur in the summer of 2005, 2006 and 2007, which may affect the analysis of CDOM variability. Some sudden anomalous variability of CDOM cannot be observed because of the missing data. Even though, the obvious seasonal variabilities of CDOM and chlorophyll a are still identified from time series. In addition, in order to minimize the influences of the missing data on the results, we apply the nearest neighbor interpolation to compute the missing data. The nearest neighbor interpolation selects the nearest two points for a non-given point and set weight according to the distance between the nearest points and the non-given point. The value at non-given point is then calculated using the weight and values of the nearest points. Accumulated anomaly of the time series shows that the variability of CDOM absorption coefficients undergoes apparent fluctuations almost every year during the last 7 years. The majorities of accumulated anomalies are negative values. There is a decreased tendency of the CDOM accumulated anomaly from 2003 to late 2005 and an increased tendency after late 2005, which means that the CDOM levels are lower before the end of 2005 and after that the CDOM levels become higher. Note that increase of CDOM in summer after the end of 2005 has a dominant contribution to the increase of the CDOM accumulated anomaly (Figure 2, upper panel). The accumulated anomalies of chlorophyll a decrease rapidly from 2003 to the beginning of 2006, and then increase with some fluctuations.This means that the majority of chlorophyll a concentration is lower than the average levels before 2006. After 2006, the majority of chlorophyll a concentration is higher than the average levels. The accumulated anomalies of precipitation are relatively stable. The higher positive anomalies in 2008 lead to an apparent increase of the precipitation accumulated anomalies.

The wavelet power spectra of area C are demonstrated in the Figure 3. There is an equally significant peak in the one-year band throughout 2003–2009 in the area C, which indicates that CDOM presents a significant one-year period. The secondary peak is in around 6-month in 2008. High power in the about 3–6 month band during the early 2004 also passes 95% significance test.

The river discharge has been thought to be a driver for the material changes in the coastal water and estuary and it can explain more than 70% of the variability for some dissolved organic material (Seitzinger et al., 2005). Hu et al. (2004) have demonstrated that high coastal runoff is caused by heavy rainfall rate in the central coast of Florida. Ni et al. (2008) thought that the monthly material fluxes from discharge are dependent upon the monthly variation in rainfall rate in the PRE. Kieber et al. (2006) have implied that the anthropogenic and terrestrial sources are important contributors to CDOM levels in precipitation. Since eighty percent of the annual discharge volume occurs during the wet season in the PRE, the satellite derived rainfall rate as an indicator of river discharge could be utilized to analyze the relationship between the river discharge and CDOM due to the absence of river discharge information. Using the rainfall intensity has its limitations as a descriptor of the runoff. For example, urban growth also alters the relationship between rainfall rate and runoff through potential maximum storage (Weng, 2001). However, urban growth has a greater effect on the long trends rather than anomalious variability in certain year.

The cross wavelet transform (XWT) is considered a valid method to examine the relationship in time frequency space between two time series (Grinsted et al., 2004). The results of the relationship between rainfall and CDOM absorption coefficient are shown in the Figure 4. A high one year common oscillation occurs throughout the time series in area C. The XWT phase angle of one year period within the 5% significant regions has the mean phase angle 176 ± 10º for area C. The rainfall and CDOM are therefore in anti-phase, which means that the heavy rainfall coincides with low CDOM. Obviously, there are some factors leading to the decrease of CDOM in wet season. These factors overwhelm the increase of CDOM driven by runoff. During the beginning of 2008, there are significant common powers in the period of less than 6-month for rainfall and CDOM, which we will discuss later.

The bacterial degradation of phytoplankton may be another source of CDOM. Kowalczuk et al. (2006) have raised empirical relationships between CDOM absorption, salinity and chlorophyll a in Baltic Sea, ag(375) = A × Salinity + B × [Chl] + C (0.01 < B < 0.07). Chen et al. (2004a) found that CDOM produced by bacterial degradation of phytoplankton dominates the subsurface waters in Mississippi River plume. The cross wavelet analysis between chlorophyll a and CDOM shows that there is a one year common oscillation since the end of 2004, and the chlorophyll a and CDOM present anti-phase relationship (Figure 4). It is indicated that the bacterial degradation of phytoplankton may not have a dominant impact on the CDOM variability during annual period. The in situ chlorophyll a and CDOM also do not show significant correlation (not shown here). So the phytoplankton-derived CDOM may be a minor cause of the CDOM variability. The covariation between chlorophyll a and CDOM in 2008 may be induced by the occurrence of flood equivalent to a worst in 50 years due to anomalous precipitation which may overwhelms the factors of decreasing CDOM (Fu et al., 2010; Wang et al., 2011). This anomalous high CDOM in the summer of 2008 can also be found in the Figure 2.

Overall, the inverse relationships of rainfall rate and phytoplankton with CDOM for annual period indicate that the increased CDOM driven by the CDOM sources (including rainfall rate and degradation of phytoplankton) is completely consumed during sink process such as photobleaching. In the Chinese estuary, main material discharge is nutrients rather than organic matter (Chen et al., 2004b). Moreover, the phytoplankton dose not result in an increase of organic matter since that the phytoplankton growing is inhibited due to high loads of suspended matter (Chen et al., 2004b).

To further verify that the increased CDOM has been consumed through some CDOM sink processes, we analyze the relationship between CDOM absorption coefficients and salinity in coastal area, which may reflect the potential of CDOM absorption to trace the freshwater input and search other factors influencing CDOM absorption except physical mixing (Ferrari and Dowell, 1998). If there is a conservative mixing of CDOM, the physical mixing has a dominate influence on the CDOM variability. For others (non-conservative mixing), the estuary acts as a reaction vessel in which biological and chemical removal takes place, substantially reducing the amount of constitute that reaches the ocean (Loder and Reichard, 1981). Kowalczuk et al. (2003) found a strong inverse relationship between salinity and CDOM absorption and this relationship depended on river flow in South Atlantic Bight. D’Sa et al. (2002) derived salinity field from CDOM. The salinity field showed feature of coastal runoff in the Florida Bay and West Florida Shelf. Chen et al. (2003a) found the CDOM decreases with increasing salinity following an apparent non-linear mixing line in the PRE. We also analyze the relationships between salinity and CDOM absorption coefficients in the dry season and wet season and we find that the CDOM decreases and the CDOM spectral slope increases with the increase of salinity. There is non-conservative mixing at both seasons (Figure 5). The non-conservative mixing of CDOM absorption coefficient in the surface waters gives a signal that some other effects interfere with the conservative CDOM-salinity relationship. Most CDOM absorption coefficients in the wet season are below the conservative mixing. We suggest that some CDOM sinks may have an impact before CDOM mixes with ocean water.

Many research findings indicated the loss of CDOM is mainly due to the photobleaching in many coastal areas and estuaries (D’Sa and DiMarco, 2009). Along with the photobleaching of CDOM, there is an increase in spectral absorption slope and 96% of CDOM was destroyed by solar radiation (Coble, 2007). Intense photobleaching process favored by the high level of sunshine plausibly results in abrupt decrease of CDOM in spring and minimum in summer in the Mediterranean Sea (Morel and Gentili, 2009). Vähätalo et al. (2004) have also found that 96% of freshwater CDOM is decomposed by solar radiation, but only 41% of DOC during 70-day exposure. Callahan et al. (2004) thought that the CDOM/DOC ratio decreases with increase of salinity could be attributed to the photo transformation of CDOM into DOC. We use noontime downward irradiance and CDOM absorption coefficient of three cruises (PRE20091122, PRE20091213 and PRE20100319) to analyze the photobleaching influence on CDOM. The downward irradiance represents radiative flux per unit area incident on a surface with the unit of W m−2 nm−1. The downward irradiance could be a useful indicative of the light energy reaching the surface. The integral spectrum of noontime downward irradiance just above water from 350 nm to 420 nm is calculated and compared with the CDOM absorption coefficient at 440 nm (Figure 6). The results present negative correlations between the noontime downward irradiance just above water and CDOM absorption coefficient (R = −0.6559). The photobleaching may be an important sink of CDOM, which is also supported by the increasing spectral absorption slope (S) with the increase of salinity in Figure 5.

Conclusions

MERIS CDOM product in the PRE is compared with in situ CDOM data. The result indicates that it possesses capability to analyze the temporal variability of CDOM in the PRE. The time series of CDOM in the specified area present higher CDOM levels in winter and lower CDOM levels in summer, but there is an anomalous high CDOM in the summer of 2008. The accumulated anomaly is used to obtain the abrupt variability of CDOM. The result indicates that CDOM accumulated anomaly has a decrease trend before late of 2005 and after then positive CDOM anomalies lead to an increase trend of CDOM accumulated anomaly. In order to determine all dominant modes of variability and how those modes vary in time, we introduce wavelet analysis. The result indicates that there are significant annual periods over the entire time series, a significant 3-month period from the beginning of 2004 to the middle of 2004 and a significant 6-month period in 2008.

The influences of rainfall, phytoplankton, salinity and light intensity on the variability of CDOM are analyzed. Photobleaching may contribute to the lower CDOM in summer. However, when the anomalous strong precipitation occurs, there is significant high CDOM for the period 3-6 months. An example is that anomalous rainfall in 2008 overwhelms the influence of photobleaching and results in anomalous high CDOM levels.

Acknowledgements

We thank the European Space Agency for providing the MERIS data.

Supplemental material

Supplemental data for this article can be accessed on the publisher’s website.

Funding

The research work was supported by the project from Guangzhou City for the Pearl River New Star on Science and Technology (No. 2011J2200022), the National 863 Program (No. 2009AA12Z135) and the National Natural Science Foundation of China (40976106).

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

References

Bailey, S. W., Werdell, P. J.,
2006
.
A multi-sensor approach for the on-orbit validation of ocean color satellite data products
.
Remote Sens. Environ.
102
,
12
23
.
Baliunas, S., Frick, P., Sokoloff, D., Soon, W.,
1997
.
Time scales and the trends in the Central England Temperature data (1659–1990): A wavelet analysis
.
Geophys. Res. Lett.
24
,
1351
1354
.
Bricaud, A., Morel, A.,
1999
.
MERIS potential for ocean color studies in the open ocean. Int. J
.
Remote Sensing
20
(
9
)
1757
1769
.
Callahan, J., Dai, M., Chen, R. F., Li, X., Lu, Z., Huang, W.,
2004
.
Distribution of dissolved organic matter in the Pearl River Estuary, China
.
Mar. Chem.
89
,
211
224
.
Chen, C., Shi, P., Yin, K., Pan, Z., Zhan, H., Hu, C.,
2003a
.
Absorption coefficient of yellow substance in the Pearl River estuary
.
Proceedings of SPIE
4892
,
215
221
.
Chen, C., Shi, P., Zhan, H.,
2003b
.
A local algorithm for estimation of yellow substance (gelbstoff) in coastal waters from SeaWiFS data: Pearl River estuary, China
.
Int. J. Remote Sens.
24
,
1171
–1176.
Chen, R. F., Bissett, P., Coble, P., Conmy, R., Gardner, G. B., Moran, M. A., Wang, X., Wells, M. L., Whelan, P., Zepp, R. G.,
2004a
.
Chromophoric dissolved organic matter (CDOM) source characterization in the Louisiana Bight
.
Mar. Chem.
89
,
257
272
.
Chen, Z., Li, Y., Pan, J.,
2004b
.
Distributions of colored dissolved organic matter and dissolved organic carbon in the Pearl River Estuary, China
.
Cont. Shelf Res.
24
,
1845
1856
.
Coble, P. G.,
2007
.
Marine optical biogeochemistry: the chemistry of ocean color
.
Chem. Rev.
107
,
402
418
.
D’sa, E., Steward, R., Vodacek, A., Blough, N., Phinney, D.,
1999
.
Determining optical absorption of colored dissolved organic matter in seawater with a liquid capillary waveguide
.
Limnol. Oceanogr.
44
,
1142
1148
.
D’sa, E. J., Dimarco, S. F.,
2009
.
Seasonal variability and controls on chromophoric dissolved organic matter in a large river-diminated coastal margin
.
Limnol. Oceanogr.
54
,
2233
2242
.
D’sa, E. J., Hu, C., Muller-Karger, F. E., Carder, K. L.,
2002
.
Estimation of colored dissolved organic matter and salinity fields in case 2 waters using SeaWiFS: Examples from Florida Bay and Florida Shelf
.
J. Earth Syst. Sci.
111
,
197
207
.
Ferrari, G. M., Dowell, M. D.,
1998
.
CDOM absorption characteristics with relation to fluorescence and salinity in coastal areas of the Southern Baltic Sea
.
Estuar., Coast. Shelf S.
47
,
91
105
.
Fu, S., Zhao, S., Sun, J., Li, W.,
2010
.
One kind of vortex causing heavy Rainfall during pre-rainy season in South China
.
Chin. J. Atmos. Sci.
34
,
236
252
.
(In Chinese)
Grinsted, A., Moore, J. C., Jevrejeva, S.,
2004
.
Application of the cross wavelet transform and wavelet coherence to geophysical time series
.
Nonlinear Proc. Geoph.
11
,
561
566
.
Hong, H., Wu, J., Shang, S., Hu, C.,
2005
.
Absorption and fluorescence of chromophoric dissolved organic matter in the Pearl River Estuary
,
South China. Mar. Chem.
97
,
78
89
.
Hu, C., Muller-Karger, F. E., Vargo, G. A., Neely, M. B., Johns, E.,
2004
.
Linkages between coastal runoff and the Florida Keys ecosystem: A study of a dark plume event
.
Geophys. Res. Lett.
31
,
L15307.15301-L15307.15304
.
Kahru, M., Mitchell, B. G.,
2001
.
Seasonal and nonseasonal variability of satellite-derived chlorophyll and colored dissolved organic matter concentration in the California Current
.
J. Geophys. Res.
106
,
2517
2529
.
Keith, D. J., Yoder, J. A., Freeman, S. A.,
2002
.
Spatial and temporal distribution of coloured dissolved organic matter (CDOM) in Narragansett Bay, Rhode Island: Implications for phytoplankton in coastal waters
.
Estuar.Coast. Shelf S.
55
,
705
717
.
Kieber, R. J., Whitehead, R. F., Reid, S. N., Willey, J. D., Seaton, P. J.,
2006
.
Chromophoric dissolved organic matter (CDOM) in rainwater, southeastern North Carolina, USA
.
J. Atmos. Chem.
54
,
21
41
Kowalczuk, P., Cooper, W. J., Whitehead, R. F., Durako, M. J., Sheldon, W.,
2003
.
Characterization of CDOM in an organic-rich river and surrounding coastal ocean in the South Atlantic Bight
.
Aquat. Sci.
65
,
384
401
.
Kowalczuk, P., Stedmon, C. A., Markager, S.,
2006
.
Modeling absorption by CDOM in the Baltic Sea from season, salinity and chlorophyll
.
Mar. Chem.
101
,
1
11
Kowalczuk, P., Ston-Egiert, J., Cooper, W. J., Whitehead, R. F., Durako, M. J.,
2005
.
Characterization of chromophoric dissolved organic matter (CDOM) in the Baltic Sea by excitation emission matrix fluorescence spectroscopy
.
Mar. Chem.
96
,
273
292
.
Kumar, P., Foufoula-Georgiou, E.,
1997
.
Wavelet analysis for geophysical applications
.
Rev.Geophys.
35
,
385
412
.
Loder, T., Reichard, R.,
1981
.
The dynamics of conservative mixing in estuaries
.
Estuar.
4
,
64
69
.
Mobley, C. D.,
1999
.
Estimation of the remote-sensing reflectance from above-surface measurements
.
Appl. Opt.
38
,
7442
7455
Morel, A., Gentili, B.,
2009
.
The dissolved yellow substance and the shades of blue in the Mediterranean Sea
.
Biogeosciences
6
,
2625
2636
.
Ni, H., Lu, F., Luo, X., Tian, H., Zeng, E.,
2008
.
Riverine inputs of total organic carbon and suspended particulate matter from the Pearl River Delta to the coastal ocean off South China
.
Mar. Pollut. Bull.
56
,
1150
1157
.
Retamal, L., Vincent W. F., Martineau, C., Osburn C.,
2007
.
Comparison of the optical properties of dissolved organic matter in two river-influenced coastal regions of the Canadian Arctic
.
Estuar., Coast. Shelf S.
72
,
261
272
.
Seitzinger, S. P., Harrison, J. A., Dumont, E., Beusen, A. H. W., Bouwman, A. F.,
2005
.
Sources and delivery of carbon, nitrogen, and phosphorus to the coastal zone: An overview of Global Nutrient Export from Watersheds (NEWS) models and their application
.
Global Biogeochem.Cy.
19
,
GB4S01
.
Torrence, C., Compo, G. P.,
1997
.
A Practical Guide to Wavelet Analysis
.
Bulletin of the American Meteorological Society
79
,
61
78
.
Vähätalo, A. V., Wetzel, R. G.,
2004
.
Photochemical and microbial decomposition of chromophoric dissolved organic matter during long (months-years) exposures
.
Mar. Chem.
89
,
313
326
.
Vecchio, R. D., Blough, N. V.,
2004
.
Spatial and seasonal distribution of chromophoric dissolved organic matter and dissolved organic carbon in the Middle Atlantic Bight
.
Mar. Chem.
89
,
169
187
.
Wang, D., Xia, R., Liu, Y.,
2011
.
A preliminary study of the flood-causing rainstorm during the first rainy season in South China in 2008
.
Acta Meteorologica Sinica
69
,
137
148
.
(In Chinese)
Wei, F.,
2007
.
Statistic Analysis and Forecasting of Climate
.
Beijing
:
China Meteorological Press
.
(In Chinese)
Weng, Q.,
2001
.
Modeling Urban Growth Effects on Surface Runoff with the Integration of Remote Sensing and GIS
.
Environ. Manage.
28
,
737
748
.
Wong, L., Chen, J., Dong, L.,
2004
.
A model of the plume front of the Pearl River Estuary, China and adjacent coastal waters in the winter dry season
.
Cont. Shelf Res.
24
,
1779
1795
.
Yin, K., Lin, Z., Ke, Z.,
2004
.
Temporal and Spatial Distribution of Dissolved Oxygen in the Pearl River Estuary and Adjacent Coastal Waters
.
Cont. Shelf Res.
24
,
1935
1948
.

Supplementary data