Light-absorbing components in the Great Lakes
Introduction
Absorption is a light attenuating process that is a fundamental determinant of the character of the underwater and emergent light fields. The magnitude of this process is quantified by the absorption coefficient (a, m− 1; listing of symbols and abbreviations in Table 1); a is the fraction of the radiant flux absorbed divided by the thickness of the layer (Kirk, 1994). As an inherent optical property (IOP), a is an intrinsic property of the water medium that is independent of the geometry of the light field (i.e., time of day). The magnitude of a varies as a function of the concentrations of light absorbing optically active constituents (OACs), that all have strong wavelength (λ) dependencies (Babin et al., 2003, Binding et al., 2008, Effler et al., 2010, Kirk, 1994). The magnitude and spectral character of a are an important determinant of optical properties that depend on radiance distribution, the so called apparent optical properties (AOPs; those that depend on the geometry of the light field), such as the diffuse attenuation coefficient for downwelling irradiance and remote sensing reflectance (Rrs, sr− 1).
Commonly, a(λ) is partitioned into four additive components (ax where x can signify various components of absorption; Babin et al., 2003, Binding et al., 2008, Effler et al., 2010, Perkins et al., 2010)which include contributions from pure water (w), phytoplankton (φ), non-algal particles (NAP), and colored dissolved organic matter (CDOM). The wavelength dependencies of these constituents differ substantially. Pure water absorption [aw(λ)] increases in the ultraviolet and red wavelengths and is well defined (Pope and Fry, 1997). Phytoplankton demonstrates a broad absorption peak at ~ 440 nm and a well-defined maximum at 676 nm (Babin et al., 2003, Bricaud et al., 1995). Both aNAP and aCDOM decrease exponentially with increasing wavelengths over the visible range (Babin et al., 2003, Bricaud et al., 1981, Perkins et al., 2009). Standardized laboratory protocols for aφ, aNAP and aCDOM have been established (Mitchell et al., 2003), primarily with the goal of supporting remote sensing initiatives for marine systems. Quantification of ax according to these protocols remains relatively rare for lacustrine systems (Perkins et al., 2009, Perkins et al., 2010), but particularly for the Laurentian Great Lakes (Binding et al., 2008, Effler et al., 2010) where remote sensing of OACs is actively being pursued (Lesht et al., 2012). NAP has received the least attention of these components (Babin et al., 2003), and includes both minerogenic (Babin and Stramski, 2004, Bowers et al., 1996) and detrital (i.e., organic; Effler et al., 2010) particles.
The remote sensing signal, Rrs(λ), is regulated by a(λ), as well as other IOPs, the backscattering coefficient [bb(λ)], as described by the following semi-analytical algorithm (Morel and Gentili, 1996)where f and Q are the parameters that depend on ambient conditions and λ, though the f/Q ratio is subject to only modest variations (Morel and Gentili, 1996) over conditions of interest and will not be further dealt with here. Clearly, the magnitudes of both a(λ) and bb(λ) are important in regulating the magnitude of Rrs(λ). The Rrs(λ) spectra are rich in their structure, as illustrated for a site in Lake Ontario with predictions based on measured a and bb spectra [Eq. (2)] that matched observations well (scenario 1, Fig. 1a). Such structure presents the opportunity for resolution of OACs in these signals (Lubac and Loisel, 2007). The spectral structure is driven primarily by a(λ), as bb(λ) is much less spectrally variant (Figs. 1b and c; O'Donnell et al., 2010, Snyder et al., 2008). This is illustrated here by comparison of predictions of Rrs(λ) (Fig. 1a) for hypothetical scenarios of spectrally invariant (at the average) bb(λ) and a(λ) (scenarios 2 and 3 respectively, Fig. 1a). The spectrally uniform bb(λ) case (scenario 2) does not deviate greatly from the successful simulations of scenario 1, but making a(λ) invariant (scenario 3) eliminates most of the structure in Rrs(λ) (Fig. 1a). This demonstrates that resolution and partitioning of a(λ) are fundamental to pursue remote sensing of OACs. The characterization task for ax(λ) is less critical for case 1 waters (open oceans) where NAP and CDOM covary with, and are subordinate to, phytoplankton (Morel and Prieur, 1977). However, lacustrine systems, including the Laurentian Great Lakes (Binding et al., 2008, Effler et al., 2010, O'Donnell et al., 2010), are case 2 waters, where NAP and CDOM do not necessarily covary with phytoplankton and may not be subordinate (Morel and Prieur, 1977). Resolution of ax(λ) and their dependence on OACs are important to support mechanistic remote sensing initiatives, including those that use mechanistic algorithms (Mouw et al., 2013).
This paper documents magnitudes and spectral features of ax in the Laurentian Great Lakes, and aCDOM in selected tributaries. The open waters of each lake during late summer are characterized and contrasted, as well as selected embayments and near-shore waters. Temporal variation is described for limited cases. Dependencies of ax on OACs (cross-sections) are evaluated; these and the basic characterizations are considered in the context of findings for other cases 2 systems and remote sensing initiatives. The credibility of a subset of ax measurements is tested by comparing the values of a determined as the summation of the components to paired in situ observations.
Section snippets
Study sites
We have conducted a comparative analysis of ax for the Great Lakes, which includes the open waters of each of the five lakes, and selected near-shore areas (within 10 km of shore), embayments, and river mouths (Fig. 2; Table 2, Table 3). The spatial coverage for the entire system in this paper is uneven, and far from complete. Sampling of the open waters (> 10 km offshore) was conducted over the 2006–2008 interval that supported resolution of all of the components of [Eq. (1)] a. Surveys of
Spectral characteristics
Generally representative spectra for the three measured components of a (ax) are presented for a Green Bay site that depicts recurring characteristic spectral features (Fig. 3). Exponential decreases in aCDOM (Fig. 3a) and aNAP (Fig. 3b) with increasing wavelengths were observed for all samples. The corresponding expressions [Eqs. (3), (4)] and non-linear regression fitting performed well in representing these spectral patterns. The phytoplankton spectra all demonstrated two maxima; a broader
CDOM
It is widely acknowledged that CDOM has both autochthonous and allochthonous sources in marine (Bricaud et al., 1981) and lacustrine (Davies-Colley and Vant, 1987) systems. Autochthonous inputs dominate in case 1-waters, as manifested in the covariation of this absorbing component with phytoplankton biomass and production (Bricaud et al., 1981). However, CDOM levels in lacustrine waters are regulated primarily by terrigenous inputs (Davies-Colley and Vant, 1987, Kirk, 1994). The importance of
Synthesis, in a remote sensing context
This paper has presented the first system-wide characterization of the magnitudes and spectral features of ax in the Laurentian Great Lakes, to support mechanistic approaches in remote sensing of OACs. Characterizations of the open waters of all of the lakes in late summer have been presented, along with representations of selected rivers (aCDOM only), embayments and near-shore areas. Resolution of dynamics was limited by comparison.
A robust range of conditions with respect to magnitudes and
Acknowledgements
This research was supported by NASA under award no. NNX09AV54G. The authors wish to thank the captains and crew of the EPA's R/V Lake Guardian, the Wisconsin Department of Natural Resources R/V Gaylord Nelson, and Michigan Technological University's R/V Agassiz for their help and support. In addition, the authors thank Dr. ZhongPing Lee (University of Massachusetts, Boston), George Leshkevich (NOAA), Dr. Alan Weidemann (Naval Research Laboratory), Dr. Martin T. Auer (Michigan Technological
References (52)
- et al.
Spectral absorption properties of dissolved and particulate matter in Lake Erie
Remote Sens. Environ.
(2008) - et al.
An analysis of MODIS-derived algal and mineral turbidity in Lake Erie
J. Great Lakes Res.
(2012) - et al.
Estimation of organic and inorganic matter in inland waters: optical cross sections of Lakes Ontario and Ladoga
J. Great Lakes Res.
(1991) - et al.
Great Lakes chloride trends: long-term mass balance and loading analysis
J. Great Lakes Res.
(2009) - et al.
Light-absorbing components in Lake Superior
J. Great Lakes Res.
(2010) - et al.
Backscattering of light and its components in the Great Lakes
J. Great Lakes Res.
(2013) - et al.
Observations of sediment transport in Lake Erie during the winter of 2004–2005
J. Great Lakes Res.
(2007) - et al.
Satellite ocean color algorithms: a review of applications to the Great Lakes
J. Great Lakes Res.
(2012) - et al.
Evaluation and regional optimization of bio-optical algorithms for central Lake Superior
J. Great Lakes Res.
(2004) - et al.
Variability and classification of remote sensing reflectance spectra in the eastern English Channel and southern North Sea
Remote Sens. Environ.
(2007)
Dissolved organic matter characteristics within the Lake Superior watershed
Org. Geochem.
Optical characterizations and pursuit of optical closure for the western basin of Lake Erie through in situ measurements
J. Great Lakes Res.
Resolution of optical gradients and pursuit of optical closure for Green Bay, Lake Michigan
J. Great Lakes Res.
Characterizations of individual suspended mineral particles in western Lake Erie: implications for light scattering and water clarity
J. Great Lakes Res.
Characterizations of the light-scattering attributes of mineral particles in Lake Ontario and the effects of whiting
J. Great Lakes Res.
Modeling the spectral shape of absorption by chromophoric dissolved organic matter
Mar. Chem.
Variations in the mass-specific absorption coefficient of mineral particles suspended in water
Limnol. Oceanogr.
Variations in the light absorption coefficients of phytoplankton, nonalgal particles, and dissolved organic matter in coastal water around Europe
J. Geophys. Res.
Towards a new generation of sediment traps and a better measurement/understanding of settling particulate flux in lakes and oceans: a hydrodynamic protocol
Aquat. Sci.
Absorption spectra of inorganic particles in the Irish sea and their relevance to remote sensing of chlorophyll
Int. J. Remote Sens.
Absorption by dissolved organic matter of the sea (yellow substance) in the UV and visible domains
Limnol. Oceanogr.
Variability in the chlorophyll-specific absorption coefficients of natural phytoplankton: analysis and parameterization
J. Geophys. Res.
Variations of light absorption by suspended particles with the chlorophyll a concentration in oceanic (case 1) waters: analysis and implications for bio-optical models
J. Geophys. Res.
Surface Water-quality Modeling
Standard Methods for the Examination of Water and Wastewater
Quantifying absorption by aquatic particles: a multiple scattering correction for glass-fibre filters
Limnol. Oceanogr.
Cited by (15)
Spatial and temporal variability of inherent and apparent optical properties in western Lake Erie: Implications for water quality remote sensing
2019, Journal of Great Lakes ResearchWater clarity of the Upper Great Lakes: Tracking changes between 1998–2012
2017, Journal of Great Lakes ResearchCharacterization of bulk and chromophoric dissolved organic matter in the Laurentian Great Lakes during summer 2013
2016, Journal of Great Lakes ResearchRemote sensing of physical cycles in Lake Superior using a spatio-temporal analysis of optical water typologies
2015, Remote Sensing of EnvironmentLight-absorbing properties of mineral particles in the Great Lakes
2015, Journal of Great Lakes Research