Deconvolving Inherent Optical Properties and Estimating Suspended Particulate Mass and Particulate Organic Carbon in Coastal Waters During an Upwelling Event

By Joe Grzymski & Jay Cullen
Presented at ASLO-99, on February 5, 1999, in Santa Fe, NM
Deconvolving the various optical signals that originate in case II waters and contribute to remote sensing reflectance is difficult. However, before the constituents of the reflectance data are understood we first must be able to sample and deconvolve in situ, absorption and attenuation. With this in mind one of our goals has been and will continue to be to investigate the relationship between the inherent optical properties (IOPs) and particulate organic carbon (POC) in the complex waters surrounding LEO-15. More specifically by breaking up measurements of the IOPs into the particulate and dissolved phases we can better understand the contributions of phytoplankton, dissolved organic matter (DOM) and other inorganic particles to the overall in situ optical signatures of complex case II waters. Eventually this will lead to better remotely sensing algorithms for inshore waters.

Here we summarize our results:

1) in situ measurements of beam-c (Transmissometer, c660nm) and c650 nm, c676 nm (AC-9) can effectively monitor particle concentration in coastal surface waters during episodic upwelling events on the inner shelf

2) In situ c660, c650, and c676 provide robust estimates of POC in optically complex coastal waters

3) The optical signature of surface waters off the southern New Jersey coast was dominated by organic particles during 07/98 upwelling events (largest estimated inorganic component was <8% by mass)

4) Remote measurements of beam-c can be used to construct high resolution temporal and spatial maps of particle and POC distributions in the upwelling

5) CDOM spectra showed variability in space, time, and depth, compromising the utility of a fixed s-value in modeling CDOM absorption

6) Upwelling leads to enhanced particulate optical loads within the upwelling eddy

7) During upwelling the relative importance of particulate to CDOM absorption is increased by 10-25 %

8) Particulate spectral signatures had blue:red ratios consistent with the phytoplankton absorption

9) Particulate optical loads were correlated with particulate organic carbon

Figure 1. Water column average absorption and attenuation at 440nm for 12 and 22 July 1998. 12 July represents a day before the onset of upwelling while 22 July 1998 is five days after upwelling began.



Figure 2.Depth profiles of A440 and C440 for 12 and 22 July 1998 for a nearshore and an offshore station. Symbols as in figure 1.



Figure 3.Depth profiles of the total, dissolved and particulate absorption at 440nm for a nearshore and an offshore station on 12 July 1998. The particulate and dissolved absorption is expressed as the percent of the total absorption. Dissolved absorption was measured by placing a 0.2 um filter inline with the AC-9. Green symbols are for particulate absorption, blue symbols are for dissolved and the red symbols are the total absorption in units of m-1.



Figure 4.Depth profiles of the total, dissolved and particulate absorption at 440nm for a nearshore and an offshore station on 22 July 1998. The particulate and dissolved absorption is expressed as the percent of the total absorption. Dissolved absorption was measured by placing a 0.2 um filter inline with the AC-9. Symbols as in figure 3.



Figure 5.Regression of beam-c (c660 nm) vs SPM for surface waters around LEO-15 sampled during upwelling events of July 1998 (n=15). SPM was corrected for the mass contribution of inorganic clays and Fe(OH)3 (correction always < 8% by mass). Regression equations relating beam-c to SPM for Sargasso Sea and north-east Atlantic slope waters are shown for comparison (Bishop 1986).



Figure 6.Regression of attenuation (c650 nm and c676 nm) measured with an AC-9 vs corrected SPM for surface waters during July 1998 upwelling events off southern New Jersey.



Figure 7. Regression of beam-c (c660 nm) vs POC (>0.8 µm) for surface waters around Leo-15 during July 1998 (n=15). Regression equations for equatorial Pacific and warm core rings in the north-east Atlantic shown for comparison (Bishop 1998).



Figure 8.Regression of attenuation (c650 nm and c676 nm) measured with and AC-9 vs POC (>0.8 µm) for surface waters during July 1998 upwelling events off southern New Jersey (n=34).



Figure 9.Regression of attenuation with the contribution due to DOM removed. Pay particular attention to the fact that while the slope of the line does not change significantly the offset changes from positive to negative; this change represents the overestimation of POC due to the DOM attenuation sigmal.
Questions???

oscar@ahab.rutgers.edu