Insights from long-term trends in Secchi depth and light attenuation by rih47632

VIEWS: 27 PAGES: 23

									 Insights from long-term trends in
Secchi depth and light attenuation
in shallow systems of Chesapeake
                Bay

                Chuck Gallegos
  Smithsonian Environmental Research Center
               CERF'09, SCI-057
        Portland, OR November 4, 2009
          Edited by Peter Bergstrom
      For SAV Workgroup, 2-1-10 (Italics)
         Two methods for Long Term
          Monitoring of Water Clarity
                                                         • Attainment of water
                                                           clarity sufficient to
                                                           restore 125,000 acres
                                                           of submerged aquatic
                                                           vegetation (SAV) is a
                                                           key element of the
                                                           Chesapeake 2000
                                                           Agreement.
                                                         • In Chesapeake Bay,
                                                           light attenuation
                                                           coefficient (Kd) is
                                                           measured in
                                                           mainstem, Secchi
                                                           depth in tributaries
Problem: Secchi depth responds more to scattering,
Kd responds more to absorption, so the 2 measures
are correlated, but the relationship can CHANGE in
time at the same place, or over space at the same time
 The Kd-Secchi depth Relationship
Can we use Secchi depth to estimate attenuation
                coefficient?
• SAV Technical Synthesis-2 group used Kd=1.45/ZSD for
  Chesapeake Bay, ca. 1992-2000
• Original report (Poole & Atkins 1929) estimated
  Kd=1.7/ZSD
• Kd·ZSD known to be higher in humic, and lower in turbid
  lakes (Koenings & Edmondson 1991)


 Q: What determines the product Kd·ZSD and
 what does it tell us about suspended solids?
                                  Observations - 1
The Kd·ZSD product is declining bay-wide in mainstem
                         Main Stem Chesapeake Bay



           2.6
                        NB: n=1 in1987
           2.4                                                  CB2OH
           2.2
                                                                CB3MH
                                                                CB4MH
           2.0                                                  CB7PH

           1.8
  Kd·ZSD




           1.6

           1.4

           1.2

           1.0

           0.8

           0.6
                 1985      1990     1995        2000         2005      2010

                                         Year          Error bars =+2 s.e.
            Observations - 2
The Kd·ZSD product did NOT change in Magothy creeks




                                     Data collected
                                     by Peter Bergstrom,
                                     MRA (all in Magothy
                                     creeks, ends 2006,
                                     BEFORE Secchi
                                     really declined—next
                                     slide)
                                       Observations - 3
Secchi depth (ZSD) is falling in the Magothy (but only in mainstem?)

                                   Magothy Aquatic Health Components, 2002‐2009
                                                             ***DRAFT 1‐15‐10***
                         100%
                                                                                      DO (water column)
                                                                                      Clarity
                         80%
                                                                                      SAV
  % attainment of goal




                         60%



                         40%                                                                Mahogany
                                                                                            tide‐‐2008

                                                Mussels‐‐2004
                         20%

                                       Wet‐‐2003           Hot & calm‐‐2005

                          0%
                                2002     2003       2004        2005      2006     2007      2008        2009
               DO & clarity data from creeks by Peter Bergstrom & mainstem by Dick Carey; SAV from VIMS
Observations – 4
 (Chris Jones, GMU, CERF 2009)




                         (Show same trends over time)
                              Observations – 5
                               (Kd*Sd, Chris Jones, GMU, 1-31-10)
            Gunston Cove Station 7: midcove


       0                                              Gunston Cove Station 9: Potomac Mainstem Channel


                                                                 0




       -1
                                                                 -1
KDSD




                                                          KDSD
       -2                                                        -2




                                                                 -3
                                                                 1,980   1,990          2,000   2,010
       -3                                                                        YEAR
       1,980         1,990          2,000     2,010
                             YEAR
  (Sidebar for SAV Workgroup)
• If Kd and Secchi trends are not tightly
  linked, which is more important to SAV
  survival?
                           Magothy, 2002-2009
• Magothy data:                     80%


Secchi depth?                       60%    y = 0.7957x + 0.0059
                                                R2 = 0.5805
                       SAV status

                                    40%



                                    20%
                                                                        2008


                                    0%
                                          0%       20%            40%          60%   80%
                                                      Clarity (Secchi) status
                                          Theory
Equation for Secchi depth, zSD*

                      Γ                 ΓK d
      Z SD       =        ⇒ K d Z SD =
                   c + Kd              c + Kd
      Γ = "Coupling constant", depends on measurement conditions
         (e.g. sea state, sun glint, reflectivity of disk, distance above
         water, etc.)
      c = Beam attenuation coefficient, sum of absorption +
          scattering, i.e.,
                                                   ⎛ b⎞
                                     c = a + b = a ⎜1 + ⎟
                                                   ⎝ a⎠
  Preisendorfer, R. W. 1986. Secchi disk science: visual optics of natural waters. Limnology and Oceanography
  *

  31:909-926.
                                  Theory
Equation for diffuse attenuation coefficient*
                                                                                  1

                   [a                      ]             a ⎡              b⎤          2
                            + G (μ 0 )ab                    ⎢1 + G (μ 0 ) a ⎥
               1                               1
   Kd =                 2                          2
                                                       =
              μ0                                         μ0 ⎣               ⎦

                   µ0        = cosine of refracted solar angle of incidence
                   G(µ0) = function that scales scattering/absorption
                         interaction



    *Kirk,
         J. T. O. 1984. Dependence of relationship between apparent and inherent optical
    properties of water on solar altitude. Limnology and Oceanography 29:350-356.
                                      Theory
Kd·ZSD Product*
                                                                             1
                                             ⎛            b⎞                     2
                                            Γ⎜1 + G (μ 0 ) ⎟
       ∴ K d Z SD =                          ⎝            a⎠
                                                              1
                                 ⎛            b⎞                  2
                                                                            ⎛ b⎞
                                 ⎜1 + G (μ 0 ) ⎟                      + μ 0 ⎜1 + ⎟
                                 ⎝            a⎠                            ⎝ a⎠

Bottom line: Kd·ZSD is a function of the scattering-to-
absorption ratio (b/a), and a few other parameters.
   *
   Effler, S. W. 1985. Attenuation versus transparency. Journal of the Environmental Engineering Division,
   ASCE 111:448-459.
   Davies-Colley, R. J., and W. N. Vant. 1988. Estimation of optical properties of water from Secchi disk depths.
   Water Resources Bulletin 24:1329-1335.
        Inherent Optical Properties of
                  Particles
 TSS = [Phytoplankton dry wt.] + [POM] + Clays + Fine Silt
                             References for Particulate IOPs
(1) Bowers, D. G., and C. E. Binding. 2006. The optical properties of mineral suspended
   particles: A review and synthesis. Estuarine, Coastal and Shelf Science 67:219-230.
(2) Mobley, C. D. and L. K. Sundman. 2008. Hydrolight 5, Ecolight 5 Technical
   Documentation. Sequoia Scientific. Bellevue, Washington.
(3) Stramski, D., A. Bricaud, and A. Morel. 2001. Modeling the inherent optical
   properties of the ocean based on the detailed composition of the planktonic
   community. Applied Optics 40:2929-2945,
(4) Babin, M., A. Morel, V. Fournier-Sicre, F. Fell, and D. Stramski. 2003. Light
   scattering properties of marine particles in coastal and open ocean waters as related to
   the particle mass concentration. Limnology and Oceanography 48:843-859.
(5) Gallegos, C. L., unpublished
   Inherent Optical Properties of
             Particles
  Mass-specific Coefficients at Reference Wavelengths
 Particle Type       Mass-         Mass-specific   Backscattering
                    specific        absorption      -to-scattering
                   scattering                            ratio
Phytoplankton &    2.25Chl-0.38        0.04        0.0096Chl-0.253
 assoc. detritus       (2)              (5)             (2)
"Small Detritus"        1.4           0.106             0.004
                    (strong, 4)    (moderate, 2)      (weak, 3)
     Clays             0.35            0.05              0.03
                   (moderate, 4)     (weak, 1)     (very strong, 2)
    Fine Silt         0.15             0.025            0.01
                    (weak, 4)        (weak, 1)      (moderate, 2)
               Particulate IOP Model
  b p (555) = 1.4[Small detritus] + 0.3Chl 0.62 + 0.35[Clays] + 0.15[Fine silt ]


adet (440 ) = 0.106[Small detritus] + 0.04(0.133Chl ) + 0.05[Clays] + 0.025[Silt ]

 bbp        0.004[Small detritus] + 0.0013Chl 0.747 + 0.03[Clays] + 0.01[Silt ]
    (555) =
  b                                      TSS
TSS = [mg L-1 Small detritus] + 0.133Chl + [mg l -1 Clays] + [mg L-1 Silt]

               + Spectral Shape Functions


                     Monte Carlo Simulation of Equations for
                     Secchi depth and Kd
          Monte Carlo Simulation
          Inputs from Random Uniform Distributions
                     Input           Range        Units

                "Small detritus"      0-10        mg L-1

                      Chl             0-30        µg L-1

                     Clays            0-10        mg L-1

                      Silt            0-2         mg L-1

                    CDOM              0-1          m-1


•Simulate 20,000 realizations
•Output: Kd(PAR), ZSD, Kd·ZSD, (+ various remote sensing parameters)
•Sort output by Kd·ZSD descending—a surrogate for time
•Select data for ranges observed at a location of interest
•Q: What structure does this sorting and selecting impose on the inputs?
Selection Example—CB3.3MH
   Parameter    Range     Remaining
               (~annual    (20,000
                mean)       initial)
      Chl        5-25      13,297
     TSS        4-12        5,321
      ZSD       0.7-2       3,850
   Kd(PAR)     0.9-1.9      3,333
    Kd·ZSD      0.9-2       2,919
                                              Compare with Observations
                                                     CB3.3MH
               2.0
                                                                       Simulated       • Sort simulations on Kd·ZSD descending
               1.8                                                     Observed
                                                                       Regression
               1.6                                                                     • Use regression on observed data to
Kd·ZSD




               1.4
                                                                                         establish time axis for simulations
               1.2

               1.0                                                                     • Compare with trends (or lack thereof)
                         1985         1990     1995     2000    2005        2010         in other observations
                       2.2

                       2.0

                       1.8
    Secchi depth (m)




                       1.6

                       1.4
                                                                                    Secchi depth: Sorting and range selection
                       1.2                                                          imposes declining trend that matches
                       1.0

                       0.8
                                                                                    observations.
                       0.6
                             1985      1990     1995     2000    2005        2010


                       1.9
                       1.8
                       1.7
                       1.6
                                                                                    Diffuse attenuation coefficient (Kd): Sorting
    Kd(PAR) (m )
   -1




                       1.5
                       1.4
                       1.3                                                          and range selection imposes weak
                       1.2
                       1.1
                       1.0
                                                                                    increasing trend early in record, none
                       0.9
                               1985     1990     1995    2000    2005        2010
                                                                                    later.
    Conclusion from Secchi & Kd
        trends at CB3.3MH
• If Secchi is going down (getting worse),
  and Kd is not getting worse (not going up),
  then the Kd·ZSD product must be changing;
  it, and the scattering-to-absorption ratio
  (b/a), must be going down over time
• In some mainstem segments, it has
  dropped from 1.45-1.7 to near 1.0
• But, why is this not happening
  everywhere?
                          Compare with Observations
                     12


                     10                                                    TSS: Sorting and range selection
TSS (mg L )




                                                                           imposes weak increasing trend early in
-1




                      8
                                                                           the set.
                      6


                      4
                          1985   1990    1995    2000     2005      2010




                     26
                     24
                     22
                     20
                     18
                                                                           Chlorophyll: Sorting and range
       Chl (µg L )
     -1




                     16
                     14
                                                                           selection imposes no trend on the
                     12
                     10                                                    simulations.
                      8
                      6
                      4
                          1985   1990   1995    2000    2005     2010
                                                             Comparison with Unobserved
                                                             Inputs—Opening the TSS Box
              Particulate Organic Matter (mg L-1)




                                                    4.0
                                                    3.5
                                                    3.0
                                                                                                                  Particulate Organic Matter: Sorting
                                                    2.5                                                           and range selection imposes the
                                                    2.0
                                                    1.5
                                                                                                                  strongest increasing trend on "small
                                                    1.0
                                                    0.5
                                                                                                                  organic detritus".
                                                                                                Small detritus
                                                    0.0
                                                          1985    1990    1995    2000    2005             2010

                                                    12

                                                    10

                                                                                                                  Inorganic Suspended Solids: May
        ISS (mg L )




                                                     8
  -1




                                                     6                                                            have had mid-record maximum.
                                                     4
                                                                                                                  Recent measurements suggest upward
                                                     2

                                                     0
                                                                                                                  trend, but insufficient record length.
                                                          1985   1990    1995     2000   2005             2010


                                  1.0

                                  0.8
                                                                                                                   Fraction Organic (VSS/TSS): Possible
Fraction Organic




                                  0.6

                                  0.4
                                                                                                                   mid-record minimum, trending upward in
                                  0.2
                                                                                                                   recent years. Limited measurements
                                  0.0                                                                              confirm recent increases, from 40% to
                                                         1985    1990    1995    2000    2005            2010      60% organic at some sites.
        Modeling Implications
• Can we hindcast and forecast with the same set
  of parameters? (Probably not)
• Today's parameters would likely over-predict Kd
  in the early record
• Depending on the source of increasing
  “small organic detritus,” there may be water
  clarity benefits from nutrient reduction that
  are not being accounted (excess nutrients may
  produce more small organic detritus)
                     Summary
• Kd·ZSD product has declined over the record in the
  mainstem, signaling an increase in the bp:a or
  scattering-to-absorption ratio (but not in all tribs?)
• Most likely cause is increase in highly scattering
  organic particulates, tentatively called "small
  organic detritus.” (Change in phytoplankton size-
  distribution under investigation)
• Mass-specific optical properties (such as Kd·ZSD)
  are sensitive to these changes. Treating them as
  constants may introduce bias in predictions.
• Bio-optical model based on particle sub-
  populations produces results consistent with many
  observations

								
To top