Coastal Water Properties significant to remote sensing

1
Coastal Water Properties significant to remote
sensing

• Samir Ahmed, Alex Gilerson,
• Barry Gross, Fred Moshary
• M. Vargas, J. Zhou, K. Aran, I. Ioannou,
• R. Amin, R. Fortich, M Oo
• NOAA-NESDIS
• M. Wang, R. Stumph

Cooperative teams Florida AM
University, Fl Creighton University, NE
University of Nebraska, NE, Morgan State
University, MD
2
Contents

• Measurement undertaken in Chesapeake 2005 Mission
  (Coastal Waters)
• Breakdown of Black Pixel Approximation
• Anamolous Backscatter Behavior
• Optical Parameter Depth Scale
• Atmospheric Correction Algorithms in coastal
  waters.
• Long Band approach
• Iterative Algorithm
• Polarization Properties of waters.
• Polarized Radiative Transfer Analysis of water
  leaving radiance

3
Field Measurements in Chesapeake Bay

• Set of instruments of CCNY team included
• Wetlabs package for the measurements depths
  profiles of absorption, attenuation, scattering,
  backscattering, Chl concentrations, CDOM
  fluorescence, temperature and salinity.
• GER spectroradiometer for water reflectance
  measurements above and below water surface with
  the option of polarization components detection.

4
Chesapeake Bay Campaign
10 days, 46 stations

• Measurements
• Flyover of the Nebraska- Lincoln Piper Saratoga
  plane with the hyperspectral AISA sensor on
  board
• Laboratory and HPLC water analyses
• Water turbidity and Secchi disk diagnostics

5
Variability of parameters and reflectances
Reflectances at different stations
6
Variability of attenuation and absorptionas
measured by Wetlabs acs (82 channels) instrument
Attenuation
Absorption
Note significant variability in the magnitude of
attenuation (extinction)
7
Depth MeasurementsRapid drop of ChL Anomalous
Scattering
8
Backscattering ratio bbr(?)bb(?)/b(?)
However, some changes 10-40, which can have
strong impact on the retrieval
Generally spectrally flat
9
Fluorescence Height
Traditional method of the fluorescence height
calculation over baseline
10
Fluorescence magnitude and peak over baseline
11
No correlation between ChL measurements and FLH
seen
12
SeaWIFS intercomparison
Correlations seen only for ChL lt 4mg/m3
13
Testing of the band ratio Chl retrieval
algorithmson Chesapeake Data
Retrievals using NIR ratio are significantly
better than for blue-green ratio and will later
be shown to be less sensitive to atmospheric
correction uncertainty
14
Atmospheric Correctionover bright pixels(Results)

• Details presented in talk by Marco Vargas (PhD
  candidate)
• Bands at 1020, 1270 and 1640 are very useful for
  correcting for atmospheres in coastal waters
  since these bands are truly dark.
• These bands recommended for GOES-R through the
  algorithm working group.
• Alternative approach where the BP approximation
  is used only as a first estimate which is
  improved in an iterative approach using suitable
  regression relationships connecting the VIS and
  NIR water leaving radiances.

15
Polarization Properties of Water Leaving Radiance
- Motivation

• A sufficiently polarized water leaving (P gt 10)
  signal can be effectively processed to isolate
  the chlorophyll fluorescence from the elastic
  signal improving the data used in NIR algorithms
• Polarization (angularspectral) characteristics
  of the water leaving radiances might be useful in
  the retrieval of bio-optical parameters and
  mineral particles

16
Case 2 Water IOP Modeling
Same ocean layer particles phytoplankton,
detritus, minerals r0.05-50 µm, Junge
distribution, scattering function from Mie
calculations, but
concentrations of minerals Cs 5 80 mg/l
- Absorption coefficient
- Absorption coefficient of phytoplankton
Morel, 1991
- Absorption coefficient of CDOM
Bricaud, et al., 1981
- Absorption coefficient of minerals
Stramski, et al., 2001
17
Particle phase functions
Note reasonable agreement between Petzold and
microphysical Mie Scattering models
Phase functions of components are obtained from
Mie simulations. Actual scattering function is
obtain by the weights of components in the
mixture
18
Reflectance and degree of polarization for
different viewing angles (Case 2 waters)
Main scattering plane
Perpindicular scattering plane
solar zenith angle 40 deg Chl20mg/m3, Cs
10 mg/l
19
Effects of surface roughness (waters with high
Chl)
Degree of polarization for the Chl concentration
20 mg/m3 and surface roughness parameter
s0.01-0.1 correlates to wind speeds from 1.37
to 18.9 m/s.
Polarization retained with averaging and surface
roughness
20
Degree of polarization for different
concentrations of mineral particles and viewing
angles
Solar zenith angle 40, viewing angle 0
Solar zenith angle 40, viewing angle 30
Depolarization observed for high minerals Change
of viewing geometry suggested
21
Polar plots of reflectance and degree of
polarization
Reflectance
Degree of polarization
Solar angle 40 deg Chl20 mg/m3, Cs10 mg/l,
440 nm
22
Conclusions

• The angular variability of the water leaving
  polarization signal suggests that a multi-angle
  observation should provide improvements in
  retrieval of the phase matrix associated with the
  scattering properties of both Chl and suspended
  solids
• The degree of polarization of the water leaving
  radiance remains large ( 0.1-0.2) over a wide
  range of observation conditions sufficient to
  separate Chl fluorescence from the elastic
  background
• Effects of ocean waves shallow water bottoms
  significantly degrades the polarization
  signature. Polarization can help in shallow water
  Bathymetry
• Atmospheric correction of bright pixel coastal
  waters can be improved using either longer
  wavelength channels or iteratively using
  correlations found between the VIS NIR