Optics of particles - University of Maine Systemmisclab.umeoce.maine.edu/boss/classes/SMS_618/Optics...
Transcript of Optics of particles - University of Maine Systemmisclab.umeoce.maine.edu/boss/classes/SMS_618/Optics...
Optical properties of oceanic particles
Wh t ph si l p p ti s d t mi th pti l p p ti s f p ti l s?What physical properties determine the optical properties of particles?
Size, composition (refractive index), shape, internal structure. These properties interact…
Based, in part, on a lecture prepared by C. RoeslerWe will not address fluorescence and polarization.
What particles absorb and scatter in the ocean?
Phytoplankton:
Variable in shape size and pigment compositionVariable in shape, size and pigment composition.
Variable in scattering and absorption properties
What particles absorb and scatter in the ocean?
Non-algal particles: Organic and inorganic.
Sand Silt clay
Aggregates:
http://www.aad.gov.au/default.aspVariable in scattering and absorption properties
IOP TheoryIOP Theory
L LLo
IncidentR d
TransmittedR d
Lt
Radiance Radiance
No attenuation (recap: radiance is constant along a ray)
Loss due to absorptionLoss due to absorption
La Absorbed Radiance
L LLo
IncidentR d
Lt
TransmittedR dRadiance Radiance
Loss due to scatteringLoss due to scattering
Lb Scattered Radiance
L LLo
IncidentR d
Lt
TransmittedR dRadiance Radiance
Loss due to absorption and scattering (attenuation)
Lb Scattered Radiance
La Absorbed Radiance
LL LtLo
IncidentR d
TransmittedR dRadiance Radiance
Loss due to absorption and scattering (attenuation)
Lb Scattered Radiance
La Absorbed Radiance
LL LtLo
IncidentR d
TransmittedR dRadiance Radiance
Lo = Lt + La + Lb
Beam Attenuation M t ThMeasurement Theory
Attenuance C = fraction of incident
radiance attenuated Lb
L
La
L
C (L L )/L
LtLo
C = (Lb + La)/Lo
C = (Lo – Lt)/Lo
Beam Attenuation M t ThMeasurement Theory
c = fractional attenuance it di t c = ΔC/Δx
Lb
per unit distance c Δx = - ΔL/L
x x
L
La
L
∫0 c dx = -∫0 dL/Lx x
c(x 0) [ ln(L ) ln(L )]LtLo c(x-0) = -[ ln(Lx)-ln(L0)]
c x = -[ ln(Lt)-ln(Lo)]
c x = - ln(Lt/Lo)
c (m-1) = (-1/x) ln(Lt/Lo)( ) ( ) ( t o)
Δx Lt=Loexp(-c x)
Lt=Loexp(-c x)τ=cx ≡ optical depth
Beer(1852), Lambert(1729), Bouguer (before 1729) p p
Important characteristics:
Absorption and scattering feature similarlyAbsorption and scattering feature similarly.
c is Linear with concentration or added substances.
Assumes no internal sources.
Class demo with food color/maalox
Sinks for photons – absorbersSinks for photons absorbers.What absorbs radiation in the oceans/lakes?
How do we measure absorption?
What is absorption:
Photons ‘disappear’ due to interaction with matter.
From: http://www haverford edu/chem/Scarrow/GenChem/quantum/BohrAtom gif
Following absorption energy could be reemitted (or not) at the same or different wavelength (e.g. scattering, fluorescence, thermal emission).
From: http://www.haverford.edu/chem/Scarrow/GenChem/quantum/BohrAtom.gif
Absorption due to molecular:Translational, rotational, vibrational and electronic modes (+combinations)and electronic modes (+combinations).
Observation include also spontaneous emission.
From: http://acept.la.asu.edu/PiN/rdg/irnuv/d7.gif
Explained by quantum mechanics.
b b /l kAbsorbers in ocean/lakes
• Water• Phytoplankton• Phytoplankton• Non-algal organic particles (CPOM)• Inorganic mineral particles (CPIM)Inorganic mineral particles (CPIM)• Dissolved organic matter (CDOM)
– not addressed today
Note: most available information is in the visible (400-700nm). Why?
Absorption by (liquid) water:F t s:Features:
•Extremly complex
•Variety of vibrational modes
•Combination possible•Combination possible
Assumes no internal sources.
http://www.lsbu.ac.uk/water/vibrat.html
PhytoplanktonSpecies-specific pigment composition
Phytoplankton
3
2
3
a(67
6)
15101520
Cell size
0
1a()/a 30
4050
0400 500 600 700
Wavelength (nm)
(Morel and Bricaud 1981)
Phytoplankton vs Chlorophyll
Sosik & Mitchell 1991
http://chaitanya1.wordpress.com/2007/07/09/strawberries/
chloroplast chlorophyll
cell Packaging: a/[chl] is function of size and [chl]Duysens (1956)
a/V= QabsG/{1.33πr3}Theoretical results:
n’=0.01
n’=apureλ/4π
‘Packaging’
Cabs = areaa/V ∝ 1/D
Molecular absorption ∝ volume.
a/V 1/D
Non-algal particlesNon algal particlesHeterotrophic Ciliates and Flagellates, and bacteria
D t itDetritus
Ahn and Morel, 1983 Iturriaga and Siegel 1989Iturriaga and Siegel 1989
aCPOM(l) = aCPOM(400)*exp(-SCPOM(l-400)) SCPOM = 0.007 to 0.011Roesler et al., 1989
CDOM: chromophoric dissolved organic matter
aCDOM(l) = aCDOM(400)*exp(-SCDOM(l-400)) SCdOM = 0.011 to 0.022
Kirk 1983 Carder et al. 1989
Absorption MeasurementsAbsorption Measurementsa = (-1/x) ln[(L + L )/L ]a = (-1/x) ln[(Lt + Lb)/Lo]
Detected flux measurement must include scattered flux
La Lb
mu u f usource
Lo Lt
detector
Lo Ltx
d t t d tt d fl undetected scattered flux must be corrected for
Ex. Absorption meter: E b WETLabs ac-9 & ac-S
• Dual path, a and c• b inferred by y
difference.• Correct for photon that
are not collected
Ex Absorption meter: Ex. Absorption meter: WETLabs ac9/acS
• Detector FOV ~ 40o
• ~90% of scattered flux
Absorption Measurementp
Detected flux measurement includes all scattered flux
Integrating sphere:Detected flux measurement includes all scattered flux.
Reflectance of sphere needs to be known very accurately (problems with fouling)accurately (problems with fouling).
Absorption MeasurementAbsorption MeasurementQFT: quantitative filter-pad techniqueSeparate sample to dissolved and particulate. Concentrate particulate material on a filter.
M b f di l dMeasure beam-c for dissolved.Measure transmission through filter pad (correct for ‘filter effects’ path length etc’)filter effects , path-length, etc ).
Absorption MeasurementsAbsorption MeasurementsIn all measurements of transmission (for a or c) there is a need for a ‘reference’c) there is a need for a reference .
Need for a stable (temperature salinity) Need for a stable (temperature, salinity) predictable standard.
Currently we use the cleanest waters we can get…
/l kScatterers in ocean/lakes
• Water + salts – not today• Phytoplankton• Phytoplankton• Non-algal organic particles (CPOM)• Inorganic mineral particles (CPIM)Inorganic mineral particles (CPIM)• Dissolved organic matter (CDOM)
– mostly negligible
OPs are good predictors of PM.
Optics is a standard method to measure turbidity, a primary determinant of water quality ( ISO 7027)(e.g. ISO-7027).
|Model-PM|/PM c(660) bb(700) bs(880)
5% 2% 1% 2%
50% 16% 9% 21%50% 16% 9% 21%
95% 54% 36% 51%
Angular dependence of scattering on size:
•Near forward scattering: Strong dependence on size,
‘large’
‘small’g p ,less on n.
•b /b: Strong dependence on •bb/b: Strong dependence on n, less on size.
Roesler and Boss, 2008
c/V=Qext·G/{1.33πr3}
Resonant curve. Change of max with n (and λ).
c ∝ a and Volume
σext = 2·areac/V ∝ 1/DVo um c/V /D
Mie theory tells us that the relationship between optical properties and mass is size dependent:
bbp/Volume
bp/Volume
1/D
D3
bs /Mass
•All curves are ‘resonant’ curves
•Highest sensitivity for micron sized ti l bsp/Mass particles
•Size of max response varies
Scattering and absorptionObservations:
Mie theory:
A photon absorbed is A photon absorbed is not scattered (at any direction).
Boss et al 2001
Validation:Beam attenuation:c=a+b
Spectral attenuation and PSD:
Boss et al., 2001
Mie Theory (homogenous spheres
c a b
Mie Theory (homogenous spheres, Volz, 1954):
For non-absorbing particles of the d h b l d b same n and an hyperbolic distribution
from Dmin=0 to Dmax=∞,
If:
⎞⎛γ
N(D)=No(D/Do)-ξIf:
then:
( ) ( ) 3 ,0
0 +γ=ξ⎟⎟⎠
⎞⎜⎜⎝
⎛λλ
λ=λγ−
pp cc0 ⎠⎝
expect a relation between attenuation spectrum and PSD.
The bb enigma?
Morel and Ahn, 1991: ‘Algal cells in open ocean, and to lesser extent small heterotrophs, dominate the scattering coefficients; …On the contrary, these organisms are contrary, these organisms are definitely insignificant contributors to the backscattering coefficient.’
Stramski et al., 2001: simulating open-ocean (oligotrophic 0 18mg open-ocean (oligotrophic, 0.18mg Chl/m3), 2-3% of the backscattering coefficient is due t l kt 50% f ti l
Stramski et al., 2001to plankton. 50% from particles <0.2μm.
The bb enigma (or paradox):
Based on Mie theory, backscattering should be dominated by inorganic particles and sub-micron particles (the least known of the bunch).
Yet b correlates well with [chl] and POC (>0 7 m):Yet bbp correlates well with [chl] and POC (>0.7μm):
Huot et al., 2008 Stramski et al., 2008
Possible explanation for the bb enigma:
1. Mie results are correct. However, all particles in the open ocean covary, hence the tight relationship.
2. Mie theory is not applicable. Organic particle actually backscatter more than we ascribe to themactually backscatter more than we ascribe to them.
How do we obtain the backscattering in the ocean:
bb=2πχβ(θ), 0.7<χ<1.3
Uncertainty ~10%
Boss and Pegau, 2001
Shape approximationsfor light scattering calculations for light scattering calculations
Mie-Theory
T-matrixModerate Axis
10 oblate
Moderate Axis ratios (0.5<AR<2)
12
0.5 T-matrix
0.1
0.1 1 10
prolate0.5
Axis ratios up to convergence limit
Particle radius (μm)0.1 1 10
Slide From Volten
Measurements across the equatorial Pacific (Dall’olmo et q (al., 2009):
bbp well correlated with cp
bbp(D<0.2mm)
bbp(D>0.2mm) <0.1
No filter effect visible
Uncertainty dominated by uncertainty in bb(H2O)
Aggregate modeling:
2
Latimer (1985)
2
For marine aggregates
4mm
For marine aggregates size and solid fraction correlate.
points having size F as in -points having size-F as in Maggi, 2007, or Khelifa and Hill, 2006.
Theoretical calculations: monodispersionAggregates
Single grain
Observed range{
Single grain
Mass normalized beam attenuation for aggregates assuming a relationship Mass normalized beam attenuation for aggregates assuming a relationship between solid fraction and size as in Khelifa and Hill, 2006 (blue lines) and solid particles (red lines). Solid lines denote particles with n=1.05+i0.0001, dashed lines n=1.05+0.005 and dotted lines n=1.15+0.0001. Each data point prepresent a population of particle all of a single size.
Laboratory experiment: aggregation
Start with <D>~7μm clay Start with <D>~7μm clay
Add salt
As aggregate size As aggregate size changes from 7 70μm cp/mass and bbp/mass stay constant (<size> stay constant (<size> confirmed by microscopy).