Phytooptics Group
Determination of the light
availability in ocean water utilizing
Vibrational Raman Scattering
Tilman Dinter
V. Rozanov, J.P. Burrows, Astrid Bracher
ESA Frascati 31 Oct 2014
[email protected]
Dinter et al.
submitted
ESA Frascati
31st Oct 2014
1
Light availability




Sunlight is the main driver of biological and physical
processes in ocean water.
Quantify processes of photosynthesis, biomass and
primary production.
For calculations of heat transfer and fluxes.
→ climate modeling
ESA Frascati 31 Oct 2014
[email protected]
2
Definition of light availability or
Depth Integrated Scalar Irradiance (DISI)
Actinic flux or scalar irradiance:
Radiation energy density:
ESA Frascati 31 Oct 2014
[email protected]
3
Definition of light availability or
Depth Integrated Scalar Irradiance (DISI)



Defined as number of photons in the water column.
Directly correlated to strength of Vibrational Raman
Scattering (VRS).
Depends on solar zenith angle (sza), surface, atmosphere
and the absorption and scattering of water body.
ESA Frascati 31 Oct 2014
[email protected]
4
Vibrational Raman Scattering in liquid water
What is Vibrational RAMAN
Scattering in liquid water?
• VRS is an inelastic scattering effect
• Incoming light excites water molecules to
vibrations
• Reemission in the same wavelength
(RAYLEIGH scattering)
• Reemission in other wavelength (RAMAN
scattering)
FIZ, Berlin (http://vs-c.de)
ESA Frascati 31 Oct 2014
[email protected]
5
Vibrational Raman Scattering in liquid water
• Transspectral processes causes filling in of absorption or
FRAUNHOFER-lines
• More photons are shifted into than out of the lines
• I-: original ; I+: after filling in process
• Defining pseudo-absorber like in DOAS-method after BEERLAMBERT-Law
ESA Frascati 31 Oct 2014
[email protected]
6
Vibrational Raman Scattering in liquid water
Modeled with SCIATRAN
• Transspectral processes causes filling in of absorption or
FRAUNHOFER-lines
• More photons are shifted into than out of the lines
• I-: original ; I+: after filling in process
• Defining pseudo-absorber like in DOAS-method after BEERLAMBERT-Law
ESA Frascati 31 Oct 2014
[email protected]
7
Vibrational Raman Scattering in liquid water
Excitation at 390-444.5nm
Reemission at 450-524nm
ESA Frascati 31 Oct 2014
[email protected]
8
Model Scenario SCIATRAN (Rozanov et al. 2014)

Clear Rayleigh, Ozone Atmosphere

Reference solar zenith angle 40°

Ocean:

500m homogeneous mixed water body, black albedo

Surface: Cox & Munk (1954) approx. (wind speed 4.1 m/s)




Standard chl-a CASE-I water model (Morel et al. (1991) and Prieur &
Sathyendranath (1981)) for absorption
Molecular (Rayleigh) scattering by Buiteveld et al. (1994)
Particle scattering based on bi-modal distribution model of
Kopelevich (1983)
Conc. of small and large particles are determined by one parameter
model of Haltrin (1999)
ESA Frascati 31 Oct 2014
[email protected]
9
VRS Weighting Function from TOA radiance simulations
C = 0.1 mg/m^3
C+ΔC = 0.11 mg/m^3
ESA Frascati 31 Oct 2014
[email protected]
10
Weighting Function Differential Optical Absorption
Spectroscopy (WF-DOAS)
ESA Frascati 31 Oct 2014
[email protected]
11
Weighting Function Differential Optical Absorption
Spectroscopy (WF-DOAS)
VRS-WF
ESA Frascati 31 Oct 2014
Ocean IOPs
without VRS
Atmosphere
[email protected]
Polynom
12
Weighting Function Differential Optical Absorption
Spectroscopy (WF-DOAS)
VRS-WF
ESA Frascati 31 Oct 2014
Ocean IOPs
without VRS
Atmosphere
[email protected]
Polynom
13
Modeling subsurface radiation to determine light
availability (DISI)
- Integrated over VRS excitation (390 – 444.5nm)
- Reference solar zenith angle 40°
- The less chl-a the more light is available in the ocean water
ESA Frascati 31 Oct 2014
[email protected]
14
Modeling subsurface radiation to determine DISI: Change
with chl-a conc.
- Integrated over VRS excitation (390 – 444.5nm)
- Reference solar zenith angle 40°
- The less chl-a the more light is available in the ocean water
ESA Frascati 31 Oct 2014
[email protected]
15
Modeled relationship between DISI and VRS fitfactor
+
Fitfactor (strength) of
VRS signal has a
=>
nearly linear
relationship to DISI
ESA Frascati 31 Oct 2014
[email protected]
16
Simulating effects of different absorptions and profiles
Different specific absorption spectra
ESA Frascati 31 Oct 2014
Different profiles (Uitz et al.)
[email protected]
17
Simulating effects of different absorptions and profiles
+
=>
Leads to a maximum
error of ~10%
ESA Frascati 31 Oct 2014
[email protected]
18
DISI dependence on solar zenith angle
ESA Frascati 31 Oct 2014
[email protected]
19
Accounting for solar zenith angle in Look-up Table
ESA Frascati 31 Oct 2014
[email protected]
20
SCIAMACHY DISI from VRS WF-DOAS fit (sza corr)
[phot/sec/m]
ESA Frascati 31 Oct 2014
[email protected]
21
SCIAMACHY VRS WF-DOAS fit
ESA Frascati 31 Oct 2014
[email protected]
22
Summary and Outlook




A new method is introduced to determine the light
availability from VRS effect
Has the potential to determine PAR directly (without
former determination of a, b, and PAR(0+))
Applicable to other hyperspectral sensors like GOME-2,
OMI, or upcoming TROPOMI with different overpass
times (-> daily cycle)
Publication submitted at OS
ESA Frascati 31 Oct 2014
[email protected]
23
End
Thank You
for
Attention!
ESA Frascati 31 Oct 2014
[email protected]
24
Modeling subsurface radiation to determine DISI
Appendix
ESA Frascati 31 Oct 2014
[email protected]
25
Comparison
Comparison