Seawater inherent optical properties (IOPs) are key parameters in a wide range of applications in environmental studies and oceanographic research. In particular, the absorption coefficient (a) is the typical IOP used to obtain the concentration of chlorophyll-a in the water—a critical parameter in biological oceanography studies and the backscattering coefficient () is used as a measure of turbidity. In this study, we test a novel instrument concept designed to obtain both the absorption and backscattering coefficients. The instrument would emit a collimated monochromatic light beam into the water retrieving the backscattered light intensity as a function of distance from the center of illumination. We use Monte Carlo modeling of light propagation to create an inversion algorithm that translates the signal from such an instrument into values of a and . Our results, based on simulations spanning the bulk of natural values of seawater IOP combinations, indicate that a diameter instrument with a radial resolution of would be capable of predicting within less than 13.4% relative difference and a within less than 57% relative difference (for 90% of the inverted a values, the relative errors fall below 29.7%). Additionally, these errors could be further reduced by constraining the inversion algorithm with information from concurrent measurements of other IOPs. Such a compact and relatively simple device could have multiple applications for in situ optical measurements, including a and retrievals from instrumentation mounted on autonomous underwater vehicles. Furthermore, the same methodology could possibly be used for an out-of-water sensor.
You do not have subscription access to this journal. Cited by links are available to subscribers only. You may subscribe either as an Optica member, or as an authorized user of your institution.
You do not have subscription access to this journal. Figure files are available to subscribers only. You may subscribe either as an Optica member, or as an authorized user of your institution.
You do not have subscription access to this journal. Article tables are available to subscribers only. You may subscribe either as an Optica member, or as an authorized user of your institution.
You do not have subscription access to this journal. Equations are available to subscribers only. You may subscribe either as an Optica member, or as an authorized user of your institution.
Inversion Errors for Simulated Detector Responses to Entire Range of IOP Combinations for Different Inversion Algorithms, Obtained with the “Training Data Sets” and Applied on the “Test Data Sets”
(%) Training Data Set
(%) Test Data Set
(%) (All Data)
(%) (90% of Data)
(%) (All Data)
(%) (90% of Data)
0.5–2.5
0.5–2.5
13.4
6.9
56.9
29.7
0.5–2.5
5
18.2
15.1
95.9
68.0
0.5–5
0.5–5
15.0
8.9
66.3
31.5
0.5
0.5
8.7
5.5
40.0
25.7
1
1
6.8
3.8
39.1
23.2
1.5
1.5
6.1
3.7
38.2
23.0
2
2
6.3
4.3
38.8
22.9
2.5
2.5
8.0
5.4
38.7
24.1
5
5
9.3
5.6
37.6
22.7
Table 4
Inversion Algorithm Performance on Simulated Response of Detectors of Different Sizes to Entire Range of IOP Combinations
Detector Radius (cm)
(%) (All Data)
(%) (90% of Data)
3.1
56.9
29.7
4
55.7
29.1
10
52.2
26.5
100
28.6
13.7
Table 5
Inversion Algorithm Performance on Simulated Response of Detectors of Different Resolutions to Entire Range of IOP Combinations
Detector Resolution (mm)
(%) (All Data)
(%) (90% of Data)
10
13.4%
6.9
1
12.1%
6.6
(%) (All Data)
(%) (90% of data)
10
56.9
29.7
1
54.6
29.0
Table 6
Inversion Algorithm Performance on Simulated Detector Response: Comparison between Numerical Experiments Using Different Scattering Phase Functionsa
Experiment
()
()
Relative Error (%)
()
()
Relative Error a (%)
P1
0.44
0.45
FF1
0.44
0.45
P2
0.117
0.128
0.6
0.84
FF2
0.128
0.138
0.6
0.86
P1 and P2 use the Petzold scattering phase function () and FF1 and FF2 use the Fournier–Forand phase function () to simulate scattering by particles. P1 and FF1, , ; P2 and FF2, , .
Table 7
Inversion Algorithm Performance and Detector Response for Instruments Located on the Water Surface versus in the Water, Far from the Surface (; )
Bp
Parameter
Instrument on Water Surface
Instrument in Water
D
0.0101
0.0100
0.025
α ()
0.316
0.303
27.5%
23.3%
D
0.0060
0.0060
0.015
α ()
0.192
0.185
6.8%
2.5%
D
0.0020
0.0020
0.005
α ()
0.063
0.065
Tables (7)
Table 1
Notations
Variable
Units
Definition
a
Total absorption coefficient
Seawater absorption coefficient
Particulate absorption coefficient
Gelbstoff absorption coefficient
b
Total scattering coefficient
Seawater scattering coefficient
Particulate scattering coefficient
Total backscattering coefficient
Seawater backscattering coefficient
Particulate backscattering coefficient
c
Total beam attenuation coefficient
Particulate backscattering ratio:
VSF
Scattering phase function
Air index of refraction (relative to water) at
Seawater index of refraction (relative to water) at
λ
nm
Wavelength
R
cm
Detector radius
mm
Thickness of detector rings
w
Photon weight used in radiative transfer simulation
Photon weight count at distance r from the center
Photon weight count per unit area at distance r from the center in a time unit
Photon weight count within distance r from the center
Number of photons emitted by the source
t
s
Instrument exposure time
D
Total detector photon count
α
Signal geometry indicator (see description in Subsection 2C)
Inversion Errors for Simulated Detector Responses to Entire Range of IOP Combinations for Different Inversion Algorithms, Obtained with the “Training Data Sets” and Applied on the “Test Data Sets”
(%) Training Data Set
(%) Test Data Set
(%) (All Data)
(%) (90% of Data)
(%) (All Data)
(%) (90% of Data)
0.5–2.5
0.5–2.5
13.4
6.9
56.9
29.7
0.5–2.5
5
18.2
15.1
95.9
68.0
0.5–5
0.5–5
15.0
8.9
66.3
31.5
0.5
0.5
8.7
5.5
40.0
25.7
1
1
6.8
3.8
39.1
23.2
1.5
1.5
6.1
3.7
38.2
23.0
2
2
6.3
4.3
38.8
22.9
2.5
2.5
8.0
5.4
38.7
24.1
5
5
9.3
5.6
37.6
22.7
Table 4
Inversion Algorithm Performance on Simulated Response of Detectors of Different Sizes to Entire Range of IOP Combinations
Detector Radius (cm)
(%) (All Data)
(%) (90% of Data)
3.1
56.9
29.7
4
55.7
29.1
10
52.2
26.5
100
28.6
13.7
Table 5
Inversion Algorithm Performance on Simulated Response of Detectors of Different Resolutions to Entire Range of IOP Combinations
Detector Resolution (mm)
(%) (All Data)
(%) (90% of Data)
10
13.4%
6.9
1
12.1%
6.6
(%) (All Data)
(%) (90% of data)
10
56.9
29.7
1
54.6
29.0
Table 6
Inversion Algorithm Performance on Simulated Detector Response: Comparison between Numerical Experiments Using Different Scattering Phase Functionsa
Experiment
()
()
Relative Error (%)
()
()
Relative Error a (%)
P1
0.44
0.45
FF1
0.44
0.45
P2
0.117
0.128
0.6
0.84
FF2
0.128
0.138
0.6
0.86
P1 and P2 use the Petzold scattering phase function () and FF1 and FF2 use the Fournier–Forand phase function () to simulate scattering by particles. P1 and FF1, , ; P2 and FF2, , .
Table 7
Inversion Algorithm Performance and Detector Response for Instruments Located on the Water Surface versus in the Water, Far from the Surface (; )