Transcript Simulating the wind and sea-surface roughness effects on Aquarius

Simulating the wind and sea-surface roughness effects on Aquarius
Sea Surface Salinity Retrievals: Evaluating alternative models to
correct for the effects of the rough sea surface on L-band
radiometer emission and scattering
NASA Applied Sciences Program
Mississippi Research Consortium
Prototype Solutions from Next Generation NASA Earth Observing and
Predictive Capabilities
Investigators: S. Howden* (P.I.), D. Burrage#, J. Wesson#, D. Ko,
D. Wang#
Funding Requested: $463,271
Duration: 18 months
*Department of Marine Science
The University of Southern Mississippi
#Naval Research Laboratory
Stennis Space Center
Concept: Rapid Prototyping of Accurate SSS Retrievals
Earth Observations:
• Aquarius Mission Microwave Radiometer and Scatterometer Data
• NASA (Quick SCAT), Jason-1 (Wind Speed and Wave Height)
Predictions and Measurements:
• Coastal Sea State (NDBC, STARRS C-band radiometer)
• Coastal SSS, SST, and SSR (Roughness Model Simulations and STARRS L-band rad.)
Decision Support:
• Optimizing the accuracy of Aquarius SSS retrievals
• Selecting operational Roughness Correction models
• Monitoring the quality and accuracy of Aquarius SSS retrievals over time
Benefits:
•
•
•
•
•
•
•
Improved knowledge of SSS retrieval and SSR influence
Accurate remotely sensed SSS for ocean circulation models and assimilation
Expanded knowledge of salinity at the air-sea boundary for constraining hydrological cycle
Improved Navy, NOAA, and NASA ocean circulation models
Better information for near shore fisheries
Improved hydrographic, conductivity and sound speed information for Navy operations
Enhanced information on deep ocean and coastal salinity and temperature fronts
Project Status at 31 May, 2009
• First written report presented July, 2007.
• Last Review presentation 4 April, 2008.*
• Extension to original NRL/USM CRADA granted June, 2008.
• Roughness modeling aspects still progressing, but with
reduced emphasis on Aquarius simulation.
• New results presented on both roughness and optics aspects at
engineering and science meetings.
• Project has spawned a successful bid to NRL for base funding to
continue the roughness work as well as two ROSES proposals.
*Decisions arising from 4 April, 2008 review
• NASA Program seeking more emphasis on immediate returns and
demonstrating new applications with utility for real world problems.
• Agreement to place new emphasis on optical/SSS results.
• Possible extension discussed to allow for late initial funding transfers
(6-month delay in establishing NRL/USM CRADA).
• Administrative difficulties of funding NASA for simulation work
led to decision to perform additional roughness field work instead.
Recent accomplishments
• Analysis and reporting of results from STARRS surveys
flown off Virginia during Dec 2006.
• Analysis and reporting of field campaign in the Gulf of
Mexico in May 2007 (microwave and optical SSS retrievals
compared).
• Two papers presented at IEEE Transactions on
Geoscience and Remote Sensing (on roughness and
optical aspects).
• Virgilio Maisonet investigated optical aspects and
presented an award-winning paper on optical CDOM/SSS.
• Preliminary assessment of available roughness models
(follow on NRL base-funded project approved).
Roughness Correction Models
Wave Spectra
&
Model Evaluation
Sea Surface Roughness Components
1. Ambient Roughness: swell from distant storms
2. Wind Wave Roughness: wind-generated short waves
3. Breaking Roughness: breakers, whitecaps, foam
Slight Sea
Short Wind Waves
Rough Sea
Surface Wave Height Frequency Spectrum Observed
During Cold Front Passage in Gulf of Mexico
WindSwell
Waves
T=7s
T=2s
Swell tilts the short waves,
changing their slope.
l=1 m
1 cm
Short Waves
Reference: Pan, et al., (2005) JGR C, v110, C02020
M(1,:,:) at DelT: 1[dB V]
M(1,:,:) at DelT: 1
100and Hence
3
Roughness Changes Emissivity
Brightness Temperature,
Causing Errors in Salinity Retrievals
that
Assume Sea is Flat.
90
2.8
2.6
Brightness Temp, Tb = e Ts
70
2.4
60
2.2
50
2
40
1.8
30
1.6
E=Er+Ei 20
1.4
10
1.2
Y
e= Flat Sea Emissivity
80
q q
Flat Sea (Ts, S)
20
40
60
80
X
100
er=Rough Sea Emissivity
Tb = (e + er) Ts
q f
E=Er+Ei
Rough Sea
1
20
40

60
80
X
e(q,Ts,S)=1-Rf(q,Ts,S)
er (q) = cos(f) (1-Rr(q,f)) df
Reflection Coef., Rf(q,Ts,S)
Reflection Coef. , Rr(q,f)
Rf(q,Ts,S) =Er2/Ei2 (Klein & Swift)
Rr(q,f)=Er2 / Ei2 (Determine
S = Salinity, Ts = Temperature
using numerical experiments)
100
Roughness Correction Models considered
for Satellite Mission Processing:
Empirical models (Simple and efficient, but have limited applicability range):
A. Camps, et al., (2003) WISE model of Tb for specified Wind and/or Wave Height.
Gabarro, et al., (2003) Retrieves SSS, Wind Speed and Water Temperature
simultaneously from Multi-angle L-Band measurements (Multi-parameter retrieval).
Neural Network Model to be trained on SMOS data after launch
Asymptotic models (Questionable accuracy, but efficient for operational use):
Yueh (1997) Two-scale model - Divides wave spectrum into long and short wave parts.
Employs a Gaussian input spectrum.
SSA/SPM Voronovich (1994) - Works well only for certain types of wave spectrum
Employs optional spectra, such as Kudryavtsev et al.
[Combines Small Slope Approximation (SSA) of Voronovich (1985)
and Johnson (1999) with Small Perturbation Method (SPM) of Rice (1951)].
A High Accuracy Reference E-M Interaction Model:
Rigorous E-M scattering models (High accuracy, but computationally intensive):
Taflove and Hagness (2005) Finite Difference Time Domain Method (FDTD) Method
– Accurate and Adaptable, Rigorous Solution of Maxwell’s Equations
Reference: Reul, et al., (2005) IGARSS '05. Proc. 2005 IEEE v3, 2195 – 2198
Predicted Brightness Temps from Two-scale Model (TSM) and
SSA/SPM for Given Roughness Spectra versus Wind Direction
at Wind Speed, Ws=15 m/s Differ by Up To ~ 2K (4 psu S)!
Optimal for SSS sensing
Tb [k]
TSM V-Pol
10
8
6
4
2
0
SPM/SSA V-Pol
Tb [k]
10
8
6
4
2
0
14
12
10
8
6
4
2
0
-2
Optimal for SSR sensing
TSM H-Pol
Tb [k]
10
8
6
4
2
0
L-band
f=1.4 GHz
14
SPM/SSA H-Pol
12
Tb [k]
10
10
8
8
6
6
4
2
4
0
2
0
-2
Wave Age W-1=0.5
References: TSM (Yueh, 1997), SSA/SPM (Reul, 2007)
14
12
10
8
6
4
2
0
-2
14
12
10
8
6
4
2
0
-2
SSA/SPM Tb Predictions Based on Different WindWave Spectra Differ Significantly (dTb~1 K, dS=2 psu)
B(k)
Curvature
Spectrum
Kudryavtsev
U=10 m/s
kL
U=5 m/s
B(k)=S(k)k3
l=1 m
kC
1 cm
Elfouhaily
K (rad s-1)
Tb
Tb(K)
(K)
V-Pol
Tb (K)
H-Pol
1K
1K
U (m/s)
U (m/s)
Parameters: Inc. Angle 37 deg., Ts=298 K, Salinity=35 psu, Wave age=0.84
Compare 2 psu roughness correction error with observed
SSS difference across Gulf Stream ~ 4 psu (Wilson et al., 1999)
References: Elfouhaily, et al., 1997; Kudryavtsev, et al. 2003
Rigorous EM Scattering
Finite Difference Time Domain
(FDTD) Model
Reference Model
Development
(Early coarse resolution version)
Y
70
Procedure to Determine
Radar Cross Section (RCS) and Hence 2.4
Emissivity
2.2
Using FDTD Reference
Model and Monte Carlo simulation
60
50
Rf=|Er|2/|Ei|2
40
Ei
30
Er
1.8
Virtual1.6
Surface
q
f
20
2
1.4
E=Er+Ei
1.2
10
20
40
60
80
100
1
X
An incident plane wave (Ei) is generated at the Virtual Surface
and is reflected off the rough sea surface.
This surface is one realization of a roughness spectrum.
The reflected wave (Er) is detected above the virtual surface
(the incident wave is absent there).
The Reflectance or RCS are determined from Rr=|Er|2 / |Ei|2
Repeat for multiple incidence angles and roughness spectrum
realizations (i.e., using Monte Carlo Simulation).
Results are averaged to estimate Rough Emissivity (Integral of 1-Rr).
Configuration for Simulating Reflection from Smooth and Rough Seas
M(1,:,:) at DelT: 1
M(1,:,:) at DelT: 1
Level flat surface
90
80
Point
Source
60
50
Air
3
90
2.6
80
2.6
2.4
70
2.4
2.2
60
2.2
2
50
2.8
2
Air
40
1.8
40
1.8
30
1.6
30
1.6
1.4
20
1.4
1.2
10
1.2
0.4 m
20
10
Flat Sea
10
100
20
30
40
50
60
X
M(1,:,:) at DelT: 1
70
80
90
100
Slight Sea
1
10
20
30
40
50
60
X
M(1,:,:) at DelT: 1
70
80
90
100
100
3
Sloping flat surface
1
3
Level very rough surface
2.8
90
2.6
80
2.6
70
2.4
70
2.4
60
2.2
60
2.2
50
2
50
2
40
1.8
40
1.8
1.6
30
1.4
20
1.2
10
90
Grid Cell # 0->100
80
Y
Level slightly rough surface
2.8
Air
30
20
10
Sloping Sea
10
20
30
40
50
X
60
70
Grid Cell # 0->100
80
90
100
1
Y
Y
Grid Cell # 0->100
70
100
3
Y
100
Air
2.8
1.6
1.4
1.2
Rough Sea
10
20
30
40
50
X
60
70
Grid Cell # 0->100
80
90
100
1
FDTD Simulation of C-band Energy [dB] for Surface Backscatter
|E|2 db, E = Ei + Er
Ez(m,:,:) at DelT: 180
100
Level flat surface
90
100
0
90
Grid Cell #Y 0->100
80
|E|2 db
Ez(m,:,:) at DelT: 180
Level slightly rough surface
80
70
70
-50
60
-50
60
Y
Shadow
Zone
50
40
50
Mean Sea
Surface
40
30
-100
l=0.05 m
-100
30
20
20
10
10
0.4 m
10
100
20
30
40
50
60
70
Ez(m,:,:) at
X DelT: 180
80
90
100
-150
10
100
Sloping flat surface
90
90
80
80
20
30
40
50
60
70
X DelT: 180
Ez(m,:,:) at
80
90
100
Level very rough surface
0
70
-150
0
70
-50
-50
60
60
50
50
Y
Y
Grid Cell # 0->100
0
40
40
-100
30
20
20
10
10
10
20
30
40
50
X
60
Grid Cell # 0->100
70
80
90
100
-100
30
-150
10
20
30
40
50
X
60
70
Grid Cell # 0->100
80
90
100
-150
Field Campaigns
(VIRGO & COSSAR)
The STARRS Sampling Scheme
STARRS Airborne Microwave Radiometer System
VisibleBands
IR-Band
STARRS
L-Band
C-Band
STARRS
Piper Navajo
STARRS Sampling Scheme
Ocean (Ts, Tb, Oc)
Flight
Direction
NEDT(1s)=
0.50 K
dS=1 psu
C, IR &
Vis-Bands
(SeaWiFS Chs.)
L-Band
Incidence Angles:
+/- 7,22,37 (deg)
Scan
6.0
0.6
Pixel Altitude (km)
~1
2.6
~0.1
0.26
NRL’s Salinity, Temperature, and Roughness Remote Scanner (STARRS)
Virgo Optical Images Crossing Gulf Stream On 12 Dec., 2006:
Swell, White Caps and Foam Also Influence SSS Retrievals
Inshore
White Caps
& Foam
14:10
15:30
Swell
~100m
Offshore
15:49
NIKON D1X Digital Camera
16:10
Virgo SST and SSS Crossing Gulf Stream On 12 Dec., 2006:
Effect of Empirical Wind-Induced Roughness Corrections
Map of SST [C] for flight on 12-Dec-2006 from 11:52:45 to 15:01:57UTC in file C12dec06a
37.4
NOAA Met. Buoys:
37.2
CBBV2
CHLV2
Lat [deg]
37
36.8
44014
36.6
36.4
36.8
44014
36.6
36.4
STARRS SST
36.2
36
-76.5
NOAA Met. Buoys:
37.2
CBBV2
CHLV2
37
Lat [deg]
Map of Salt [psu] for flight on 12-Dec-2006 from 11:52:45 to 15:01:57UTC in file C12dec06a
37.4
-76
-75.5
-75
-74.5
Long [deg]
15
STARRS SSS
36.2
-74
20
Chesapeake Bay
-73.5
36
-76.5
-73
25
-76
-75.5
-75
-74.5
Long [deg]
-74
-73.5
-73
10
15
20
30
35
40
25
*Virgo 12 Dec 06 Estimated Roughness
Corrections (Inc. Angle 7 deg)
7.00
5.80
6.00
5.20
#(cf 1.1 psu
For Hs=0.7 m)
5.00
Cape
Hatteras
4.00
3.00
2.50
2.45
2.00
1.00
1.06
2.20
1.22
1.10
CHLV2
44014
0.53
0.00
CBBV2
Wind Spd m/s
Terra MODIS SST
Delta Tb (K)
Delta S (psu)
*Based on WISE wind model (# Wave model
correction is smaller by a factor of two!)
Color, Surface Salinity and Roughness (COSSAR) Wind
and Wave Data and STARRS flights 10-15 May 07
U (m/s)
Buoy NDBC 42007
q (deg)
NDBC 42040
Mississippi Outfall
Hs (m)
f (deg)
NOAA NDBC 42007
STARRS flights over Mississippi Outfall
Observations and Analysis
• R/V Pelican survey from Atchafalaya Bay to deep
ocean salinity (8-10 May 2007).
• Two aircraft surveys 10 May 2007 with STARRS and
Satlantic (SeaWifs Airborne Simulator) instruments.
• Confirm STARRS Salinity matches shipboard.
• Show that Optical measurements detect fronts and can be
used in Salinity regression.
• Compare regression Salinity with STARRS Salinity over
flight survey region.
STARRS Salinity, Morning flight
10 May 2007
29.6
29.4
Lat [deg]
29.2
29
28.8
28.6
28.4
28.2
-92.4
-92.6
5
10
-92.2
15
-92
20
-91.8
25
Long [deg]
-91.6
30
-91.4
35
Salinity Range 0-40 psu all figures
-91.2
40
STARRS Salinity, Afternoon flight
10 May 2007
29.6
29.4
Lat [deg]
29.2
29
28.8
28.6
28.4
28.2
-92.6
-92.4
5
10
-92.2
15
-92
20
-91.8
25
Long [deg]
-91.6
30
-91.4
35
-91.2
40
Shipboard underway Salinity,
STARRS Salinity, Afternoon flight
10 May 2007
29.6
29.4
Lat [deg]
29.2
29
28.8
28.6
28.4
28.2
-92.6 -92.4
5
10
-92.2
-92
-91.8 -91.6 -91.4
Long [deg]
15
20
25
30
35
-91.2
40
Ship (green) and aircraft (blue) Salinity
Afternoon Flight, 10 May 2007
40
35
Salinity (psu)
30
25
20
15
10
5
0
0
20
40
60
80
100
120
140
Distance (km) from Outbound Endpoint
160
Salinity Regression vs STARRS Salinity
Morning flight, 10 May 2007, outbound leg
40
Sal=c+a(ch5/ch2)+b(ch5/ch6)
Regresion Salinity (psu)
35
30
25
20
15
10
5
0
0
10
20
30
STARRS Salinity (psu)
40
Regression salinity, morning flight, 10 May 2007
29.6
29.4
Lat [deg]
29.2
29
28.8
28.6
28.4
28.2
-92.6
5
-92.4
10
-92.2
-92
-91.8
Long [deg]
15
20
25
-91.6
30
-91.4
35
-91.2
40
Regression salinity, afternoon flight, 10 May 2007
29.6
29.4
Lat [deg]
29.2
29
28.8
28.6
28.4
28.2
-92.6
5
-92.4
10
-92.2
-92
15
20
-91.8
25
Long [deg]
-91.6
30
-91.4
35
-91.2
40
VJ Maisonet: Student Project on
Optics and Salinity
Overview
•
•
•
•
•
•
•
Introduction
Equipment
Study Site
Algorithm used
Results
Summary
Current/Future work
Ocean Color Remote Sensing
• Light from the Sun
(irradiance ,Ed(λ))
penetrates , reacts
with the water with
a portion of the light
energy being
reflected back out
(water leaving
radiance ,Lu(λ))
Colored dissolved organic matter
• Colored dissolved organic
matter (CDOM) is the optically
measurable component of the
dissolved organic matter in
water.
• Naturally occurring substance
– When plant tissue
decomposes either in the soil
or in a body of water the
organic matter is broken down
by microbes
• The color of water will range
through green, yellow-green,
and brown as CDOM
increases
The right side of the figure is the Remote sensing
reflectance (Rrs(λ, 0-) = Lu(λ, 0-) / Ed(λ, 0-)) of CDOM,
where Lu is water leaving radiance and Ed is downwelling
irradiance.
Equipment
STARRS
Piper Navajo
IR-Band
C-Band
L-Band
OCR-507
STARRS/OCR Sampling Pattern
STARRS
Sampling Rates:
STARRS
OCR
OCR:STARRS
Ocean (Ts, Tb)
C, IR &
Vis-Band
SeaWiFS Chs.
L-Band
Incidence
Angles:
+/- 7,22,37 (deg)
Flight
Direction
~2.0 s
~.17 s
~ 11:1
NEDT(1s):
0.50 K 
dS=1 psu
Alt. 2600 m
Area of Study
Sampling Flights
Algorithm
• For ease of computation an empirical algorithm for
CDOM from D’Sa et al. 2006 was used.
– Their study was conducted in the same region and time of year
– Their study was preformed with similar optical equipment
• Below is the algorithm they developed:
– Acdom (412) = 0.227 x (Rrs510/Rrs555)-2.022
Results
R2 Value= 0.76
n= 5220
Results cont.
R2 Value= 0.90
N=1100
Results cont.
• Using the regression analysis from the morning flight
combined with the CDOM algorithm to create the
following:
• Salinity= 0.227 (Rrs 510/Rrs 555)-2.022 – 0.34
-.0082
• This Salinity model was then applied to the
afternoon flight for verification
Results cont.
R2 Value= 0.88
Summary
• This study resulted in a Ocean Color-Salinity model that can
measure with ~88% accuracy the Sea-Surface Salinity of the
Louisiana shelf
• These results come with a few caveats:
– This study is a seasonal model not a annual model
– This model is only effective in the near Coastal zone
• This model assumes :
– Photo-degradation is low in the near coastal waters
– That CDOM is behaving conservatively
Current/Future Work
•
•
•
In late 2009 early 2010 NASA will
deploy Aquarius
– L-Band Radiometer
– 100 km Resolution
Currently we are in the process of
applying the Ocean Color-Salinity
Algorithm to SeaWiFS & Modis A
for a broader view of the coastal
zone
Our next step is to develop a
‘smart’ algorithm to interpolate
between the CDOM-Salinity and
the Aquarius-Salinity
– In hopes to fill in the gaps left
by the satellite to assemble a
‘whole’ picture
Acknowledgments
Funding Agency: NASA/Mississippi Research Consortium
Project Contract Number: NNS06AA98B
Title: Simulating the Wind and Sea-Surface Roughness Effect on
Aquarius Sea Surface Salinity Retrievals: Evaluating Alternative
Models to Correct for the Effects of The Rough Sea Surface on Lband Radiometer Emission and Scattering
New Developments, Spinoffs, Publications
• Better understanding of how E-M radiation interacts with the rough
sea surface – leading to NRL New Start.
• New techniques for comparing and selecting wind/wave spectra
and roughness models for more accurate microwave open ocean
remote sensing of SSS – NRL New Start.
• New parameters and algorithms for retrieving SSS from optical
remote sensing data in Gulf of Mexico coastal seas.
• Advanced preparations for accurate retrieval of SSS from SMOS
and Aquarius satellite-borne L-band radiometers.
Sea Surface Roughness Impacts on
Microwave Sea Surface Salinity
Measurements (SRIMS)
ESA
SMOS
Derek Burrage, David Wang, Joel Wesson
(SSC) and Paul Hwang (DC)
NASA
Aquarius
Hypothesis: Small-scale roughness components generated by diverse
physical processes including wind, swell, breaking waves and foam
dominate microwave sea surface emission and scattering, and thus sea
surface salinity (SSS) retrieval accuracy.
Goal: Advance understanding of physical processes governing sea
surface roughness (SSR) and its interaction with electromagnetic (E-M)
radiation, to enhance salinity remote sensing using L-band radiometers.
Payoff: More accurate global sea surface salinities for input to navy
ocean circulation models and data assimilation systems.
NRL New Start 6.1 (FY 2010-12)
Conference Papers
Burrage, D., J. Wesson, D. Wang, and S. Howden (2007). Airborne Passive Microwave
Measurements of Sea Surface Salinity, Temperature and Roughness, and Implications for
Satellite Salinity Retrieval. IEEE Geoscience and Remote Sensing Society (IGARSS)
2007, Barcelona, Spain, 23-27 July, (Poster Paper).
Burrage, D., J. C. Wesson, D. W. Wang, S. D. Howden, and N. Reul (2008). Sea Surface
Roughness Influence on Salinities Observed with an Airborne L-Band Microwave
Radiometer: Model Inter-Comparisons, Validation and Implications for Satellite Salinity
Retrieval. IEEE Geoscience and Remote Sensing Society (IGARSS), Boston,
MA.(Poster Paper), July 7-11.
Wesson, J., D. Burrage, C. Osburn, V.J. Maisonet, S. Howden, and X. Chen (2008).
Aircraft and In Situ Salinity and Ocean Color Measurements and Comparisons in the
Gulf of Mexico. IGARSS, Boston, MA (GRSS 2008 IGARSS IEEE Int’l Vol 4, pp.383386).
Maisonet, V. J., J. Wesson, C. Osburn, D. Burrage and S. Howden (2009) Using Ocean
Color to Measure Coastal Sea-Surface Salinity of the Louisiana Shelf. Virgilio (Oral
presentation) Mississippi Academy of Sciences (MAS) annual meeting, Feb. 26-27,
2009, Olive Branch, MS. (Published abstract: Journal of the Mississippi Academy of
Sciences, 54, 1, 83-84., Outstanding Oral Presentation Award in Division of Marine and
Reports
Atmospheric Science.
Howden, S., D. Burrage, J. Wesson and D. Ko. Simulating Wnd and Sea-Surface
Roughness Effects on Aquarius Retrievals. (First progress Report submitted to NASA
Applied Sciences Program and Mississippi Research Consortium, July 2007)
Refereed Papers by Team Members
(Arising from related projects)
Burrage, D. M, J. Wesson and J. Miller (2008), Deriving Sea Surface Salinity and Density
Variations from Microwave Radiometer Measurements: Application to Coastal River
Plumes using STARRS, Transactions on Geoscience and Remote Sensing, SMOS Special
Issue, 46, 3, 765-785.
Burrage D. M., J. Wesson, M. A. Goodberlet and J. L. Miller (2008). Optimizing
performance of a microwave salinity mapper: STARRS L-band radiometer enhancements,
J. Atm. & Oc. Tech. 25, 776-793.
Burrage, D., J. Wesson, C. Martínez, T. Perez, O. Moller, Jr. and A.Piola (2008). Patos
Lagoon outflow within the Rio de la Plata plume using an Airborne Salinity Mapper:
Observing an embedded plume, Cont. Shelf Res. PLATA project special issue, 28, 16251638.
Gabarro, C., J. Font, J. Miller, A. Camps, J. Wesson, D. Burrage and A. Piola (2008) Use of
empirical sea surface emissivity models to determine sea surface salinity from an airborne
L-band radiometer, Scientia Marina, June, 72, 2, 329-336.
Jerry L. Miller, David W. Wang, Paul A. Hwang and Derek M. Burrage (2007) Small-scale
Rogue Waves in the Ocean (In Revision).
Final Steps
• Execute roughness field campaign off Chesapeake
Bay (Virgo II) in late 2009, if possible coinciding with
SMOS over flights (Piggyback with NRL 6.1 project).
• Continue development of Rigorous Reference
model and L-band Scatterometer Simulation (NRL
6.1 Project).
• Complete roughness model evaluation and
selection process.
• Finalize papers on roughness and optical SSS
retrieval.
• Compile and submit final report.