iss hyperspectral imager for the coastal ocean (hico ... · 16.07.2013 · for 5% albedo scene (10...

Office of Research and DevelopmentFull Name of Lab, Center, Office, Division or Staff goes here. <Go to View, Master, Slide Master to change>Office of Research and DevelopmentNational Health and Environmental Effects Research Laboratory/ Atlantic Ecology Division July 16, 2013

EPA/ORD/NHEERL/NERL

Naval Research Laboratory/StennisSpace Center

ISS Hyperspectral Imager for the CoastalOcean (HICO): Application of Space-basedHyperspectral Imagery for the Protection of

the Nation’s Coastal Resources

Jessica Aukamp USEPA/NHEERL/Gulf Ecology DivisionDonald Cobb USEPA/NHEERL/Atlantic Ecology DivisionRobyn Conmy USEPA/National Risk Management

Research Laboratory/Land Remediationand Pollution Control

George Craven USEPA/NHEERL/Gulf Ecology DivisionRichard Gould Naval Research Laboratory/ Bio-optical

Physical Processes and Remote SensingSection

James Hagy USEPA/NHEERL/Gulf Ecology DivisionDarryl Keith* USEPA/NHEERL/Atlantic Ecology DivisionJohn Lehrter USEPA/NHEERL/Gulf Ecology DivisionRoss Lunetta USEPA/National Exposure Research

Laboratory/Environmental Services DivisionMichael Murrell USEPA/NHEERL/Gulf Ecology DivisionKenneth Rocha USEPA/NHEERL/Atlantic Ecology DivisionBlake Schaeffer USEPA/NHEERL/Gulf Ecology Division

*email: [email protected]

Office of Research and DevelopmentFull Name of Lab, Center, Office, Division or Staff goes here. <Go to View, Master, Slide Master to change>Office of Research and DevelopmentNational Health and Environmental Effects Research Laboratory/ Atlantic Ecology Division

1

The U.S. Environmental ProtectionAgency’s mandate to protect human healthand the environment requires innovativeand sustainable solutions for addressingthe Nation’s environmental problems.

HICO offers EPA and the environmentalmonitoring community an unprecedentedopportunity to observe changes in coastaland estuarine water quality across a rangeof spatial scales not possible with field-based monitoring.


2

What is ocean/estuary color?

Rrs(0+,λ) = Lw (0+,λ)/ Es(0+,λ)

Rrs = remotely sensed reflectance (1/sr)

Lw (0+, λ) = water leaving radiance measured above the air/water interface(W m-2 sr-1),

Es ((0+, λ) = downwelling irradiance measured above the air/water interface(W m-2 sr-1)

Definition of Remotely SensedReflectance (Rrs)

Remotely sensedreflectance

Remotely sensedreflectance


3

Use HICO imagery to develop a novel space-basedenvironmental monitoring system that providesinformation for the sustainable management ofcoastal ecosystems.

To demonstrate that water quality informationderived from HICO could be incorporated into aprototype smart phone application to disseminatedata to managers in the EPA Office of Water.

Program Objectives

Water quality news reports may change the social and economicdynamics for the Nation to not only be aware of its water qualityconditions but support sustainable practices to maintain orimprove conditions at their favorite recreational areas.


4

• HICO is intended as a pathfinder for follow-on sensors designed with betterresolution and for routine observations of coastal, riverine, and estuarinewaters (Lucke et al., 2011).

• Built on the legacy of the NRL Ocean Portable Hyperspectral Imager for Low-LightSpectroscopy (Ocean PHILLS) airborne imagers and funded as an InnovativeNaval Prototype by the Office of Naval Research

• January, 2007: HICO selected to fly on the International Space Station (ISS)

• November, 2007: Construction began following the Critical Design Review

• September, 2008: Space-qualified instrument delivered to the DOD Space TestProgram for integration and spacecraft-level testing

• April, 2009: Shipped to Japan Aerospace Exploration Agency (JAXA) for launch

• September 10, 2009: HICO launched on JAXA H-II Transfer Vehicle (HTV)

• September 24, 2009: HICO installed on ISS Japanese Module Exposed Facility

Hyperspectral Imager for the CoastalOcean (HICO).... Intent and a bit of history


5HICO Viewing Slit

spectrometer

Hyperspectral Imager for the Coastal Ocean (HICO)

1st spaceborne imaging spectrometer specifically made for theenvironmental characterization of the coastal ocean from space

HICO is installed in the HICO-RAIDS experiment payload(HREP) on the Japanese Experiment Module - Exposed Facility

camera


6

To increase scene access frequency+45 to -30 degCross-track pointing

Data volume and transmission constraints1 maximumScenes per orbit

Large enough to capture the scale ofcoastal dynamics

Adequate for scale of selected coastalocean features

Sensor response to be insensitive topolarization of light from scene

Provides adequate Signal to Noise Ratioafter atmospheric removal

Derived from Spectral Range andSpectral Channel Width

Sufficient to resolve spectral features

All water-penetrating wavelengths plusNear Infrared for atmospheric correction

Rationale

~100Number of Spectral

Channels

~50 x 200 kmScene Size

~100 metersGround Sample

Distance at Nadir

< 5%Polarization Sensitivity

> 200 to 1for 5% albedo scene

(10 nm spectral binning)

Signal-to-Noise Ratiofor water-penetrating

wavelengths

5.7 nmSpectral Channel Width

380 to 960 nmSpectral Range

PerformanceParameter

Arnone et al., (2009) HICO for NASA Gulf Workshop

HICO Tech Specs


7

Program Implementation

Collected 49 images fromfour estuaries along NWFlorida during April 2010 toMay 2012 during ISSExpeditions 24 - 31

Pensacola, Choctawhatchee,and St. Andrew Bays areshallow (~4.0 m), microtidal(tidal range ~ 1.0 m) brackishwater estuaries. St. JosephBay is slightly deeper ( ~8 m)and not influenced by theinflow of fresh water.

Pensacola Bay ChoctowhatcheeBay

St. AndrewBay

St.Joseph

Bay


Field Validation Program

Pensacola Bay

Choctawhatchee Bay

St. Joseph Bay

St. Andrew Bay

Samplestations

Green = waterquality moorings(chl a, turbidity,

and CDOM)

Yellow = abovewater radiance,

temp, salinity, andoptical properties

Red = same asyellow + discretewater samples


Using small boat surveys within each estuary, collected andprocessed water column profile and above-water

hyperspectral data (Rrs) and water quality samples (Chl a,TSS, CDOM, salinity)

Field Validation Program – Part 1

Attempted to conduct sampling within 1-3 days of an ISS overflight

Laboratory analysisAbove-water radiometryWater column profiling


10

Field Validation Program – Part 2

Autonomous underwater vehicles (AUVs) were deployed concurrently with HICOoverpasses by the NRL Stennis Space Center Detachment (NRL/SSC) and the USEPA

Atlantic Ecology Division (AED)

NRL/SSC deployed a Slocum electricglider off of Pensacola Bay along the 15-30 m bathymetric contours whichrecorded:

temperaturesalinity (conductivity)

chlorophyll and CDOM florescence

SlocumG2 Glider

Designed andmanufactured

by Webb Research

AED deployed a REMUS (RemoteEnvironmental Monitoring Unit) AUV inPensacola and Choctowhatchee Bayswhich recorded:

temperaturesalinity (conductivity)

chlorophyll and turbidityat approximately one meter depth

Photo courtesy of Kongsberg Maritime


11

HICO Image Processing: Optical Components of aCoastal Scene

Multiple light paths

• Scattering due to:– atmosphere– aerosols– water surface– suspended particles– bottom

• Absorption due to:– atmosphere– aerosols– suspended particles– dissolved matter


12

HICO Image Processing Steps

Processing done with Excelis VIS ENVI 4.7 software

Open HICO Level 1b Image with calibrated TOAat-sensor radiances

Recover true at-sensor radiance valuesStep 2

Step 7

Steps 8 -9

Create GLT + Geolocate + WarpImages

Step 10Clip estuary shapefile to Warped

image

Step 11Export clipped Rrs data to EXCEL

CDOMTurbidity

TSM Chl a

Step 3

Step 1

Sun glint correction

Cloudremoval (ifneeded)

Step 5 Convert from W/m2/um/sr touW/cm2/nm/sr

Step 4

Step 6

Offshore pixel subtraction

Convert at-sensor Reflectance toRemote sensing Reflectance

Convert at-sensor Radiance toReflectance

Determine MeanSolar Irradiance

(F0)

Determine diffusetransmittance from

sun to pixel (t0)


Spectral match-ups between HyperSAS (dottedline) and HICO spectral signature (solid line).

iss2012016.0116.193730.L1B_PensacolaBay_des.v03.9170.20120117211554.100m.hico.bil

Step 1: Level 1b calibrated radiance Step 4:Offshore pixel subtraction (W/m2/um/sr)

Step 7: Remotely sensed Reflectance (1/sr)

0.000

0.002

0.004

0.006

0.008

0.010

405

428

451

474

497

520

543

565

588

611

634

657

680

703

726

Rrs

(1/s

r)

Wavelength (nm)

PB 18: Redfish Cove


14

0

0.001

0.002

0.003

0.004

0.005

0.006

0.007

0.008

0.009

40

543

947

450

854

357

761

164

668

071

4

Rrs

(sr-

1)

Wavelength (nm)

PB06: June 2, 2011

0

0.001

0.002

0.003

0.004

0.005

0.006

0.007

0.008

0.009

0.01

40

5

43

9

47

4

50

8

54

3

57

7

61

1

64

6

68

0

71

4

Rrs

(sr-

1)

Wavelength (nm)

PB 16: June 2, 2011

0

0.001

0.002

0.003

0.004

0.005

0.006

0.007

0.008

40

543

947

450

854

357

761

164

668

071

4

Rrs

(sr-

1)

Wavelength (nm)

PB 15: June 2, 2011

0

0.001

0.002

0.003

0.004

0.005

0.006

0.007

0.008

0.009

0.01

40

5

43

9

47

4

50

8

54

3

57

7

61

1

64

6

68

0

71

4

Rrs

(sr-

1)

Wavelength (nm)

PB04: June 2, 2011

HICO HYPERSAS

PB04

PB06

PB16

PB15

Spectral Match-ups

Pensacola Bay, FL


15

0

2

4

6

8

10

-0.200 -0.100 0.000 0.100 0.200HICO[ (1/Rrs674-1/Rrs703)*Rrs726]

R2 = 0.68

0

2

4

6

8

10

12

0 2 4 6 8 10 12

Measu

red

ch

la

(ug

/L)

HICO predicted chl a (ug/L)

R2 = 0.82RMSE = 1.7 ug/L

Chl a = 19.264 X [a] + 6.614a = [1/Rrs(674) – 1/Rrs(703)]x Rrs(726)

approach of Gitelson et al., 2011

Chl a: indicator of phytoplankton abundance and eutrophication


16

0.0

0.5

1.0

1.5

2.0

2.5

3.0

0.0 1.0 2.0 3.0

CT

DN

TU

HICO NTU

R2 = 0.67RSME = 0.56 NTU

n = 14

Turbidity = 2*106 X [Rrs(646)2.7848]

R2 = 0.72n = 18

0.10

1.00

10.00

0.003 0.005 0.007 0.009

Tu

rbid

ity

(NT

U)

HICO Rrs(646)

approach of Chen et al., 2007

Turbidity is an indicator of the distribution of totalsuspended matter in estuaries


17

Colored Dissolved Organic Matter Absorption: indicator of lightattenuation in estuaries

R² = 0.93RMSE = 0.21 /m

n = 180.0

0.5

1.0

1.5

2.0

2.5

3.0

0.0 1.0 2.0 3.0 4.0

Meas.C

DO

Mab

so

rpti

on

HICO CDOM absorption

R² = 0.60n = 17

0.0

0.5

1.0

1.5

2.0

2.5

0.0 1.0 2.0 3.0

ag

(412)

(1/m

)

HICO Rrs(670)/Rrs(490)

ag412 (m-1) = 0.8426 x [Rrs(670)/Rrs(490)] – 0.032

approach of Bowers et al., 2004


HICO/MODIS ComparisonMODIS scene 19:00 GMT, 7 minutes before HICOoverpass

Adj R2 = 0.83

HICO scene

Pixelscommon tobothimages andused in thematch-up

October 28,2011

HICO andMODIS

image date

Dashed line =HyperSAS Rrs

Dotted line = HICORrs at MODISbandwidths

1:1


19

Using atmospherically corrected HICO imagery and a comprehensivefield validation program, regionally-tuned algorithms were developedto estimate the spatial distribution of chlorophyll a, colored dissolvedorganic matter, and turbidity for four estuaries along the northwestcoast of Florida from April 2010 – May 2012.

Program Summary

The HICO-derived water quality data fromthis project have been uploaded to ainternal EPA HICO website for review anduse by the EPA Office of Water and aprototype mobile application has beencompleted.


20

Conclusions

HICO helped us to show that it is possible for a hyperspectral space-basedsensor to produce products that meet the needs of EPA.

While the potential benefits are many, there are several issues that must beresolved before HICO images and data can be incorporated into routinemonitoring programs of EPA

• HICO is currently limited to only one image per orbit.

• ISS overpass times are difficult to precisely predict. Theseuncertainties led to the rescheduling of planned image acquisitions which attimes impacted the effective deployment of crews for field validation activities.

HICO has the potential to be a valuable monitoring tool, if transitioned to aconstellation of sensors.


21

Special Thanks to:

Crews on ISS Expeditions 24 - 31Naval Research Laboratory Remote Sensing Division

Oregon State UniversityNASA ISS Program

American Astronautical Societyand

EPA Office of Research and DevelopmentEPA Pathfinder Innovation Program (Grant 2011)

EPA Safe and Sustainable Waters Research Program

iss hyperspectral imager for the coastal ocean (hico ... · 16.07.2013 · for 5% albedo scene (10...

Documents