O g nS te re o ta

UNIVERSITY

Hyperspectral Imager for the Coastal Ocean (HICO): Overview and Coastal Ocean Applications

O g nS te re o ta
UNIVERSITY

Introduction and Outline

Why Hyperspectral Imaging The Hyperspectral Imager for the Coastal Ocean (HICO) How it came to be The challenge of operating on the International Space Station (ISS) What we can see with HICO? Collecting and processing HICO data Comparison to other ocean color data Scientists and scenes from around the World Access to HICO data via CEOAS HICO website http://hico.coas.oregonstate.edu

O g nS te re o ta
UNIVERSITY

Optical Components of a Coastal Scene

Multiple light paths Scattering due to: atmosphere aerosols water surface suspended particles bottom Absorption due to: atmosphere aerosols suspended particles dissolved matter Scattering and absorption are convolved

Physical and biological modeling of the scene is often required to analyze the hyperspectral image.

O g nS te re o ta
UNIVERSITY

Introduction to Hyperspectral Imaging

A hyperspectral imager records a spectrum of the light from each pixel in the scene Hyperspectral image analysis exploits this extra spectral information For an open land scene: Total spectrum for a pixel is a Hyperspectral imager weighted sum of the spectra of what is in that pixel
Spectral Decomposition for an open land scene
k

Grass

+
Dry Soil

+
Leaves

S= i E i N
i=1

Spectrum for each pixel

+
Camouflage

The imager and method of exploitation must be tailored to the scene and the desired products.

O g nS te re o ta
UNIVERSITY

HICO based on experience with PHILLS

PHILLS image of shallow water features near Lee Stocking Island, Bahamas used to develop and validate hyperspectral algorithms for bathymetry, bottom type and water clarity.

AN-2 Aircraft

Airborne Experiments with the Portable Hyperspectral Imager for Low-Light Spectroscopy (PHILLS) demonstrated: Sensor design. Processing algorithms. Shallow water bathymetry, hazards to navigation, and beach trafficability from hyperspectral remote sensing data.

PHILLS Sensor

O g nS te re o ta
UNIVERSITY

NRL Airborne Coastal Environmental Hyperspectral Program

Sensor Performance Modeling
Signal to Noise Ratio 10 nm Spectral Bins
350 300 f/2 f/2.8
GSD = 100m Albedo = 5% GMC = 1

20 years end-to-end development of airborne coastal environmental hyperspectral imaging

Signal to Noise

250 200 150 100 50 0 400

f/4

500

600

700

800

900

1000

Wavelength (nm)

Sensor Development Requirements Evaluation Sensor Calibration Spiral Development Flight Campaigns

Nonlinear Manifold Analysis

PURSUIT Pattern Recognition / Classification

Product Evaluation

Product Extraction

ORASIS Spectral Identification

Data Processing

Ground / Water Truth Georectification

300

250

200

TAFKAA Atmospheric Removal Algorithm

ectance X 104

150

O g nS te re o ta
UNIVERSITY

What is the Hyperspectral Imager for the Coastal Ocean (HICO)?

HICO is an experiment to see what we gain by imaging the coastal ocean at higher resolution from space. The HICO sensor: first spaceborne imaging spectrometer for coastal oceans samples coastal regions at <100 m (400 to 900 nm: at 5.7 nm) high signal-to-noise ratio to resolve the complexity of the coastal ocean Sponsored as an Innovative Naval Prototype (INP) by the Office of Naval Research: Goal to reduce cost and a greatly shortened schedule. Start of Project to Sensor Delivery in 16 months Launched to the ISS September 10, 2009
HICO image of Hong Kong, October 2, 2009. HICO is integrated and flown under the direction of DoDs Space Test Program

O g nS te re o ta
UNIVERSITY

HICO Flight Sensor - Stowed position

spectrometer camera

lens

View port

O g nS te re o ta
UNIVERSITY

HICO meets Performance Requirements

Parameter Spectral Range Performance 380 to 960 nm 5.7 nm 102 Rationale All water-penetrating wavelengths plus Near Infrared for atmospheric correction Sufficient to resolve spectral features Derived from Spectral Range and Spectral Channel Width

Spectral Channel Width Number of Spectral Channels Signal-to-Noise Ratio for water-penetrating wavelengths Polarization Sensitivity Ground Sample Distance at Nadir Scene Size Cross-track pointing Scenes per orbit

> 200 to 1 Provides adequate Signal to Noise Ratio for 5% albedo scene after atmospheric removal (10 nm spectral binning) < 5% (430-1000 nm) 92 meters 42 x 192 km +45 to -30 deg 1 maximum Sensor response to be insensitive to polarization of light from scene Adequate for scale of selected coastal ocean features Large enough to capture the scale of coastal dynamics To increase scene access frequency Data volume and transmission constraints

O g nS te re o ta
UNIVERSITY

HICO Installed on the ISS on September 24, 2009

HICO

Japanese Module Exposed Facility

O g nS te re o ta
UNIVERSITY

HICO docked at ISS Now What?

HICO Viewing Slit

O g nS te re o ta
UNIVERSITY

HICO Image Locations

- Currently ~300 locations identified - New sites can be added which may mean fewer observations of each site due to competition between sites

Locations chosen based on: 1. Location within latitude limits of ISS orbit 2. Type ocean, coast, land 3. Use CalVal, Science, Navy, etc

O g nS te re o ta
UNIVERSITY

HICO Processing
Level 0 Level 01a Navigation Level 1bCalibration

Vicarious Calibration

Multispectral

Level 1b : Calibrated data

Hyperspectral

Level 1c Modeled Sensor bands MODIS MERIS OCM SeaWIFS

Level 2a: Sunglint Level 2c: Standard APS Multispectral Algorithms Products
QAA, Products At, adg, Bb, b. CHL (12) NASA: standards OC3, OC4, etc (9) Navy Products Diver Visibility
Laser performance

Level 2b TAFKAA Atmospheric Correction

Level 2f: Cloud and Shadow Atm Correction

Level 2c- : Hyperspectral L2genAtm Correction

Atmospheric Correction Methods

Level 2d: Hyperspectral Algorithm Derived Product Hyperspectral QAA

At, adg, Bb, b. CHL (12)

K532 Etc (6)

HOPE Optimization
(bathy, optics, chl, CDOM ,At, bb ..etc

CWST - LUT Bathy, Water Optics Chl, CDOM

Coastal Ocean Products Methods

Level 3: Remapping Data and Creating Browse Images

O g nS te re o ta
UNIVERSITY

Radiometric Comparison of HICO to MODIS (Aqua)

Nearly coincident HICO and MODIS images of turbid ocean off Shanghai, China demonstrates that HICO is well-calibrated

HICO Date: 18 January 2010 Time: 04:40:35 UTC Solar zenith angle: 53 Pixel size: 95 m
East China Sea off Shanghai

MODIS (Aqua) Date: 18 January 2010 Time: 05:00:00 UTC Solar zenith angle: 52 Pixel size: 1000 m
Top-Of-Atmosphere Spectral Radiance
120 HICO 100 MODIS 80 60 40

Image location

50 km

20 0

R.-R. Li, NRL

0.4

0.5

0.6

0.7

0.8

0.9

Wavelength (microns)

O g nS te re o ta
UNIVERSITY

Chlorophyll Comparison of HICO to MODIS (Aqua)

Nearly coincident MODIS and HICOTM images of the Yangtze River, China taken on January 18, 2010. Left, MODIS image (0500 GMT) of Chlorophyll-a Concentration (mg/m3) standard product from GSFC. The box indicates the location of the HICO image relative to the MODIS image. Right, HICOTM image (0440 GMT) of Chlorophyll-a Concentration (mg/m3) from HICOTM data using ATREM atmospheric correction and a standard chlorophyll algorithm. (Preliminary Results by R-R Li and B-C Gao.)

O g nS te re o ta
UNIVERSITY

Andros Island, Bahamas, Oct 22, 2009

RGB image

bathymetry

absorption

O g nS te re o ta
UNIVERSITY

Internal waves at the Straits of Gibraltar

Generation of the Internal Wave

Camarinal Sill And the Tidal Boar. Internal Wave packets

Straits of Gibraltar HICO Image Dec 5, 2009 (R. Arnone analysis)

O g nS te re o ta
UNIVERSITY

Microcystis bloom in Lake Erie

HICO Image of a massive Microcystis bloom in western Lake Erie, September 3, 2011 as confirmed by

O g nS te re o ta
UNIVERSITY

Birth of a New Island, Canary Islands

HICO Image of the new underwater volcano off the small Canary Island of El Hierro, December 22, 2011.

O g nS te re o ta
UNIVERSITY

HICO Data Distribution at OSU

Developed HICO Public Website at OSU using published and approved for distribution data, publications and presentations. Includes some example HICO data that are approved for distribution. OSU HICO Web site will be portal for data requests and distribution Data requests require a short proposal and data agreement signed by the requestor and their institution and approved by NRL. Example data and data requested by that user will be available to them. http://hico.coas.oregonstate.edu Full description of the data and directions for use on the website

O g nS te re o ta
UNIVERSITY

Current HICO users

Besides the work ongoing at NRL and OSU and their collaborators a number of US and international Scientists have submitted proposals and been approved to work with HICO data. Projects are listed under the Current Project Tab on the HICO website.

O g nS te re o ta
UNIVERSITY

The HICO Team

NRL SSC Bob Arnone Rick Gould Paul Martinolich Will Hou David Lewis Ronnie Vaughn Theresa Scardino Adam Lawson Alan Weidemann Ruhul Amin Academic Curt Davis, OSU, Project Scientist Jasmine Nahorniak, OSU Nick Tufillaro, OSU Curt Vandetta, OSU Ricardo Letelier, OSU Zhong-Ping Lee, U. Mass Boston Industry John Fisher, Brandywine Photonics

NRL DC Michael Corson, PI Robert Lucke, Lead Engineer Bo-Cai Gao Charles Bachmann Ellen Bennert Karen Patterson Dan Korwan Marcos Montes Robert Fusina Rong-Rong Li William Snyder

Special thanks to our sponsors the Office of Naval Research, the Space Test Program, and to NASA and JAXA who made this program possible.

O g nS te re o ta
UNIVERSITY

HICO Summary (HICO Docked on the Space Station)

Japanese Exposed Facility

HICO

Built and launched in 28 months Over 5000 scenes so far One more year of operations Data from OSU HICO website

O g nS te re o ta
UNIVERSITY

Are we there Yet? Need beyond HICO

A new Free flying Hyperspectral Imager (CWR) would offer several key advantages over HICO: Broader bandwidth. CWR will collect data from 380 to 1650 nm. The longer infrared wavelengths will make it possible to characterize beaches and plants along the shoreline and snow and ice in alpine regions and in the Arctic. Finer spatial resolution. CWR will collect data with 30-m GSD, which will be able to re-solve ocean bottom, coastal, glacier and arctic features. Optimized orbit. CWR will fly on a smallsat in a near-noon equatorial crossing time polar orbit. It will be able to image locations in high latitudes every day and locations in mid to low latitudes every 2nd or 3rd day. Wide field of regard. Agile spacecraft to rotate CWR through 60 degrees, from -30 to +30 from nadir. Allows rapid revisit (1-3 days) to study sites. Optimized radiometric calibration. CWR will calibrate signal brightness on-orbit by scanning the full moon once a month. Optimized spectral calibration. CWR will calibrate signal wavelengths on-orbit to within 0.2 nm by comparing observed and known spectral features including Fraunhofer absorption lines and atmospheric absorption features. Complete data processing system. HICO data processed to level 1B, need routine full processing to geolocated products, including reprocessing as needed.