Hyperion
  

Ordering EO1 Hyperion and ALI Data
If you would like to acquire EO1 Hyperion and/or ALI data for your site or would like to purchase archived data please go to:

Detailed description of EO1sensors and data products, price lists and online ordering forms are available there.


ACRES  
NASA  
CSIRO  


Evaluation and Validation of the

NASA Hyperion Hyperspectral Satellite Sensor

 

Page Links

Title
Abstract
Technical Plan
    Introduction
Description of Research Approach
    Proposed Research Sites
        Calibration Validation Sites
           Tinga Tingana
           Uardry
           Lake Argyle
           Kioloa
         Application Sites
           Kunoth
           Mt Fitton
           Panorama
           Cape Tribulation
Instrumentation and Airborne Data
Site Objectives and Outcomes
    EO-1 Data Level Requirements
    Data Plan
References

Data Acquisition Gantt

.             . 

 Hyperion Study Site Links

*  = Hyperion validation site

 

 

Site Meterological Data 

Misc. Australian Maps 



Title

Evaluation of Hyperion Performance at Australian Hyperspectral Calibration and Validation Sites
(NRA-99-0ES-01)

Abstract

CSIRO (Australia) has a range of active programs in earth observation and in hyperspectral research leading to high quality inputs to many integrated applications in Australia, the USA and the Asia-Pacific region. These include large area land rehabilitation and monitoring as well as resources inventory and management. The CSIRO Office of Space Science & Applications (COSSA) seeks to coordinate and support these activities across Divisions in CSIRO. The Australian Centre for Remote Sensing (ACRES) is an accredited data reception and distribution organization with two (potentially three) fully operational X-Band stations and involvement in the Landsat-7 series. The CSIRO and ACRES are partners in the proposed ARIES-1 spaceborne hyperspectral system. In this proposal, an Australian Team coordinated by CSIRO Earth Observation Centre (EOC) (refer Appendix 1) and ACRES proposes to use existing Australian calibration and validation sites and specific existing hyperspectral test sites with range of existing and ongoing data collection and analysis programs to evaluate Hyperion instrument and product performance. The Team proposes to monitor Hyperionâs performance against benchmarks of ground and airborne data at Australian sites and support Data Collection Events (DCEs) by providing capacity to download current data holdings or Australian data using the Australian X-Band stations that are operated by ACRES. Reception stations at Alice Springs in central Australia and Hobart in the very south of Australia are currently able to carry out this activity. These stations will be involved in collecting Landsat-7 data. It is possible a third station in the West of Australia will be established during the lifetime of the EO-1 mission. If the EO-1 mission is launched during the northern hemisphere winter as planned, the Australian Team has offered its expertise and sites to engage in early mission proving and vicarious assessment of the EO-1, especially Hyperion and to provide capacity to unload stored data from previous DCEs. The CSIRO hyperspectral research program has an established set of sites and a growing base of environmental information allowing spectral modelling and measurement. The Australian HyMap airborne system (Cocks et al., 1998 and Appendix 3) is already being used and we are proposing to undertake a benchmarking and cross-calibration of HyMap as part of the activity. A new Australian hyperspectral instrument, OARS, will also be offered in support of this activity. We are also proposing to engage in a joint NASA/CSIRO AVIRIS mission to establish cross-calibration and performance synchronization between the Australian systems (HyMap and OARS plus ground data systems) and the NASA benchmark appropriate calibration sites in the US or (preferably) Australia. When the mission has become established we have proposed that specific DCEs focus on five key applications areas where product performance is the major issue. These are two geologically oriented sites, one rangelands site, one rain-forest and tropical coastal site and one temperate forest site. In all of these cases there are external resource or environmental drivers for the evaluation of the applications products and comparisons of performance in this between the current airborne data and the Hyperion space-based data.
 
 
 

Technical Plan

Introduction

In Australia there has been a considerable interest, investment and activity involving hyperspectral remote sensing for applications to resource and environmental areas. The Australian Team involved in this proposal is comprised of researchers from CSIRO and ACRES and will use existing Australian calibration and validation sites and specific existing hyperspectral test sites to evaluate Hyperion performance. The locations of these sites are provided in Figure 1 and listed in Table 1.

If the EO-1 mission is launched during the northern winter as planned, the Australian Team has offered to engage in early mission proving and vicarious assessment of the EO-1. This involves a selection of the sites listed as Summer Sites as December, 1999 and January 2000 are months of the Australian Summer and in Summer the north of Australia has a tropical monsoon based wet season.

The majority of the sites listed have been established in the CSIRO Hyperspectral Research Program. They provide a growing base of environmental and field spectral information allowing spectral modelling and measurement. The Australian HyMap airborne system (see Cocks et al., 1998, Appendix 3 or http://www.intspec.com/) is being used periodically as part of this program and we propose to undertake a benchmarking and cross-calibration of HyMap as part of the activity. A new Australian hyperspectral instrument, OARS, is also being offered in support of this activity. We propose that the Team evaluate Hyperion data against these benchmarks using the current base of data as well as by integration with Landsat-7, ALI and LEISA-AC data at the selected sites and times in the year. At two sites it is proposed to analyze the data in combination with ASTER data from the TERRA platform.

To allow flexible scheduling of the sites, we propose that data for the selected sites and specific missions be acquired in Australia by making full use of the X-Band network run by ACRES. Stations at Alice Springs in central Australia and Hobart in the very south of Australia are currently able to carry out this activity and are available for both Australian and previously recorded DCEs as a means of providing capacity to record the areas associated with this proposal or downloading data from other parts of the world to create capacity to record and store data from the Australian DCEs. Data would be send directly from Australia to NASA Goddard. These stations will be involved in collecting Landsat-7 data. It is possible a third station in the West of Australia will be established during the lifetime of the EO-1 mission. ACRES will ensure that at least one site (most likely Cape Tribulation) there is an opportunity to assess integration capabilities between Landsat 7, ALI and Hyperion.
 
 

 Table 1: CSIRO EOC Sites for Hyperion Missions (also refer EOC Hyperspectral Sites and associated links).

Site
State
Location
Type
CSIRO PI
Data
SA
29°00âS  139°50âE
Desert with regular dunes (calibration)
Dean Greatz
Ross Mitchell
DATM, CIMEL, Nephelometer, Aeronet site, ground data and BRDF characterization. Cross-sensor studies with AVHRR/ASTER.
SA 
29°55âS  139°25âE 
Mining, Geology
Tom Cudahy
Jon Huntington
HyMap Missions conducted by ISPL. Major study by CSIRO, TIMS, ASTER, MIRACO2LAS, Landsat TM, SPOT, ARIES simulation, field spectra, 1:50,000 scale mapping
NSW
34°23â30''S 145°18â17"E
Grassland, Calibration
Fred Prata
Bob Cechet 
DATM, various satellites (eg AATSR) Tower based SRB and meteorological data, site BRDF measurements
NSW
35°35âS  150°18âE 
Coastal forestry and forest environment 
Nicholas Coops
Darius Culvenor 
Forestry Data, University Field Station,
Casi, HyMap, airborne Daedalus & Video. , see (www link). Winter/Spring Sites (Jul-Oct)
QLD
16°05âS  145° 20âE 
Land use, tropical forests, Daintree rainforests 
Alex Held
HyMap mission will occur in September 1999. CSIRO Field Station for logistics and support.
WA 
16°14.5âS  128°48âE
Man made lake as dark target, agricultural site
Ross Mitchell
Dean Graetz
Denis OâBrien
CIMEL data site (aeronet site) with local research station support.
WA
21°21âS  119°10âE 
Mining, Geology
Tom Cudahy
Jon Huntington
HyMap mission, TIMMS, various data types, planned ASTER site, ground data.
NT
23°32âS  133°50âE
Arid grazed rangeland with long term monitoring 
Vanessa Chewings
Geoff Pickup
HyMap, Video, meteorological, ground data and site monitoring.(www link

* = Hyperion validation sites.


 
 
 

Description of Research Approach

Proposed Study Sites


The approach taken in this proposal is to analyze Hyperion and LAC data at a number of sites and in the context of investigations being undertaken in Australia either for:

    1. Calibration and validation activities involving characterization of atmospheric properties, data alignment and surface radiance/reflectance during satellite missions; or

    2. Applied Hyperspectral research in Australia based currently on HyMap airborne data and (in some cases) casi airborne hyperspectral data as well as extensive ground based information.


The sites listed in Table 1 can, on this basis, be grouped into two major categories:
 

1. Calibration & Validation Sites

Tinga Tingana, together with Lake Argyle have been set up to provide information on Australian aerosols and have CIMEL instruments located and running to acquire long-term data (see Mitchell et al., 1996, 1997; OâBrien et al., 1998). These data are part of Aeronet. Daedalus and other airborne as well as some ground based data have been taken to estimate the BRDF of the site and its use in AVHRR calibration and AVHRR/ATSR cross-calibration studies is documented in OâBrien and Mitchell (1999). A background aeolian dune system is at a scale that will be resolved by Hyperion allowing geometric studies as well as spectral calibration studies. If this site is accepted it will be characterized spectrally using HyMap as well as ground based measurements over the full 0.4-2.5 ?m range in late 1999. It is not easy to access.

Uardry is a grassland site which has been used over five years as a surface radiation balance site as reported by Prata et al. (1995, 1996, 1999). It has a permanent tower and continuous readings of solar radiation, albedo and meteorological data. It has a suite of instruments forming a standard site for the Australian Bureau of Meteorology. This site, together with another near Alice Springs and one in the Northwest of Australia near Broome, forms part of the CIGSN or Continental Integrated Ground Site Network and is part of the MODIS set of sites. However, since the Broome and Alice Springs sites may be affected by the northern rainy season only Hay has been selected for the early analysis.

Lake Argyle is a very large man-made lake in the Kimberley region of Australia. It is serving as a dark target to Tinga Tingana as a bright target and has a continuously reading CIMEL aerosol and atmosphere monitoring station in place as well as a Nephelometer. The use of this dark target will be made in August or September to avoid the summer rains and late winter fires. At Tinga Tingana and Lake Argyle it is proposed to use Hyperion and LEISA to investigate aerosols and atmospheric correction. If ths site is selected it will be characterized optically and flown by a Hyperspectral instrument (casi or HyMap) in late 1999. Lake Argyle is well served with an airstrip and research facilities.

The Kioloa State Forest is a patch of well mapped and monitored forest on the coast of NSW not far from Canberra. There is a field station there owned by the Australian National University. It has had an extensive set of data flown, including video, Daedalus and Hyperspectral casi and HyMap data.. It has a distinctive coastline with sharp cliffs and beach contrasts at one side offering some significant contrasts and geometric testing in addition to a stable vegetation signature. It can also be seen as an applications site for forest studies and providing a contrast to the tropical forests at Cape Tribulation.
 
 

2. Applications Sites

2.1 Kunoth

Kunoth Paddock is the site of long-term monitoring of semi-arid grassland and woodland areas. The area has been mapped and monitored using a variety of airborne and satellite data. The site was a major site for the Japanese GRNS projects. It has also been extensively mapped using airborne video as well as Daedalus data and Hyperspectral casi and HyMap data. The objective at Kunoth is to assess how well the products based on airborne data can be translated to the satellite platform.

2.2 Mt Fitton

Mt Fitton is a geological area with high spectral contrasts and applications to mineralogy and geochemistry. It has been extensively flown and measured spectrally and is an ASTER site for which it has been flown by TIMMS. Mt Fitton provides a very significant test for spectral resolution and integrity of the Hyperion in the SWIR. It is well mapped and has distinct features for precise geometric registration to allow the data comparison and integration with (eg) ASTER data. The patches of different spectral material in Mt Fitton occur in large units making it an ideal study site.

2.3 Panorama

Panorama provides a second geological test site. It is in an iron ore province and also has significant spectral contrasts and a wide range of data flown and measured during field campaigns taking place over a number of years. It is an ASTER site and has been characterized for thermal region emissivity. It has had HyMap data flown and has produced a range of useful products. The effect of the space platform and Hyperion sensitivity on these products will be a major outcome of the proposal.

Cape Tribulation in North Queensland provides a tropical rainforest site contrasted with various land uses. The run will take in the coastal areas with fringing mangroves and rainforests to the waterâs edge. HyMap data will be flown for this site in late 1999. At this time it has not been characterized for BRDF and end-member spectra. A CSIRO Research Station provides logistical support for investigations here as well as facilities for above canopy studies and measurements.
 
 
 

Instrumentation and Airborne Data


Missions to the listed sites have been supported by CSIRO and involve a range of either continuous or mission specific measurements. In many cases, use of airborne Video data, Daedalus data and aerial photography has led to a good base characterization of cover type, distribution and temporal changes. More detailed studies at some of the sites have resulted in BRDF characterization or spectral characterization.

In field missions involving HyMap airborne data, ground based spectra, meteorological data and spectral direct and diffuse irradiances as well as atmospheric profiles have been normally collected and used to help atmospherically correct and characterize the airborne data. Increasing investment in field logistical support and instrumentation is being pursued for the late 1999 site missions over Hyperspectral sites. The calibration sites have been variously characterized for surface radiation balance and aerosol properties. The Hay site has a 5 year record of continuous solar radiation and albedo measurements as well as being a thermal data validation site used for AVHRR, ATSR and as a primary site for the TERRA mission.

We propose to establish cross-calibration between HyMap and AVIRIS in 1999. The options for this task are to carry out the comparison in the US or to bring AVIRIS to Australia. The advantage of the second is that the comparison would be made at selected Australian sites. The Hyperion images will then be able to be directly related with AVIRIS testing and calibration studies in the US. The previous version of the current HyMap (the PROBE scanner) has been flown with AVIRIS and it is expected that very good comparisons of calibration and instrument performance will ensue.

Current processing of the site data involves standard processing of meteorological data and its use as input to base atmospheric models which are currently available such as Modtran or Atrem. The airborne data are corrected for atmosphere and geometry and the resulting reflectance data used to assess the site applications. Input and involvement of other Science Team members in evolving standards, alternative instrumentation and methods for site data analysis is welcome as are suggestions for comparison with higher levels of Hyperion processed data from the Goddard facility. For each mission, all base data are made available and (in the case of the current Hyperspectral sites) made available on CD ROM for other researchers. All data collected at the sites is open file.

To give an example of the data collection for specific missions, in a recent HyMap mission at Kunoth the base data collected were:
 

  • Weather station data (solar radiation, air temperature, humidity, wind speed/direction)

  • Radiosonde data

  • Solar spectral diffuse and direct solar radiation

  • Ground target spectra (canvas sheets)

  • Natural target and background spectra (using CDI Vis/NIR spectroradiometer)


A shared ASD was not available for the site at that time but equipment purchases in train will ensure that full range spectra (0.4-2.5 ?m) will be available at every site for future missions. Mt Fitton and Panorama data currently include full range spectral data.

At the Hay, Tinga Tingana and Lake Argyle sites, a continuous monitoring regime is in place for a variety of aerosol and surface radiation balance measurements.
 
 
 

Site Objectives and Outcomes


The first year of the Hyperion mission will see the DCEs occur and the data brought together with the Hyperspectral and Calibration site data and enter its analysis stream. In this phase, we expect there to be close collaboration between NASA Goddard and ACRES to schedule data acquisition in Australia to allow Australian DCEs to occur within the power limitations and scheduling requirements of other earth observation missions.

The specific objectives and outcomes we are proposing from the Hyperion data acquisition are as follows:
 
 
 
Site
Objectives
Outcomes
Calibration Sites
 
 
Compare ground and airborne aerosol and water vapor determinations with Hyperion and LAC estimates. Determine surface leaving radiance for calibration. 
Reported comparisons between LAC and Hyperion Levels 1a and 2 with CIMEL, ground based radiation and airborne characterizations based on a single scheduled DCE.
Comparison of combined ASTER and Hyperion data with TIMMS and HyMap data for assessment of space platform capabilities for geological mapping and mineral abundance modelling. 
Direct assessment of the performance and future potential of space platform data for geochemical species and concentrations at Level 2 data format with differences evaluated at Level 1a.
Early and Mid Mission assessment of surface SRB and radiance/ reflectance modelling and measurements in uniform region.
Assessment of radiometric and calibration performance of Level 1a products based on the uniform and well characterized site. Potential to cross check Level 2.
Early mission performance of spectral estimation and radiometric and geometric alignment with high contrast targets as well as dark vegetated targets. 
Assessment of instrument Level 1a performance based on a single early mission DCE. Assessment of general forest cover performance based on a well characterized site. Optional Level 2 assessment.
APPLICATION SITES
 
 
Assessment of the change to a space platform and space derived atmospheric data on assessments of rain-forest health that are being derived from airborne data. Close by coastal areas provide a wide variation of environmental signatures. 
Direct comparison between selected products (eg stress, water content and canopy signatures) derived from the different sources (airborne, Landsat 7, ALI and Hyperion Level 2) and analysis of differences.
Mid and Later mission dark target characterization for calibration and aerosol and water vapor assessment.
Comparison of aerosol and water vapor estimates using Hyperion and LAC with ground and airborne assessments. Assessment of the future role of dark targets and stable water bodies in vicarious calibration. Primary data are Level 1a and provision of validation for Level 2.
Comparison of combined ASTER and Hyperion data with combined TIMMS and HyMap data for assessment of space platform capabilities for geological mapping and mineral abundance modelling. 
Direct assessment of the performance and future potential of space platform data for geochemical species and concentration mapping at Level 2 with discussion of differences using Level 1a.
Comparison of change mapping in two DCEs at selected times of year for a well characterized semi-arid rangeland.
Assessment of the performance of higher spatial resolution and HyMap data with the space platform Level 2 results for the same period. Outcomes include green and dry cover changes. Outcomes include registration and signature changes.


 
 
 

EO-1 Data Level Requirements


All of the sites are being mapped and characterized at a scale suitable for a single DCE involving a single cube of Hyperion data. An exception is proposed for Cape tribulation where a longer run is being proposed to include coastal and shallow sea benthic studies. The extended studies near Cape Tribulation include coastal and shallow water environmental signatures and their acquisition will depend on scheduling and environmental conditions.

It has been assumed that data for each DCE will include ALI, LAC and Hyperion data at Level 1a (calibrated and navigated). This assumption is based on the interest of the Australian researchers in algorithms and development issues as well as the fact that there will be questions of difference between Level 2 products (atmospherically corrected to apparent surface reflectance) and field data that will only be resolvable by checking back at Level 1a.

However, the primary ãproductä evaluations at Mt Fitton, Kioloa, Cape Tribulation, Panorama and Kunoth require access to the Goddard Level 2 products for Hyperion that come from these areas. In order for this comparison to be made effectively, we are assuming that we will interact with the Science Team to ensure the processing of the airborne and field data is to a level where they are consistent and comparable with the Level 2 Hyperion product. At most sites, both the airborne data and the Level 1a EO-1 data will be processed in Australia to apparent surface reflectances but the field based evaluation of Goddard Level 2 is a primary aim of this proposal.

At Cape Tribulation it is intended to investigate the integration and alignment of ALI and Landsat 7 data as well as Hyperion using the facilities of ACRES. ACRES is an accredited reception site for Landsat 7 data and will schedule the relevant acquisitions. Selected known environmental signatures will be used to compare cross-sensor and platform consistency.
 
 
 

Data Plan (summary)


The base site data associated with EO-1 DCEs will, as for the Hyperspectral sites, be made available through CD ROM or Web site access to researchers involved in EO-1. The wider community will also have access as the data are open file and free of restrictions for research and other non-commercial purposes. The integrated data from the EO-1, airborne and specific EO-1 mission ground based data collections will be freely available to other investigators in the EO-1 program.

Data products needs form NASA are the Level 1a LAC and Hyperion data as well as a unique opportunity in the Australian plan is to enable Australian DCEs to be scheduled by either downloading currently stored data from previously acquired missions at an Australian X-Band station or downloading the Australian DCEs to provide capacity for missions to be acquired in later orbits in other pars of the world without limitation. The Australian stations are located at Alice Springs in Central Australia and Hobart in the south of the island of Tasmania to the South of Australia. Data from the north of Australia and the Pacific may be acquired this way. Data would be sent to NASA Goddard by various means (depending on turn around requirements). This option is being investigated for the EO-1 Checkout period as well as for the man operations period.
 
 
 

References (selected)

 

Cocks, T., Jenssen, R., Stewart, A., Wilson, I., Shields, T. (1998). The HyMap airborne hyperspectral sensor: the system, calibration and performance. 1st EARSeL Workshop on Imaging Spectroscopy, 6-8 october 1998, Zurich, Switzerland, EARSeL, Paris, p 37-42.

Grant, I.F., Prata, A.J., Rondeaux, G., Steven M.D. (1997). Australian sites for the validation of satellite retrievals of the radiative properties of land surfaces. Nov 1997.

Mitchell, R.M., O'Brien, D.M., Edwards, M., Elsum, C.C., Graetz, R.D. (1997). Selection and initial characterization of a bright calibration site in the Strzelecki Desert, South Australia, Canadian J. Remote Sens., 23, 342-353.

Mitchell, R.M., O'Brien, D.M., Graetz R.D., Simpson, J. J. (1996). Identification and characterization of a bright calibration site in the Strzelecki Desert, South Australia, AGU Fall meeting, NOAA workshop on calibration and characterization of satellite sensors I, San Francisco, California.

O'Brien, D. M., Mitchell, R. M. (1999). Feasibility of cross-calibration of AVHRR shortwave channels against ATSR-2, Remote Sensing of Environment, submitted May 1999.

O'Brien, D. M., Mitchell, R.M., Edwards M., Elsum, C. C. (1998). Estimation of BRDF from AVHRR short-wave channels: tests over semiarid Australian sites, Remote Sensing of Environment, 66, 71-86.

Prata, A.J. (1995). The Continental Integrated Ground Site Network (CIGSN) - Report: 1994/95.

Prata, A.J., Cechet R.P. Grant, I.F. (1996). Validation of ATSR retrieved surface reflectance for Hay, NSW, using in-situ data. 1st Australian ERS Symposium - 6 February 1996.

Prata, A.J., Grant, I.F., Cechet R.P., Rutter, G.F. (Preprint). Five years of shortwave radiation budget measurements at a continental land site.


 

 

Return to Hyperspectral Task Home Page

EOC Hyperspectral Task Sites                             
Principal Investigator Contact Addresses
COSSA/EOC Hyperion Science Pages
EOC Home

CSIRO Home
International Hyperspectral Links
 
 

Mail to: Dr. David L B Jupp

Earth Observation Centre, CSIRO, AUSTRALIA.

 

Use of this web site and information available from it is subject to our
Legal Notice and Disclaimer