Summary Report of the EOS Test Sites Meeting -- March 18-19, 1996
--Chris Justice (justice@kratmos.gsfc.nasa.gov), University of Maryland/NASA Goddard Space Flight Center
--Diane Wickland (dwickland@mtpe.hq.nasa.gov), NASA Headquarters
--David Starr (starr@climate.gsfc.nasa.gov), NASA Goddard Space Flight Center
--Tim Suttles (suttles@ltpmail.gsfc.nasa.gov), Hughes STX Corp.
The EOS Test Sites Meeting was held on March 18-19, 1996 at NASA Goddard Space Flight Center under the sponsorship of the EOS Project Science Office Validation Program. The meeting focused on land-based test sites involving measurements for land, atmosphere, and vicarious calibration studies and was co-chaired by Diane Wickland, NASA Headquarters, and Chris Justice, University of Maryland/NASA Goddard. Attendees included 57 participants from government and university research organizations and from private industry.
The meeting was motivated by several long-term test site activities underway within the instrument and interdisciplinary science (IDS) teams as part of preparations for EOS AM-1 algorithm development and data product validation. It was deemed appropriate to convene a meeting to allow communication of existing activities between the teams, to communicate other EOS and non-EOS activities to the teams, and to identify areas for coordination, potential collaboration, and cost sharing. Specific objectives of the meeting were to summarize in an informal report the requirements, plans, and timelines for test site development in the early EOS AM-1 time-frame and to build the foundations for coordinated inter-instrument and instrument-IDS test site activities. These foundations were to be built upon in subsequent validation planning including the EOS Science Data Validation Workshop in May 1996.
The meeting was conducted in a workshop format. Summary reviews of pre-meeting materials provided by various EOS teams, and brief status reports on ongoing community activities were presented and provided a basis for subsequent breakout group sessions. The first round of breakout sessions included discipline groups for Vegetation and Land Cover; Radiation; Aerosols, Chemistry and Meteorology; and Vicarious Calibration. These groups were charged with developing the basis for a test-site measurement implementation plan, including specification of required measurement packages and potential measurement synergy from their discipline viewpoints. A second round of breakout group sessions was designed to develop synergies between the EOS measurement suites identified in the previous breakout group sessions and further develop a strawman implementation plan. For this second round of sessions, six groups were established including: Measurement Package Synergies, Scoping a Test Sites Initiative, Validation and Data Assimilation Activities, Data Management and Standards, Calibration Sites, and Organizing a Test Sites Initiative. Each of these breakout groups reported results of its deliberations in plenary sessions.
This article presents a summary of the meeting in the form of short reports from the four discipline breakout groups followed by the findings of the meeting which incorporate the discipline and synergy group findings. A detailed report of the meeting is available as a Validation Document from the EOS Project Science Office homepage on the World Wide Web (http://spso.gsfc.nasa.gov/validation/valpage.html).
From the beginning of the EOS program, it has been recognized that use of satellite, aircraft, and surface-based observations is essential to achieving the principal scientific objective of increased understanding of the Earth as an integrated system. The global nature of Earth system processes dictates a sampling strategy that includes coverage of all important climatological zones of the globe including pristine regions as well as areas impacted by human activities such as biomass burning and industrial production. With satellites, global coverage is relatively straightforward; however, sufficient global sampling with aircraft and surface-based observations presents a major strain on both financial and human resources. In addition to their important role in scientific studies, aircraft and surface-based observations are required to provide correlative measurements and validation for the global satellite observations. Validation of the satellite observations is extremely important since global measurements of high accuracy spanning the full dynamic range of phenomena are required to achieve the program goals.
Aircraft and surface-based observations using both in situ and remote sensing techniques play a key role for scientific studies and for satellite data validation. Thus, the EOS Instrument Science Teams (ISTs) and Interdisciplinary Science (IDS) Teams have included such observations as elements of their investigations. Individually, the teams can accomplish limited objectives for their investigations, but the synergies of a coordinated, EOS-wide approach can produce much greater scientific payoff for the program. This is especially true for land-based test sites, since economies of scale and improved coordination with the many existing non-EOS land-based test site programs can be realized. Significant benefits can be realized in the EOS program by coordinating and integrating these activities to establish an EOS-wide, Land Test Sites Program.
Chris Justice gave the charge to the meeting participants. He began with a proposed definition of EOS Test Sites: EOS Test Sites are Community sites or locations where multiple surface and/or atmosphere measurements are taken for use in calibrating sensors or validating multiple EOS sensor data products and models. When the individual sites are combined as a network of sites, they provide an important step toward global representation.
The specific charge was:
Summary Report of the Vegetation and Land Cover Group - Warren Cohen, Chair and Stephen Prince, Rapporteur
For EOS, validation is needed for several vegetation and land cover parameters, including land cover type, land cover change, leaf area index (LAI), fraction of incident photosynthetically active radiation absorbed (FPAR), net primary productivity (NPP), and albedo and directional reflectance. Validation of these parameters may require measurements of additional parameters such as canopy and surface optical properties, digital elevation model (DEM) data, site biogeochemistry, biomass, percent vegetation cover, meteorology, CO2 fluxes, and emissivity.
The group discussed in detail requirements and potential approaches for validating land cover and cover change and NPP. Also, requirements were discussed for the suite of variables including LAI, FPAR, reflectance, and albedo. Results from these discussions and those of the other groups have been incorporated into the section on findings of the meeting given later in this article. Details of the group discussions are given in the complete report, which will be made available on the Validation segment of the EOS Project Science Office homepage.
The discussion of measurement strategies considered requirements for Intensive Validation sites and Extensive Validation sites.
The Intensive Validation network is aimed at accuracy and multi-temporal measurements. As such, this network need not be globally representative, rather existing facilities such as the remaining sites from the Boreal Ecosystem-Atmosphere Study (BOREAS), the NOAA Climate Monitoring and Diagnostics Laboratory (CMDL) tower sampling sites, Long-Term Ecological Research (LTER) sites, agricultural experiment stations, and established physical environment monitoring facilities should be utilized and augmented. Large and uniform areas are essential; sites larger than a contiguous block of 3 x 3 fields of view of the sensors are needed to ensure correspondence of satellite sensor observations and field measurements.
The Extensive Validation network should aim to represent the ranges of the globally occurring values of the EOS parameters. The need for cross-calibration of measurement instruments and methods is paramount. Many of the sites will need to be outside the USA, and existing networks such as the IGBP Global Change and Terrestrial Ecosystem (GCTE) transects should be used where possible.
Summary Report of the Radiation Group -- Alan Strahler, Chair and Thomas Charlock, Rapporteur
This group considered validation of fluxes at the surface and within the atmosphere, including radiative fluxes, sensible and latent heat fluxes, and chemical fluxes and validation of surface characterization, including land type, spectral BRDF, and description of vegetation canopy. Representatives from MODIS, MISR, CERES, ASTER and the Data Assimilation Office attended this group. Radiative fluxes and surface characterization were represented; however, microwave interests, the ocean, and moist processes were not represented thoroughly.
The Radiation Group recommended several classes of continuous surface validation sites:
Comprehensive Surface Tower sites to be constructed by EOS including a full complement of in situ observations for remote sensing validation. Six types of surfaces should be covered: barren, grassland, brush, broadleaf crop, deciduous forest, and needle leaf forest. There should be a comprehensive suite of surface instruments and periodic aircraft measurements at these sites.
Remote Sensing Physics sites at which measurements of atmosphere variables would be made to test remote sensing physics. These sites should have extensive suites of instruments and periodic aircraft campaigns should be conducted at these locations. Examples of such sites are the three DOE-Atmospheric Radiation Measurement (ARM) sites : Southern Great Plains (SGP) in the U.S., now operating; Tropical Western Pacific (TWP), planned; and North Slope of Alaska (NSA), planned. DOE supports extensive instrumentation at these sites, and periodic aircraft campaigns are also planned, some of which will require EOS support.
Regional Climate Trend sites to validate EOS "atmospheric subtraction" methods and to differentiate trends in surface, aerosol, and cloud properties. Taking advantage of existing and planned national and international networks, approximately 5 sites currently exist, and 40+ are planned through collaboration with the NOAA Surface Radiation Budget (SURFRAD) and GEWEX Baseline Surface Radiation Network (BSRN) projects. The objective is to have co-located radiometers and aerosol sunphotometers at these sites. EOS support will be needed to augment some sites and to allow aircraft campaigns at selected sites.
Discrete Validation sites that have a limited scope of measurements, but can be used to validate individual (discrete) EOS products. Existing and planned national and international networks provide targets of opportunity. Examples include: BSRN, ISIS (Integrated Surface Irradiance Study at NOAA), and GEBA (Global Energy Balance Archive in Zurich, Switzerland) type surface radiometer sites, and laser beam ceilometers at airports for cloud base height.
Summary Report of the Aerosol, Chemistry & Meteorology Group -- Jinxue Wang, Chair and Eric Vermote, Rapporteur
Members of the MISR, MODIS, MOPITT, and SAGE III teams were represented in the Aerosol, Chemistry & Meteorology group. Requirements for sites, instruments, and measurements for the validation of data processing algorithms and geophysical data products of each instrument were presented and discussed. Examples of how EOS could build on existing monitoring systems are listed here:
DOE/ARM sites (SGP, NSA, and TWP) include aerosol measurements using lidars and sunphotometers and can be enhanced for measurement of CO, CH4, and O3 by using automated flask samplers and surface trace gas samplers. These enhancements would provide comprehensive sites for algorithm and geophysical data products validation. Aircraft overflights over the ARM sites are very important, and EOS AM-1 coordinated aircraft campaigns should be planned.
NOAA/CMDL Cooperative Flask Sampling Network (~ 60 sites worldwide) with profiling capability at most, if not all sites, are identified as potential long-term, trace gas sampling sites for geophysical products validation and correlative measurements. Profiling capability can be achieved by using the automated flask sampling system and small airplanes. The group strongly encourages the early implementation of the trace gas measurement program with automated flasks and small airplanes proposed by the Carbon Cycle Group of the NOAA/CMDL.
AERONET (Aerosol Sunphotometer Network) & AEROCE (Atmosphere/Ocean Chemistry Experiment), with enhanced measurement capability gained by including instruments to measure the downward and upward angular radiance distribution at the surface, are identified as long-term sites for geophysical products validation and correlative measurements.
NDSC (Network for Detection of Stratospheric Change) sites with IR Fourier transform interferometry, microwave radiometers, laser heterodyne spectrometers, UV/visible spectrometers, and lidars are identified as long-term sites for geophysical products validation and correlative measurements.
Summary Report of the Vicarious Calibration Group - Phil Slater, Chair and Jim Butler, Rapporteur
The Calibration Breakout Group met on Monday, March 18, and on Tuesday, March 19. The March 18 meeting addressed the following charges: 1) design and scope the required/desired EOS Test Site activity to meet EOS investigator data needs; and 2) determine appropriate measurement packages suited to multiple products and instruments.
The discussion of the first charge began with a clear statement of the goal of vicarious calibration at calibration test sites, namely to predict top-of-the-atmosphere (TOA) radiances in the spectral bands of sensors and to validate the geometric registration of sensor radiometric scenes. A list of some candidate calibration test sites for predicting TOA radiances was presented by Phil Slater. Each test site was examined with respect to its use by EOS instruments and its calibration benefits and liabilities. Tables were developed to summarize the test sites and their characteristics and to outline the required measurements and instruments.
The second meeting of the Calibration Breakout Group was held in the morning of March 19. The charge to the group was delivered by Chris Justice and included examining geometric calibration test sites, coordinating international participation in vicarious calibration, and examining test sites for thermal infrared calibration. Because of time limitations this last topic was not discussed.
Geometric test sites were discussed with regard to the calibration of EOS AM-1 instrument footprints. In the case of the ASTER instrument, candidate sites include Iowa road/field patterns and linear features such as bridges. For the MODIS instrument, possible use of the edges of playas and lakes was postulated, and for the ETM+ instrument, Jim Storey presented a summary of Landsat geometrical test sites.
The final topic addressed by the Calibration Breakout Group was the coordination of international groups interested in the vicarious calibration of satellite sensors. It was generally agreed by the group that the Committee on Earth Observations Satellites (CEOS) and its working groups and subgroups in calibration and validation offer a vehicle to promote coordinated international comparison campaigns
The findings of the meeting incorporate results of both the discipline and the synergy breakout groups in terms of consensus on test site characteristics, measurement groups, test site classifications, measurement suites, and data management and standards.
Consensus on Test Sites Characteristics
The meeting revealed considerable overlap between the needs and approaches identified by different disciplines for test site characteristics.
Homogeneity: Measurements are needed from sites that are homogeneous over areas larger than the footprints of instruments to be validated; 4-9 km2 appears minimal.
Diversity: Measurement streams should be acquired from a diversity of land surface cover types, paying special attention to vegetation structure; 6-10 basic surface types (biomes).
Synergy: Data acquisition will be most effective with respect to both cost and scientific value if data are acquired in synergy with other measurements and measurement programs.
Ramp-up Strategy: Instrument costs are substantial and there needs to be a balance struck between the amount and cost of instrumentation and the number of sites. A balanced suite should prevail. The most costly (but most valuable) sites need to be expanded in number on a regular annual basis, adding a new increment each year.
Locations: Some possible existing and planned locations for instrumentation have already been identified at this meeting, including BOREAS, ARM sites, Brazil, etc. Existing networks, such as AERONET, AEROCE, BSRN, and LTER, should be utilized by adding value wherever appropriate to provide new data streams with proper characteristics. Other well instrumented long-term monitoring sites that are well distributed in the primary climatic zones within the U.S. are the U.S. Department of Agriculture-Agricultural Research Service experimental watersheds and the U.S. Geological Survey - WEBB (Water Energy and Biogeochemical Balance) sites.
Measurement Groups
The principal measurement groups are: Atmospheric Optical Measurements, BRDF Radiometry, Chemistry, Vegetation Structure, and Hydrology. Very little discussion was devoted to hydrology and existing watershed monitoring systems at this meeting. It was generally agreed that, in the future, additional emphasis will be needed to clarify the in situ data requirements for the hydrological aspects of the EOS platforms and program.
EOS Integrated Test Site Classifications
The requirements for EOS test site measurements can be met through an implementation plan consisting of different types of test site instrumentation. Discussion of classes of test sites developed from existing concepts formulated by the Global Terrestrial Observing System (GTOS) and the MODIS and CERES Instrument Teams. The concept of a hierarchical system (in terms of tiers) was developed for EOS Test Sites based on the required functionality, distribution, and level of instrumentation. It was recognized that individual measurement programs will continue and will be of use to the EOS community. Emphasis here was given to integration of land and atmosphere measurements for EOS Validation with an emphasis on EOS products.
Tier 1 -- Intensive Field Campaign Sites. These sites are developed as part of the International Intensive Field Campaign Program supported in part by NASA such as the International Satellite Land-Surface Climatology Project (ISLSCP) - First ISLSCP Field Experiment (FIFE), the Boreal Ecosystem-Atmosphere Study (BOREAS), the Global Tropospheric Experiment (GTE) - Transport and Chemistry near the Equator in the Atlantic (TRACE-A) experiment, the International Satellite Cloud Climatology Project (ISCCP) - First ISCCP Regional Experiment (FIRE), and the planned Large Scale Biosphere-Atmosphere Experiment in Amazonia (LBA). The sites have comprehensive multi-disciplinary ground-based instrumentation and repeated aircraft and satellite coverage. The field campaigns are intensive, lasting a month to a season and sometimes spanning successive years. The campaigns have an experimental focus and there is a large cost to supporting the field activities. They have been located in major biomes or climate regions. The multidisciplinary nature of the research is usually stressed. Such campaigns will be very useful for EOS Validation. In the context of long-term measurements and time series analysis, it may be desirable to maintain one or two of the test sites (as Tier 3 sites) beyond the duration of the intensive field campaigns, taking advantage of the capital investment in the site infrastructure and providing the possibility of long-term monitoring. There will probably be around ten of these international campaigns during the entire life of the EOS program.
Tier 2 -- Super Sites. These sites are designed for long-term monitoring with a central focus on establishing a full suite of radiation and flux measurements, including broadband and spectral radiation fluxes, continuous carbon dioxide, temperature and moisture sounding, aerosol optical thickness and absorbing properties measurements, meteorological data, and surface characteristics data. Tall tower measurements are desired at these sites. The collocation of ground-based radar and cloud lidar measurements is highly desirable. Aircraft data will also need to be acquired at these sites. An example of this type of site is the DOE ARM SGP site. Given the full suite of measurements, it is unlikely that there will be more than 5 of these fully instrumented sites globally during the EOS timeframe. Currently, two additional sites are planned as part of this DOE ARM network (NSA and TWP). Interagency collaboration will be necessary to provide the full suite of instruments. International participation may provide the means to increase the number of such sites.
Tier 3 -- Biome Tower Sites. These sites will provide long-term monitoring using instrumented towers at locations representing major biomes. These sites will be less well instrumented than the Tier 2 Sites, but will be at a larger number of locations. The sites could include eddy-correlation tower measurements of carbon dioxide and water vapor fluxes, selected radiation measurements, aerosol optical thickness, vegetation structure and phenology, land cover and land use characterization, soil fluxes, and physics and meteorology. Flask sampling of stable carbon isotopes will be a useful synergistic addition to these measurements. Site locations will represent major ecosystems and climatic regions. Emphasis will be given to process studies at these sites. An example is the Harvard Forest - Temperate Deciduous Forest Site. The BOREAS Thompson Site may also be continued to provide a long-term monitoring site in this category. The planned LBA Tower sites might also fall within this category. At present these sites have a strong bias towards the land community needs. The EOS Validation Program would benefit from the synergism resulting from collecting radiation and atmospheric data at these sites. Interagency coordination will be needed to increase the distribution of such sites. Internationally there are already strong indications that a global network could be established, for example, through IGBP coordination. The EUROFLUX network is an important step in this direction. With international participation there could be as many as 20-30 of these long-term sites in the EOS time frame.
Tier 4 -- Globally Distributed Test Sites. These sites provide an extensive site network aimed at a broader and more global representation of surface land cover, radiation, and atmospheric conditions. The sites will be permanent, for example the LTER network, the NOAA CMDL Flask Network, the BSRN, and the SURFRAD network. These different networks are each currently focusing on specific measurement sets and communities. Emphasis for EOS Validation will be to strengthen such networks by building multi-measurement components and to broaden the global extent of the sites. Emphasis at these sites will be on surface and atmosphere characterization for a limited number of variables, such as LAI, canopy optics, vegetation structure, land cover fraction, soil characteristics, land cover, surface radiation, and aerosol optical thickness. The primary purpose of these test sites within EOS will be global data product validation, e.g., land aerosols, atmospheric correction, land cover, LAI, and surface radiation data products. Of particular importance will be sites that can be used to address a broad range of parameters. Measurements may be continuous or may be taken at various intervals during the year and extend over a number of years. The sites will be used to capture seasonal and interannual variability and develop climatologies for the location. The instrumentation complement is likely to be less than at the Tier 2 and 3 sites, however, this will likely allow a greater number and much wider distribution of the sites. Occasional portable flux tower measurements may be possible. The minimal instrumentation needed for these sites means that there can be a greater number and wider distribution than for the sites in the previous categories. It may be that regional teams can be developed to provide consistent measurement and monitoring between these sites, which may, for example, fall along the IGBP Transects addressing ecology, hydrology, and atmospheric chemistry. Interagency and international cooperation will be needed to secure the network. However, interaction will be primarily between PIs. It is envisioned that there may be as many as 60 of these sites globally.
Tier 5 -- Instrument Calibration Sites. A separate category of test sites is needed by EOS for instrument calibration. These sites will require unique properties of reflectance and emittance, with an emphasis on uniformity, typically non-vegetated. Examples of this category in the U.S. are the White Sands and Railroad Playa sites. There will be few test sites in this category (less than 5), and these sites will be well instrumented for vicarious calibration. It is recommended that international coordination between the various space agencies provide a network of these sites for use by multiple space-based platforms and instruments. Characterization of the atmosphere as well as the surface is critical. Aircraft overflights will be needed in association with the vicarious calibration campaigns. Geometric calibration sites were not discussed in any detail at this meeting, although it was recognized that an additional type of site may be needed for vicarious geometric calibration.
The role of EOS in the above activities will be to develop a network of sites to focus the EOS validation activities. Emphasis will be given to augmenting existing networks with measurements needed to validate EOS data products rather than developing new networks. Interagency and international coordination will be essential to develop the necessary global representation. The large number of in situ data collection programs currently in place and the apparent overlap with EOS objectives makes coordination a high priority. The spatial scale of the satellite data will place stringent needs for spatial sampling at the sites, and considerable emphasis will be needed in developing the appropriate methodologies. Coordination of aircraft overflights at the sites is needed.
Measurement Suites
Individual measurements identified for different instruments and disciplines were grouped into the following example of measurement suites which indicate synergy between measurements.
| "A" Measurements: High priority, long term: |
|
|---|---|
| "B" Measurements: In-depth, specific focus, rotating from site to site, and acquiring data for 3-6 month periods at a site: |
|
| "C" Measurements: Field measurements needed to support radiometry: |
|
| "D" Measurements: Aircraft measurements: |
|
| "E" Measurements: Add-ons for IDS products: |
|
Data Management and Standards
EOS test site data will need careful management to ensure open and timely availability, ease of accessibility, and archiving. Within the EOS Data and Information System (EOSDIS) DAAC system, the Oak Ridge DAAC is currently responsible for field data and, for example, the DOE funded ARM Data Archive is located at Oak Ridge. Test site data are also well suited to Principal Investigator (PI)-generated data systems using internet or CD-ROM distribution. The federated system currently being developed for EOSDIS might be well suited to validation data management and distribution. As part of the EOS Validation program, PIs will be responsible for managing their data effectively and in keeping with EOS data policy.
It will be important to ensure that measurements at different sites are made following set standards and guidelines. Specific findings and recommendations are:
Standards
Data Formats
Quality Assurance/Quality Control
Data Integration and Packaging
User Feedback
Recommendations from the Meeting are:
The co-chairs of the meeting, Diane Wickland and Chris Justice, thanked the participants for their diligence, enthusiasm, and excellent contributions, and stated that the accomplishments surpassed their expectations.
| Table of Contents | | Previous | | Next |