The NEON Imaging Spectrometer: Airborne Measurements of Vegetation Cover and Biochemistry for the Continental-scale NEON Observatory

Transcription

1 The NEON Imaging Spectrometer: Airborne Measurements of Vegetation Cover and Biochemistry for the Continental-scale NEON Observatory Thomas U. Kampe, Brian R. Johnson, Michele Kuester, Joel McCorkel National Ecological Observatory Network, Inc. Abstract The National Ecological Observatory Network (NEON) will fly an imaging spectrometer as part of the instrument payload aboard the Airborne Observation Platform (AOP) to obtain spectroscopic information on terrestrial ecology at sub-meter to meter scale ground resolution. The NEON imaging spectrometer will measure surface reflectance over the continuous wavelength range from 380 to 2500 nm sampled at 5 nm with high spatial and spectral uniformity. The NEON imaging spectrometer represents a significant advancement in ecological research by providing high-resolution remote sensing data of land cover and lands use, plant biochemistry and biophysical properties, and detection of invasive plant species. NEON, funded by the National Science Foundation, is a continental-scale ecological observatory for discovering, understanding, and forecasting the impacts of climate change, land-use change, and invasive species on ecology. NEON will observe both the human drivers of climate change and the biological consequences of environmental change. Local flux tower and field measurements at sites within the 20 NEON ecoclimatic domains distributed across the contiguous United States, Alaska, Hawaii, and Puerto Rico will be coordinated with high resolution, regional airborne remote sensing observations. In addition to the NEON imaging spectrometer, AOP remote sensing instrumentation consists of a scanning, small footprint waveform LiDAR for 3-D canopy structure measurements and a high-resolution airborne digital camera which, in combination, provide a unique data set for bridging scales from organisms and stand scales to the scale of satellite based remote sensing. The AOP science objectives, key mission requirements, and development status are presented including an overview of near-term activities associated with the development of NEON Imaging Spectrometer Design Verification Unit. Introduction The National Ecological Observatory Network (NEON) is a planned facility of the National Science Foundation for discovering and understanding the impacts of climate change, land use change, and invasive species on continental scale ecology (Keller et al., 2008). A wide range of biotic and physical processes link the biosphere, geosphere, hydrosphere, and the atmosphere. However, our understanding of the biosphere does not match our increasingly sophisticated understanding of Earth s physical and chemical systems at regional, continental, and global scales. Because many responses and feedbacks within the biosphere are large-scale, they cannot be investigated effectively with disconnected studies on individual sites or over short periods of observation. The NEON science focuses on questions that are relevant to large regions, and that cannot be addressed solely with traditional ecological approaches (Field et al., 2006). NEON is based on a multi-scaled sampling strategy, employing systematically deployed ground-based sensors, high-resolution airborne sensors and integration with national geospatial information. The objective of the NEON observatory is two-fold: infrastructure will be developed to provide systematic, long-term, large-scale data sets to scientists, students, educators and decision-makers. NEON will also serve as a research and educational platform for investigator-initiated sensors, observations, and experiments. NEON s educational and outreach program will include numerous 1 of 1 6

2 physical and virtual capabilities to enable educational and public use of the facility, including a central web portal to provide on-line learning experiences, tools for decision makers, professional developmental opportunities to prepare educators to use NEON data and educational tools, and research and internships opportunities. NEON will support workshops, seminars, and courses to provide training and learning experiences for individuals to more effectively use and contribute to NEON data, tools, and learning experiences. In this paper, we discuss the role of airborne remote sensing in the NEON design. The airborne instrumentation currently under development for NEON will provide the capability for obtaining detailed, regional measurements of ecosystem structure and function. NEON partitions the United States into 20 ecoclimate domains (Fig. 1) based on a statistical analysis of ecoclimate state variables. Each domain contains one fully instrumented core site located in a wildland area and two relocatable sites which have been selected to address ecological gradients. By bringing together observations from field observations, core and relocatable sites, airborne sensors, and mobile ground-based observing systems, as well as assimilating information from satellite and national data sets, the observatory aims to capture the ecological and climate variability at the continental scale over a 30-year period. Fig. 1. The NEON Domains distributed across the continental United States, Alaska, Hawaii, and Puerto Rico. AOP s Role in NEON The NEON Airborne Observation Platform (AOP) will, for the first time, enable routine meter-scale remote sensing measurements of vegetation structure and biochemistry, and land-use over more than 2 million hectares surrounding the 60 NEON sites on a yearly basis (Johnson et al., 2009). The major functional elements of the AOP are three aircraft platforms, 3 identical remote sensing instrument payloads, a sensor calibration facility, a data processing and distribution facility, and flight operations. Each remote sensing instrument payload consists of an imaging spectrometer, a small footprint waveform-lidar, a high-resolution digital camera, a dedicated Global Positioning System (GPS) and Inertial Measurement Unit (IMU) subsystem, and an instrument controller and data capture subsystem. 2

3 The requirements for the AOP aircraft are largely driven by the characteristics of the sensors to be flown and the need to provide observations over the 60 NEON ground sites. In particular, the maximum flight altitude is set by the required signal-to-noise ratio necessary for retrieving vertical structure from the waveform LiDAR. The low flight altitude limit and range of ground speeds are set by a desire to achieve 1 to 3-meter ground resolution, and a need to maximize signal integration time for the spectrometer, respectively. Together, the waveform LiDAR and spectrometer requirements drive the desired aircraft platform capability towards a low and slow performance range with survey speeds of 160 to 200 km/hr and survey altitudes between 1,000 and 3,000 m. Instrumentation The NEON imaging spectrometer is a pushbroom imaging spectrometer that measures the upwelling radiance of the Earth in 426 narrow spectral bands from 380 to 2510 nm at a spectral sampling of 5 nm. The hyperspectral data provided by the imaging spectrometer provides the capability to assess vegetative species diversity and classify vegetation to plant functional types or species levels (Ustin et al., 2004). Shortwave infrared bands provide the capability for discriminating tropical and temperate tree species and discrimination of senesced plant materials, wood, or bark from background soils (Roberts et al., 2004; Clark et al., 2005). Visible to near-infrared bands provide the capability for characterizing canopy chemistry, physiology, and type. The NEON imaging spectrometer uses a single spectrometer module and focal plane array to achieve the required spatial and spectral uniformity. The entire imaging spectrometer, including telescope, spectrometer and focal plane array will be housed in a vacuum chamber cryogenically cooled to 150 K to minimize background and dark noise. This is required to meet the high signal-to-noise ratio needed to support the science measurements, in addition to providing a controlled thermal environment for the spectrometer during flight operations. An on-board calibration subsystem integral to each imaging spectrometer provides the capability for flat fielding every imaging spectrometer data set, traceability to laboratory calibration standards, and monitoring of imaging spectrometer performance over time. The on-board calibrator, in conjunction with the NEON laboratory calibration facility also allows for cross calibration between replacement sensors over the 30-year lifetime of the NEON observatory and between sensors flying on separate airborne platforms. The choice of a small footprint waveform LiDAR was driven by the desire to record the entire timevarying intensity of the returned energy from each laser pulse and obtain a record of entire height distribution of the objects illuminated by the laser pulse (Lefsky et al., 2002) as well as the desire to resolve individual plant canopies and vegetation clusters. The waveform LiDAR requires a high pulse repetition frequency; high scan frequency; and measurements over a wide scan angle of 35 degrees with a ground resolution of 1 to 3 meters to match the observing characteristics of the imaging spectrometer. High-resolution imagery from the digital camera is useful for determining land use and allows for full visualization of the morphology of site locations. Panchromatic imagery from the camera will be provided at a resolution at least three times finer than the spectrometer resolution over a field of view matching the swath of the other AOP sensors. Remote sensing measurements from all instruments on the payload must be accurately registered in a common geographic coordinate system during ground data processing. This requires the relative alignment of the optical sensors be accurately known and remain stable during flight. The integrated GPS/IMU provides precision measurements of instrument payload position and attitude during remote sensing data collection. This information will be combined with knowledge of the relative orientation of the spectrometer, LiDAR, and camera in the GPS/IMU reference frame to compute the line-of-sight 3

4 trajectory of each laser shot and spectrometer detector element at a specific time. Data collected over each site will be stored on removable hard drives and sent to the NEON Headquarters for processing at the NEON Cyber Infrastructure facility. Operations The AOP flight plan for each year will include the standard observations of the core and relocatable sites of all 20 NEON domains, as well as directed flights to planned targets or unplanned flights in response to unanticipated events (e.g. response to wildfire). The flight season will extend from April to September for the sites located in the contiguous 48 states. Flights over the Alaskan sites will occur in a relatively narrow time window in July and August. Data collections over sites in Hawaii and Puerto Rico have relatively wide time windows of opportunity. Puerto Rico will be flown early in the year to avoid the hurricane season. Hawaii will be the last domain flown as not to significantly impact the campaign year since equipment must be shipped over the Pacific Ocean. The baseline mission flight plan is optimized so that sites are over-flown during peak productivity and at a time with the best chances for cloud-free weather or minimal cloud cover. These considerations are balanced with consideration for distances between domains in order to efficiently cover all the sites with the least amount of transit time. The notional baseline mission-plan for the year will be established in February and March for the upcoming season and initial flight plans will be released for all sites to be flown in that year. Payload 3 will be used for directed flights and as a "hot spare" for the first two payloads. These directed flights could include additional flights over NEON domain sites for phenology, transects over scientifically important regions, areas impacted by wildfire or other natural disasters, rapid deployment for significant natural or unnatural disasters, or as part of joint campaigns with other agencies. The mission plan for each year will be posted on the NEON website ( Near-Term Development Activity Early development of the remote sensing payload provides an opportunity to reduce the design and fabrication risk prior to actual spectrometer builds during NEON construction. By building and testing an imaging spectrometer design verification unit and demonstrating that key performance and operational goals have been met, the majority of technical risk can be retired prior to science operations. In September 2009, NEON Inc. was awarded a 2-year grant by the National Science Foundation for the development of the NEON Imaging Spectrometer Design Verification Unit. In addition to hardware development, this program includes scientific software design and prototype science algorithm development associated with Level-1 science data processing. As part of this development effort, a series of test flights are also planned. These include vicarious calibration flights to verify procedures and validate radiometric laboratory calibration and instrument boresight co-registration, and a flight campaign targeted at advancing LiDAR-imaging spectrometer data fusion. AOP s Role in NEON Science By taking advantage of the powerful synergy between imaging spectrometer and waveform LiDAR measurements (Asner et al., 2007), the AOP provides the capability to quantitatively measure biochemical and biophysical properties of vegetation at regional scales. It will therefore play a key role in bridging scales from organism and stand scales, as captured by field and tower observations, to the scale of satellite-based remote sensing. Unlike traditional field measurements which provide dozens to perhaps hundreds of samples over a region, the AOP will provide thousands of high resolution (1-3 meters) observations over hundreds of square kilometers at each of the NEON sites distributed across the contiguous United States, Alaska, Hawaii, and Puerto Rico. While multi-spectral remote sensing 4

5 instruments operating from space can provide global coverage at daily or multi-day intervals, the spatial resolution obtainable from these instruments (e.g., Landsat; 30 m, MODIS; 250 m) is barely sufficient to resolve even the largest tree crowns, and insufficient to detect small scale features or disturbances which are critical for monitoring land use change. As an example, Asner et al. (2005) showed that selective logging in the Amazon was not detectable in Landsat images but was captured from an airborne platform. This study showed that selective logging roughly doubled previous estimates of the total amount of forest degraded by human activities while increasing estimated greenhouse gas emissions by 25%. Similarly, imaging spectrometer/lidar measurements from low-flying aircraft have demonstrated the capability to quantify biomass (Wulder et al., 2004; Tollefson, 2009) and invasive species (Asner et al., 2008). The high spectral resolution and broad spectral coverage of the NEON imaging spectrometer provides flexibility in selection of spectral features used to map plant functional types based on their unique spectral reflectance signatures, as well as supporting a broader range of measurements including pigment, water, nitrogen, and carbon chemistry of plants (Ustin et al., 2004). The biochemical and physiological properties measured by the imaging spectrometer are greatly affected by vegetation structure and shadows that occur within and between vegetation canopies (Asner et al., 2007). The fusion of waveform LiDAR data (canopy height, crown shape, biomass estimates) with spectroscopic data allows for a broad range of products to be produced ranging from estimates of photosynthetic and non-photosynthetic fractional coverage, vegetation indices, pigment concentrations, light-use efficiency, and canopy water content to higher-level products such as ecosystem productivity and estimates of biomass at regional to continental scales that will require the assimilation of data from multiple scales into ecological models. AOP s detailed mapping of sites distributed across the contiguous United States as well as Alaska, Hawaii, and Puerto Rico, in conjunction with other NEON data sources, will provide researchers, educators and decision makers with an unprecedented view of ecological change at regional scales over the next several decades. All data products generated by the observatory, Level-1 data (calibrated spectral reflectances and LiDAR waveforms), and tools for analyzing data will be freely available from the NEON web portal. Acknowledgements The National Ecological Observatory Network is a large facility project sponsored by the National Science Foundation and managed under a cooperative agreement by NEON, Inc. References Asner, G. P., D. E. Knapp, E. N. Broadbent, P. J. C. Oliveira, M. Keller, J. N. Silva (2005). Selective logging in the Brazilian Amazon. Science 310, Asner, G. P, D. E. Knapp, T. Kennedy-Bowdoin, M. O. Jones, R. E. Martin, J. Boardman, and C. B. Fields (2007). Carnegie Airborne Observatory: in-flight fusion of hyperspectral and waveform light detection and ranging (wlidar) for three-dimensional studies of ecosystems. J. Appl. Remote Sens. 1, [doi: / ]. Asner, G. P., R. F. Hughes, P. M. Vitousek, D. E. Knapp, T. Kennedy-Bowdoin, J. Boardman, R. E. Martin, M. Eastwood, R. O. Green (2008). Invasive plants transform the three-dimensional structure of rain forests. Proc. Natl. Acad. Sci. USA 105(11), [doi /pnas ]. 5

6 Clark, M., D. A. Roberts, and D. B. Clark (2005). Hyperspectral discrimination of tropical rain forest tree species at leaf to crown scales. Remote Sens. Environ. 96(3-4), Field, C., R. DeFries, D. Foster, M. Grove, R. Jackson, B. Law, D. Lodge, D. Peters, and D. Schimel (2008). Integrated science and education plan for the National Ecological Observatory Network, (23 Oct 2006) < Johnson, B. R., Kampe, T. U., M. Kuester, and M. Keller (2009). NEON: The First Continental-Scale Ecological Observatory with Airborne Remote Sensing of Vegetation Canopy Biochemistry and Structure. Proc. SPIE 7454, [doi: / ]. Keller, M., D. S. Schimel, W. W. Hargrove, and F. M. Hoffman (2008). A continental strategy for the National Ecological Observatory Network. Front. Ecol. Environ. 6(5), [doi: / (2008)6[282:ACSFTN]2.0.CO;2]. Lefsky, M., W. B. Cohen, G. G. Parker, and D. J. Harding (2002). Lidar remote sensing for ecosystem studies. BioScience 52, [doi: / (95)0039-4]. Roberts, D. A., S. L. Ustin, S. Ogunjemiyo, J. Greenberg, S. Z. Dobrowski, J. Q. Chen, and T. M. Hinckley (2004). Spectral and structural measurements of northwest forest vegetation at leaf to landscape scales. Ecosystems 7, [doi: /s ]. Tollefson, J. (2009). Counting carbon in the Amazon. Nature 461(22), Ustin, S. L., D. A. Roberts, J. A. Gamon, G. A. Asner, and R. O. Green (2004). Using imaging spectroscopy to study ecosystem processes and properties. Bioscience 54(6), Wulder, M. A., R. J. Hall, N. C. Coops, S. E. Franklin (2004). High Spatial Resolution Remotely Sensed Data for Ecosystem Characterization. Bioscience 54(6),