LI-COR 1800 Spectroradiometer OTTER Data Documentation and Description INVESTIGATOR: Susan L. Ustin Assistant Professor of Resource Science Department of Land, Air, and Water Resources University of California Davis, CA 95616 (916) 752-0621 email: [SUSTIN/NASA]NASAMAIL/USA slustin@ucdavis.edu Requested acknowledgment: Co-Author or Citation INTRODUCTION: The objective of this study was to characterize the spectral variability in conifer foliage between age classes, species, and sites. The bi- directional spectral reflectance factor of cut branches (removed from the trees) was measured between 400 to 800 nanometers. The purpose was to build a spectral library and the expected analysis tool was to use them in a spectral mixture model (Smith et al., 1990a,b; Ustin et al. 1993). The principle of the library is to use them to identify the major components of landscape variation found in the AVIRIS imagery. The library spectra should represent all major classes of spectral diversity found within the image data. In this study, we were relying on other OTTER investigators to measure reflectance in other categories of landscape components (e.g., soil, bark, and litter samples) in the field to complete the library. Our spectrometer did not measure in the SWIR region, which is most critical for distinguishing soils and dry plant materials. EQUIPMENT: Instrument Description: Platform: Laboratory Key Variable: Bi-directional Reflectance Principle of operation: Dispersive grating spectroradiometer Instrument Geometry: The spectrometer head was mounted on a tripod approximately 0.5 meters above the stage, nadir-looking, with a 15- degree field of view. Two stabilized halogen lamps (Smith-Victor Corp., Model 750; Griffith, Indiana) were used to illuminate the stage, each with a emittance-angle of approximately 45 degrees. The lamps and the stage thus formed a triangle with the telescope receptor mounted in the center in a nadir orientation. The instrument used was a LI-COR 1800 spectroradiometer with 1800-10 fiber optic link to the 1800-06 telescope optics head. Owned by S. Ustin/UCD. Manufacturer: LI-COR 4421 Superior Street PO Box 4425 Lincoln, NE 68504 (402) 467-3576 Calibration: A spectral calibration was performed by J. Dungan on this spectrometer on June 17, 1990 in Sisters, Oregon, using a helium neon laser with light output at 632.8 nm which indicated that the instrument performance was within the range of variation of other OTTER spectrometers (Dungan, personal communication and Dungan, 1991). Measurements are reported in terms of reflectance made by comparison to a highly reflective reference panel. One spectrum was taken of a reference panel made from powdered haylon (polytetrafluoroethylene), ground and packed to NBS specification (Weidner and Hsia, 1981; Weidner et al., 1985). Calibration scans were taken from the stage at approximately 30 min. intervals. Scans were made before and after each set of foliage measurements. These values were examined for drift by ratioing them to the first calibration spectra in the file and visually examining the ratio for deviance from a straight line and for albedo offsets. PROCEDURES: All spectrometer measurements for the Juniper, Metolius, and Santiam sites were taken under artificial illumination in the hotel room in Sisters, Oregon during a 3 day period between June 17-19, 1990. The next two days were spent with the system set up in a hotel room in Eugene, Oregon, measuring field samples obtained from the Scio and Warings Wood sites en route between Sisters and Eugene. Branch samples were collected from selected trees by shotgun or pruning poles by the OSU participants. All spectral measurements were made within 15 hours of collection. Samples were immediately tagged and placed into plastic bags and stored in a cooler until spectral measurements were made. After reflectance measurements were made using the LI-COR 1800 on the branches, the sample spectra were immediately measured under the same lamp arrangement by Dr. Carol Wessman (U. Colorado) using a GER SIRIS spectroradiometer. Following her measurements, the branch samples were bagged and were taken to Dr. Wessman's laboratory for biochemical analyses. A stage was constructed to measure the foliage while minimizing reflectance from background objects. A cardboard box, approximately 3 ft. x 2 ft. x 1.5 ft, was painted inside, around the opening, and over the upper surface with Krylon flat-black spray paint. A 18 cm diameter rectangular hole was cut in the center of the upper surface and conifer branch samples were placed across the hole for measurement. The hole edges were spray painted black. Corrugated cardboard dividers were also spray painted black and then assembled into an interwoven net and placed inside the box to act as baffles. The resulting stage produced a flat spectral response approximately 2% that of the reflectance panel. Two tripods were placed on either side of the box to hold the power stabilized (PH Stabilizer, Model AVR-1000W) 1000 watt halogen lamps at 45 degree angles. This resulted in overlapping beam spots at the stage. The power output of the lamps was checked at the stage surface to determine the stability of light intensity across the FOV of the sensor and box opening. The LI-COR telescope lens was placed approximately 0.5 meter above the stage and the 15 degree (aprox. circular) FOV entrance aperture was used. At 0.5 m height the FOV was aprox. 13.1 cm. The lens aperture in the telescope was opened to permit visual inspection that the location of the FOV was within the opening of the box and for aligning the calibration panel. Five branches of each major tree species/site were cut from the trees used for the physiological measurements at the OTTER sites. The branch samples were selected, cut, and tagged (naming and numbering) by OSU participants. The five branches of each tree were separately measured by stacking them individually over the hole in the stage and each one was scanned three times. The branches were rearranged and rotated 90 and 180 degrees and re-scanned. This procedure was done to decrease bias introduced by the architecture of the stacking. For ponderosa pine (Metolius fertilized and control sites), foliage was cut into five annual age class segments (stem area between each years needles was cut) and the spectra was recorded age class. This age class separation was not possible for other species and where possible, branches were arranged over the stage to maximize foliage from current years needles (for three replicates with rearranging their orientation between measurements) and regrouped to measure all older needle age classes in the FOV. Age class numbering is indicated as zero (current) year, one (one year old needles) year, to n-year age classes. DATA MANIPULATION: Percent reflectance was calculated from each foliage sample by ratioing the values to the mean calibration standard. Samples were then grouped by estimating the average and standard deviation of the species and age class categories. The wavelength range was between 400nm to 800nm. The nominal wavelength resolution was 3 nm. ERRORS: Sources of Error / Quality Assessment: It is possible that branch spectra changed during the storage time before measurements due to changes in pigment concentrations. Previous studies of ours on ponderosa pine have shown that the spectra and chlorophyll contents remain stable (within measurement error) for more than 48 hours when stored using methods employed in this study. We visually noticed changes in color and loss of chlorophyll by the end the set of scan measurements (ca. 1.5 minutes) on some samples, thus we believe that it was possible for the light/heat of the lamps to have caused some deterioration and spectral changes in the samples during the measurement period. To evaluate these errors and others (e.g., measurement and/or recording errors), we displayed the spectra on the computer and visually examined them for obvious problems. The scans included in the data set seem reasonable. The mean and standard deviations of all spectra were examined for especially large standard deviations. NOTES: There are no known problems with the data. REFERENCES: Dungan, J., Field Spectroradiometer Calibration Progress Report, Interoffice Memorandum, NASA/ Ames Research Center, 5 January 1991. Smith, M. O., S. L. Ustin, J. B. Adams, and A. R. Gillespie, 1990a, Vegetation in deserts I. A regional measure of abundance's from multispectral images, Remote Sensing of Environment, 29, pp. 1-26. Smith, M. O., S. L. Ustin, J. B. Adams, and A. R. Gillespie, 1990b, Vegetation in deserts II. Environmental influences on regional abundance, Remote Sensing of Environment, 29, pp. 27-52. Ustin, S. L., M. O. Smith, and J. D. Adams, 1993, Remote Sensing of Ecological Processes: A strategy for Developing Ecological Models Using Spectral Mixture Analysis. In J. Ehlringer and C. Field (Eds.) Scaling Processes Between Leaf and Landscape Levels. Academic Press, p.339-357. (in press) Weidner, V. R. and J. J. Hsia, 1981, Reflection properties of pressed polytetrafluoroethylene powder. J. Opt. Soc. Am. 71: 856-861. Weidner, V. R., J. J. Hsia, and B. Adams, 1985, Laboratory intercomparison study of pressed polytetrafluoroethylene powder reflectance standard. Appl. Optics 24:2225-2230.