From jay Mon Feb 8 12:15:44 1993 >From jay Mon Feb 8 12:15:43 1993 To: gary@pldsa1.arc.nasa.gov Hemispherical Photographs and LAI Estimates: 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 stand mean and variability in Leaf Area Index. These properties are important to both direct estimates of gas, exchange parameters and for validation of remotely sensed image interpretation. Indirect methods for estimating light extinction and relating this property to gap fraction and leaf area index are well established since direct measures are impractical if not impossible. The basic assumption of these methods is that the one- sided canopy leaf area can be inferred from measurements of canopy gap area. Several instruments have been designed to measure this canopy property, including line quantum sensors, hemispherical light sensors, and hemispheric photographs. Computer software for analysis of digitized photographs is available (Rich 1989, 1990). The program CANOPY computes the "unweighted canopy openness" from the zenith and azimuth cells in the digitized fish-eye image. This matrix can then be analyzed using a one-dimensional extinction model (e.g., Norman and Campbell, 1989) or more simply by using the Beer-Lambert Law and assuming an extinction coefficient. Jarvis and Leverenz (1983) found that extinction coefficients ranged from 0.28-0.65 for 13 needle- and broad-leaved tree species, with an average of 0.47. EQUIPMENT: Platform: Field Measured Photographs Key Variable: Leaf Area Index (LAI) Principle of operation: Camera Instrument Geometry: Camera with hemispheric lens is mounted on a tripod 0.5 m above the ground, in a direction oriented vertically up. The camera body is oriented with a compass so that the top is pointing toward the north and is leveled using a bubble level. The focal is set at infinity, exposure set with light meter, and no back filter was used. The assumption of measurement is that the sky is uniformly bright relative to canopy elements and can be readily separated by differences in DN values in digitized images. Diffuse lighting is required to avoid specular reflections from stems and foliage to permit good separation between canopy and sky elements. Calibration: Blank exposed film was used to calibrate the film in the digitizing process. PROCEDURES: All photographs for the Juniper, Metolius, and Santiam sites were taken were taken between sundown and sunset during a 3 day period between June 17-19, 1990. The Scio and Warings Wood sites were measured on a second ,trip to the OTTER sites in August, 1990. One roll per site of 20 hemispherical photographs (180 degrees FOV) were made on black and white film (Kodak Tri-X, ISO 400) with a Canon 9 mm fisheye lens. Locations for making the hemispherical photographs were made randomly by randomly selecting an azimuth and walking a transect line in that orientation. Sites for photos were determined by pacing and stopping at 20 m intervals. After 10 slides, the transect direction was changed 90 degrees, paced off 40 m and returned to the origin along the heading 180 degrees opposite the initial transect. The transects were continued until 20 slides were obtained, or if the stand depth did not permit sufficient photos with 20 m spatial separation, then we again turned 90 degrees, paced 40 m and reversed direction 180 degrees from the previous orientation. DATA MANIPULATION: Normally developed negatives were digitized using hardware and software (CANOPY program) described by Rich (1990). Software to compute LAI from the unweighted openness produced by the CANOPY program was written by Scott N. Martens. Values for unweighted openness in each photograph were ratioed to those for a blank negative to obtain gap fraction for each of 160 segments (8 azimuth sectors and 20 zenith classes). When using Norman and Campbell's inversion model, LAI was calculated for each of the eight azimuth sectors and then these values were averaged to provide the LAI for the image. This is essentially the same as using a log-average of the gap fractions about the azimuth as suggested by the work of Lang and Yuequin (1986). A second method of computing LAI using a Beer-Lambert Extinction model was also computed in the program. The two estimates do not give equivalent predictions and the user needs to determine which analytical method they prefer to accept. A comparison of methods for calculating gap fraction can be found in Martens et al. (1993). Finally an estimate of the mean tip angle (leaf inclination) was derived from the analysis. The values of LAI computed by the two models and the mean tip angle are recorded by slide number. Values were initially inspected to determine whether they fell within physically reasonable limits (>0 and less than LAI=8). The LAI for the stands were averaged and standard deviations computed. ERRORS: Sources of Error / Quality Assessment: There are sources of error throughout all procedures. Probably the largest error is in the selection and number of field photographic measurements and the non-homogeneity of the forests. The next largest source of error is due to the violation of assumptions implicit in the analysis by conifers which have both highly non-random foliage distributions and small elongated foliage units (relative to distance from camera) that are subject to penumbra and resolution errors. Although the estimate of "mean tip angle (MTA)" or the leaf inclination angle is computed in the gap fraction analysis, it is particularly adversely affected by the non-homogeneity of the canopy and these values are probably substantially in error. We visually inspected the film before and after digitizing. The threshold to discriminate sky vs canopy DNs is selected by the interpreter and subject to bias. A student with considerable experience, Ms. Huoa Lo, digitized all the slides for this study. NOTES: There are no known problems with the data. REFERENCES: Lang, A. R. G. and X. Yuequin, 1986, Estimation of leaf area index from transmission of direct sunlight in a discontinuous canopies. Agri. Forest Meteor. 37: 220-243. Martens, S. N., S. L. Ustin, and R. A. Rousseau, 1993, Estimation of tree canopy leaf area index by gap fraction analysis. Forest Ecology and Management (in press). Norman, J. M. and G. S. Campbell, 1989, Canopy Structure. In: R. W. Pearcy, J. Ehleringer, H. A. Mooney and P. W. Rundel (Eds.), Plant Physiological Ecology: field methods and instrumentation. Chapman and Hall, New York, pp. 301-325. Rich, P. M., 1989, A manual for analysis of hemispherical canopy photography. Los Alamos National Laboratory Report LA-11733-M. Rich, P. M., 1990, Characterizing plant canopies with hemispherical photographs. Remote Sensing of Environment 5: 13-29.