National Aeronautics and Space Administration


Date of this Version



Published in Agricultural and Forest Meteorology 148 (2008).


Validation of Leaf Area Index (LAI) derived from moderate resolution remote sensing observations generally involves optical technique to measure ground LAI. As the current validation datasets are derived using multiple optical retrieval techniques, assessment of the consistency between these techniques is required. In this study the effective Plant Area Index (PAIeff) retrievals by three major optical instruments, LAI-2000, AccuPAR, and Digital Hemispherical Photographs (DHPs), were analyzed over 10 crops (soybean, corn, alfalfa, sorghum, peanut and pasture) at Manfredi site in Cordoba province, Argentina. The focus of research was on quantifying PAIeff sensitivity to the type of instrument, retrieval parameters and gap fraction inversion methods as well as environmental conditions (canopy heterogeneity, senescent vegetation, illumination conditions). Results indicate that sensitivity of DHP method to illumination conditions is low (14% compared to 28% and 86% for LAI-2000 and AccuPAR, respectively). The intercomparison of PAIeff retrievals indicates large discrepancies between optical techniques for short canopy over which downward-pointing DHP technique performs better than LAI-2000 and AccuPAR. Better agreement was found for tall canopy without senescent vegetation and low spatial heterogeneity. Overall, discrepancies in PAIeff between instruments are mainly explained by differences in spatial sampling of transmittance between instruments (over short and heterogeneous canopies) caused by variations in instrument footprint, azimuthal range, and zenith angle spatial resolution (coarser for LAI-2000 than DHP). Our results indicate that DHP is the most robust technique in terms of low sensitivity to illumination conditions, accurate spatial sampling of transmittance, ability to capture gap fraction over short canopy using downward-looking photographs, independence from canopy optical ancillary information, and potential to derive clumping index. It can thus be applied to a large range of canopy structures, and environmental conditions as required by validation protocols.