Canopy structural attributes derived from AVIRIS imaging spectroscopy data in a mixed broadleaf/conifer forest

Article history: Received 23 July 2015 Received in revised form 15 April 2016 Accepted 28 April 2016 Available online xxxx There is awell-established needwithin the remote sensing community for improved estimation and understanding of canopy structure and its influence on the retrieval of leaf biochemical properties. The main goal of this research was to assess the potential of optical spectral information fromNASA's Airborne Visible/Infrared Imaging Spectrometer (AVIRIS) to discriminate different canopy structural types. In the first phase, we assessed the relationships between opticalmetrics and canopy structural parameters obtained fromLiDAR in terms of different canopy structural attributes (biomass (i.e., area under Vegetation Vertical Profile, VVPint), canopy height and vegetation complexity). Secondly,we identified and classifieddifferent “canopy structural types”by integrating several structural traits using Random Forests (RF). The study area is a heterogeneous forest in Sierra National Forest in California (USA). AVIRIS optical propertieswere analyzed bymeans of several sets of variables, including single narrow band reflectance and 1st derivative, sub-pixel cover fractions, narrow-band indices, spectral absorption features, optimized normalized difference indices and Principal Component Analysis (PCA) components. Our results demonstrate that optical data contain structural information that can be retrieved. The first principal component, used as a proxy for albedo, was themost strongly correlated opticalmetricwith vegetation complexity, and it also correlatedwellwith biomass (VVPint) and height. In conifer forests, the shade fraction was especially correlated to vegetation complexity, while water-sensitive optical metrics had high correlations with biomass (VVPint). Single spectral band analysis results showed that correlations differ inmagnitude and in direction, across the spectrumandby vegetation type and structural variable. This research illustrates the potential of AVIRIS to analyze canopy structure and to distinguish several structural types in a heterogeneous forest. Furthermore, RF using opticalmetrics derived fromAVIRIS proved to be a powerful technique to generate maps of structural attributes. The results emphasize the importance of using the whole optical spectrum, since all spectral regions contributed to canopy structure assessment. © 2016 Elsevier Inc. All rights reserved.