Canopy Cover Estimation in Semiarid Woodlands: Comparison of Field-Based and Remote Sensing Methods

Abstract

Abstract Canopy cover is a widely used measurement for characterizing forest structure. Numerous comparative studies have been conducted for a variety of canopy cover field methods. However, comparisons of canopy cover methodology are lacking for semiarid woodland, characterized by short-statured trees, simple vertical canopy structure, and relatively low canopy cover. Furthermore, there is limited knowledge on the compatibility between field and remote sensing methods, despite the increasing interest in scaling up canopy cover estimates from forest stands to large landscapes. In this study we investigated the use and functional similarity of various canopy cover field methods and a remote sensing method in semiarid pinyon-juniper woodlands in the Great Basin of the western United States. We compared line intercept, vertical densitometer, spherical densiometer, and crown radius methods with each other and with a remote sensing method using object-oriented image classification of digital orthophotography. Our results show that methods using a wider angular view generated higher canopy cover estimates. This difference may be greater in semiarid woodland that in mesic forest because of differences in canopy architecture. Canopy cover from remote sensing showed a nonlinear relationship with that from field-based methods. Such nonlinearities need to be considered when remote sensing is used to extrapolate canopy cover from stand to landscape scales or when field-based canopy cover estimates are used to calibrate or validate remote sensing classifications.