The use of high-resolution, active airborne remote sensing technologies to support precision forestry

Hans-Erik Andersen (Precision Forestry Cooperative, University of Washington ,,

Robert J. McGaughey (USDA Forest Service Pacific Northwest Research Station, rjm),

Stephen E. Reutebuch (USDA Forest Service Pacific Northwest Research Station)

The development and application of advanced digital technologies has enabled the implementation of more precise, site-specific and efficient forest management systems. One of the primary requirements of a site-specific approach to forest management is accurate and detailed three-dimensional spatial data relating to the type and condition of forest stands and characteristics of the underlying terrain surface. A new generation of high-resolution, active remote sensing technologies, including airborne laser scanning (LIDAR) and interferometric synthetic aperture radar (IFSAR) have the capability to provide direct, three-dimensional measurements of forest canopy structure and topography. Airborne laser scanning is an optical remote sensing technology providing high-resolution, precise measurements representing the location of laser reflections from the vegetation and ground surface. LIDAR has been shown to provide highly accurate digital terrain models, even under dense forest canopy. In addition, metrics based upon the lidar height distribution are highly correlated with critical forest structure variables, such as dominant height, basal area, and volume. The intensity of near-infrared LIDAR reflections can also be used to classify by species type. IFSAR is a microwave remote sensing technology that is also capable of providing three-dimensional positions of backscattering elements within a forest scene. While IFSAR typically provides measurements of lower resolution and accuracy than LIDAR, it has an all-weather capability and is acquired at a much lower cost per unit area. In addition, the use of multiple-frequency radar systems allows for the collection of information on different scene components. For example, long-wavelength P-band energy penetrates through the canopy and reflects mainly from the terrain surface, while short-wavelength X-band energy reflects from the first reflective surface. Therefore, the difference between the X-band (canopy) surface measurements and the P-band (terrain) surface will yield vegetation height information. This height information, along with other radar observables such as interferometric coherence and backscatter magnitude, has been shown to be highly correlated to critical forest structure variables. In this talk, I will describe the basic principles of these active remote sensing technologies in the context of forest canopy inventory and terrain mapping, and present an example of their application within a Pacific Northwest conifer forest.

  LIDAR canopy model (top) and terrain model (bottom), Capitol Forest, Washington, USA   IFSAR canopy model (top) and terrain model (bottom), Capitol Forest, Washington, USA