Pecora 16 Global Priorities in Land Remote Sensing

October 23-27, 2005, Sioux Falls, SD

Early Detection of Oak Wilt Disease: A Hyperspectral Approach

Blake Weissling and Hongjie Xie

Department of Earth and Environmental Science

University of Texas at San Antonio

6900 N. Loop 1604 W., San Antonio, TX 78249

Keywords: oak wilt, forestry pathogens, hyperspectral

Abstract:
  Disease pathogens such as Dutch Elm Disease and Chestnut Blight have wreaked havoc on North American forests during the 20th century.  A less virulent pathogen, but potentially as devastating in its impact, is the pathogen responsible for the destructive vascular disease in Quercus ssp., the fungus Ceratocystis fagacearum.  Commonly known as oak wilt, this disease claims many thousands of trees annually from the upper midwest to central Texas, resulting in significant property, ecological, and aesthetic losses.  Multispectral and high spatial resolution remote sensing have been shown to be effective in identifying oak wilt mortality centers but only in middle to late stages of disease pathology (stages of defoliation and/or leaf discoloration).  Of particular importance in disease management is the ability to detect disease centers in their earliest stages of development. Our approach, using in-situ hyperspectral technology, is to construct a spectral library of the pathology of oak wilt in the most commonly affected oak species native to central Texas.  Filtered by a spectral baseline of healthy trees as they progress through their phenological stages, our goal is to look for a hyperspectral signature of the disease, data that will ultimately be convolved to airborne and satellite hyperspectral sensors, such as Hyperion and AVIRIS.  Success in this project has ramifications for early detection of numerous forestry pathogens, invasive species, and disease vectors.