Tuesday, August 5, 2008
Exhibit Hall CD, Midwest Airlines Center
Background/Question/Methods
Amur honeysuckle (Lonicera maackii) is an Asian shrub that invades the understory of deciduous forests in North America. L. maackii has become widespread, but the characteristics of invaded sites and the nature of its spread are not well understood. As part of a larger project to map historical spread and determine which forest stands are most susceptible to invasion, we have examined the relationship between L. maackii abundance and spectral characteristics as derived from Landsat Thematic Mapper (TM) data. The leaf phenology of the shrub differs from native tree species, allowing this satellite-based sensor to "view" the invaded stands when trees are leafless, but L. maackii is still photosynthetically active. Using a November 2005 Landsat TM scene, we compared L. maackii cover to a range of spectral vegetation indices including the normalized difference vegetation index (NDVI), Tasseled Cap, Simple Ratio (SR), and the Enhanced Vegetation Index (EVI) for 35 plots in southwestern Ohio.
Amur honeysuckle (Lonicera maackii) is an Asian shrub that invades the understory of deciduous forests in North America. L. maackii has become widespread, but the characteristics of invaded sites and the nature of its spread are not well understood. As part of a larger project to map historical spread and determine which forest stands are most susceptible to invasion, we have examined the relationship between L. maackii abundance and spectral characteristics as derived from Landsat Thematic Mapper (TM) data. The leaf phenology of the shrub differs from native tree species, allowing this satellite-based sensor to "view" the invaded stands when trees are leafless, but L. maackii is still photosynthetically active. Using a November 2005 Landsat TM scene, we compared L. maackii cover to a range of spectral vegetation indices including the normalized difference vegetation index (NDVI), Tasseled Cap, Simple Ratio (SR), and the Enhanced Vegetation Index (EVI) for 35 plots in southwestern Ohio.
Results/Conclusions
L. maackii cover was best predicted by a quadratic function of NDVI (R2=0.75). This model was verified using an additional 15 plots (R2=0.56), confirming that satellite-derived spectral vegetation indices have potential for predicting L. maackii cover in the understory of forest stands.
