Current focus on forest conservation and forest sustainability has increased the level of attention given to measures of ground flora in forest ecosystems. Traditionally, such data are collected via time- and resource-intensive methods of field identification, clipping, and weighing. With increased focus on community composition and structure measures of forest ground flora, the manner in which these data are collected must change. This project uses color and color infrared digital cameras to proximally sense forest ground flora and to develop regression models to predict green and dry biomass (g/m^) from the proximally sensed data. Traditional vegetative indices such as the Normalized Difference Vegetative Index (NDVI) and the Average Visible Reflectance Index (AVR) explained 35-45% of the variation in forest ground flora biomass. Adding individual color band variables, especially the red and near infrared bands, to the regression model allowed the model to explain 66% and 58% of the variation in green and dry biomass, respectively, present.
Doruska, Paul F.; Weih, Robert C. Jr.; Lane, Matthew D.; and Bragg, Don C.
"Quantifying Forest Ground Flora Biomass Using Proximal Sensing,"
Journal of the Arkansas Academy of Science: Vol. 57
, Article 7.
Available at: https://scholarworks.uark.edu/jaas/vol57/iss1/7