Enhancing a eucalypt crown condition indicator driven by high spatial and spectral resolution remote sensing imagery

Individual crown condition of Eucalyptus gomphocephala was assessed using two classification models to understand changes in forest health through space and time. Using high resolution (0.5 m) digital multispectral imagery, predictor variables were derived from textural and spectral variance of all pixels inside the crown area. The results estimate crown condition as a surrogate for tree health against the total crown health index. Crown condition is derived from combining ground-based crown assessment techniques of density, transparency, dieback, and the regrowth of foliage. This object-based approach summarizes the pixel data into mean crown indices assigned to crown objects which became the carrier of information. Models performed above expectations, with a significant weighted Cohen's kappa (K< 0.60 and P< 0.001) using 70% of available data. Using in situ data for model development, crown condition was predicted forwards (2010) and backwards (2007) in time, capturing trends in crown condition and identifying decline in the healthiest between 2008 and 2010. The results confirm that combining spectral and textural information increased model sensitivity to small variations in crown condition. The methodology provides a cost-effective means for monitoring crown condition of this or other eucalypt species in native and plantation forests.