Research is ongoing at an abandoned acid mine site near Timmins, Ontario to evaluate relations between forest structural damage symptoms and spectral, textural, and radiometric fraction information in high-resolution airborne digital camera imagery. Image measures such as the mean spectral band brightness, co-occurrence texture, texture variation, semivariance range, and the areal proportion of multispectral shadow fractions have been shown to be significantly related to individual measures of forest structure such as effective leaf area index (LAI), crown closure and stand basal area. The objectives of the study presented here were to: 1. conduct a multivariate analysis of the relations between data sets of soil, forest structure, forest structure relative to standing live volume, and airborne multispectral digital camera image variables, and 2. using the most significant relations of forest-image variables, develop an integrated image-based forest health index. To study relations between any two of the four data sets, each consisting of a multitude of variables, canonical correlation analysis was conducted using the principal components of each data set. Results show that some forest structural variables exhibit strong relations with soil properties. For example, greater tree blowdown is associated with sandier soils and less soil organic matter. From relations between the image and relative forest structure variables, it was found that a decrease in leaf area, canopy closure and stem size, an increase in dead and blowdown basal area, and an increase in dead and blowdown numbers of stems are associated with an increase in image texture, texture variation, and within-crown shadow fractions, and a decrease in average subscene brightness, deep shadow fraction and brightness. From these results, an image-based forest health index was developed that combines spectral and spatial information and is sensitive to the wide range of forest health conditions at the site.

Additional Metadata
Journal Canadian Journal of Remote Sensing
Olthof, I., & King, D. (2000). Development of a forest health index using multispectral airborne digital camera imagery. Canadian Journal of Remote Sensing, 26(3), 166–176.