Image tonal - spatial variation to model the structure and diversity of a tropical forest

What is it about?

An evaluation was made of the potential of gray level co-occurence matrix (GLCM) texture metrics, obtained from a GeoEye1 image, to estimate the vegetation community characteristics of a Tropical Dry Forest. We found that the best estimated variables with this method were structural ones, in particular, stem density, mean height and basal area.

Featured Image

Why is it important?

We found that GLCM metrics were capable of sensing the between-plots structural variation (particularly stem density, mean height and basal area) of an old-growth Tropical Dry Forest with a medium to high fit (0.66 < R2 > 0.58). Interestingly, an old-growth tropical forest's spatial variation is harder to model with remote sensed metrics in comparison with secondary forests. Therefore, this study suggests that texture metrics might be a promising option to monitor tropical forests in the future