An assessment of the effectiveness of a random forest classifier for land-cover classification