Post-classification digital change detection analysis of a temperate forest in the southwest basin of Mexico City, in a 16-year span

Changes in forest cover during a 16-year period were evaluated by means of a post-classification digital change detection process in a site to the southwest of the basin of Mexico City. Post-classification was preferred over other change detection methods since it offers the advantage of indicating the nature of changes, such as forest to shrubland, to cropland, or to other land uses. Overall classification accuracy ranges from 59.8 percent to 70.2 percent, and the multivariate measure of classification accuracy from 0.55 to 0.66 (kappa coefficient). The forest coverage maps obtained for 1973, 1985, and 1989 show an 18 percent deforestation in that period in that area. The derived annual deforestation rates, expressed as the percentage of remaining forest that is cleared per year, were 0.5 percent for the interval 1973-1985 and 3.4 percent for 1985-1989. A digital elevation model (DEM) and derived slope gradient, and slope aspect maps, were useful in the digital classification adjustment. The digital change detection performed herein only reported quantities of lost forest, but fieldwork observations indicate some regions of the remaining forest are already severely affected. Hence, further research on the type of changes, and general vigor of the forest or its degradation level, are required to evaluate the full impact of forest destruction in other areas, such as aquifer recharge.