Simulating Forest Cover Changes of Bannerghatta National Park Based on a CA-Markov Model: A Remote Sensing Approach

Establishment of protected areas (PA) has been one of the leading tools in biodiversity conservation. Globally, these kinds of conservation interventions have given rise to an increase in PAs as well as the need to empirically evaluate the impact of these PAs on forest cover. Few of these empirical evaluations have been geared towards comparison of pre and post policy intervention landscapes. This paper provides a method to empirically evaluate such pre and post policy interventions by using a cellular automata-Markov model. This method is tested using remotely sensed data of Bannerghatta National park (BNP) and its surrounding, which have experienced various national level policy interventions (Indian National Forest Policy of 1988) and rapid land cover change between 1973 and 2007. The model constructs a hypothetical land cover scenario of BNP and its surroundings (1999 and 2007) in the absence of any policy intervention, when in reality there has been a significant potential policy intervention effect. The models predicted a decline in native forest cover and an increase in non forest cover post 1992 whereas the actual observed landscape experienced the reverse trend where after an initial decline from 1973 to 1992, the forest cover in BNP is towards recovery post 1992. Furthermore, the models show a higher deforestation and lower reforestation than the observed deforestation and reforestation patterns for BNP post 1992. Our results not only show the implication of national level policy changes on forest cover but also show the usefulness of our method in evaluating such conservation efforts.