<![CDATA[Deforestation is now becoming a global concern due to its effect on the global warming. This paper describes a dynamic change of deforestation and spatial modeling for predicting deforestation in North Konawe and Konawe Districts, Southeast Sulawesi Porvince. The study objective is to examine and analyze the variety of explanatory variables related to the process of deforestation at each deforestation typology. The data used for the analysis include Multitemporal Landsat images acquired in 1997, 2000, 2005, 2010 and 2013, the existing land cover maps published by the Ministry of Forestry, statistical data and ground truth. All district within the study area were classified into two typologies on the basis of social and economic factors by using clustering approaches, i.e., low-speed and high-speed deforestation district. To analyze model and predictions using land cover data in 2005, 2010 and 2013. The study found that the spatial model of deforestation for low-speed deforestation area is Logit (Deforestation) =– 1.0998 – 0.017031*Kpd05(population density) – 0.000095*JJ(distance from road) – 0.000419*JS(distance from the river) – 0.002057*JH05(distance from forest edge) – 0.00001*JPmk05(distance from settlements) – 0.000019*JPlc05(distance to the mixture of dry land agriculture)+0.016305*S(slope)+0.084348*E(elevation), high-speed deforestation area is Logit (Deforestation) =– 1.2361– 0.062622*Kpd05(population density) – 0.000008*JJ(distance from road) – 0.00001*JS(distance from the river) – 0.005443*JH05(distance from forest edge) – 0.000077*JPmk05(distance from settlements) – 0.000067*JPlc05(distance to the mixture of dry land agriculture)+0.469883*S(slope)+0.300739*E(elevation). The low-speed and high-speed deforestation models had ROC (Relative Operating Characteristics) of 93.48% and 97.71%, respectively. The study concludes that typology could be made on the basis of population density and the amount of dry land with wetland. The results of this study showed that there are eight explanatory variables that significantly affect deforestation probability, namely population density, distance from road, distance to the river distance from the forest edge, distance to settlement, distance to the mixture of dryland agriculture, slope, elevation and. Keywords: deforestation, konawe, logistic model, spatial model, typology]]>
