Modeling forest extent change and its influencing factors, using logistic regression model in GIS environment, (case study: Vaz and Lavij basins)

Abstract

Forests degradation and their convert to pasture, agriculture and urban or rural land use, occurs annually at high extent, in which make high damages to natural resources. In the current research, the logistic regression model was applied to identify effective factors of forest degradation and zoning possible forest degradation at Vaz and Lavij basins. Landsat Satellite images were used to identify land use variation within 1998 to 2006. Results showed that forest area was reduced from 81.97% in 1998 to 77.43% in 2006. Distance from village, river and road and also slope gradient were considered as independent variables, whereas land use change (forest degradation) was considered as dependent variable in the model. The logistic regression model was applied at two cases: classified and standardized variables. The obtained coefficient from the applied model at classified case indicated that forest degradation is more likely to happen at 0-100 m. distance from village, whereas the obtained coefficient from the applied model at the standardized case, indicated that there was negative relationship between the independent and dependent variables. Overall, zoning map of possible forest degradation at the studied area was developed at five levels: severe, high, moderate, low and very low, based on the calculated coefficients. According to the results achieved from modeling at classified case, 11.06 % of the studied area has potential for sever forest degradation.

Keywords

• Forest deforestation
• zoning map
• Land use
• dependent and independent variables
• Landsat Satellite images