Assessing Local Variations of Deforestation Processes in Mexico Using Geographically Weighted Regression

This study identifies drivers of deforestation in Mexico by applying Geographically Weighted Regression (GWR) models to cartographic and statistical data. A wall-to-wall multitemporal GIS database was constructed incorporating digital land use/land cover maps for 2002 and 2007; along with ancillary data (road network, settlements, topography and socioeconomical parameters). The database analysis allowed assessing the spatial distribution of forests and deforestation at the municipal level. The statistical analysis of deforestation drivers presented here was focused on the proportion of anthropogenic cover in 2007 as dependent variable. In comparison with the global model, the use of GWR increased the strength in the relationship in terms of the goodness-of-fit (adjusted R2) from 0.69 (global model) to 0.72 (average R2 of GWR local models), with individual GWR models ranging from 0.48 to 0.81. The GWR model highlighted the spatial variation of the relationship between the percentage of anthropogenic cover and its drivers. Factors identified as having a major impact on deforestation were related to topography (slope), accessibility (road and settlement density) and marginalization. Results indicate that some of the drivers explaining deforestation vary over space, and that the same driver can exhibit opposite effects depending on the region. Based on local regression model coefficients, a cluster analysis allowed the aggregation of municipalities with similar patterns of deforestation into homogeneous regions. A deforestation model for the entire country will be developed further, using these regions to divide the model procedures into sub-regions with specific deforestation patterns.