Incorporating Social and Policy Drivers into Land-Use and Land-Cover Projection

Abstract

Forestlands in the southeastern U.S. generate a great variety of ecosystem services that contribute to the well-being of humans and nonhumans alike. Despite their importance, forests continue to be lost to other land uses such as agricultural production and urban development. Advancements in remote sensing and machine learning techniques have facilitated land use/land cover (LULC) change projections, but many prior efforts have neglected to account for social and policy dimensions. We incorporated key socio-economic factors, conservation policies, societal preferences, and landscape biophysical features into LULC projection techniques under four different development scenarios. We applied this approach in the Upper Flint watershed, which flows south from the Atlanta, Georgia metropolitan area and is characterized by extensive urbanization and associated deforestation. Our results suggest that incorporating social and policy drivers in future LULC projection approaches leads to more realistic results with higher accuracy levels, offering decision-makers, development planners, and policymakers better opportunities to forecast the effects of anticipated changes on the availability of ESs in the future. Conservation organizations and public agencies can benefit from such analysis to identify regions requiring conservation interventions for prioritizing their conservation efforts. We used publicly available data for the conterminous U.S., hence our approach can be replicable in other study regions within the nation.