Green infrastructure, such as water harvesting basins and urban vegetation, provide valuable ecosystem services and play an essential role in the overall functioning of urban habitats. Despite recent efforts to promote sustainable urban habitats in arid regions, the spatial distribution of water harvesting basins and urban vegetation may not match the need for green infrastructure across the urban landscape. For example, socioeconomic factors such as annual income may influence the spatial distribution of green infrastructure and thus ecosystem services. This study uses a spatially explicit model to assess the current distribution and potential capacity of green infrastructure in a semi-arid urban environment in Tucson, AZ. By doing so, this study (1) identifies effective areas that can serve as regional models, (2) identifies priority areas that, due to their lack of vegetation, topography, soil geomorphology, and proximity to watershed lines, are critical areas for concentrating future green infrastructure efforts, and (3) tests whether barriers to implementing green infrastructure as an adaptation and sustainability solution are largely socioeconomic. In order to create a spatially explicit model of the current distribution and future capacity of green infrastructure capacity at the city-scale, this study integrates remote sensing data, including: LiDAR, NDVI, and thermal satalite data obtained from local city and county governmental offices. We also integrate socioeconomic data from the US Census to create an index of socioeconomic risk to environmental vulnerability
Results/Conclusions
We compared the vegetation cover using LiDAR and NDVI with from 2000 and 2010 to assess how the urban forest canopy provided ecosystem services, primarily mitigation of the urban heat island effect. We observed significant varation in surface thermal temperatures, including an elevation consistent with urban heat islands. We found that income was the best predictor of surface temperature in 2000, but vegetation density was the best predictor in 2010. The socio-economic risk indicator as a whole provided poor results to predict vegetation cover and surface temparature, but elements of the risk indicator were more successful. For example, in 2000 the proportion of household who completed high school was correlated with surface temperatures, and in 2010, population density and the number of household memebers under 18 were good predictors of surface temperature. We discuss these results in the context of social-ecological systems and environmental resilience. Ultimately, this these data indicate their utility for supporting environmental planning and afforestation efforts to address socio-environmental challenges in Tucson, AZ.