A Remote Sensing-based Characterization of the Urban Heat Island and its Implications for Modeled Estimates of Urban Biogenic Carbon Fluxes in Boston, MA.

Urban land use occupies a small but critical proportion of global land area for the carbon cycle, and in the coming decades, urban land area is expected to nearly double. Conversion of natural land cover to urban land cover imposes myriad ecological effects, including increased land surface and air temperatures via the urban heat island effect. In this study, we characterize the seasonal and spatial characteristics of the urban heat island over Boston, MA and estimate its consequences on biogenic carbon fluxes with a remote sensing-based model. Using a 12-year time series of emissivity- and atmospherically-corrected land surface temperatures from Landsat TM and ETM+ imagery, we find a high degree of spatial heterogeneity and consistent seasonal patterns in the thermal properties of Boston, controlled mainly by variations in vegetative cover. Field measurements of surface air temperature across an urbanization gradient show season- and vegetation-dependent patterns consistent with those observed in the Landsat data. With a fused data set that combines surface air temperature, MODIS, and Landsat observations, we modify and run the Vegetation Photosynthesis and Respiration Model (VPRM) to explore 1) how elevated temperatures affect diurnal and seasonal patterns of hourly urban biogenic carbon fluxes in Massachusetts in 2013 and 2014 and 2) to what extent these fluxes follow spatial patterns found in the urban heat island. Model modifications simulate the ecological effects of urbanization, including empirical adjustments to reanalysis-driven air temperatures (up to 5 K) and ecosystem respiration reduced by impervious surface area. Model results reveal spatio-temporal patterns consistent with strong land use and vegetation cover controls on biogenic carbon fluxes, with non-trivial biogenic annual net ecosystem exchange occurring in urban and suburban areas (up to -2.5 MgC/ha/yr). We specifically consider the feedbacks between Boston's urban heat island and landscape phenology and explore the role of phenology in biogenic carbon uptake and release in urban areas. While most remote sensing-based models assume zero biogenic fluxes over urban, suburban, and exurban areas, our results suggest that live biomass in urban areas plays an important role in regional carbon budgets and should not be ignored.