Robustness to Atmospheric Transport Error in Urban Greenhouse Gas Flux Inversion: a Los Angeles Case Study

Megacities with populations of ten million or more produce a dominant share of the world's anthropogenic greenhouse gas emissions. Increasingly, flux inversion methods originally developed for the global or regional scale are being applied to estimate those emissions on the urban scale using concentration data from in situ or remote sensing measurements. Such inversions require meteorological information to simulate atmospheric transport and ultimately link concentrations to fluxes. Custom-built meteorological models can have high spatial and temporal resolution and may be tuned to a specific urban environment, but are available for only a fraction of global megacities. By contrast, reanalysis products are readily available, with wide coverage, but lack locally tuned dynamics and are typically provided at coarser resolution. Using Los Angeles, for which both a purpose-built model and a dedicated in situ measurement network (the Megacities Carbon Project network) already exist, as a test case, we investigate the impacts of using reanalysis products as a substitute for a detailed meteorological model in an urban-scale flux inversion. We test the effects of estimating fluxes at coarser spatiotemporal scales, and we identify scientific questions, such as relative change in flux over time, that are less sensitive to transport error and may therefore be addressed adequately using reanalysis products. These investigations aim toward the development and verification of a widely applicable method for observational greenhouse gas flux estimation, which would allow for monitoring across the world's increasing number of megacities.