Effects of the spatial resolution of climate data on estimates of biogenic isoprene emissions

Volatile organic compounds of biogenic origin (BVOCs) in the atmosphere are a key component of the Earth system, influencing ozone and secondary organic aerosol formation, and the global oxidant budget. Global BVOC emissions are dominated by terrestrial vegetation, in particular the compound isoprene, whose emission rate, in common with many BVOCs, is strongly and non-linearly dependent on temperature and the photosynthetically active radiation (PAR) flux. Detailed models of BVOC emission are now starting to be deployed in global chemistry-transport and chemistry-climate models. By necessity, the spatial resolution of these models is coarse (of the order of a few degrees), and spatial averaging removes information about areas of high temperature and PAR which contribute disproportionately to the isoprene flux. By comparing output from a BVOC emission model driven by both high- and low-resolution meteorological data, we show that this averaging effect does not lead to substantial discrepancies in simulated isoprene emissions (~2%) when considering fluxes averaged over regional scales, but can lead to large discrepancies of up to ~150% at much finer scales (e.g., 10 × 10 km). These smaller scale results have implications for highly coupled chemistry-climate simulations. The application of such models for the assessment and prediction of air quality, and subsequent decisions regarding the potential for mitigation or the need for adaptation, should be conducted using climate data with the highest possible spatial resolution. In particular, isoprene emissions calculated for topographically-heterogeneous regions, including coasts, should be treated with increased caution.