Robustness of terrestrial carbon and water cycle simulations against variations in spatial resolution

Abstract

Dynamic Global Vegetation Models (DGVMs) of the terrestrial carbon and water cycle have been developed and validated at specific spatial resolutions (mostly 0.5°) but are increasingly being coupled to climate models at coarser spatial resolutions. Is this permissible? We ran the LPJ‐DGVM at different spatial resolutions (0.5 × 0.5° to 10.0 × 10.0° in 0.5° intervals) to assess the robustness of terrestrial carbon and water flux simulations to changes in spatial resolution. We show that global model results are robust with only small deviations in the single‐digit percent range from a benchmark run at 0.5°. The magnitude of the deviation increases with grid coarseness. Temporal dynamics are largely unaffected by grid cell size. The deviations from the benchmark are mostly spread evenly in space and are otherwise concentrated in areas with strong environmental gradients. We conclude that for coarse‐resolution model coupling (such as with climate models) as well as for specific global‐scale applications (such as global agroeconomic modeling or integrated assessment modeling) the spatial resolution of DGVMs can be reduced to coarser grids with little biogeochemical error.