Accurate representation of leaf longevity is important for simulating ecosystem carbon cycle

Leaf longevity and leaf turnover rate are important plant traits relating to plan: growth, leaf photosynthesis and respiration, plant canopy dynamic and many other factors, but dynamic global vegetation models only set inaccurate values for these factors, In this study, we firstly investigated the leaf longevities at major vegetation types based on 418 field measurements from around the world. By replacing the default leaf longevities in the Lund-Potsdam-Jena model (LPJ) with the revised values, we examined the changes in simulated carbon cycle caused by the revised parameters. The results suggested that the compiled observations of leaf longevity were significantly different from the default values in LPJ. Both the vegetation production and respiration of the simulated natural ecosystems showed significant changes based on the revised leaf longevity compared with the predictions developed using default model parameters. Of all of the variables, the aboveground and belowground litter pools showed the largest changes (about 10% globally). Globally, the default parameters induced a significant overestimation of terrestrial carbon sink by 3%, compared with a simulation using the revised parameters. Furthermore, the uncertainties in leaf longevity caused various uncertainties (5-30%) in the simulated carbon fluxes and carbon pools. The offset of biases in intermediate variables might result in rational anal model outputs, Overall, more accurate leaf longevities are critical for simulating the vegetation distribution and ecosystem carbon cycle.