Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
4375520 | Ecological Modelling | 2016 | 21 Pages |
•Inclusion of N cycling in CLASS-CTEMN+ model helped to improve simulation of temporal and spatial patterns of regional C budgets.•N responses showed a strong influence on the simulated C budgets in high latitude regions during summer months.•C-N coupled model showed strong N attenuation effect on C budgets over the last two decades (1980–2010), when compared with the C-only model.•CLASS-CTEMN+ model provided a realistic simulation of plant productivity and C budgets at the global scale.
A carbon–nitrogen (C–N) coupled terrestrial ecosystem model, CLASS-CTEMN+, was used to simulate global terrestrial carbon, water, and energy cycles at 0.5-degree spatial resolution from 1901 to 2010. Forcing and initializing data sets from the North American Carbon Program (NACP)—Multi-Scale Synthesis and Terrestrial Model Intercomparison Project (MsTMIP) were used. Two versions of the model (the C–N coupled and the C-only version) were compared. The C–N (C) version of the model simulated global annual total gross ecosystem productivity (GEP) of 122.7 (128.2) Pg C yr−1, ecosystem respiration (Re) of 119.1 (122.4) Pg C yr−1, net ecosystem productivity (NEP) of 3.64 (5.82) Pg C yr−1, net primary productivity (NPP) of 62.7 (67.4) Pg C yr−1, latent heat (LE) of 146.2 (167.4) ZJ yr−1, sensible heat (H) of 194.0 (162) ZJ yr−1, soil organic carbon (SOC) of 1230 (1328) Pg C, and total vegetation biomass (Tvg) of 608 (629) Pg C, over the 1980–2010 period were comparable to scaled observations and modelling studies in the literature. Our results suggest that altough inclusion of N cycling has not had a large effect on the simulated total global terrestrial ecosystem C budget over this period; however the temporal evolution and spatial distribution of the simulated C budget revealed important influences of the N cycle. Analysis of long-term annual variations in response to evolving climate, CO2 concentration levels, and N deposition over the 1901–2010 period showed largely similar increasing trends in global GEP, Re, NEP and NPP until about 1970, after which the rate of increase was suppressed in the C–N coupled version of the model relative to the C-only version. Spatial plots showed that GEP and Re were reduced in the C–N coupled version of the model over the temperate and boreal regions and in high latitudes, where N limitation is well documented. The strongest response in NEP was in the boreal forests of North America, Northern Europe and Northwestern Asia, which showed decreases in response to the inclusion of a N cycle. Commensurate with the reductions in GEP and NEP, leaf area index (LAI) and total vegetation biomass (Tvg) also showed decreases in the boreal region, while SOC showed wider decreases in temperate, boreal and northern regions.This study helps to identify and diagnose the impact of N controls on the simulated spatiotemporal patterns of terrestrial C sinks and sources, and to evaluate and quantify the variability of simulated N cycle effects on different plant functional types (PFTs). These results have important implications for future climate change scenarios because the response to increasing CO2, N deposition, and climate patterns shows diverging model behavior as compared with the C-only version, when a N cycle is incorporated in the model.