Summary: | The carbon and nitrogen coupled dynamic vegetation model, CLASS-CTEMN+ combines process-based, large-scale representations of terrestrial vegetation dynamics and land-atmosphere carbon and water exchanges in a modular framework. It prognostically simulates the principal processes of the terrestrial biosphere carbon and kinetic energy exchanges at the soil surface and plants, as well as the dynamic soil-plant nitrogen cycles. In this study, improvements made in parameterization of different plant functional types (PFTs) were evaluated, and then, model was used to assess the effects of nitrogen controls on simulated terrestrial carbon, water and energy exchanges and carbon pools from the site-level to regional and global scales. Prior to global simulations, standardized hourly meteorological forcing data, eddy covariance (EC) fluxes, and other site-specific observations from 39 FLUXNET sites from the North American Carbon Program (NACP) and the Large Scale Biosphere-Atmosphere Experiment in Amazonia (LBA) projects, spanning 194 site-years and covering 8 major PFTs across the North America and the Amazonian basin, were used to evaluate model performance. Two versions of the model, carbon and nitrogen coupled (C-N) version and carbon-only (C) version were used to simulate diurnal, daily, seasonal and annual values of carbon, water and energy fluxes at each site. Carbon pools and key nitrogen cycling variables were compared to investigate nitrogen controls on carbon, water and energy exchanges at each site. On the global scale, gridded forcing and initializing data sets developed by the North American Carbon Program (NACP)-Multi-Scale Synthesis and Terrestrial Model Intercomparison Project (MsTMIP) were used in CLASS-CTEMN+ simulations at 0.5 × 0.5 degree spatial resolution from 1901 to 2010. Exploratory and diagnostic assessment of the model was conducted at the global multi-decade scale, by comparing results from both versions of the model with observational and modeled estimates from literature to determine the impact of nitrogen availability on spatiotemporal dynamics and distributions of terrestrial carbon, water and energy fluxes and C pools. Model results revealed satisfactory performance of the model in simulating carbon, water and energy fluxes and carbon stocks, when compared to observations, especially in summer, and at evergreen needleleaf forest ecosystems. In contrast, simulation-observations agreement declined in winter and early spring, and at non-forested sites (crops and grassland), especially in dry periods during the growing season. The C-N coupled model simulated global total mean annual estimates from 1980-2010 for Gross Ecosystem Productivity (GEP, 122.7 Pg C yr−1), Ecosystem Respiration (Re, 119.1 Pg C yr−1), Net Ecosystem Productivity (NPE, 3.46 Pg C yr−1), Net Primary Productivity (NPP, 57.1 Pg C yr−1), Latent Heat (LE, 146.2 ZJ yr-1), Sensible Heat Flux (H, 194.0 ZJ yr-1), Soil Organic Carbon (SOC, 1230.0 Pg C) and Total Vegetation Biomass (Tvg, 608.0 Pg C) were similar to reported values in the literature. Evaluation of nitrogen limitation impacts on global carbon sink and sources dynamics showed considerable variability between and within forest types due to non-linearity of N effects and spatiotemporal heterogeneity of C-N interactions. For the recent 1970-2010 period, the C-N model estimated annual increase rate in the global mean terrestrial carbon uptake, was 0.05 Pg C yr-1, which was less than the 0.12 Pg C yr-1 simulated by C-only version of the model, suggesting a strong N attenuation effect compared to the C-only over this period. The consideration of N dynamics in the CLASS-CTEMN+ simulations reduces the terrestrial C uptake compared with that of the C-only counterpart in some regions, where N might not be always be sufficiently available for plants to grow, particularly in mid to high latitude regions of boreal forests, tundra and some temperate forest regions, where N is a primary limiting nutrient. While a smaller N limitation effect was observed in the southern temperate and tropical regions where ecosystem production is most likely to be limited by phosphorus (P) rather than N. Overall, the inclusion of the nitrogen cycle in the CLASS-CTEMN+ model improved its prediction accuracy, in particular for forests. This study gives us confidence that CLASS-CTEMN+ can predict carbon, water and energy fluxes and carbon stocks quite well in multiple vegetation ecosystems. The inclusion of nitrogen cycle in the model helped in its application at regional and global scales to evaluate nitrogen availability impacts on carbon cycle in terrestrial ecosystems and to determine nitrogen cycle feedbacks on Earth’s climate system. This study also suggested the need for a network of long-term monitoring sites to measure changes in the vegetation and soil carbon biomass at the local and regional levels. Thesis Candidate in Philosophy
|