The overarching goals of this project are: 1) to constrain the global budget of reactive nitrogen (Nr) through a combination of numerical simulations, data synthesis and analysis; and 2) to produce initial estimates of the overall interaction between reactive nitrogen cycling and climate. The nitrogen cycle is a key regulator of the Earth's climate system, linking terrestrial, marine, photochemical, and industrial processes, and modulating the carbon cycle. Over the last century and a half, expansion and intensification of agriculture and fossil fuel combustion have led to a more than doubling of Nr emissions to the atmosphere with profound impacts on the earth system. A closed global Nr cycle will be simulated within the CCSM (Community Climate System Model) by tracking Nr across three model domains: (1) atmosphere, (2) land, including native and agroecosystems, and (3) fresh and oceanic waters. While the basis for much of this work has already been developed within the CCSM, the nitrogen cycle has not been coupled across the different model domains. In the fully coupled system, each model domain will simulate the transport and production of Nr within its domain and the chemistry and loss of Nr from its domain, with the requirement that the nitrogen fluxes between the domains be self-consistent. This mass-balanced approach will avoid the untracked losses of Nr that occur when the nitrogen cycle is modeled in isolation within a single domain. Moreover, it will consider the diverse and opposing impacts of Nr on terrestrial carbon sinks and on the radiatively important species nitrous oxide, ozone, methane and aerosol ammonium sulfate. It is hypothesized that the overall effect of a changing Nr cycle on these four atmospheric species will lead to a warming sufficient to offset the cooling associated with increased Nr availability and increased terrestrial carbon uptake.
The synthesized datasets for evaluating the nitrogen cycle in Earth System Models, the coupled nitrogen-carbon-climate model developed here, and the simulations with the CCSM will be made broadly available to university and national laboratory communities and will contribute to the upcoming IPCC 5th Assessment. Most of the principal investigators in this project are actively engaged in graduate student education and training, and most participate in undergraduate activities as well. The research will support four additional graduate students and one postdoctoral associate as well as a number of undergraduates. All will become involved in the research activities both at their respective local institutions, as well as across all institutions and disciplines involved in the project.
In this project, a model of river nitrogen (N) transport was developed within the Community Earth System Model (CESM). The development of the river transport model (RTM) for nitrogen was a subcomponent of a larger effort to comprehensively model the global N cycle within the CESM. By transporting N leached from soils, the RTM links the Community Land Model (CLM) and ocean modules of CESM, and by estimating in-river production of the greenhouse gas nitrous oxide, the RTM links the land and atmospheric modules of CESM. This project marked the first time that a cradle-to-grave accounting of N was attempted within the CLM, which previously treated leached N as a terminal sink for nitrogen, which was simply lost from the model. RTM predictions of N concentrations and export rates in major world rivers are generally the right order of magnitude when evaluated against available observations. In rivers impacted by human activity and agriculture, N export to the coast tends to be overestimated relative to observations, suggesting excessive leaching from these watersheds. Conversely, river N export is generally low compared to observations in "natural" rivers, due to the tendency of CLM to underestimate the natural N leaching flux. Overall, when the model is run from 1850 to the present, N export from to the coast has increased dramatically (more than 3-fold) due to increasing N inputs from human agriculture. The model predicts a corresponding increase over this time in nitrous oxide production, which is associated mainly with bottom sediment denitrification in streams and rivers. The large increase in river N export over time has also been derived from diagnostic models of river N transport, which use empirical correlations to estimate the fraction of watershed N inputs delivered to rivers. However, the RTM is unique in that the river N model is linked to a comprehensive earth system model with the capability of simulating two-way feedbacks between river N export and climate. The RTM shows that there is substantial year-to-year fluctuation in river N export due to interannual variability in climate, which is superimposed on the secular increase due to expanding human agriculture. In most river basins, N export is positively correlated to mean watershed rainfall due to increased N runoff and leaching in wet years. Aquatic nitrous oxide production in contrast has a more complex response across different watersheds to climatic drivers like rainfall, due to competing influences from increased leaching in wet years but reduced sediment denitrification associated with deeper mean flow.
