|
|
Ecosystems & Biogeochemical ModelsAtmospheric Climate and Chemistry ModelsMathematical and Statistical Techniques of Model and Data Assimilation |
Ecosystems & Biogeochemical ModelsThe Terrestrial Ecosystem Model The Terrestrial Ecosystem Model (TEM) is a process-based ecosystem model that describes carbon and nitrogen dynamics of plants and soils for terrestrial ecosystems. The TEM uses spatially referenced information on climate, elevation, soils, vegetation and water availability as well as soil- and vegetation-specific parameters to make monthly estimates of important carbon and nitrogen fluxes and pool sizes of terrestrial ecosystems. The TEM now operates on global scale, at a monthly time step and a 0.5 degrees latitude/longitude spatial resolution.
Following is a structure diagram of TEM
|
Soil thermal model (STM) The STM was developed based on the Goodrich model, the vertical profile is divided into snow cover, moss, upper organic soil, lower organic soil, and mineral soil layers (Figure 1a). Appli-cation of the model for a site requires specification of the thickness of each layer and simulation depth steps within each layer. The thermal properties of each layer also need to be prescribed. In addition, the dynamics of phase changes in the soils depend on the phase temperature, which we set to 0oC for applications of the model in this study. Specification of the upper boundary condition includes the temperature at the top of the moss layer during the summer and at the surface of snow during the winter. In this study we prescribed the depth, density, and thermal properties of snow cover. For the lower boundary condition we assumed a constant heat flux. Alter-natively, the lower boundary condition can be specified as a temporally varying function of temperature or heat flux. Ap-plication of the model requires the prescription of initial con-ditions, which include specifying the initial soil temperatures of the system and the presence or absence of permafrost (Zhuang et al., 2001).
Following is a diagram of the STM
|
Water Balance Model 1.0
|
Fire version of TEM
In this version of TEM, we integrated more effective algorithms of biogeochemistry after fire with the soil thermal dynamics simulated by STM and the hydrology simulated by HM. After fire, the STM and HM require information from TEM on how the thickness of moss and leaf area index change. Therefore, we modified TEM by including formulations to simulate changes in the thickness of moss, canopy biomass, and leaf area index as the stand recovers from disturbance. The TEM requires information from STM on soil temperature and from HM on soil moisture of the humic organic layer and on estimated actual evapotranspiration (EET). We modified TEM so that soil temperature and soil moisture of the humic organic soil layer influences the simulation of heterotrophic respiration, nitrogen mineralization, and nitrogen uptake by the vegetation. Similar to previous versions of TEM, EET influences the simulation of gross primary production (GPP) (Zhuang et al, 2002).
Following is a structure diagram of the FireTEM
|
Updated Water Balance Model (WBM)
The hydrological module [HM, Zhuang et al., 2002] was further developed including: 1) the consideration of surface runoff when determining infiltration rates from rain throughfall and snow melt; 2) the inclusion of the effects of temperature and vapor pressure deficit on canopy water conductance when estimating evapotranspiration based on Waring and Running [1998] and Thornton [2000]; 3) a more detailed representation of water storage and fluxes within the soil profile of upland soils based on the use of the Richards equation in the unsaturated zone [Hillel, 1980]; and 4) the development of daily estimates of soil moistures and water fluxes within the soil profile instead of monthly estimates. As the original version of the HM is designed to simulate water dynamics only in upland soils, algorithms have also been added to simulate water dynamics in wetland soils.
For wetlands, the soil profile is divided into two zones based on the water table depth: 1) an oxygenated, unsaturated zone; and 2) an anoxic, saturated zone. The soil water content and the water table depth in these wetland soils are determined using a water-balance approach that considers precipitation, runoff, drainage, snow melt, snow sublimation, and evapotranspiration. We assume that wetland soils are always saturated below 30 cm, which represents the maximum water table depth [Granberg et al., 1999]. Daily soil moisture at each 1 cm depth above the water table is modeled with a quadratic function and increases from the soil surface to the position of the water table [Granberg et al., 1999]. Infiltration, runoff, snowmelt, snow sublimation and evapotranspiration are simulated in wetlands using the same algorithms as for uplands. Drainage from wetlands is assumed to vary with soil texture, but does not exceed 20 mm day-1 (Zhuang et al., 2004GBC).
Following is a structure digram of the Updated WBM
|
Land-Use and Land-Change version of TEM
|
Methane Model
|
Methane Modeling Framework
|
Atmospheric Climate and Chemistry Model (GEOS-Chem) The GEOS-chem model is a global 3-D model of atmospheric composition driven by assimilated meteorological observations from the Goddard Earth Observing System (GEOS) of the NASA Global Modeling and Assimilation Office. It is applied by research groups around the world to a wide range of atmospheric composition problems, including future climates and planetary atmospheres using general circulation model meteorology to drive the model. Central management and support of the model is provided by the Atmospheric Chemistry Modeling Group at Harvard University. See web below to get more information: http://www-as.harvard.edu/chemistry/trop/geos/
|
Global sensitivity analysis and Bayesian inference framework
Our framework was based on Bayes' theorem: Pr(θ|V) Pr(V|θ) Pr(θ) where Pr(θ|V) is the posterior after Bayesian inference conditioned on available observations V (hereafter the bold letter indicates a matrix). θis the matrix of parameters and TEM outputs (e.g., GPP) and V is the matrix of observation or the matrix of the differences between prior simulations and the corresponding observations, whose element Vij denotes the type j data V(·)j at time step i. Pr(V|θ) is the likelihood function, which will be calculated as a function of TEM Monte Carlo simulations and the available eddy flux data. Pr(θ) is the prior of the TEM parameters and our estimated C fluxes (e.g., GPP, RESP and NEP) and EET. To address our research questions, we first conducted TEMensemble simulations with parameter priors. Second, the likelihood function Pr(V|θ) was calculated based on model simulations and observations. Third, the global sensitivity analysis was applied, and fourth, the Bayesian inference was conducted.
|
