Assessing parameter variability in a photosynthesis model within and between plant functional types using global Fluxnet eddy covariance data

Agricultural and Forest Meteorology (Impact Factor: 3.89). 01/2011; 151(1):22-38. DOI: 10.1016/j.agrformet.2010.08.013
Source: OAI

ABSTRACT The vegetation component in climate models has advanced since the late 1960s from a uniform prescription of surface parameters to plant functional types (PFTs). PFTs are used in global land-surface models to provide parameter values for every model grid cell. With a simple photosynthesis model we derive parameters for all site years within the Fluxnet eddy covariance data set. We compare the model parameters within and between PFTs and statistically group the sites. Fluxnet data is used to validate the photosynthesis model parameter variation within a PFT classification. Our major result is that model parameters appear more variable than assumed in PFTs. Simulated fluxes are of higher quality when model parameters of individual sites or site years are used. A simplification with less variation in model parameters results in poorer simulations. This indicates that a PFT classification introduces uncertainty in the variation of the photosynthesis and transpiration fluxes. Statistically derived groups of sites with comparable model parameters do not share common vegetation types or climates. A simple PFT classification does not reflect the real photosynthesis and transpiration variation. Although site year parameters give the best predictions, the parameters are generally too specific to be used in a global study. The site year parameters can be further used to explore the possibilities of alternative classification schemes. © 2010 Elsevier B.V.

Download full-text


Available from: Han (A.J.) Dolman, Aug 17, 2015
  • Source
    • "In parallel, two independent efforts simultaneously used data constraints from several sites to assess the degree of improvement of the simulated fluxes depending on the " generic criterion " used for the optimized parameters (Groenendijk et al., 2011; Kuppel et al., 2012). The study of Groenendijk et al. (2011), conducted at over a hundred locations across several PFTs, found that the cross-site parameter variability after optimization explained the poorer performances of grouping sites by PFT, while no such discrepancy appeared in Kuppel et al. (2012), whose study was however limited to temperate deciduous broadleaf forests. "
    [Show abstract] [Hide abstract]
    ABSTRACT: This study uses a variational data assimilation framework to simultaneously constrain a global ecosystem model with eddy covariance measurements of daily net ecosystem exchange (NEE) and latent heat (LE) fluxes from a large number of sites grouped in seven plant functional types (PFTs). It is an attempt to bridge the gap between the numerous site-specific parameter optimization works found in the literature and the generic parameterization used by most land surface models within each PFT. The present multisite approach allows deriving PFT-generic sets of optimized parameters enhancing the agreement between measured and simulated fluxes at most of the sites considered, with performances often comparable to those of the corresponding site-specific optimizations. Besides reducing the PFT-averaged model–data root-mean-square difference (RMSD) and the associated daily output uncertainty, the optimization improves the simulated CO2 balance at tropical and temperate forests sites. The major site-level NEE adjustments at the seasonal scale are reduced amplitude in C3 grasslands and boreal forests, increased seasonality in temperate evergreen forests, and better model–data phasing in temperate deciduous broadleaf forests. Conversely, the poorer performances in tropical evergreen broadleaf forests points to deficiencies regarding the modelling of phenology and soil water stress for this PFT. An evaluation with data-oriented estimates of photosynthesis (GPP – gross primary productivity) and ecosystem respiration (Reco) rates indicates distinctively improved simulations of both gross fluxes. The multisite parameter sets are then tested against CO2 concentrations measured at 53 locations around the globe, showing significant adjustments of the modelled seasonality of atmospheric CO2 concentration, whose relevance seems PFT-dependent, along with an improved interannual variability. Lastly, a globalscale evaluation with remote sensing NDVI (normalized difference vegetation index) measurements indicates an improvement of the simulated seasonal variations of the foliar cover for all considered PFTs.
    Geoscientific Model Development 11/2014; 7(6):2581–2597. DOI:10.5194/gmd-7-2581-2014 · 6.09 Impact Factor
  • Source
    • "This model, originally formulated by Farquhar et al. (1980) (here termed the FvCB model) allows for inferences to be made about biochemical limitations to leaf and canopy functioning, overlain by environmental constraints (Long & Bernacchi 2003). The original FvCB model (Farquhar et al. 1980) and its subsequent modifications (Sharkey 1985; von Caemmerer 2000; Sharkey et al. 2007) successfully predicts photosynthesis under a very wide range of conditions and has been applied to scales ranging from the chloroplast (von Caemmerer 2013) to forest canopies (Groenendijk et al. 2011) and biomes (Kattge et al. 2009; Bonan et al. 2011). A unique element of the FvCB model is the ability to estimate photorespiratory CO2 efflux concurrent with photosynthetic CO2 influx as component processes contributing to the net CO2 assimilation rate of leaves (Sage & Sharkey 1987; Busch 2013). "
    [Show abstract] [Hide abstract]
    ABSTRACT: Leaf photosynthetic CO2 responses can provide insight into how major nutrients such as phosphorus (P) constrain leaf CO2 -assimilation rates (Anet ). However, triose-phosphate limitations are rarely employed in the classic photosynthesis model and it is uncertain to what extent these limitations occur in field situations. In contrast to predictions from biochemical theory of photosynthesis, we found consistent evidence in the field of lower Anet in high [CO2 ] and low [O2 ] than at ambient [O2 ]. For ten species of trees and shrubs across a range of soil P availability in Australia, none of them showed a positive response of Anet at saturating [CO2 ] (i.e. Amax ) to 2 kPa O2 . Three species showed >20% reductions in Amax in low [O2 ], a phenomenon explained by orthophosphate (Pi ) savings during photorespiration. These species, with largest photosynthetic capacity and Pi > 2 mmol P m(-2) , rely the most on additional Pi made available from photorespiration rather than species growing in P-impoverished soils. The results suggest that rarely-used adjustments to a biochemical photosynthesis model are useful for predicting Amax , and give insight into the biochemical limitations of photosynthesis rates at a range of leaf P concentrations. Phosphate limitations to photosynthetic capacity are likely more common in the field than previously considered.
    Plant Cell and Environment 10/2014; 38(6). DOI:10.1111/pce.12468 · 5.91 Impact Factor
  • Source
    • "This is similar to the conventional parameterization of ecosystem models that use a single value for each parameter and do not consider the uncertainty and/or variability of the parameter. Despite numerous studies on parameter estimation, few studies have assessed the variability of parameters within a given PFT and across PFTs (Groenendijk et al., 2011; Xiao et al., 2011). The Upper Midwest region of northern Wisconsin and Michigan , USA, is a highly heterogeneous mixture of upland forests and lowland wetlands (Fig. 1). "
    [Show abstract] [Hide abstract]
    ABSTRACT: Models have been widely used to estimate carbon fluxes at regional scales, and the uncertainty of mod-eled fluxes, however, has rarely been quantified and remains a challenge. A quantitative uncertaintyassessment of regional flux estimates is essential for better understanding of terrestrial carbon dynamicsand informing carbon and climate decision-making. We use a simple ecosystem model, eddy covariance(EC) flux observations, and a model-data fusion approach to assess the uncertainty of regional carbonflux estimates for the Upper Midwest region of northern Wisconsin and Michigan, USA. We combine netecosystem exchange (NEE) observations and an adaptive Markov chain Monte Carlo (MCMC) approachto quantify the parameter uncertainty of the Diagnostic Carbon Flux Model (DCFM). Our MCMC approacheliminates the need for an initial equilibration or “burn-in” phase of the random walk, and also improvesthe performance of the algorithm for parameter optimization. For each plant functional type (PFT), weuse NEE observations from multiple EC sites to estimate parameters, and the resulting parameter esti-mates are more representative of the PFT than estimates based on observations from a single site. Aprobability density function (PDF) is generated for each parameter, and the spread of the PDF providesan estimate of parameter uncertainty. We then apply the model with parameter PDFs to estimate NEEfor each grid cell across our study region, and propagate the parameter uncertainty through simulationsto produce probabilistic flux estimates. Over the period from 2001 to 2007, the mean annual NEE ofthe region was estimated to be −30.0 Tg C yr−1, and the associated uncertainty as measured by standarddeviation was ± 7.6 Tg C yr−1. Uncertainty in parameters can lead to a large uncertainty to estimates ofregional carbon fluxes, and our model-data approach can provide uncertainty bounds to regional carbonfluxes. Future research is needed to apply our approach to more complex ecosystem models, assess theusefulness, validity, and alternatives of the PFT and vegetation type concepts, and to fully quantify theuncertainty of regional carbon fluxes by incorporating other sources of uncertainty.
    Agricultural and Forest Meteorology 04/2014; 189-190:175-186. DOI:10.1016/j.agrformet.2014.01.022 · 3.89 Impact Factor
Show more