Bayesian Methods for Quantifying and Reducing Uncertainty and Error in Forest Models

Purpose of reviewForest models are tools for analysis and prediction of productivity and other services. Model outputs can only be useful if possible errors in inputs and model structure are recognized. However, errors cannot be quantified directly, making uncertainty inevitable. In this paper, we aim to clarify terminological confusion around the concepts of error and uncertainty and review current methods for addressing uncertainty in forest modelling.Recent findingsModellers increasingly recognize that all uncertainties—in data, model inputs and model structure—can be represented using probability distributions. This has stimulated the use of Bayesian methods for quantifying and reducing uncertainty and error in models of forests and other vegetation. The Achilles’ heel of Bayesian methods has always been their computational demand, but solutions are being found.SummaryWe conclude that future work will likely include (1) more use of Bayesian methods, (2) more use of hierarchical modelling, (3) replacement of model spin-up by Bayesian calibration, (4) more use of ensemble modelling and Bayesian model averaging, (5) new ways to account for model structural error in calibration, (6) better software for Bayesian calibration of complex models, (7) faster Markov chain Monte Carlo algorithms, (8) more use of model emulators, (9) novel uncertainty visualization techniques, (10) more use of graphical modelling and (11) more use of risk analysis.

[1]  Andreas Huth,et al.  Connecting dynamic vegetation models to data – an inverse perspective , 2012 .

[2]  Ron Smith,et al.  Bayesian calibration of process-based forest models: bridging the gap between models and data. , 2005, Tree physiology.

[3]  N. Metropolis,et al.  Equation of State Calculations by Fast Computing Machines , 1953, Resonance.

[4]  D. Cameron,et al.  Process-based simulation of growth and overwintering of grassland using the BASGRA model , 2016 .

[5]  A. O'Hagan,et al.  Bayesian calibration of computer models , 2001 .

[6]  Jonathan Rougier,et al.  Probabilistic Inference for Future Climate Using an Ensemble of Climate Model Evaluations , 2007 .

[7]  Ke Zhang,et al.  Variation in stem mortality rates determines patterns of above‐ground biomass in Amazonian forests: implications for dynamic global vegetation models , 2016, Global change biology.

[8]  Ranga B. Myneni,et al.  Impact of droughts on the carbon cycle in European vegetation: a probabilistic risk analysis using six vegetation models , 2014 .

[9]  R. Lark,et al.  Communicating the uncertainty in estimated greenhouse gas emissions from agriculture , 2015, Journal of environmental management.

[10]  Ivan A. Janssens,et al.  Bayesian comparison of six different temperature-based budburst models for four temperate tree species , 2012 .

[11]  P. Levy,et al.  Estimation of cumulative fluxes of nitrous oxide: uncertainty in temporal upscaling and emission factors , 2017 .

[12]  James S. Clark,et al.  The impacts of increasing drought on forest dynamics, structure, and biodiversity in the United States , 2016, Global change biology.

[13]  S. Kellomäki,et al.  Evaluation of past and future changes in European forest growth by means of four process-based models , 2008 .

[14]  Francesco Minunno,et al.  Selecting Parameters for Bayesian Calibration of a Process-Based Model: A Methodology Based on Canonical Correlation Analysis , 2013, SIAM/ASA J. Uncertain. Quantification.

[15]  A. Rammig,et al.  Modelling CO2 Impacts on Forest Productivity , 2015, Current Forestry Reports.

[16]  M. Salam,et al.  Comparing Simulated and Measured Values Using Mean Squared Deviation and its Components , 2000 .

[17]  Mark A. Sutton,et al.  Uncertainties in the relationship between atmospheric nitrogen deposition and forest carbon sequestration , 2008 .

[18]  C. Reyer.  Forest Productivity Under Environmental Change—a Review of Stand-Scale Modeling Studies , 2015, Current Forestry Reports.

[19]  Margarida Tomé,et al.  Using stand-scale forest models for estimating indicators of sustainable forest management , 2012 .

[20]  Margarida Tomé,et al.  Using a Bayesian framework and global sensitivity analysis to identify strengths and weaknesses of two process-based models differing in representation of autotrophic respiration , 2013, Environ. Model. Softw..

[21]  T. Bayes.  An essay towards solving a problem in the doctrine of chances , 2003 .

[22]  Catherine A Calder,et al.  Accounting for uncertainty in ecological analysis: the strengths and limitations of hierarchical statistical modeling. , 2009, Ecological applications : a publication of the Ecological Society of America.

[23]  S. Richardson,et al.  Soil–climate interactions explain variation in foliar, stem, root and reproductive traits across temperate forests , 2016 .

[24]  Thomas Rötzer,et al.  Models for supporting forest management in a changing environment , 2011 .

[25]  J. Yeluripati,et al.  A Bayesian framework for model calibration, comparison and analysis: Application to four models for the biogeochemistry of a Norway spruce forest , 2011 .

[26]  Rob Kooper,et al.  On improving the communication between models and data. , 2013, Plant, cell & environment.

[27]  M. G. Ryan,et al.  The likely impact of elevated [CO2], nitrogen deposition, increased temperature and management on carbon sequestration in temperate and boreal forest ecosystems: a literature review. , 2007, The New phytologist.

[28]  M. Flechsig,et al.  Integrating parameter uncertainty of a process-based model in assessments of climate change effects on forest productivity , 2016, Climatic Change.

[29]  R. B. Jackson,et al.  Forest biogeochemistry in response to drought , 2016, Global change biology.

[30]  Richard E Chandler,et al.  Exploiting strength, discounting weakness: combining information from multiple climate simulators , 2011, Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences.

[31]  Kiona Ogle,et al.  Bayesian Data—Model Integration in Plant Physiological and Ecosystem Ecology , 2008 .

[32]  Gerard B. M. Heuvelink,et al.  Environmental change impacts on the C- and N-cycle of European forests: a model comparison study , 2012 .

[33]  Martin Wattenbach,et al.  Bayesian calibration as a tool for initialising the carbon pools of dynamic soil models , 2009 .

[34]  P. Levy,et al.  The Effect of Nitrogen Enrichment on the Carbon Sink in Coniferous Forests: Uncertainty and Sensitivity Analyses of Three Ecosystem Models , 2004 .

[35]  Kiona Ogle,et al.  Hierarchical bayesian statistics: merging experimental and modeling approaches in ecology. , 2009, Ecological applications : a publication of the Ecological Society of America.

[36]  R. Milne,et al.  Integrating remote sensing datasets into ecological modelling: a Bayesian approach , 2008 .

[37]  Rognvald I. Smith,et al.  Extending a Bayesian Belief Network for ecosystem evaluation , 2012 .

[38]  David Cameron,et al.  Correcting errors from spatial upscaling of nonlinear greenhouse gas flux models , 2017, Environ. Model. Softw..

[39]  James S. Clark,et al.  Capturing diversity and interspecific variability in allometries: A hierarchical approach , 2008 .

[40]  Benjamin Poulter,et al.  Emergent climate and CO2 sensitivities of net primary productivity in ecosystem models do not agree with empirical data in temperate forests of eastern North America , 2017, Global change biology.

[41]  George Kuczera,et al.  Bayesian analysis of input uncertainty in hydrological modeling: 1. Theory , 2006 .

[42]  Amanda M. Thomson,et al.  Toward Bayesian uncertainty quantification for forestry models used in the United Kingdom Greenhouse Gas Inventory for land use, land use change, and forestry , 2010 .

[43]  Mike Pearson,et al.  Visualizing Uncertainty About the Future , 2022 .

[44]  Noel A Cressie,et al.  Statistics for Spatio-Temporal Data , 2011 .

[45]  Atul K. Jain,et al.  Using ecosystem experiments to improve vegetation models , 2015 .

[46]  A. Mäkelä,et al.  Bayesian calibration, comparison and averaging of six forest models, using data from Scots pine stands across Europe , 2013 .