Quantifying uncertainty from large‐scale model predictions of forest carbon dynamics

Linking environmental computer simulation models and geographic information systems (GIS) is now a common practice to scale up simulations of complex ecosystem processes for decision support. Unfortunately, several important issues of upscaling using GIS are rarely considered; in particular scale dependency of models, availability of input data, support of input and validation data, and uncertainty in prediction including error propagation from the GIS. We linked the biogeochemical Forest-DNDC model to a GIS database to predict growth of Eucalyptus globulus plantations at two different scales ( � 0.045 ha plot � 1 scale and � 100 ha grid � 1 scale) across Victoria, in south-eastern Australia. Results showed that Forest-DNDC was not scale dependent across the range of scales investigated. Reduced availability of input data at the larger scale may introduce severe prediction errors, but did not require adjustment of the model in this study. Differences in the support of input and validation data led to an underestimation of predictive precision but an overestimation of prediction accuracy. Increasing data support, produced a high level of prediction accuracy (¯e% ¼� 3:54%), but a medium level of predictive precision (r 2 5 0.474, ME 5 0.318) after statistical validation. GIS error contribution could be detected but was not readily or reliably quantified. In a regional case study for 2653 ha of E. globulus plantations, the linked model GIS system estimated a total standing biomass of 95 260 t C for mid-2003 and a net CO2 balance of � 45 671 t CO2Cy r � 1 for the entire year of 2002. This study showed that regional predictions of forest growth and carbon sequestration can be produced with greater confidence after a comprehensive assessment of upscaling issues.

[1]  Carl C. Trettin,et al.  Modeling Impacts of Management on Carbon Sequestration and Trace Gas Emissions in Forested Wetland Ecosystems , 2004 .

[2]  Yude Pan,et al.  BIOMASS AND NPP ESTIMATION FOR THE MID-ATLANTIC REGION (USA) USING PLOT-LEVEL FOREST INVENTORY DATA , 2001 .

[3]  Steve Frolking,et al.  A model of nitrous oxide evolution from soil driven by rainfall events: 2. Model applications , 1992 .

[4]  László Pásztor,et al.  Comparison of approaches for estimating carbon sequestration at the regional scale , 2002 .

[5]  K. Willson.  Climate and Soil , 1985 .

[6]  Changsheng Li,et al.  An integrated model of soil, hydrology, and vegetation for carbon dynamics in wetland ecosystems , 2002 .

[7]  J. Nash,et al.  River flow forecasting through conceptual models part I — A discussion of principles☆ , 1970 .

[8]  M. Battaglia,et al.  Coarse root biomass for eucalypt plantations in Tasmania, Australia: sources of variation and methods for assessment , 2003, Trees.

[9]  B. Zierl,et al.  A water balance model to simulate drought in forested ecosystems and its application to the entire forested area in Switzerland , 2001 .

[10]  K. Butterbach-Bahl,et al.  Quantifying the regional source strength of N-trace gases across agricultural and forest ecosystems with process based models , 2004, Plant and Soil.

[11]  W. Ju,et al.  Spatial distribution of carbon sources and sinks in Canada’s forests , 2003 .

[12]  R. Mickler,et al.  Regional estimation of current and future forest biomass. , 2002, Environmental pollution.

[13]  Stefano Tarantola,et al.  Sensitivity and uncertainty analysis in spatial modelling based on GIS , 2000 .

[14]  P. Sands.  Modelling Canopy Production. I. Optimal Distribution of Photosynthetic Resources , 1995 .

[15]  Klaus Butterbach-Bahl,et al.  A process‐oriented model of N2O and NO emissions from forest soils: 2. Sensitivity analysis and validation , 2000 .

[16]  Paula Soares,et al.  A critical look at procedures for validating growth and yield models. , 2003 .

[17]  Nicholas C. Coops,et al.  Assessing forest productivity at local scales across a native eucalypt forest using a process model, 3PG-SPATIAL , 2001 .

[18]  Changsheng Li,et al.  A model of nitrous oxide evolution from soil driven by rainfall events: 1. Model structure and sensitivity , 1992 .

[19]  Edzer Pebesma,et al.  Spatial aggregation and soil process modelling , 1999 .

[20]  Changsheng Li,et al.  A process-oriented model of N2O and NO emissions from forest soils: 1. Model development , 2000 .

[21]  Gerard B. M. Heuvelink.  Uncertainty analysis in environmental modelling under a change of spatial scale , 1998 .

[22]  Changsheng Li,et al.  Assessing productivity and carbon sequestration capacity of Eucalyptus globulus plantations using the process model Forest-DNDC: Calibration and validation , 2006 .

[23]  Stefano Tarantola,et al.  Uncertainty and sensitivity analysis: tools for GIS-based model implementation , 2001, Int. J. Geogr. Inf. Sci..

[24]  P. Reich,et al.  Extrapolating leaf CO2 exchange to the canopy: a generalized model of forest photosynthesis compared with measurements by eddy correlation , 1996, Oecologia.

[25]  P. Heuberger,et al.  Calibration of process-oriented models , 1995 .

[26]  D. Legates,et al.  Evaluating the use of “goodness‐of‐fit” Measures in hydrologic and hydroclimatic model validation , 1999 .

[27]  Scott V. Ollinger,et al.  Estimating regional forest productivity and water yield using an ecosystem model linked to a GIS , 1998, Landscape Ecology.

[28]  Reiner Wassmann,et al.  Modeling greenhouse gas emissions from rice‐based production systems: Sensitivity and upscaling , 2004 .

[29]  M. Battaglia,et al.  An application of terrain and environmental modelling in a large-scale forestry experiment , 1999 .

[30]  A. Camacho.  Comparison of Approaches to Social Choice , 1982 .

[31]  M. Jansen,et al.  Prediction error through modelling concepts and uncertainty from basic data , 1998 .

[32]  M. Battaglia,et al.  Data input quality and resolution effects on regional and local scale Eucalyptus globulus productivity predictions in north-east Tasmania , 2002 .

[33]  Ge Sun,et al.  Linkage of MIKE SHE to Wetland-DNDC for carbon budgeting and anaerobic biogeochemistry simulation , 2005 .

[34]  Jean-Francois Lamarque,et al.  Variations in the predicted spatial distribution of atmospheric nitrogen deposition and their impact on carbon uptake by terrestrial ecosystems , 1997 .

[35]  Gerard B.M. Heuvelink,et al.  Uncertainty analysis in environmental modelling under a change of spatial scale , 1998, Nutrient Cycling in Agroecosystems.

[36]  Changsheng Li,et al.  Regional inventory of nitric oxide and nitrous oxide emissions for forest soils of southeast Germany using the biogeochemical model PnET‐N‐DNDC , 2001 .

[37]  Gerard B. M. Heuvelink,et al.  Error Propagation in Environmental Modelling with GIS , 1998 .

[38]  W. Parton,et al.  TERRESTRIAL NPP: TOWARD A CONSISTENT DATA SET FORGLOBAL MODEL EVALUATION , 1999 .

[39]  Richard T. Conant,et al.  BEST PRACTICES IN PREDICTION FOR DECISION‐MAKING: LESSONS FROM THE ATMOSPHERIC AND EARTH SCIENCES , 2003 .

[40]  John O. Carter,et al.  Using spatial interpolation to construct a comprehensive archive of Australian climate data , 2001, Environ. Model. Softw..

[41]  M. Battaglia,et al.  Applying PROMOD spatially across Tasmania with sensitivity analysis to screen for prospective Eucalyptus globulus plantation sites , 2001 .

[42]  CHANG SHENG,et al.  Regional application of PnET-N-DNDC for estimating the N 2 O source strength of tropical rainforests in the Wet Tropics of Australia , 2004 .

[43]  Klaus Butterbach-Bahl,et al.  Regional application of PnET‐N‐DNDC for estimating the N2O source strength of tropical rainforests in the Wet Tropics of Australia , 2005 .

[44]  D. Mummery,et al.  Application of process-based models to forest management: experience with PROMOD, a simple plantation productivity model. , 2000, Tree physiology.

[45]  Philip Smith,et al.  Regional estimates of carbon sequestration potential: linking the Rothamsted Carbon Model to GIS databases , 1998, Biology and Fertility of Soils.