Parameter optimization, sensitivity, and uncertainty analysis of an ecosystem model at a forest flux tower site in the United States

Abstract

Ecosystem models are useful tools for understanding ecological processes and for sustainable management of resources. In biogeochemical field, numerical models have been widely used for investigating carbon dynamics under global changes from site to regional and global scales. However, it is still challenging to optimize parameters and estimate parameterization uncertainty for complex process-based models such as the Erosion Deposition Carbon Model (EDCM), a modified version of CENTURY, that consider carbon, water, and nutrient cycles of ecosystems. This study was designed to conduct the parameter identifiability, optimization, sensitivity, and uncertainty analysis of EDCM using our developed EDCM-Auto, which incorporated a comprehensive R package-Flexible Modeling Framework (FME) and the Shuffled Complex Evolution (SCE) algorithm. Using a forest flux tower site as a case study, we implemented a comprehensive modeling analysis involving nine parameters and four target variables (carbon and water fluxes) with their corresponding measurements based on the eddy covariance technique. The local sensitivity analysis shows that the plant production-related parameters (e.g., PPDF1 and PRDX) are most sensitive to the model cost function. Both SCE and FME are comparable and performed well in deriving the optimal parameter set with satisfactory simulations of target variables. Global sensitivity and uncertainty analysis indicate that the parameter uncertainty and the resulting output uncertainty can be quantified, and that the magnitude of parameter-uncertainty effects depends on variables and seasons. This study also demonstrates that using the cutting-edge R functions such as FME can be feasible and attractive for conducting comprehensive parameter analysis for ecosystem modeling.

Ecosystem carbon stocks and sequestration potential of federal lands across the conterminous United States

“…The endpoints and paths are tightly coupled with and regulated by nitrogen and water cycles, land uses, management activities, etc. For model evaluation, we selected the most sensitive parameter (PRDX) to calibrate the plant production using observed grain yield for croplands (e.g., corn and soybean) and the moderate-resolution imaging spectroradiometer (MODIS) NPP for noncroplands (e.g., forest and grasslands) (22). The models have been well calibrated and validated at regional and national scales (18,19,22 (23, 24) was used to produce spatially explicit LULC maps consistent with the IPCC SRES scenarios.…”

Quantitative attribution of major driving forces on soil organic carbon dynamics

“…This carbon pool model (EDCM) focuses on tracking the dynamics of carbon storage in each pool. EDCM was updated to include a generic autocalibration package (Shuffled Complex Evolution (SCE) [ Duan et al ., ] and R‐based Flexible Modeling Environment (FME) [ Soetaert and Petzoldt , ] for site and regional model calibration, and known as EDCM‐Auto [ Wu et al ., ]. Driven by its interface—General Ensemble Modeling System (GEMS), EDCM‐Auto has been used to assess the carbon stocks and fluxes under changing climate and land covers for the baseline and projection periods across the conterminous United States [ Liu et al ., ; Zhu , ].…”

Projection of corn production and stover-harvesting impacts on soil organic carbon dynamics in the U.S. Temperate Prairies

“…From this dispute, however, it is clear that quantifying the spatially explicit effects of soil erosion/deposition and the resulting carbon movement on SOC dynamics is a good topic and deserves intensive investigations in our further work. Application of numerical models usually involves parameter optimization, and thus the analysis of parameter uncertainty and quantification of its effects on model output (e.g., SOC) are attractive and challenging in the modeling field 49 especially for large-scale studies 50 51 . From our experience with parameter uncertainty analysis 49 52 53 , the R packages such as Flexible Modeling Environment (FME) could be a potential tool for this purpose.…”

“…Application of numerical models usually involves parameter optimization, and thus the analysis of parameter uncertainty and quantification of its effects on model output (e.g., SOC) are attractive and challenging in the modeling field 49 especially for large-scale studies 50 51 . From our experience with parameter uncertainty analysis 49 52 53 , the R packages such as Flexible Modeling Environment (FME) could be a potential tool for this purpose. In addition to the parameter uncertainty, there are other uncertainties to consider such as those existing in the initial SOC data (from SSURGO and STATSGO), the scenario climate data (from GCMs), crop management data, and the implicit assumption of similar occurrence of pests and diseases in the future.…”