Votre recherche
Résultats 5 ressources
-
It is increasingly being recognized that global ecological research requires novel methods and strategies in which to combine process‐based ecological models and data in cohesive, systematic ways. In process‐based model applications, inherent spatial and temporal heterogeneities found within terrestrial ecosystems may lead to the uncertainties of model predictions. To reduce simulation uncertainties due to inaccurate model parameters, the Markov Chain Monte Carlo (MCMC) method was applied in this study to improve the estimations of four key parameters used in the process‐based ecosystem model of TRIPLEX‐FLUX. These four key parameters include a maximum photosynthetic carboxylation rate of 25°C (Vmax), an electron transport (Jmax) light‐saturated rate within the photosynthetic carbon reduction cycle of leaves, a coefficient of stomatal conductance (m), and a reference respiration rate of 10°C (R10). Seven forest flux tower sites located across North America were used to investigate and facilitate understanding of the daily variation in model parameters for three deciduous forests, three evergreen temperate forests, and one evergreen boreal forest. Eddy covariance CO 2 exchange measurements were assimilated to optimize the parameters in the year 2006. After parameter optimization and adjustment took place, net ecosystem production prediction significantly improved (by approximately 25%) compared to the CO 2 flux measurements taken at the seven forest ecosystem sites. Results suggest that greater seasonal variability occurs in broadleaf forests in respect to the selected parameters than in needleleaf forests. This study also demonstrated that the model‐data fusion approach by incorporating MCMC method is able to better estimate parameters and improve simulation accuracy for different ecosystems located across North America.
-
Our current understanding of terrestrial carbon processes is represented in various models used to integrate and scale measurements of CO 2 exchange from remote sensing and other spatiotemporal data. Yet assessments are rarely conducted to determine how well models simulate carbon processes across vegetation types and environmental conditions. Using standardized data from the North American Carbon Program we compare observed and simulated monthly CO 2 exchange from 44 eddy covariance flux towers in North America and 22 terrestrial biosphere models. The analysis period spans ∼220 site‐years, 10 biomes, and includes two large‐scale drought events, providing a natural experiment to evaluate model skill as a function of drought and seasonality. We evaluate models' ability to simulate the seasonal cycle of CO 2 exchange using multiple model skill metrics and analyze links between model characteristics, site history, and model skill. Overall model performance was poor; the difference between observations and simulations was ∼10 times observational uncertainty, with forested ecosystems better predicted than nonforested. Model‐data agreement was highest in summer and in temperate evergreen forests. In contrast, model performance declined in spring and fall, especially in ecosystems with large deciduous components, and in dry periods during the growing season. Models used across multiple biomes and sites, the mean model ensemble, and a model using assimilated parameter values showed high consistency with observations. Models with the highest skill across all biomes all used prescribed canopy phenology, calculated NEE as the difference between GPP and ecosystem respiration, and did not use a daily time step.