Assessing model sensitivity and uncertainty across multiple Free-Air CO2 Enrichment experiments.

Abstract:

As atmospheric levels of carbon dioxide levels continue to increase, it is critical that terrestrial ecosystem models can accurately predict ecological responses to the changing environment. Current predictions of net primary productivity (NPP) in response to elevated atmospheric CO₂ concentrations are highly variable and contain a considerable amount of uncertainty. It is necessary that we understand which factors are driving this uncertainty. The Free-Air CO₂ Enrichment (FACE) experiments have equipped us with a rich data source that can be used to calibrate and validate these model predictions.

To identify and evaluate the assumptions causing inter-model differences we performed model sensitivity and uncertainty analysis across ambient and elevated CO₂ treatments using the Data Assimilation Linked Ecosystem Carbon (DALEC) model and the Ecosystem Demography Model (ED2), two process-based models ranging from low to high complexity respectively. These modeled process responses were compared to experimental data from the Kennedy Space Center Open Top Chamber Experiment, the Nevada Desert Free Air CO₂ Enrichment Facility, the Rhinelander FACE experiment, the Wyoming Prairie Heating and CO₂ Enrichment Experiment, the Duke Forest Face experiment and the Oak Ridge Experiment on CO₂ Enrichment.

By leveraging data access proxy and data tilling services provided by the BrownDog data curation project alongside analysis modules available in the Predictive Ecosystem Analyzer (PEcAn), we produced automated, repeatable benchmarking workflows that are generalized to incorporate different sites and ecological models. Combining the observed patterns of uncertainty between the two models with results of the recent FACE-model data synthesis project (FACE-MDS) can help identify which processes need further study and additional data constraints. These findings can be used to inform future experimental design and in turn can provide informative starting point for data assimilation.