Understanding variability and temporal trends in biosphere-atmosphere CO2 exchange through integrating models with data

March 19, 2012
Trevor Keenan | Harvard University

Process-based models of atmosphere-biosphere interactions, along with empirical data mining techniques, are the primary tools used for scaling disparate observations through space and time. In the past few decades they have been developed in tandem with our understanding of ecological theory, resulting in models of various levels of complexity and detail. Model intercomparisons, however, show a large range in model performance, with no clear consensus as to whether model structural error (process mis-representation) or mis-parameterization is to blame. One potential reason for this lies in difficulties in using data sources at different scales to adequately test model performance and the common reliance on uni-variate model validations. Another is the lack of quantification of the uncertainty associated with model projections. In order to advance our ability to make policy-actionable projections of the future evolution of the earth system, we need to address these issues.

In this talk I will assess our ability to model land-atmosphere CO2 exchange at different spatial and temporal scales, both in the present climate and under future climate change. The analysis makes heavy use of model-data fusion techniques, which constitute a powerful framework by which to combine models with various data streams. Model benchmarking tools, such as empirical data mining techniques, also provide a strong alternative model evaluation. To illustrate the potential benefits of such an approach, we assess the performance of 17 process-based models of atmosphere-biosphere interactions, and two data mining tools, across 11 long-term eddy covariance forest sites. The results highlight details of model performance often overlooked by conventional model-data comparisons, and quantify the degree of coupling of terrestrial carbon sequestration to climate anomalies at multiple sites and time scales.

Speaker Details

Born and bred on a farm outside a small town called Longford, in the heart of Ireland. I started my young career with a get rich quick plan, studying financial maths in Dublin. The good pay wasn´t worth it though, and I found myself far more motivated by the complex questions of the terrestrial biosphere. I then studied a really nice Masters at the University of York (/www.), aimed at applying mathematical techniques to the living environment.

I made my second home at CREAF, in Barcelona, Spain, where I undertook my PhD. There I learned the intricacies of modelling terrestrial carbon and water cycles in a drought prone region of the world. After a year as a Post-Doc at CREAF I moved to the Richardson Lab., Harvard University. Here I have focused on quantifying uncertainty in model projections through time and space by informing terrestrial biosphere models with data. It is my firm conviction that we need to merge models and data in a common framework if we are to give future projections of earth-system dynamics that can inform policy decisions. The data is there, the models are there, we just need to put them all together.