Does Your Terrestrial Model Capture Key Arctic-Boreal Relationships?: Functional Benchmarks in the ABoVE Model Benchmarking System

Eric Stofferahn, Jet Propulsion Laboratory / Caltech, ericstofferahn@gmail.com (Presenter)
Joshua B. Fisher, NASA JPL, jbfisher@jpl.nasa.gov
Daniel Hayes, University of Maine, daniel.j.hayes@maine.edu
Christopher R Schwalm, Woods Hole Research Center, schwalm.christopher@gmail.com
Deborah Nicole Huntzinger, Northern Arizona University, deborah.huntzinger@nau.edu
Wouter Hantson, University of Maine, wouter.hantson@maine.edu

The Arctic-Boreal Region (ABR) is a major source of uncertainties for terrestrial biosphere model (TBM) simulations. These uncertainties are precipitated by a lack of observational data from the region, affecting the parameterizations of cold environment processes in the models. Addressing these uncertainties requires a coordinated effort of data collection and integration of the following key indicators of the ABR ecosystem: disturbance, vegetation / ecosystem structure and function, carbon pools and biogeochemistry, permafrost, and hydrology. We are continuing to develop the model-data integration framework for NASA’s Arctic Boreal Vulnerability Experiment (ABoVE), wherein data collection is driven by matching observations and model outputs to the ABoVE indicators via the ABoVE Grid and Projection. The data are used as reference datasets for a benchmarking system which evaluates TBM performance with respect to ABR processes. The benchmarking system utilizes two types of performance metrics to identify model strengths and weaknesses: standard metrics, based on the International Land Model Benchmarking (ILaMB) system, which relate a single observed variable to a single model output variable, and functional benchmarks, wherein the relationship of one variable to one or more variables (e.g. the dependence of vegetation structure on snow cover, the dependence of active layer thickness (ALT) on air temperature and snow cover) is ascertained in both observations and model outputs. This in turn provides guidance to model development teams for reducing uncertainties in TBM simulations of the ABR.

Associated Project(s): 

• Fisher-01

Poster Location ID: 66

Session Assigned: Modeling