Sensitivity of Vegetation Structure and Dynamics to Climate Data

Steve Flanagan, University of Maryland Geography Dept, saf204@gmail.com (Presenter)
George Hurtt, University of Maryland, gchurtt@umd.edu
Justin Fisk, University of Maryland, fisk@umd.edu
Oliver Rourke, University of Maryland, oliverrourke@gmail.com
Louise Parsons Chini, University of New Hampshire, louise.chini@unh.edu
Amanda Armstrong, UMD, aha@umd.edu

Climate data sets drive ecosystem models that can be exceptionally sensitive to those inputs. Identifying which climate inputs have the strongest correlation to which model outputs, the sensitivity of these inputs, and the effect of spatially averaging these inputs aids the identification of which ecosystem model outputs are most sensitive to changes in climate and the effects of averaging on model prediction error. Appropriate spatial scale for climate data sets and which inputs exert control over model outputs is an emerging issue. An advanced ecosystem model, the Ecosystem Demography Model (ED), was initialized with multiple climate data sets to determine the sensitivities of the data sets and their correlations to vegetation model outputs in North America over a two-year time period. Climate data sets consisting of temperature, precipitation, soil moisture, photosynthetically active radiation (PAR), and dew point were used as inputs and compared to model generated outputs of above and below ground biomass, leaf area index (LAI), pasture biomass, plant functional type (PFT), average height, and maximum height to determine which climate inputs effected which vegetation outputs. A subset of the climate data was then examined at resolutions ranging from 0.25 degree by 0.25 degree to 5.0 degree by 5.0 degree to explore the effects of spatially averaging climate variables on vegetation structure and dynamics. As current research demonstrates the importance of forests in slowing climate change through carbon sequestration, our findings help highlight which climate data set variables have the greatest sensitivity on vegetation structure. Knowing this allows for improved future forecasting in ecosystem models in regards to the climate data of interest and the scale of the inputs.

Presentation Type:  Poster

Session:  Global Change Impact & Vulnerability   (Tue 11:30 AM)

Associated Project(s): 

• Hurtt, George: Modeling the Impacts of Major Forest Disturbances on the Earth's Coupled Carbon-Climate System, and the Capacity of Forests to Meet Future Demands for Wood, Fuel, and Fiber ...details
• Hurtt, George: Using NASA Remote Sensing and Models to Advance Integrated Assessments of Coupled Human-Forest Dynamics for North America ...details

Poster Location ID: 177