“Assessing the costs and benefits of increasing ecosystem model complexity using data assimilation”
Dr. Marjy Friedrichs
Center for Coastal Physical Oceanography
Monday, April 11, 2005
3:30 PM
Room 109, Crittenton Hall
Abstract
Numerical models developed to examine marine ecosystems and biogeochemical cycles are
becoming more complex as they include increasing numbers of organisms and biological processes.
Under the banner of “biocomplexity,” these developments are being embraced and encouraged as we
explore the emerging plethora of ecological phenomena. Unfortunately, however, our knowledge of
ecosystem complexities is growing faster than the availability of appropriate validation data,
resulting in the development of models with an increasing number of unconstrained parameters. Although
these models are becoming better able to reproduce the available data, their predictive ability remains
unresolved. Here we examine this problem by applying the variational adjoint method of data assimilation
to several mechanistic ecosystem models including varying degrees of complexity. After optimizing the
parameters of each model under identical conditions, we assess the performance of the mechanistic models
using a variety of methods, and compare their ability to describe the data with that of empirical models
with equivalent degrees of freedom.
Biography
Marjorie Friedrichs earned a B.A. in physics from Middlebury College, a M.S. in physical oceanography
from the MIT/WHOI Joint Program, and a Ph.D. in oceanography from Old Dominion University. After two
years as a post-doctoral investigator, she became an Assistant Research Professor (SSRP) at Old Dominion
University in 2001. Her primary research interests include biological-physical interactions on regional to
basin-scales, ecosystem modeling, data assimilation, and estimating primary production from ocean color
observations.
Reception before seminar at 3:00 PM
Crittenton Hall
Old Dominion University
Norfolk, VA 23529
757-683-5548
Last updated 4/6/2005.
This page is maintained by
Gabriel Franke
Copyright Info: Old Dominion University 2004
