SS1.05 How Will Aquatic Ecosystems Respond to Climate Change?
Date: Tuesday, June 11, 2002
Time: 4:30:00 PM
Location: Oak Bay
 
BirdTJ, Dept. of Biology, University of Victoria, Victoria, Canada, tbird@uvic.ca
Denman, K, , Canadian Centre for Climate Modelling and Analysis, University of Victoria, Victoria, Canada, ken.denman@ec.gc.ca
 
THE PERILS OF ARTIFICE: SEEKING WAYS TO MAXIMIZE THE SCOPE AND ACCURACY OF BIOLOGICAL MODELS
image
Mathematical modeling has become a popular way to pose questions about interactions in large scale ocean ecosystems which would be impossible to test in the field or in the lab. However, limitations to and variability in our modeling abilities can lead to results that are limited or inaccurate. We investigate the following common problems in model systems: 1) Attempts to add variables to increase the level of complexity in a model (and thereby better mimic reality) usually decrease its robustness. 2) Interactions between even simple equations can behave in ways that go against what we know biologically and intuit mathematically, yet still produce apparently satisfactory results. 3) The way a model is built influences its results, so that models differing in construction can produce dissimilar results based on the same initial conditions. We look into instances of these problems in models of oceanic ecosystems and discuss solutions at different stages in the modeling process. Findings of our investigations will be discussed.