Planktonic ecosystem models: Perplexing parameterizations and a failure to fail

Abstract

Planktonic ecosystem models have been used for many decades; modern models are only subtle variations on model structures established in the 1970s or earlier. Here I explore two problems that I see with these models: (i) their formulation and parameterization and (ii) their use. Using nutrient uptake by the phytoplankton as an example, I trace the history of why we use Michaelis-Menten kinetics in our models, and show that this functional form may not be the best representation of nutrient uptake by a diverse and changing phytoplankton community. I then discuss how models are used - not as the hypotheses they are, but more like toasters. I make the point that treating models as hypotheses could lead to much stronger and more robust insights into planktonic dynamics. However, this requires better statistical comparisons of models and data, including the "right" kinds of data. Finally I suggest some ways to move forward, to make planktonic ecosystem models much more powerful tools in our investigations of ocean dynamics.