Analytical methods for parameter-space delimitation and application to shallow-lake phytoplankton-dynamics modeling (Q1070187)

The first step in parameter estimation is to reduce the dimensionality of the problem by deriving estimates from independent experimentation and from the literature. In addition, insensitive parameters are either removed or fixed. In the remaining lower-dimensional problem, parameter- space delimitation is possible by analytical means. Three conjunctive methods are derived: period-average analysis, extremum analysis, and quasisteady-state analysis. The basic idea is to find conditions for the parameters that must be fulfilled in order to comply with average and extreme values in the observations. The approach is applied to the modeling of the phytoplankton dynamics of Lake Balaton. The analytical techniques prove to supply valuable insight into parameter interrelationships and model adequacy, and can serve as satisfactory substitutes for formal parameter-estimation techniques in the early stages of model development.