Research interests

Structured errors in modeling fisheries population dynamics and stock assessment Two types of structured errors affect results when modeling population dynamics and stock assessment of living marine resources: non-random variability in biological parameters (e.g. trends in the intrinsic population growth rate) and non-random variability in data used to infer resource condition (e.g. trends or changes in errors for total catch reported by commercial fishers). The former are structured “process” errors and the latter are structured “measurement” errors. These process errors complicate efforts to understand the dynamics of living marine resources, make predictions about status of fish populations, and provide management advice. The measurement errors violate the randomness assumption on errors associated with fisheries data, which are required by almost all fisheries stock assessment models. In my study I will use simulation analyses to understand how structured errors affect the results from models used in fisheries stock assessment and management. I will also develop data collection, modeling, and policy approaches that can reduce the uncertainty and risk caused by structured errors. Simulated systems will be based on the American lobster fishery, the Atlantic cod fishery, and the Atlantic herring fishery in the Gulf of Maine, which have different biological and industrial characteristics. Using realistic simulation analyses, I will evaluate the effects of structured process and measurement errors on the full range of events in natural resource assessment: choice of models, parameter estimation (including assumptions about inestimable parameters and assumptions about error terms), and choice of management policies (risk aversion, choice of management goals, economic discount factors, etc.).

[Edit my profile]