Advancements to State-Space Models (SSMs) for Fisheries Science

Why Study SSMs for Fisheries?

State-Space Models (SSMs) are a modeling framework, commonly used in ecology for time series data analytics. These models are used to show population dynamics, animal movement and behaviour, capture-recapture data, and other ecological processes.

State-space modeling is an important part of science-based fisheries management. This team aims to advance SSMs for fisheries and facilitate the adoption of these novel approaches by Fisheries and Oceans Canada (DFO).

Areas of Exploration

Methodologies for SSMs

Includes building robust statistical methodologies for SSMs used in fisheries science and management. These methods will allow for both errors and misspecification in the observation process and in the state/dynamic process.

Model Selection Mechanisms

Includes developing a formal model-selection mechanism to choose reliably among candidate SSMs. This is done by accounting for the variability associated with the model criterion measure itself.

Evaluation Diagnostics

Includes developing relevant model diagnostics that can be used to evaluate model goodness-of-fit, for both the observation and state process. In practice, stock assessment scientists and fisheries managers need to know if their advice is sensitive to reasonable alternative assumptions.

Reliability Testing

Includes evaluating the reliability of the SSM approach in terms of accuracy and precision, for measuring model outputs which are key for fisheries. This is to gain a better understanding of the data requirements and limitations of the models.

Solving Global Challenges

Research Team’s Goals

1. To develop general-purpose, statistically-rigorous SSM methodologies for fisheries science.
2. To demonstrate the utility of these new approaches for addressing pressing fisheries issues in Canada.