Robust state space models for estimating fish stock maturities

Abstract

Here, we formulate robust state space models (SSMs) and develop inference tools in the context of fisheries science and management. Our prototype model concerns the maturity of fish by age over time, knowledge of which is fundamental to understanding the dynamics and productivity of fish stocks, a key component in fish stock assessment. Our SSM incorporates dynamics over time and yields robust estimates of the proportion of fish mature at various ages for a collection of cohorts of interest. The estimates are obtained using an iterative weighted likelihood where the high-dimensional unobserved dynamics of the SSM and the optimization over the fixed parameters are handled with the Automatic Differentiation Model Builder. The data used to both demonstrate and validate our new approach comes from the annual sampling program of Fisheries and Oceans Canada, the primary scientific and regulatory body responsible for stock assessments. In addition, we carried out a comprehensive simulation study to demonstrate the robustness of our approach to realistic contamination

Robust reflections

“…It was necessary to use appropriate winsorization to limit the influence of some very large bycatch trips and to build in robustness. In two papers, Xu et al (2015) and Aeberhard et al (2020), the estimation of fish stock biomass or maturity status used state space models with robustly reweighted likelihoods to downweight and identify unusual years in terms of fish biomass or maturity from research surveys. There are certainly challenges in determining how to do the downweighting keeping in mind computational feasibility and ensuring that we obtain reliable estimates of variability for the parameters of interest.…”

Review of State-Space Models for Fisheries Science

“…Many assumptions required by state-space models are unverifiable in practice, potentially casting doubt on any inference drawn from stock assessment models. In response to this concern, robust estimation techniques which remain reliable even when distributional assumptions are not entirely satisfied have been recently developed (Aeberhard et al, 2017;Xu et al, 2015).…”

Dynamic data science and official statistics

“…Other examples of the use of state space models for purposes important to government decision‐makers include the seasonal forecasting of crop yield (Newlands et al, ; Tam, ) and the estimation of fish stock maturities (Xu et al, ).…”