Dynamic Benthic Pelagic Model (DBPM) outputs: Evaluating multiple sources of uncertainty in a global marine ecosystem model from a Southern Ocean perspective

Updated: 2026-11-19

Global marine ecosystem models (MEMs) are increasingly being used for assessing climate impacts at various spatial scales (global, regional, countries, etc.), but their outputs are influenced by uncertainties linked to the ocean models used as forcings. We run simulations following Track A of the Fisheries and Marine Ecosystem Model Intercomparison Project 2.0 using the Dynamic Benthic Pelagic Model (DBPM). We then evaluated the impact of three sources of uncertainty influencing the accuracy of simulated fishing catches in the Southern Ocean (1961–2010): (1) coarsening horizontal resolution of environmental forcings (from 0.25° to 1°), (2) restricting spatial distribution of fishing effort using a sea ice mask, and (3) using regional versus global fishing effort forcing data. Our results indicate that DBPM captured observed spatial catch distributions, with minimal differences between forcings resolutions. DBPM overestimated observed fishing catches to a similar degree in both horizontal resolutions. Temporal trends aligned more closely with observations prior to the 1990s,but catch overestimation grew larger thereafter. Overestimation appears to be linked to an unrealistic spatial distribution of effort by DBPM, which was improved by the inclusion of a sea ice mask, and to errors and biases in the underlying fishing effort forcing data, which continues to represent a major hurdle for model skill improvement. Our findings suggest that coarsening inputs from an eddy-permitting ocean model maybe an effective approach to improve MEM performance without increasing computational costs.

Data