FishMIP Simulation Experiments and Protocols


Marine ecosystem models use very different basic structures to represent food-web or ecosystem components and the links between them. Ecosystem components can be represented by size classes, functional groups, trophic levels, species groups, and life-history stages, whereas links can be represented by who-eats-whom networks, diet composition or energy transfer. Another challenge is that our models do not all use the same input and output variables. Finally, there are limited observational data for historical runs and model validation, and limited data on spatially resolved fishing effort.


FishMIP Model Evaluation Protocol [ISIMIP 3a]

The goal of the FishMIP Model Evaluation Protocol [3a] is to understand and reduce uncertainty associated with FishMIP models through model evaluation under historical climate and fishing effort forcings.

This information will allow FishMIP to better target policy initiatives such as IPCC and IPBES by providing more robust uncertainty assessment, as well as advancing the state of FishMIP models for informing vulnerability, impact, and adaptation plans of coastal sea ecosystems and fisheries (requested by the FAO).

The focus of papers produced from this simulation round will be a significant advance in at least two ways:

  1. forcing models with higher spatial resolution global climate models and reconstructed historical fishing effort and

  2. assessment of models against reconstructed historical catches (as well as biomass and other fisheries and ecological metrics in regions where additional data are available).

Uploading simulations by January 30th 2023 is essential to ensure enough time for analysis and writing of manuscripts in time for submission to the FishMIP 10 Year Anniversary Special Issue planned for 2023. If you are able to upload results sooner, that would be helpful. If you need more time, please let us know as soon as possible.

Progress towards this deadline will be supported and facilitated through monthly global and regional modeller workshops from to:

  • Ensure correct ESM model inputs and access

  • Ensure fishing drivers work (separate global and regional breakaway groups)

  • Tool sharing & troubleshooting

  • Check model outputs/issues

  • Brainstorming and drafting papers (initially methods and preliminary results)

The FishMIP Model Evaluation protocol can be downloaded from here while the official ISIMIP3 protocol is located on this website.

Earlier and ongoing simulation rounds:

FishMIP CMIP6 Phase 1

The goal of FishMIP CMIP6 Phase 1 (commenced in 2020) is to provide input to the IPCC WGII AR6 report. This necessitates submitting a manuscript by November 1st 2020. The focus of this simulation round will be on (a) new FishMIP outputs using climate scenarios from the CMIP6 earth-system model forcings; (b) exploring fishing impacts in addition to those of climate; (c) comparison to a no-climate change pre-industrial control baseline (which was not available for the last round of simulations), and (d) if possible, validation against observed catches.

Note that this FishMIP CMIP6 Phase 1 protocol represents only a subset of the full ISIMIP3b protocol with a smaller set of scenarios considered. The FishMIP CMIP6 Phase 2 protocol will build upon these experiments in 2023.

The FishMIP CMIP6 Phase 1 (contributing to ISIMIP3b) protocol can be downloaded here while the official ISIMIP3 protocol is located on this website.


The first two rounds of FishMIP simulation protocols were aligned with CMIP5 under the ISIMIP 2 Simulation Round. The ISIMIP2a simulation protocol represents the choices we have made in order to compare as many global and regional models as possible. These include common inputs with respect to climate, physical and chemical data (temperature, currents, salinity, pH, oxygen), biological data (e.g. primary production, phyto- and zooplankton), and fisheries data (effort, catch, fisheries mortality) as well as common outputs including fish biomass and catch as well as other ecosystem parameters.