Refining sea ice models to capture fractal features reshapes estimates of Arctic marine production

The Arctic is not a uniform white sheet; it is a fractured, dynamic mosaic where the geometry of ice dictates the rhythm of life. The finding that sea ice model rheology, specifically, how models represent mechanical damage like leads and ridges, shifts the timing of spring blooms and carbon export in the Chukchi Sea is a critical calibration point for ocean intelligence. This is not a technical footnote; it is a correction to how we measure the biological pulse of the polar oceans. When a model's representation of ice physics changes when and where phytoplankton bloom, it changes our confidence in every downstream climate indicator derived from those simulations. The study's comparison of the Los Alamos CICE model with the brittle-rheology neXtSIM framework reveals that the mechanical damage encoded in sea ice models directly influences simulated light availability and, consequently, the timing of spring carbon export. The difference is not marginal. In the neXtSIM(light) simulation, the parameterised flux of light through young ice altered sea ice concentration and extended the window of biological productivity into late spring and summer. This is a measurable, empirical signal that the sea ice model itself, not just the biogeochemical component, shapes what we infer about Arctic marine production. For researchers working with coupled models, the takeaway is clear: rheology is not a mechanical detail but a first-order control on simulated ecosystem dynamics. The choice of sea ice model can shift the timing of the spring bloom and carbon export, which has consequences for how we validate these models against observations. This matters beyond model intercomparison exercises. As we refine estimates of Arctic primary production and carbon export, the representation of leads and ridges, fractal features that dominate the mechanical damage of the ice pack, becomes a question of empirical accuracy, not numerical preference. The study's finding that light flux through young ice changes simulated phytoplankton growth echoes a broader pattern: how we treat physical processes at fine scales propagates through the entire simulated food web. That same logic applies to other emerging techniques, such as validating DNA metabarcoding for fragile Arctic gelatinous zooplankton, where methodological choices determine which components of the ecosystem we can actually see and measure. And just as melting ice and shifting microalgae in Antarctica's seasonal ice-ocean boundary reveal the tight coupling between ice structure and biological response, this work shows that the mechanical representation of sea ice, not just its total extent, shapes the light field that drives phytoplankton blooms. The study's central insight is that rheology matters. An elastic-viscous-plastic model like CICE smooths over the leads and ridges that a brittle rheology resolves, and those differences are not cosmetic. In the Chukchi Sea, the choice of sea ice model shifted the timing of the spring bloom and the subsequent pulse of carbon export. When light flux was allowed through young ice, simulated productivity rose during late spring and summer. That is not a subtle model tweak; it is a signal that our current estimates of Arctic marine production may be systematically biased by how we represent mechanical damage in sea ice. For researchers working in biogeochemical modelling, the practical implication is direct: if sea ice models cannot resolve the fracture features that dominate a thinning pack, then the light fields driving phytoplankton growth, and the carbon export estimates built on them, will remain incomplete. The comparison between the elastic-viscous-plastic CICE and the brittle-rheology neXtSIM is particularly telling. Brittle rheology reproduces the leads and ridges that are not just structural details but ecological gateways, controlling when and where light enters the water column. The difference in spring-bloom timing and carbon export between the models is not a minor calibration issue; it is a signal that our current estimates carry structural uncertainty tied to how we represent mechanical sea ice damage. As we have seen with validating DNA metabarcoding for fragile Arctic gelatinous zooplankton, methodological choices in polar observation can change what we think we know about ecosystem dynamics. Likewise, melting ice and shifting microalgae in Antarctica's seasonal ice-ocean boundary remind us that the ice-ocean interface is where the most consequential biological signals emerge first.