Explicit silicate cycling in the Kiel Marine Biogeochemistry Model,version 3 (KMBM3) embedded in the UVic ESCM version 2.9

Abstract. We describe and test a new model of biological marine silicate cycling, implemented in the University of Victoria Earth System Climate Model (UVic ESCM) version 2.9. This new model adds diatoms, which are a key aspect of the biological carbon pump, to an existing ecosystem model. The new model performs well against important ocean biogeochemical indicators and captures the large-scale features of the marine silica cycle. Furthermore it is computationally efficient, allowing both fully-coupled, long-timescale transient simulations, as well as "offline" transport matrix spinups. We assess the fully-coupled model against modern ocean observations, the historical record since 1960, and a business-as-usual atmospheric CO2 forcing to the year 2300. The model simulates a global decline in net primary production (NPP) of 1.3 % having occurred since the 1960s, with the strongest declines in the tropics, northern mid-latitudes, and Southern Ocean. The simulated global decline in NPP reverses after the year 2100 (forced by the extended RCP CO2 concentration scenario), and NPP returns to pre-industrial rates by 2300. This recovery is dominated by increasing primary production in the Southern Ocean, mostly by calcifying phytoplankton. Large increases in calcifying phytoplankton in the Southern Ocean offset a decline in the low latitudes, producing a global net calcite export in 2300 that varies only slightly from pre-industrial rates. Diatoms migrate southward in our simulations, following the receding Antarctic ice front, but are out-competed by calcifiers across most of their pre-industrial Southern Ocean habitat. Global opal export production thus drops to 50 % of its pre-industrial value by 2300. Model nutrients phosphate, silicate, and nitrate build up along the Southern Ocean particle export pathway, but dissolved iron (for which ocean sources are held constant) increases in the upper ocean. This different behaviour of iron is attributed to a reduction of low-latitude NPP (and consequently, a reduction in both uptake and export and particle, including calcite, scavenging), an increase in seawater temperatures (raising the solubility of particle forms), and stratification that "traps" the iron near the surface. These results are meant to serve as a baseline for sensitivity assessments to be undertaken with this model in the future.