MESMO 2: a mechanistic marine silica cycle and coupling to a simple terrestrial scheme

Abstract

Abstract. Here we describe the second version of Minnesota Earth System Model for Ocean biogeochemistry (MESMO 2), an earth system model of intermediate complexity, which consists of a dynamical ocean, dynamicthermodynamic sea ice, and energy moisture balanced atmosphere. The new version has more realistic land ice masks and is driven by seasonal winds. A major aim in version 2 is representing the marine silica cycle mechanistically in order to investigate climate-carbon feedbacks involving diatoms, a critically important class of phytoplankton in terms of carbon export production. This is achieved in part by including iron, on which phytoplankton uptake of silicic acid depends. Also, MESMO 2 is coupled to an existing terrestrial model, which allows for the exchange of carbon, water and energy between land and the atmosphere. The coupled model, called MESMO 2E, is appropriate for more complete earth system simulations. The new version was calibrated, with the goal of preserving reasonable interior ocean ventilation and various biological production rates in the ocean and land, while simulating key features of the marine silica cycle.

Inverse-model estimates of the ocean's coupled phosphorus, silicon, and iron cycles

“…However, there is no literature consensus on a mechanistic formulation of the iron dependence of silicic acid uptake. For example, Matsumoto et al (2013) assume that the Si : N uptake ratio is inversely proportional to the dFe concentration (capped at a minimum), while Jin et al (2006) assume the Si : N ratio to depend only on the Si(OH) 4 concentration. Others suggest that the dFe concentration only impacts the diatom growth rate and not the cellular Si : C ratio, while the Si(OH) 4 concentration impacts the cellular Si : C ratio and not growth rate (e.g., Marchetti et al, 2009b;Brzezinski et al, 2011a).…”

“…For χ obs P and χ obs Si we use WOA13 fields interpolated to our grid, and for χ obs Fe we used the GEO-TRACES intermediate data product (Mawji et al, 2015) and the data set compiled by Tagliabue et al (2012).…”

“…For PO 4 and Si(OH) 4 , the WOA13 observations were interpolated to our model grid. The dFe observations were interpolated to our model grid from the data compilation of Tagliabue et al (2012) and from the GEOTRACES Intermediate Data Product 2014 (Mawji et al, 2015). have difficult-to-quantify temporal and spatial sampling biases, and dFe being a trace element is sensitive to episodic events in the aeolian source (e.g., Croot et al, 2004), and possibly to internal episodic events such as submarine volcanism (e.g., Massoth et al, 1995).…”

“…The observations are too sparse to form a reliable climatology, and it is remarkable that we can fit the available observations as well as we do. The larger biases in the Pacific could well be due to the absence of Pacific transects in the GEOTRACES Intermediate Data Product 2014 (Mawji et al, 2015), which means that mismatches in the Pacific incur a relatively smaller penalty in our cost function.…”

Coupled climate-carbon cycle simulation of the Last Glacial Maximum atmospheric CO<sub>2</sub> decrease using a large ensemble of modern plausible parameter sets

“…This range roughly accommodates the impact of error in the model's process representations and the impact of certain potentially important ∆CO 2 mechanisms that are missing from the model, such as changing marine bacterial metabolic rate, wind speed (via its effect on gas transfer) and Si fertilization (Kohfeld and Ridgwell, 2009). This is not a comprehensive assessment, however, as our model also does not include processes such as the effect of changing winds on 10 ocean circulation (Toggweiler et al, 2006), Si leakage (Matsumoto et al, 2002(Matsumoto et al, , 2013(Matsumoto et al, , 2014, the effect of decreasing SSTs on CaCO 3 production (Iglesias-Rodriguez et al, 2002), or changing oceanic PO 4 inventory (Menviel et al, 2012). The fact that it is difficult for our model to achieve a ∆CO 2 of ~-90 ppmv without the missing processes is a significant result because of our extensive exploration of model uncertainty when building the model ensemble.…”