We study the impact of synthesizing ocean and sea ice concentration data with a global, eddying coupled sea ice-ocean configuration of the Massachusetts Institute of Technology general circulation model with the goal of reproducing the 2004 three-dimensional time-evolving ice-ocean state. This work builds on the state estimation framework developed in the Estimating the Circulation and Climate of the Ocean consortium by seeking a reconstruction of the global sea ice-ocean system that is simultaneously consistent with (1) a suite of in situ and remotely-sensed ocean and ice data and (2) the physics encoded in the numerical model. This dual consistency is successfully achieved here by adjusting only the model’s initial hydrographic state and its atmospheric boundary conditions such that misfits between the model and data are minimized in a least-squares sense. We show that synthesizing both ocean and sea ice concentration data is required for the model to adequately reproduce the observed details of the sea ice annual cycle in both hemispheres. Surprisingly, only modest adjustments to our first-guess atmospheric state and ocean initial conditions are necessary to achieve model-data consistency, suggesting that atmospheric reanalysis products remain a leading source of errors for sea ice-ocean model hindcasts and reanalyses. The synthesis of sea ice data is found to ameliorate misfits in the high latitude ocean, especially with respect to upper ocean stratification, temperature, and salinity. Constraining the model to sea ice concentration modestly reduces ICESat-derived Arctic ice thickness errors by improving the temporal and spatial evolution of seasonal ice. Further increases in the accuracy of global sea ice thickness in the model likely require the direct synthesis of sea ice thickness data.
CITATION STYLE
Fenty, I., Menemenlis, D., & Zhang, H. (2017). Global coupled sea ice-ocean state estimation. Climate Dynamics, 49(3), 931–956. https://doi.org/10.1007/s00382-015-2796-6
Mendeley helps you to discover research relevant for your work.