Circum-Antarctic shoreward heat fluxes derived from eddy- and tide-resolving simulation

• Other Authors: Stewart, Andrew (hasPrincipalInvestigator), Klocker, Andreas (coInvestigator), Klocker, Andreas (pointOfContact), University of California Los Angeles (UCLA) (hasAssociationWith), Institute for Marine and Antarctic Studies (IMAS), University of Tasmania (UTAS) (hasAssociationWith)
• Format: Dataset
• Language: unknown
• Published: University of Tasmania, Australia
• Subjects: oceans; OCEAN CURRENTS; EARTH SCIENCE; OCEAN CIRCULATION; OCEANOGRAPHY; EARTH SCIENCES; Geophysical Fluid Dynamics; GEOPHYSICS; Physical Oceanography; Antarctic; Antarctic Peninsula; The ''Y''; The Antarctic; Antarc*; Sea ice
• Online Access: https://researchdata.edu.au/circum-antarctic-shoreward-resolving-simulation/1201261; https://metadata.imas.utas.edu.au:443/geonetwork/srv/en/metadata.show?uuid=99ccd395-af52-4a36-bca4-edd7991435bd; https://thredds.imas.utas.edu.au/thredds/catalog/IMAS/AKlocker_CircumAntarctic_heatflux/catalog.html; http://wwwcvs.mitgcm.org/viewvc/MITgcm/MITgcm_contrib/llc_hires/llc_4320/

Model data used to quantify the roles of mean and transient flows in transporting heat across almost the entire Antarctic continental slope and shelf using an ocean/sea-ice model run at eddy- and tide-resolving (1/48 degree) horizontal resolution. The LLC_4320 simulation was conducted using the MIT general circulation model [Marshall et al., 1997a,b] on a global Latitude-Longitude-Cap (LLC) grid [Forget et al., 2015] with 1/48 degree horizontal grid spacing and 90 vertical levels. The horizontal grid spacing is less than 1 km over the entire Antarctic continental shelf and slope, with the exception of the tip of the Antarctic Peninsula, which is sufficient to simulate shoreward heat transfer by eddies [St-Laurent et al., 2013; Stewart and Thompson, 2015b]. The model includes the 16 largest-amplitude tidal components globally and partially resolves the internal wave spectrum [Rocha et al., 2016] and thus cross-slope excursions associated with baroclinic tides [Padman et al., 2009]. The LLC_4320 simulation is not directly constrained by observational data, but is a high-resolution continuation of the ECCO2 0.14 degree reanalysis product. The simulation spans September 2011–October 2012, and derives surface fluxes from ECMWF 1/6 degree operational analysis [ECMWF, 2011] combined with an annual climatology of continental runoff [Fekete et al., 2002]. An identically configured simulation was run using 1/24 degree horizontal grid spacing, spanning January 2011 – April 2013, and is referred to as LLC_2160. The model configuration files to understand the model grid layout, integration procedure, parameters etc can be found at http://wwwcvs.mitgcm.org/viewvc/MITgcm/MITgcm_contrib/llc_hires/llc_4320/. The data are stored in .mat files separated by LLC face number (1, 2, 4 or 5), grid resolution (LLC2160 vs LLC4320), and averaging procedure (averaging all 6-hour snapshots together, '_6hsnapshots', or first averaging 6-hour snapshots into days and then averaging the days, '_dailyavg6h'). The LLC2160 model output is averaged from 03/06/2011 to 04/22/2013 (model time step 92160 to 1586400), while the LLC4320 output is averaged from 09/12/2011 to 11/15/2012 (model time step 10368 to 1495008). Each file contains the MITgcm grid vectors and time-averaged products of model output variables. Where necessary, the LLC grids have been rotated so that the x-coordinate points east, or approximately east, while the y-coordinate points north, or approximately north. For products between velocities and tracers, the tracers were interpolated linearly onto velocity points before the product was taken.