An Idealised Barotropic Ocean Gyre Model Code, Based on MITgcm

This dataset is supplementary to The Cryosphere paper "Reversal of ocean gyres near ice shelves in the Amundsen Sea caused by the interaction of sea ice and wind". A full model description and the application can be found in the paper. Here is the modified version of the model set-up written in the paper in the chapter 3.

This model is based on the MIT general circulation model (MITgcm; MITgcm’s user manual 4.1; Marshall et al., 1997) with an idealised barotropic set-up. The model has an ocean domain with a size of 60 × 60 km and a horizontal grid spacing of 1 km. It has one 1-km-thick vertical layer with a free surface. All simulations are run for six model months, which allows all of them to spin up to be sufficiently close to a steady state. The spin-up time of the simulations varies from 51 days to 91 days, assessed as the time at which the daily change of the total kinetic energy of the ocean is less than 0.1% of the total kinetic energy of the ocean at the final model day of the 6 model months. 

The wind field is the only external forcing applied to the model ocean. We generate a simplified wind forcing field based on the key features of the climatological wind in the south-eastern Amundsen Sea to include the ice conditions for both Pine Island Bay and around the Thwaites Ice Tongue. The ERA5 climatological 10-m wind (Hersbach et al., 2018) above the PIB and Thwaites gyres blows from the ice shelves to the ocean, with a speed decreasing from the southeast to the northwest. The maximum wind speed (10 m s^-1) occurs in the southwestern corner of the model domain. The meridional gradient of wind speed (−1.667 × 10^-6 s^-1) is one-fifth of the zonal gradient of wind speed (−8.333 × 10^-6 s^-1). We vary the strength and sign of the wind stress curl to generate four wind forcing fields: strong or weak, cyclonic or anticyclonic wind stress curl. The average wind speed over the whole ocean model domain is kept the same for all four wind fields.

Inside the main folder, there are five subfolders,

- exp_ada_Flat: for the simulations with sea ice covering the top-left of the model domain, shown as "Seaice-\(\frac \pi2\)"

- exp_ada_Topleft: for the simulations with sea ice covering the top-left side of the model domain, shown as "Seaice-\(\frac \pi4\)"

- exp_ada_Topright: for the simulations with sea ice covering the top-right sideof the model domain, shown as "Seaice-\(\frac {3\pi}4\)"

- exp_ada_Weaktopleft: for the simulations with sea ice covering the half part of the top-left side of the model domain, shown as "Seaice-\(\frac {3\pi}8\)"

- exp_ada_Weaktopright: for the simulations with sea ice covering the half part of the top-right side of the model domain, shown as "Seaice-\(\frac {5\pi}8\)"

- exp_ada_Discussion: for the simulations mentioned in the Discussion section, including those for the 1.5-layer baroclinic model (folders with "_grav" in the title, "grav" stands for the reduced gravity), for different widths of the marginal ice zone ("exp_MIZ") and for the model with a rectangle domain ("exp_rec").

Inside the subfolders, there are fourty-four subfolders, with the combination of the following components in the title that stand for different wind forcings,

- Maxeast stands for "maximum wind speed occurs in the eastern side of the model domain", so "StrongCyclonic"

- Maxwest stands for "maximum wind speed occurs in the western side of the model domain", so "StrongAnticyclonic"

- Weakmaxeast stands for "maximum wind speed occurs in the eastern side of the model domain, but with a weaker wind stress curl", so "WeakAnticyclonic"

- Weakmaxwest stands for "maximum wind speed occurs in the western side of the model domain, but with a weaker wind stress curl", so "WeakAnticyclonic"

- numbers 0, 20, 40, ... 200 stand for the percentage of the wind stress that is transferred to the ocean through the sea ice.