These data represent the results of the first study to use Earth System Model (ESM) outputs of SST and chlorophyll-a to simulate circumpolar krill growth potential for the recent past (1960-1989) and future climate change scenarios (2070-2099). Growth potential is obtained using an empirically-derived krill growth model (Atkinson et al. 2006, Limnol. Oceanogr.), where growth is modeled as a function of SST and chlorophyll-a. It serves as an approximation of habitat quality, as areas that support high growth rates are assumed to be good habitat (see Murphy et al., 2017, Sci Rep). To increase confidence in the future projections, ESMs were selected and weighted for each season based on their skill at reproducing observation-based krill growth potential for the recent past. First, eleven ESMs which provided SST and chlorophyll-a outputs were obtained from the Coupled Model Inter-comparison Project 5 archive. These included: CanESM2, CMCC-CESM, CNRM-CM5, GFL-ESM2G, GFDL-ESM2M, GISS-E2-H-CC, HadGEM2-CC, IPSL-CM5A-LR, MPI-ESM-MR, MRI-ESM1 and NorESM1-ME. For each ESM, seasonal surface averages of SST and chlorophyll-a were used to calculate growth potential for the historical scenario (1960-1989), which was then bilinearly interpolated on to the same 1°x1° grid. Satellite observation-based datasets for SST and chlorophyll-a were used to calculate observation-based growth potential for the recent past (1997-2010). These comprised seasonal surface averages of SST (from the OISST v2 daily dataset, 1/4⁰ horizontal resolution) and chlorophyll-a (the mean of the SeaWiFS and Johnson et al. (2013) corrected estimate of SeaWiFS daily datasets, 1/12⁰ horizontal resolution). Observation-based growth potential was then bilinearly interpolated onto the same grid as the ESMs. ESM skill for each season was subsequently assessed against observation-based growth potential using a Taylor Diagram. The ESMs were selected and weighted according to their performance to produce a weighted subset (see "ESM_weighting_method.pdf" file). Of the netcdfs provided, "hist_mean_ensemble.nc" represents the unweighted mean of seasonal growth potential, calculated from the initial ensemble of eleven ESMs for the historical scenario. The "hist_mean_subset.nc" file represents the analogous output of the weighted subset. Future projections of seasonal growth potential for Representative Concentration Pathways (RCPs) 4.5 and 8.5 were obtained using the weighted subset for the period of 2070-2099. These projected seasonal surface averages are provided in the "rcp45_mean_subset.nc" and "rcp85_mean_subset.nc" files. RCPs represent standard climate change scenarios developed by the Intergovernmental Panel on Climate Change, with 4.5 reflecting some mitigation of carbon emissions, and 8.5 being the "business as usual" scenario. Analogous netcdfs for the weighted subset outputs of chlorophyll-a (chl) and SST (tos) for the historical and RCP scenarios are also provided in the "chl_tos_netcdfs.zip" file so that the driving environmental variables underlying growth potential can be examined.

The configuration of the East-Antarctic Atlantis model included several pages of documentation from SOKI (the Southern Ocean Knowledge and Information wiki, hosted at soki.aq in 2019). There is also an associated software package angstroms was used to couple Atlantis models to Regional Ocean Model System (ROMS). The documentation pages have been archived in PDF and WORD format here in 11 separate documents. The software package is included as an archive of the Australian Antarctic Division github repository as at 2019-07-22 with commit '2b7b10e86963195df049ded6ca842255b2335de1'. https://github.com/AustralianAntarcticDivision/angstroms

Direct Numerical Simulations are carried out at the ice ocean interface of 1.8 m long, inclined at angles, 50 degree, 65 degree and 90 degree from the horizontal where external source buoyancy is added as a boundary conditions with relative buoyancy B* 5, 7 and 10 times the wall buoyancy. The data set contains 1. Time averaged temperature, salinity and velocity fields of the flow at steady state where averaging windows are several times the respective buoyancy frequency for 90 degree, B* =1, 5,7,10; 50 degree, B*=1, 5, 7 respectively. 2. Tabulated, time averaged along-slope profiles of a) temperature, b) salinity, c) meltrate, d) plume velocity for 90 degree, B* =1, 5,7,10; 65 degree, B* =1, 5,7,10 and 50 degree, B*=1, 5, 7 respectively. 3. Tabulated, domain averaged meltrate, plume velocity for 90 degree, B* =1,3, 5,7,10; 65 degree, B* =1,3, 5,7,10 and 50 degree, B*=1,3, 5, 7 respectively.