August 2018, Milan Kloewer (milan.kloewer@physics.ox.ac.uk, www.milank.de) This directory contains all data needed to reproduce Kloewer et al 2018, Energy budget-based backscatter in a shallow water model of a double gyre basin, Ocean Modelling. The shallow water model that was used to create theses data sets can be found on github github.com/milankl/swm with the master branch being the control setting and the backscatter branch contains the backscatter implementation as described in the manuscript. Version 1.0 of the github repository also has a doi https://doi.org/10.5281/zenodo.1403130 The model has three prognostic variables u,v,eta (saved as u.nc, v.nc, eta.nc) that entirely determine the prognostic state. For backscatter runs an additional variable e saved as e.nc exists. Further information can be found in the netcdf header. Additionally every run folder named runXXXX also contains file scripts.zip which is an exact copy of the github repository at the time when the model run was initialised. Run folders might contain a folder called "analysis", which contains .npy files that were created with the analysis scripts in the github repository (github.com/milankl/swm/calc) "plots", which contain pictures of the last state of the model run, only used for feedback The files info.txt served feedback purposes while the model was running. The files param.npy, param.txt contain the same parameter information as in scripts.zip, and in the netcdf headers and were only created for convenience. Contents swm_control/ run0000 This is the low resolution model run as described in the manuscript. run0001-run0006 This is the high resolution model run as described in the manuscript. The data is split into 5-year chunks to avoid large file sizes. Note: The last time step of a given folder is the initial condition of the next. run0007, run0009 This is a very high resolution model run (dx=3.75km, 1024x1024 grid points) not described in the manuscript. This model run covers only two years of data and was performed for supplementary analysis. run0008 This is a medium resolution model run (dx=15km, 256x256 grid points) not described in the manuscript and performed for supplementary analysis. Only five years of data are covered. Free-slip boundary conditions are used. run0010 This is a very low resolution model run (dx=60km, 64x64 grid points) not described in the manuscript and performed for supplementary analysis. run0011 This is a medium resolution model run (dx=15km, 256x256 grid points) not described in the manuscript and performed for supplementary analysis. Only ten years of data are covered. In contrast to run0008, no-slip boundary conditions are used. run0012 This is a coarse resolution model run (dx=120km, 32x32 grid points) not described in the manuscript and performed for supplementary analysis. Only ten years of data are covered due to weak variability. swm_backscatter/ The resolution is fixed to dx=30km, 128x128 grid points. The backscatter diffusion nu_e is 540m^2/s^2 unless otherwise stated. The variable backscatter strength Rdiss is described in the manuscript. The comments contain either the name of this model run as described in the manuscript or runs have not been mentioned in the manuscript but were performed for supplementary analysis. Run id Rdiss nu_e comment 0000 1 LR + weak backscatter 0001 2 not used in the manuscript 0002 6 not used in the manuscript 0003 8 LR + moderate backscatter 0004 infinity not used in the manuscript 0005 8 5400m^2/s^2 only mentioned in Chapter 3.2 0006 16 not used in the manuscript 0007 32 not used in the manuscript 0008 64 LR + strong backscatter