This dataset contains outputs generated using the Gorgon Magnetohydrodynamic (MHD) code, for simulations of the magnetosphere-ionosphere system during impact by a series of interplanetary shocks with different solar wind conditions and dipole magnetic field orientations.

The MHD equations were solved in the magnetosphere on a regular 3-D cartesian grid of resolution 0.25 RE (Earth radii), covering a domain of dimensions (-30,90) RE in X, (-40,40) RE in Y and (-40,40) RE in Z with an inner boundary at 3 RE. In this coordinate system the Sun lies in the negative X-direction, the Z axis is aligned to the dipole in the 0 degree tilt case (where positive tilt points the north magnetic pole towards the Sun), and Y completes the right-handed set. The ionospheric variables were calculated on a separate 2-D spherical grid of dimensions 128x256 in latitude and longitude (with the north pole at 90 degrees latitude and the Sun at 180 degrees longitude), coupled to the magnetospheric domain at the inner boundary.

5 different shocks were simulated in total, with the following solar wind jump conditions injected at the sunward edge at 7200s simulation time:

Shocks 1-4: n = 5 /cm^3 -> 10 /cm^3 (number density)
v = 400 km/s -> 600 km/s (velocity)
T = 5 eV -> 417 eV (temperature)
B = 2 nT -> 4 nT (interplanetary magnetic field)

Shock 5: n = 5 /cm^3 -> 20 /cm^3 (number density)
v = 400 km/s -> 1000 km/s (velocity)
T = 5 eV -> 1250 eV (temperature)
B = 2 nT -> 4 nT (interplanetary magnetic field)

Shocks 1, 3 and 5 had an interplanetary magnetic field (IMF) clock angle of 180 degrees, i.e. B = Bz = -2 nT, whereas Shocks 2 and 3 had IMF clock angles of 135 degrees and 90 degrees, respectively. In addition, Shocks 1, 2, 3 and 5 had zero dipole tilt, whereas Shock 4 had a tilt angle of 30 degrees. These simulations employed zero electrical resistivity. The simulations of Shocks 1, 3 and 5 were then repeated utilising an explicit resistivity eta with value of eta/mu_0 = 5e10 m^2/s. The full set of 8 simulations are labelled 'Shock1', 'Shock2', 'Shock3', 'Shock4', 'Shock5' for the zero resistivity runs and 'Shock1_res', 'Shock3_res' and 'Shock5_res' for those with explicit resistivity.

Output grid data are timestamped in seconds and are defined at the centre of the grid cells, stored as .hdf5 files for each timestep. Output time-series data are for a single variable over a simulation time range, stored in .csv files. The simulation data corresponding to each shock are stored in separate directories, according to the simulation labels listed above.

The magnetospheric variables are stored in the files 'Gorgon_[YYYYMMDD]_[RUN]_MS_params_[XXXX]s.hdf5' where RUN is the simulation label and XXXX is the simulation time in seconds. The magnetospheric data are in SI units and include the magnetic field vector ('Bvec_c'), electric current density vector ('jvec') and ion thermal pressure ('P') for multiple timesteps following initialisation at 7200s of simulation time. The dataset for each magnetospheric variable is of shape (480,320,320,3) for vectors and (480,320,320) for scalars, where the first 3 dimensions are the grid indices in (X,Y,Z) indexed from negative to positive, and the final dimension is the cartesian vector components in (i,j,k).

Similarly, the ionospheric data are stored as 'Gorgon_[YYYYMMDD]_[RUN]_IS_params_[XXXX]s.hdf5', containing the field-aligned current density ('FAC') in SI units for multiple timesteps following initialisation at 7200s of simulation time. The dataset for each ionospheric variable is of shape (130, 256) where the first dimension is the grid index in colatitude, indexed from the north towards the south (i.e. 0 to 180 degrees), and the second dimension is the grid index in longitude, indexed from midnight towards noon via dawn (i.e. 0 to 360 degrees).

Finally, the time-series data are stored as 'Gorgon_[YYYYMMDD]_[RUN]_[XXX].csv' where 'X' is the simulation label and 'XXX' is the time-series variable. These include the subsolar magnetopause standoff distance 'RMP' in RE, and the North (and South) polar cap flux content 'FPC' in Wb*RE^2; in each case the first column contains the simulation time in seconds, with the variables in the second (and third) column(s). NE/P017142/1