Grid quantities in OSIRIS share a common interface that allows the user to output several different types of grid diagnostics for a given grid quantity. The types of general diagnostics available are the following:

  • Full dump of grid quantity to disk
  • Data reduction operation: spatial averaging, spatial envelope, lineouts and slices
  • Time averaging

Time averaging can be combined with spatial reduction – for example, the user may select to report a lineout of a time average of a given quantity.

Types of Diagnostics

In each diagnostics section, the selection of grid diagnostics to output is defined by a list of strings in the form:

reports = diag1, diag2, ...,

Each of the diagnostics must be specified in the form:

quant [, tavg] [, savg | , senv | line, xj, j1(, j2) | slice, xj, j1 ]

where

  • quant specifies the quantity to be saved. See each specific diagnostics section to find out which grid quantities are available for diagnostics.
  • tavg reports a time average of the selected quantity. The number of time steps averaged is controlled by the n_tavg parameter described below. Time averages can be used together with the data reduction operations described below.
  • savg reports a spatial average of the selected quantity. The number of cells to average in each direction are specified using the n_ave parameter described below. The original grid size doesn’t need to be a multiple of n_ave, but the diagnostic will be faster in this case.
  • senv reports a spatial envelope of the selected quantity, calculated by finding the maximum absolute value of the quantity in a given number of neighboring grid cells. The number of cells to sample in every direction is specified using the n_ave parameter described below. The original grid size doesn’t need to be a multiple of n_ave, but the diagnostic will be faster in this case.
  • line (not available in 1D) extracts a single lineout from the selected grid quantity. The parameters specifying the line direction and position must follow:
    • xj must be “x1”, “x2”, or “x3” specifying the direction of the line to be extracted.
    • j1, j2 - (j2 in 3D only) are integers specifying the grid cell coordinates (1-based indexing) crossed by the line, and refer to the coordinates other than the line direction.
  • slice (3D only) extracts a 2D slice from the selected grid quantity. The parameters specifying the slice direction and position must follow:
    • xj must be “x1”, “x2”, or “x3” specifying the normal direction of the slice to be extracted.
    • j1 is an integer specifying the grid cell coordinates (1-based indexing) of the slice in the normal direction.

The savg, senv, line, and slice options are mutually exclusive and only one can be used in each diagnostic. A detailed description of the spatial average/envelope diagnostics can be found in the average and envelope grids section of the documentation. Note that the j1 and j2 indices refer to integer indices in the global grid, which start at 1 (Fortran-based indexing).

Frequency of Diagnostics

The frequency of data dumps is controlled by three parameters for both normal and time-averaged diagnostics:

  • ndump_fac - controls the frequency of full grid diagnostics.
  • ndump_fac_ave - controls the frequency of spatial average/envelope grid diagnostics.
  • ndump_fac_lineout - controls the frequency of lineout/slice diagnostics.

These values are multiplied by the global ndump parameter defined in the time_step section to get the actual diagnostic frequency. For example, if ndump = 2, ndump_fac = 10, and ndump_fac_lineout = 1, then the full grid will be output every 20 timesteps and a lineout will be output every 2 timesteps.

Time and spatial filter parameters

The time average, spatial average, and spatial envelope diagnostics are controlled by the following parameters:

  • n_ave specifies the number of gridpoints in each direction to average over/sample for spatially averaged/envelope dumps.
  • n_tavg specifies the number of time steps to be used when calculating the time-averaged diagnostics.

Example

Here’s an example of a diag_emf section using grid diagnostics for a 3D simulation:

 diag_emf 
 {
   reports = "e3",                     ! full E3 field
             "e2, line, x1, 64, 96",   ! lineout of E2 field along x1 at postion ix2 = 64, ix3 = 96
             "e3, senv",               ! spatial envelope of E3 
             "b1, tavg",               ! time average of B1
             "b2, tavg, savg",         ! Spatial average of the time average of B2
             "b3, slice, x2, 32",      ! Slice of B3 along x1-x3, at position ix2 = 32
      
   ndump_fac = 20,                     ! do full grid diagnostics at every 20 timesteps (assuming ndump = 1): 
   ndump_fac_ave = 10,                 ! do average/envelope grid diagnostics at every 10 timesteps
   ndump_fac_lineout = 1,              ! do lineouts/slices at every timestep
    
   n_ave(1:3) = 2,2,2,                 ! average/envelope 8 cells (2x2x2)
   n_tavg = 5,                         ! average 5 iterations for time averaged diagnostics 
 }