Masked data output#

This feature allows to output quantities at different mask locations, i.e. arbitrary subsets of the total model domain. Subsets can be 3d volumes, 2d cross sections, or even 0d or 1d data at any position and of any amount. NetCDF4 parallel I/O is not available for masked data output. Default quantities (instantaneous and time averaged) are declared with the parameter data_output_masks, user-defined quantities can be output with data_output_masks_user (see also user-defined output quantities).

Terrain-following masked output allows to output masked quantities at a specified height above the surface (see mask_k_over_surface).

Output Steering#

The runtime parameters that are listed below steer the output of those quantities that have been set via data_output_masks and/or data_output_masks_user:

Following parameters define the masks. Each mask can be steered with a separate set of x-, y- and z-parameters. By default all gridpoints along the respective direction are output.


mask_x	all x-coordinates of mask locations (in multiples of mask scale)
mask_y	all y-coordinates of mask locations (in multiples of mask scale)
mask_z	all z-coordinates of mask locations (in multiples of mask scale)
mask_x_loop	loop begin, end and stride for x-coordinates of mask locations for masks (in multiples of mask scale)
mask_y_loop	loop begin, end and stride for y-coordinates of mask locations for masks (in multiples of mask scale)
mask_z_loop	loop begin, end and stride for z-coordinates of mask locations for masks (in multiples of mask scale)
mask_k_over_surface	vertical grid index above a surface to use for terrain-following masked output

The scaling lengths can be used to scale the parameters which defines the masks (those listed above). The scaling lengths apply for all masks.


mask_scale_x	scaling length for masked data output in x-direction
mask_scale_y	scaling length for masked data output in y-direction
mask_scale_z	scaling length for masked data output in z-direction

The time intervals of the output times for each mask are determined with dt_domask. An individual time interval for output of temporally averaged data can be assigned via parameter dt_data_output_av. The length of the averaging interval is controlled by averaging_interval. Output of masked data can be switched off until the interval defined via skip_time_domask has passed.

No particular parameters are available for steering the time averaged output of each of the masks separately.

By default, a maximum number of 300 different masks can be defined. Each mask is output to one separate local file:

Instantaneous data of mask 1 is output to file DATA_MASK_NETCDF_M001 in the local temporary working directory (permanent file suffix _masked_M001), time averaged data to file DATA_MASK_AV_NETCDF_M001 (_av_masked_M001).
Instantaneous data of mask 2 is output to file DATA_MASK_NETCDF_M002 in the local temporary working directory (permanent file suffix _masked_M002), time averaged data to file DATA_MASK_AV_NETCDF_M002 (_av_masked_M002).
...

If more than 300 masks shall be defined, the Fortran parameter max_masks that is defined in source code file modules.f90 needs to be increased manually and the model has to be re-compiled via palmbuild. This way, max_masks can be increased to a maximum value of 999.

Examples#

The following examples are given based on the example setup for a convective boundary layer example_cbl_p3d with a model domain size of 2000 m x 2000 m x 2000 m and a spatial resolution of 50.0 m.

Output of one mask
Output of two different masks
Output of three different masks
Output of three different masks with scaling length
Output of four different masks with user-defined quantities
Output of terrain-following mask

1. Output of one mask#

Mask 1: 3d volume data with an extension of 500 m x 200 m x 2000 m from x = 0 m to x = 500 m, from y = 800 m to y = 1000 m, both at every 50 m, and from the bottom to the top of the model domain. Output is for instantaneous data of the three wind components.

&initialization_parameters

    nx = 39, ny = 39, nz = 40,
    dx = 50.0, dy = 50.0, dz = 50.0,
    .../

&runtime_parameters

    ...
    data_output_masks(1,:) = 'u','v','w',

    mask_x_loop(1,:) = 0., 500. ,50. ,
    mask_y_loop(1,:) = 800., 1000., 50. ,/

If mask_x , mask_y , mask_z or mask_x_loop, mask_y_loop, mask_z_loop are not set, all gridpoints along the corresponding direction are output. Output is done at time intervals given by dt_data_output, as dt_domask has not been set in this example.

2. Output of two different masks#

Mask 1: xy cross section from x = 500 m to x = 1000 m and from y = 1000 m to 2000 m, both at every 50 m, and at heights 0 m, 50 m, 100 m, 500 m, 1000 m, and 1200 m. Output of instantaneous and time averaged data of the three wind components.

Mask 2: Every second gridpoint in all directions. Output of instantaneous data of the potential temperature.

&initialization_parameters

    nx = 39, ny = 39, nz = 40,
    dx = 50.0, dy = 50.0, dz = 50.0,
    .../

&runtime_parameters

    ...
    data_output_masks(1,:) = 'u','v','w','u_av','v_av','w_av',
    data_output_masks(2,:) = 'theta',

    mask_x_loop(1,:) = 500.,1000.,50. ,
    mask_y_loop(1,:) = 1000.,2000.,50. ,
    mask_z(1,:) = 0.,50.,100.,500.,1000.,1200.,

    mask_x_loop(2,:) = 0.,2000.,100.,
    mask_y_loop(2,:) = 0.,2000.,100.,
    mask_z_loop(2,:) = 0.,2000.,100.,

    dt_domask = 600.,1800.,
    skip_time_domask = 3600.,3600.,/

Output starts after 1h simulated time (see skip_time_domask) for both masks and is done every 10 minutes for mask 1 and every 30 minutes for mask 2 (see dt_domask).

3. Output of three different masks#

Mask 1: 1d data along x-direction at y = 50 m, 200 m, 500 m and 1000 m and every fifth grid point in z-dircetion. Output of instantaneous data of the w-velocity component.
Mask 2: Output of the potential temperature at position x = 0 m, y = 500 m, z = 200 m, 300 m and 400 m.
Mask 3: xz cross section at y = 400 m, 450 m, 600 m. Output of time averaged data of the three wind components.

&initialization_parameters

    nx = 39, ny = 39, nz = 40,
    dx = 50.0, dy = 50.0, dz = 50.0,
    .../

&runtime_parameters

    ...
    data_output_masks(1,:) = 'w',
    data_output_masks(2,:) = 'theta',
    data_output_masks(3,:) = 'u_av','v_av','w_av',

    mask_y(1,:) = 50., 200., 500., 1000.,
    mask_z_loop(1,:) = 0.,2000.,250.,

    mask_x(2,:) = 0.,
    mask_y(2,:) = 500.,
    mask_z_loop(2,:) = 200.,400.,100.,

    mask_y(3,:) = 400., 450., 600.,

    dt_domask = 1800.,
    skip_time_domask = 0.,3600.,3600.,/

Output starts from the beginning for mask 1 and after 1h simulated time (see skip_time_domask) for mask 2 and 3. It is done every 30 minutes for mask 1 and at time intervals of dt_data_output for mask 2 and 3 (see dt_domask).

4. Output of three different masks with scaling length#

Mask 1, 2 and 3 as in example 3.

&initialization_parameters

    nx = 39, ny = 39, nz = 40,
    dx = 50.0, dy = 50.0, dz = 50.0,
    .../

&runtime_parameters

    ...
    data_output_masks(1,:) = 'w',
    data_output_masks(2,:) = 'theta',
    data_output_masks(3,:) = 'u_av','v_av','w_av',

    mask_scale_x = 10.0,
    mask_scale_y = 10.0,

    mask_y(1,:) = 5., 20., 50., 100.,
    mask_z_loop(1,:) = 0.,2000.,250.,

    mask_x(2,:) = 0.,
    mask_y(2,:) = 50.,
    mask_z_loop(2,:) = 200.,400.,100.,

    mask_y(3,:) = 40., 45., 60.,

    dt_domask = 1800.,
    skip_time_domask = 0.,3600.,3600.,/

Since mask_scale_x and mask_scale_y are assigned, the parameters for steering the mask locations along x- and y-direction (here mask_x and mask_y) has to be given in multiples of mask_scale_x and mask_scale_y.

5. Output of four different masks with user-defined quantities#

&initialization_parameters

    nx = 39, ny = 39, nz = 40,
    dx = 50.0, dy = 50.0, dz = 50.0,
    .../

&runtime_parameters

    ...
    data_output_masks(1,:) = 'w',
    data_output_masks(2,:) = 'theta',
    data_output_masks(3,:) = 'u_av','v_av','w_av',

    mask_y(1,:) = 50., 200., 500., 1000.,
    mask_z_loop(1,:) = 0.,2000.,250.,

    mask_x(2,:) = 0.,
    mask_y(2,:) = 500.,
    mask_z_loop(2,:) = 200.,400.,100.,

    mask_y(3,:) = 400., 450., 600.,

    mask_z(4,:) = 100., 250., 500.,

    dt_domask = 1800.,
    skip_time_domask = 0.,3600.,3600.,/

&user_parameters

    data_output_masks_user(3,:) = 'u2',
    data_output_masks_user(4,:) = 'u2',

Output starts from the beginning for mask 1 and after 1h simulated time (see skip_time_domask) for mask 2, 3 and 4. It is done every 30 minutes for mask 1 and at time intervals of dt_data_output for mask 2, 3 and 4 (see dt_domask).

6. Output of terrain-following mask#

Mask 1: xy cross section from x = 500 m to x = 1000 m and from y = 1000 m to 2000 m at the second, fith and tenth grid point above the surface. Output of instantaneous and time averaged data of the three wind components.

&initialization_parameters

    nx = 39, ny = 39, nz = 40,
    dx = 50.0, dy = 50.0, dz = 50.0,
    .../

&runtime_parameters

    ...
    data_output_masks(1,:) = 'u','v','w','u_av','v_av','w_av',

    mask_x_loop(1,:) = 500.,1000.,50. ,
    mask_y_loop(1,:) = 1000.,2000.,50. ,
    mask_k_over_surface(1,:) = 2, 5, 10,

    dt_domask = 600.,
    skip_time_domask = 3600.,/

Output starts after 1h simulated time (see skip_time_domask) for both masks and is done every 10 minutes for mask 1 and every 30 minutes for mask 2 (see dt_domask).

Keys	Action
`?`	Open this help
`n`	Next page
`p`	Previous page
`s`	Search