3.7) The PIA Mapping Simulation System

We have developed a mapping simulation system within PIA, capable of generating sky images with several point / extended sources on  a flat / gradient background and simulating what ISOPHOT would have recorded under certain instrument and spacecraft raster configurations.  While the benefits of performing simulations for accessing the efficiency, accuracy, confusion level, etc., on different mapping algorithms and deconvolution techniques in and outside PIA are mostly of interest to calibrators and instrument specialists, it is also very important for a general observer because this highly user friendly system provides the possibility of simulating his/her observation by matching the selected observing mode.

The system can be used for testing the following:

  • validity of theoretical/experimental beam profiles
  • detection limits and confusion levels
  • transients and glitches simulated into SRD level data
  • different algorithms in relation to different observation configurations (improving data analysis)
  • reliability of obtained results by real observations
  • The system is divided into three main stages:
  • sky simulation : a portion of the sky as well as the user defined number of sources with their respective positions in the sky, fluxes / brightness and sizes (circular and bar structures are included ) are defined.  The choice between a flat or a gradient background is also available.  Convolution with the footprint of a selected filter yields an image of the convolved sky in addition to the pure sky.  These images together with the parameter set used to create them are the products of this portion of the simulation, which can be saved as data entities for reuse at any time.
  • observing mode specification : the observing mode is defined together with the raster parameters (number of raster legs and points per leg, distance between legs and points, central position of the raster).  Both the orientation of raster vs sky as well as the orientation of spacecraft vs sky are necessary for establishing the exact positioning of the detectors (and hence the "measured" flux) on the sky at any step of the observation.  The result is an AAP level data structure which can be used within the PIA environment as any real observation, e.g. deriving the corresponding image.
  • instrumental effects : the expansion of every spatially "measured" point into a time series of signals as would have been recorded by a real observation is done together with the inclusion of instrumental effects, like the transient behavior of the ISOPHOT detectors, the disturbances produced by cosmic rays, different detector pixel response and a random noise contribution.  The product, a SRD data structure, can also be ingested into PIA for virtually simulating the data analysis.
  • The parameters included in all three stages of the simulation system are kept in configurable tables, so that changes can be easily made from dedicated GUIs.  The model for the transient behavior (by default the so called "offset exponential function" as explained in the section 3.3.6) can be replaced by any other model function.

    The PIA mapping simulation system includes a widget based graphical user interface, with dedicated menus for each of the three parts of the system, as mentioned above.  Visualization of the created sky images with superimposed raster and detector pixel positions is also accessible from the main menu.  The visualization tools attached to this GUI are common to the PIA mapping system, thus offering the user the main IDL capabilities with respect to image display and manipulation from comprehensive menus.


    3.7.1) The PIA Map Simulation Window

    This section describes in detail the PIA Map Simulation window shown in Figure 1 and its pull-down items.  It appears when you click on the
    "Simulator" button on the PIA Top Window.  This window shown below will be referred to as the "main window" in the following sections of 3.7.

    FIGURE 1: The PIA Map Simulation Window


    Parameters

    This pull-down item offers the following options:
    Edit/change sky map parameters

    This brings up the PIA Map Simulation Parameters window shown in Figure 2.

    Edit/change data parameters

    This brings up the PIA Raster Simulation Parameters window shown in Figure 6.


    Sky Simulation

    This pull-down item offers the following options:
    Using parameter table

    This will run the sky simulation using the parameters specified with the PIA Map Simulation Parameters window.

    Using graphic derivation

    This starts the graphic derivation of the sky simulation using the IDL prompt.  With this option, the user is asked to specify the necessary parameters at the IDL prompt by answering series of questions.  As the inputs are made, the program shows displays and information to guide the user along.  Hence, you are encouraged to try this graphic version to familiarize yourself with the sky simulation system.


    Data Simulation

    This pull-down item offers the following options:
    Using parameter table

    This will run the data simulation using the parameters specified with the PIA Raster Simulation Parameters window.

    Using graphic derivation

    This starts the graphic derivation of the data simulation whose end products are AAP and SRD data structures.  With this option, the user is asked to specify the necessary parameters at the IDL prompt by answering series of questions.  As the inputs are made, the program shows displays and information to guide the user along.  Hence, you are encouraged to try this graphic version to familiarize yourself with the raster (AAP data) simulation system.


    Display

    This pull-down item offers the following options:
    Convolved Sky Image (Jy/beam)

    This shows the simulated sky convolved with the footprint of the selected filter in units of Jy/beam.  For example, see the Figure 4.

    Convolved Sky Image (MJy/sr)

    This shows the simulated sky convolved with the footprint of the selected filter in units of MJy/sr.  This is exactly the same as the previous image except for the units (MJy/sr instead of Jy/beam).

    Pure Sky Image

    This shows the simulated sky (ideal image) in arbitrary units, see Figure 3.  Please note that the scale is meaningless in this image.

    Image + Measured Positions

    This shows the simulated sky convolved with the footprint of the selected filter in units of Jy/beam plus the user defined raster positions.  For example, see the Figure 8.

     ... + Det.Pixel pos. (in S/C coord) (Jy/beam)

    This shows the simulated sky convolved with the footprint of the selected filter in units of Jy/beam plus the user defined raster positions in the spacecraft coordinates instead of sky coordinates.  For example, it looks like Figure 9 (except that the unit is in Jy/beam).

     ... + Det.Pixel pos. (in S/C coord) (MJy/sr)

    This shows the simulated sky convolved with the footprint of the selected filter in units of MJy/sr plus the user defined raster positions in the spacecraft coordinates instead of sky coordinates.  For example, see the Figure 9.


    Load map

    This calls the Please Select a File window as shown below.

    Please specify the name of the file containing the saved simulated map and its directory path.

    Data -> PIA

    This pull-down item offers the following options:
    AAP -> PIA

    Select this option to load the generated AAP data into the PIA buffer.

    SRD -> PIA

    Select this option to load the generated SRD data into the PIA buffer.


    Save ...

    This pull-down item offers the following options:
    Generated map

    This calls the Please Select a File for Writing window, which asks for a filename (and directory) under which the generated map will be saved (a default name is proposed).

    Generated AAP

    This calls the Please Select a File for Writing window, which asks for a filename (and directory) under which the generated AAP data will be saved (a default name is proposed) in internal PIA format.

    Generated SRD

    This calls the Please Select a File for Writing window, which asks for a filename (and directory) under which the generated SRD data will be saved (a default name is proposed) in internal PIA format.

    The proposed default name has the following format: "C1" or "C2" depending on the detector used, followed by "99" to differentiate it from real observations, and the remaining digits represent the time the data was simulated (i.e., the each set of two digits is in the order of the month, date, hour and minute).  For example, "C19906171703" can be decoded as a simulated data (looking at the "99" following the "C1") using the C100 detector created on June 17 at 17:03 or 5:03 PM.


    QUIT


    3.7.2) Sky Simulation

    To complete a sky simulation, you must first define the necessary parameters either using the PIA Map Simulation Parameters window  in Figure 2 or using the IDL prompt for graphic derivation.  Then, select the appropriate option under the "Sky Simulation" button in the PIA Map Simulation window; choose the first option, "Using parameter table" if you defined the parameters using the PIA Map Simulation Parameters window or choose the second option, "Using graphic derivation" if you would like to run the sky simulation graphically by making inputs at the IDL prompt.  Finally, you can examine the results (pure sky image and convolved sky images as shown in Figure 3 and Figure 4 respectively) using the "Display" button.

    Using the PIA Map Simulation Parameters window

    Selecting the "Edit/change sky map parameters" under the "Parameters" on the main window brings up the PIA Map Simulation Parameters window shown in Figure 2.  See the detail on each parameter below.


    FIGURE 2: The PIA Map Simulation Parameters window


  • Skybin size S (arcsec) and Sky sidelength L (arcsec):  For now, the skybin size is fixed as 7.5 arcsec, but you can choose the size of the sky by specifying the side-length in arcsec.  The simulated sky is always square in shape.
  • Nr. of sources:  Enter the number of sources you would like to create, and then you must hit the Enter button to modify this window to accommodate that number of inputs.  You can create circular or bar structures, and you also have the option of entering either the flux or brightness for each.  Specify the position of each source, and enter their sizes (Z-dim = 0 for circular sources).  Don't forget to select the correct option, Jy or MJy/sr, for each Flux/Brightness value you enter.  For a point source, simply make the size small enough.
  • Flat Background [MJy/sr]:  Enter the brightness for a flat background.  If you don't want a flat background, change the default value of 10.00 MJy/sr to 0.00 MJy/sr.
  • Gradient Bckgr. [MJy/sr] and Grad. Bckgr. Positions:  Enter the brightness values at two different locations on the sky, and the gradient background will be interpolated between these values.  We suggest that you choose the diagonally opposing corners of the sky.
  • Convolution Filter:  Choose the number corresponding to the filter you would like to use.  For example, the number 5 represents the C100_100 micron filter.
  • Click the button, "Set Default parameters" to change back to default parameters or click on the "Quit" button to use the inputs you made on this window.
  • Running the sky simulation

    If you would like to use the values specified on the parameters table, then simply click on the option, "Using parameter table" under the "Sky Simulation" button on the main window.  When the sky simulation is completed, you can view the results (i.e., the pure sky image and the convolved sky images) using the "Display" button.

    If you would like to run the sky simulation graphically at the IDL prompt, simply click on the second option, "Using graphic derivation" under the "Sky Simulation" button and answer the proposed questions at the IDL prompt.  The results will be displayed as the necessary inputs are made, so you can follow the program graphically.

    Viewing the sky simulation results

    When the sky simulation is completed regardless of the method you used (using parameter table or using graphic derivation), the results are accessible from the main window using the "Display" button (only the first three options are available).  The Figure 3 shows an example of a simulated pure sky image, the Figure 4 the convolved sky image in Jy/beam, and the Figure 5 the surface plot of the convolved image.  This example case shows three different sources: a point source at the center of the sky, a extended circular source and a bar structure.  The available visualization tools are common to the PIA mapping system, thus offering the user the main IDL capabilities with respect to image display and manipulation from comprehensive menus.


    FIGURE 3:  The Pure Sky Image of three different sources


    FIGURE 4:  The Convolved Sky Image in Jy/beam.  The pure sky image in Figure 3 was convolved with C100_100 filter as shown in the plot title.


    FIGURE 5:  The surface plot of the Figure 4.  This is produced using the "Surface" button in Figure 4.


    Now you can save the generated map by selecting the "Generated map" under the "Save ..." button on the main window, which calls the Please Select a File for Writing window.  A default name is proposed.


    3.7.3) Data Simulation

    The observing mode is specified with the raster parameters which include the number of raster legs and points per leg, distance between legs and points, central position of the raster, and different rotation angles (all these will be explained in detail later).  The aim is to establish the exact positions of the detector pixels (and hence the "measured" flux) on the simulated sky at all specified raster positions.  The result is an AAP level data (and a SRD level data if the user wants it) which can be used within the PIA environment like any real observation data.

    The steps involved in the data simulation are very similar to those involved in the sky simulation.  First, you must define the necessary parameters either using the PIA Raster Simulation Parameters window in Figure 6 or using the IDL prompt for graphic derivation.  Then, select the appropriate option under the "Data Simulation" button in the main window; choose the first option, "Using parameter table" if you defined the parameters using the PIA Raster Simulation Parameters window or choose the second option, "Using graphic derivation" if you would like to run the data simulation graphically by making inputs at the IDL prompt.  Finally, you can examine the results (the measured positions superimposed on the convolved sky image in the sky coordinates or the same in the spacecraft coordinates as shown in Figure 8 and Figure 9 respectively) using the "Display" button.  Also, you can load the generated AAP and SRD data into the PIA buffer or save it to a file.


    3.7.4) Instrumental Effects

    The production of a SRD data involves the expansion of every spatially "measured" point (AAP data) into a time series of signals as would have been recorded by a real observation done together with the inclusion of instrumental effects, like the transient behavior of the ISOPHOT detectors, the disturbances produced by cosmic rays (glitches), different detector pixel response ("unflatfielding") and a random noise contribution.  After placed into the PIA buffer, this SRD data can virtually simulate the data analysis.  The following example of a simple P22 observation shown in the Figure 12 is used to demonstrate this point visually.

    FIGURE 12:  A simple P22 observation of 3 x 5 raster


    The following Figure 13 shows the generated SRD data from the simulated P22 observation in the Figure 12.  The value of 10.0 % was used for all of the following parameters: the pixel responsivity variation, noise level, and the individual signal uncertainty.  Also, both transients and glitches were added.  The added glitches should be residual glitches that would not have been removed from the ERD level correction; hence, the glitches you see in this simulated SRD data have moderate values.

    FIGURE 13:  The PIA display of the SRD data derived from the simulated P22 observation in the Figure 12


    The following Figure 14 and Figure 15 visually demonstrate the addition of instrumental effects to the AAP data in the Figure 14 to simulate a SRD data in the Figure 15.

    FIGURE 14:  The zoomed image of the AAP data of a certain pixel.  This is from the observation shown in Figure 12.


    FIGURE 15:  The zoomed image of the SRD data of a certain pixel corresponding to the AAP data shown in the Figure 14.  This is from the observation shown in the Figure 12.


    Chapter history:
    Date  Author  Description 
    28/02/2000  Min Hur (IPAC) / Carlos Gabriel (ESA) First Version