Difference between revisions of "Protoplanetary Disk Simulation (CASA 5.4)"
(Created page with " Category: Simulations * '''This is an advanced simulation tutorial. New users are recommended to begin with reading the [http://casa.nrao.edu/casadocs/latest/ CASA Docs...")
Revision as of 17:25, 14 December 2018
- This is an advanced simulation tutorial. New users are recommended to begin with reading the CASA Docs documentation pages on "Simulations".
- This guide is applicable to CASA version 5.1. For older versions of CASA see Protoplanetary_Disk_Simulation_(CASA_4.4).
- To create a script of the Python code on this page see Extracting scripts from these tutorials.
For this CASA Guide we will use a protoplanetary disk model from S. Wolf. Get the data here. See the Simulation Inputs page for this and other sample images. If you use this FITS data for anything more than learning CASA, please cite Wolf & D'Angelo 2005.
Script with Explanation
Set simobserve as the current task and reset all parameters.
# In CASA default("simobserve")
Review the image coordinate system using task imhead.
# This reports image header parameters in the Log Messages window imhead("ppdisk672_GHz_50pc.fits")
We now use the ia (image analysis) and qa (units and quantities) tools from the CASA Toolkit to find the image center. In comparison to tasks, tools are a more advanced way of manipulating data in CASA. You can learn more about tools using the tool reference manual.
When data are being manipulated with tools the data file must be explicitly opened and closed.
# In CASA ia.open("ppdisk672_GHz_50pc.fits")
Next, get the right ascension and declination of the image center. We get the number of pixels along each axis using ia.shape. Then, we get the RA and Dec values for the center pixel using ia.toworld.
# In CASA axesLength = ia.shape() # Divide the first two elements of axesLength by 2. center_pixel = [ x / 2.0 for x in axesLength[:2] ] # Feed center_pixel to ia.toworld and and save the RA and Dec to ra_radians and dec_radians (ra_radians, dec_radians) = ia.toworld( center_pixel )['numeric'][:2] ia.close()
Use the qa tool to convert the image center from radians to sexagesimal coordinates.
ra_hms = qa.formxxx(str(ra_radians)+"rad",format='hms',prec=5) dec_dms = qa.formxxx(str(dec_radians)+"rad",format='dms',prec=5)
Let's call our project "psim2". This defines the root prefix for any output files from simobserve.
project = "psim2"
We'll set skymodel to the FITS file downloaded, above, and leave all skymodel subparameters at their default values. simobserve will create CASA image psim2.skymodel.
skymodel = "ppdisk672_GHz_50pc.fits"
We will specify the sky position for the center of the observation and set the map size to the size of the model image. Since the model image is 2/3 arcseconds across, we should only need one pointing. In this case, pointingspacing and maptype can be left at their default values.
setpointings = True direction = "J2000 18h00m00.031s -22d59m59.6s" mapsize = "0.76arcsec"
We do want to simulate an interferometric observation, so we set obsmode accordingly. We'll set totaltime to a 20-minute snapshot observation.
obsmode = "int" totaltime = "1200s"
We want to use the appropriate antenna configuration for the desired angular resolution. Configuration 20, alma.out20.cfg, is the largest "compact" configuration. A list of configuration files available in CASA 4.1 is available here.
antennalist = "alma.out20.cfg"
We do not want to simulate observations with thermal noise in this example, so we set:
thermalnoise = ''
Now run simobserve.
Now that we've simulated the visibility measurements, we want to generate an image from the simulated data. simanalyze makes this easy. We begin by setting project to the same prefix used in simobserve and setting image to True.
default ("simanalyze") project = "psim2" image = True
We set modelimage to use the input FITS image when cleaning the simulated visibilities. We set the image size to 192 pixels square.
modelimage = "ppdisk672_GHz_50pc.fits" vis = project + ".alma.out20.ms" imsize = [192, 192]
Specify the number of iterations for cleaning, with proper threshold and weighting.
niter = 10000 threshold = "1e-7Jy" weighting = "natural"
We'd like to calculate a difference and fidelity image, but we don't need to see the uv coverage.
analyze = True showuv = False showresidual = True showconvolved = True
Plot both to the screen and PNG files with lots of messages.
graphics = "both" verbose = True overwrite = True