The RTDC
Processing SMA Data
1.2 m Telescopes
AST/RO
Extra

SMA Data Reduction with CASA


CASA introduction

CASA (Common Astronomy Software Applications package) is a software package developed to calibrate and image data from ALMA and other arrays. Calibrating SMA data with CASA has not been rigorously tested and it is possible some issues remain with the conversion. However, it is becoming common to use MIR for calibration and CASA for imaging.


Converting SMA data to CASA format

If you wish you do any reduction using CASA you will need to convert SMA data into a CASA compatible format. You can find a description of the options at Converting SMA data to CASA MS format.

There are three possible paths for converting data to CASA format: MIR, MIR & MIRIAD, and the tailored scripts sma2casa.py & smaImportFix.py. You can find detailed instructions on using sma2casa.py and smaImportFix.py at sma2casa Details.


CASA on RTDC computers

The link below lists where to find the different versions of CASA on the RTDC and CF computers. The latest release will be aliased to the casa command unless you overwrite that in your own setup files.

Accessing CASA on RTDC computers

Be sure you do not run CASA on NFS mounted disks as this will slow it down considerably. CASA is partly multi-threaded so can make use of more than one CPU. However NRAO state that currently clock speed is more important than a large number of cores, and their highest-performance recommendation only specifies 16 x 2.6GHz cores.


Parallelization

With CASA becoming the software of choice for imaging SMA data, there is a demand to speed up the process. As such we are conducting a series of tests to assess the parallelization of CASA using their inbuilt mpicasa framework.

Checking the output of parallelized tclean with CASA 5.3.0
Checking the output of parallelized tclean with CASA 5.1.2
Testing the best way to run CASA on RTDC
Instructions for running CASA in parallel


The future

CASA is increasingly the software of choice for interferometry reduction and we also expect this to the future of SMA data reduction.

The SMA is working on a reliable script to transform raw data to CASA format without requiring MIR or MIRIAD. The task is complicated by the ongoing changes and upgrades to the way SWARM data is collected, along with the wideband-SMA (wSMA) upgrades.

Once the data can be successfully converted, we are confident that existing CASA tasks can be used to calibrate SMA data. The exception will be poliaization data which is very SMA specific.

The long processing time compared to MIR has proved off-putting to some users, however we hope that full utilization of parallel processing will solve this issue.

CENTER FOR ASTROPHYSICS | HARVARD & SMITHSONIAN
60 GARDEN STREET, CAMBRIDGE, MA 02138