CARMA Sensitivity Calculator


The CARMA Sensitivity Calculator (aka SCalc) calculates the expected RMS noise for an observation with the CARMA array, given inputs about weather, frequency, array configuration, number of tracks, mosaic pattern, and visibility weighting. It should be used when preparing CARMA proposals to estimate the time required to achieve a desired sensitivity. The time estimate produced by this program includes overhead; the value in the field Proposal Request Hours should be used for the time request in your CARMA proposal. When preparing your CARMA proposal, you must select "Typical" weather for your sensitivity estimate.

Click to Launch CARMA Sensitivity Calculator from your browser


Current Version: 4.1/2014-July-29

If you get a security error, try importing the UMD/CARMA digital certificate via the Java console. Then relaunch the application.

Basic Usage

This program calculates the expected RMS noise for an observation with the CARMA array, given inputs about weather, frequency, array configuration, number of tracks, mosaic pattern, and visibility weighting. Changing a text-field entry, radio button, or slider on the left hand side will update the calculated values on the right. Informational and error messages appear in the yellow box in the lower right. Array configurations are described in CARMA memos 20 and 24.

Details of The Calculation

The RMS noise is determined using the equation 6.55 of Thomson, Moran, and Swenson (1st edition) and assuming correlator efficiency 0.873 (2-bit), 0.962 (3-bit), 0.983 (4-bit), 10-m antenna aperture efficiency of 0.6 and 6-m antenna aperture efficiency of 0.65 . Typical weather at Cedar Flat is 5 mm water vapor, very good weather is 2 mm water vapor. Atmospheric phase noise is taken to be 40 degrees for typical weather, and 30 degrees for very good weather. Spillover fractions for 10-m antennas are 0.08*Tambient and 0.04*Tambient for 6-m antennas.

The Total Integration Time is the time actually spent on-source and is calculated using the S/N optimization algorithm outlined in BIMA Memo #47 and updated for CARMA in CARMA Memo #41. (see also the Hour Angle calculation applet, GetHA). By this method, the length of a track is the length required to reach 90% of signal to noise ratio that would be achieved observing from horizon to horizon (above 10 degrees elevation) or 8 hours, whichever is smaller. It assumes a 30% overhead for calibration and slewing with a fixed minimum overhead of 0.5 hours. Your time request when proposing should be Proposal Request time in hours.

Visibility Weighting

The slider on the lower left allows you to modify the weighting of the visibilities which in turn changes the synthesized beam and RMS noise. The weighting scheme uses the robustness parameter devised by Briggs. The robustness parameter can be used to down-weight excessive weight being given to visibilities in relatively sparsely filled regions of the uv plane. A robustness of -2 corresponds to uniform weighting and +2 corresponds to natural weighting. See help on the MIRIAD task invert for more details.

How Shadowing Is Handled

For all arrays configurations the synthesized beam calculated for low declination sources has shadowed visibilties removed and the total time on-source is reduced by the fraction of visibilities shadowed. Shadowing has been calculated using the MIRIAD task csflag, which properly takes into account the differing diameters of the 6-m and 10-m antennas. The shadowing calculation assumes some shadowed visibilities are kept, using 80% of the 6-m antenna diameter and 100% of the 10-m antenna diameter. For those familiar with the csflag task this means using cfraction=1.0,0.8. It is similar but not identical to using select=-shadow(5) in the MIRIAD task uvflag. Future enhancement may allow this to be configurable.

Mosaics

For mosaics, the RMS is calculated over the central region of the mosaic that has unity gain. Only standard hexagonal mosaic patterns are supported.

Notes

Running SCalc

  1. Java Webstart: To run SCalc with Java webstart, you may need to configure your browser. If, when you click on one of the above links, your browser asks you what to do with a .jnlp file, tell it to open with javaws, which should be in the same place as java (Typically, /usr/java for Linux operating systems).

  2. To Run SCalc as a desktop application: Right-click to download the JAR files:
    SCalc.jar
    astroutil.jar
    Latest Version of CGS

    Navigate to the folder where the files were downloaded and type
    java -cp astroutil.jar:CGS.jar:SCalc.jar scalc.CarmaRms
    from a command line. Alternatively, right-click and choose "Open with Java WebStart" from a graphical environment. You must have JRE 1.6 or better to run SCalc as a desktop application.

  3. For Mac users: If you have Java 1.6 installed (or a newer version) and you are still having trouble loading CGS, you may need to change your Java preferences. To do this, go to:
    Finder -> Macintosh HD -> Applications -> Utilities -> Java preferences
    Then, drag the most recent version of Java to the top of the list. The one at the top is the preferred version.

Known issues

  1. If you get a Security Error message: If you get a "security error" when trying to launch this application, you can run SCalc as a desktop application as described above. PLEASE report to Marc Pound (mpound@umd.edu) if you got this error and which OS/Browser/Java Version you were using.

  2. If Cut and Paste doesn't work: If you are unable to paste a CGS correlator script to the input window, you can run CGS as a desktop application as described above. PLEASE report to Marc Pound (mpound@umd.edu) if you got this error and which OS/Browser/Java Version you were using.