Positive signal strength vs negative signal strength for ABBA dither pattern

[astrophy...@gmail.com]

Hi Nick,
After running our MOSFIRE data through the most recent version of the MOSFIRE DRP (mosdrp_2014.06.10 ) I have noticed something that I would like to get a clarification on. For reference I am using data from all four passbands, Y, J, H, and K and the data were all taken in the ABBA dither mode of observing.

On every science mask design we include one star for flux calibration purposes, but also for sanity checks. Checking several stars on different masks in different bands I looked at the summed spatial profile of the detection. (By summed spatial profile I mean for each spatial position I summed the flux over all wavelengths from *_eps.fits image and then plot the flux versus the spatial position.) I have included a sample plot below. The overall shape of the summed spatial profile looks great but the intensities of the positive signal compared to the negative signal are what bother me. In the newest version of the pipeline these intensities are nearly equal while in the old version of the pipeline the negative signal intensities were ~0.5*positive signal intensity.

To my knowledge the ABBA dither pattern uses the B (A) dither position to accurately subtract the sky from the A (B) dither position and is done for every AB pair. Once this is done you stack all the sky-subtracted A positions as well as all the sky-subtracted B positions (independently) result in two stacks. To achieve your final signal you align the positives signal locations in the A, B stacks and then coadd the two. This would result in a summed spatial profile similar to what the previous version of the pipeline would give. Has the new version of the pipeline implemented a new reduction algorithm where I should expect the positive and negative signals to be equivalent in strength?

/Users/adegroot/Desktop/example_summedspatialprofile_star.png

[Andrew DeGroot]

Here is the actual plot file. 

--
You received this message because you are subscribed to the Google Groups "MOSFIRE Data Reduction PIpeline" group.
To unsubscribe from this group and stop receiving emails from it, send an email to mosfire-drp...@googlegroups.com.
To post to this group, send email to mosfi...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/mosfire-drp/eb7fadfe-5bba-4d4c-bb6f-acc29ea07009%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

--

===========================

Graduate Student

University of California, Riverside

Department of Physics and Astronomy

[Chuck Steidel]

Hi Andrew,

   What is actually done is that a “co-average” is made of the two spatial position spectra, since

we are always working in units of e-/second of integration time.  With this scaling, the negative

images should have the same “intensity” as the positive ones, but will have sqrt(2) lower S/N.

   The reason the old reductions behaved differently is that this averaging step was being done

incorrectly (it was doing an addition, but not changing the units appropriately, so that fluxes

were 2 times higher than they should have been, and S/N sqrt(2) smaller. It’s a long story, but

we believe the current version is the correct one!

   Note that all of this is not necessarily true for objects that are not on the slit at both

dither positions (e.g., for serendip objects) , and you can get asymmetric positive/negative 

images of those in the final stack of the As and Bs.

      -Chuck

To view this discussion on the web visit https://groups.google.com/d/msgid/mosfire-drp/CAKoWSA2bJQK4q9ibLVFq9MOjQo0qrSzD%2BtvjEDJdYvMkm35mYg%40mail.gmail.com.

For more options, visit https://groups.google.com/d/optout.

<example_summedspatialprofile_star.png>

[astrophy...@gmail.com]

Hi Chuck,
Thank you for the explanation! That makes much more sense now. And I can alter my code accordingly now that I have a better understanding of how the how the different dither positions are combined.

Best,
Andrew

[khru...@gmail.com]

Hi Chuck,

Thanks for that explanation. I have a similar question. On one slit I summed up all the spectral pixels in the dispersion direction, and plotted it vs. spatial position, exactly as what Andrew has done. The only difference is I summed up just around the emission line and not the whole row of pixels in the slit. Anyway, the weird thing is the two negative signals have very different fluxes - it looks the same on the 2D spectrum as well. Is this still normal?

A second related question is, there were a few slits where in making the same plot, the positive signal is indeed twice the height of one of the negative signals, while others (on the same mask) are roughly the same. I've verified that I reduced the data with the newest pipeline (June 10, 2014), so is it normal to still see cases where the positive and negative signals do not have the same (absolute value) intensity?

Thanks!