NHSC/PACS Webinar. Aperture and pointing corrections for point sources

Transcription

1 NHSC/PACS Webinar Aperture and pointing corrections for point sources D. Fadda December 13, 2012 page 1

2 Introduction This web-seminar will introduce you to the aperture correction and pointing/jitter correction for point sources. Pre-requisites Installing HIPE 10.0 on your computer PACS data reduction guide Spectroscopy: section page 2

3 Overview 0 Point sources 1 Aperture correction 2 Pointing/jitter correction 3 Case of non-central pixel page 3

4 What is the beam of Herschel? The beam of Herschel has been empirically derived in several key-wavelengths for all the 25 modules of the blue and red arrays. The pixel size is 9.4 x 9.4 and, even at the bluest wavelengths, approximately 30% of the flux falls out of the central pixel for a point source. In the red, the percentage could be bigger than 50%. On the top of that, pointing inaccuracies and jitter of the telescope add uncertainty on the total flux collected in the central pixel. In this case, unfortunately, a solution is available only in the case of bright sources. page 4

5 The beam The pixel size of the array is 9.4 x9.4. The FWHM varies at different wavelengths from 8 x8 in the blue to 12 x12 in the red. page 5

6 The empirical beam 75 μm 150 μm The dashed lines indicate the percentage of flux in the central spaxel and in the central 3x3 spaxels in the case of the empirical beam. Most of the flux is inside The 3x3 central spaxels. page 6

7 The PSF model The aperture correction has to take into account the totality of the flux not observed in the central spaxel. Direct observations are limited to a 30 arcmin radius and the outer parts are dominated by noise. For this reason, a PSF model is used to compute the fraction of flux seen in the central pixel. As shown, the flux fraction missed can be quite significant especially in the red. page 7

8 Aperture correction in the pipeline For point sources, the recommended way is to use the rebinned cube to measure the flux. If correctly planned, the source should be in the central pixel. You can use: extractspaxelspectrum to extract the spectrum of a spaxel in SimpleSpectrum format. Then use: pointsourcelosscorrection, to correct for the loss of flux using a PSF model. page 8

9 Aperture correction in the pipeline In the case the pixel with flux is the central one, the module extractcentralspectrum which combines extraction with aperture correction can be used: centralspectrum = extractcentralspectrum(slicedfinalcubes, \ slice=slice, nonans=1, width=0, \ applypointsourcecorrection=1, \ correct3x3=0, caltree=caltree) Choosing width=0, will apply a single median correction over the entire wavelength range. This is the recommended choice for line scan AOTs. The option correct3x3 is explained later. page 9

10 Pointing accuracy Herschel had a requirement of 1.5 arcsec pointing accuracy which was achieved in the year of observations. In the last year, the pointing has been improved to 1.0 arcsec. Pointing jitter requirement was of 0.3 arcsec/min. Occasionally, depending on the number of guide stars available, this value can be greater. Since such displacements can lead to substantial variations of the collected flux it is advisable to apply a correction. Unfortunately, it is possible to derive a correction only for bright sources with some flux in the central 3x3 spaxels. For calibration sources, already adding up the flux from the central 3x3 spaxels improve the reproducibility by a factor 2 wrt flux from the central pixel only. page 10

11 Pointing accuracy In case of point sources in the central spaxel, it is possible to Apply the so-called 3x3 correction. This takes into account the flux from the 3x3 central spaxels To correct for mispointing the flux of the central spaxel. Please, note that this correction can be applied only if there is measured flux in all the 3x3 central spaxels. centralspectrum = extractcentralspectrum(slicedfinalcubes, \ slice=slice, nonans=1, width=0, \ applypointsourcecorrection=1, \ Correct3x3=1, caltree=caltree) page 11

12 Non central pixel In the case the pixel with flux is not the central one, we have to first extract the pixel and the apply the aperture correction. This is available under scripts in HIPE 10. centralspectrum = extractspaxelspectrum( slicedfinalcubes, slice=slice, spaxelx=spaxelx, spaxely=spaxely) centralspectrum = pointsourcelosscorrection( centralspectrum, caltree=caltree) page 12