Week 10. Lab 3! Photometric quality. Stamp out those bad points. Finish it.

Transcription

1 Week 10 Lab 3! Photometric quality. Stamp out those bad points. Finish it. Lab 4! Great data. Evening sessions this week focus on Lab 3 wrap-up and Lab 4 reducgons. Exams ready for return Read the book! Chromey Chapters 9 and 10 are excellent reinforcement. Don t plan on going to graduate school without owning this informagon. Final project familiarize yourself with the available instruments at APO, browse the schedule for proposal abstracts. hvp:// Topics: CCD/Detector/Calibration wrapup. At-the-telescope techniques wrapup. Aperture photometry PSF fit photometry

2 CCD Overscan Since the number of shiys applied to a CCD during the readout phase is arbitrary it is possible to Underscan read out only a fracgon of the pixels (for a CCD largely in one corner). Doing so could reduce readout Gme between frames substangally. Overscan clock more pixels than are actually there. Doing so brings empty fake pixels to the readout amplifier. These fake pixels are useful for determining Gme dependent bias changes since they are read at nearly the same Gme as the pixels in a given row. Chromey Electrical pickup

3 Infrared Arrays Visible wavelength CCDs benefit from the fact that the primary integrated circuit construcgon material is silicon, which also serves as the detector material. Infrared detectors require material with smaller bandgap (InGaAs, InSB, HgCdTe are popular) which must be bonded to the readout electronics (sgll out of necessity made of silicon).

4 Infrared Arrays This architecture naturally leads to designing readout circuits which have readout amplifiers under EVERY pixel. The charge never moves. Pixels are addressed by their (x,y) coordinates. The collecgng image can be read out without disturbing it (non-destrucgve readout). Very easy to address mulgple and randomly distributed subapertures. Easy to establish guide areas while the rest of the chip integrates.

5 Astrophysical advantages SensiGvity to cooler objects. Infrared Arrays Wein peak is at 500 nm in the visible for 6000K objects. The Sun and hot stars Wein peak is at 2000 nm (2 microns) in the infrared for 1500K objects. Forming stars, self luminous young massive planets, brown dwarfs Wein peak is at 100,000 nm (100 microns) in the far infrared for 30K objects. Warm dust in galaxies, mass losing stars

6 Astrophysical advantages SensiGvity to cooler objects. Infrared Arrays Wein peak is at 500 nm in the visible for 6000K objects. The Sun and hot stars Wein peak is at 2000 nm (2 microns) in the infrared for 1500K objects. Forming stars, self luminous young massive planets, brown dwarfs Wein peak is at 100,000 nm (100 microns) in the far infrared for 30K objects. Warm dust in galaxies, mass losing stars

7 Infrared Arrays Astrophysical advantages SensiGvity to cooler objects. Wein peak is at 500 nm in the visible for 6000K objects. The Sun and hot stars Wein peak is at 2000 nm (2 microns) in the infrared for 1500K objects. Forming stars, self luminous young massive planets, brown dwarfs Wein peak is at 100,000 nm (100 microns) in the far infrared for 30K objects. Warm dust in galaxies, mass losing stars

8 Infrared Arrays Astrophysical advantages Increasing transparency of dust at longer wavelengths. Interstellar grains are typically sub-micron in size and good at blocking wavelengths of light their size or smaller. Red/infrared gets through, blue blocked.

9 Infrared Arrays Astrophysical advantages Increasing transparency of dust at longer wavelengths. Interstellar grains are typically sub-micron in size and good at blocking wavelengths of light their size or smaller. Red/infrared gets through, blue blocked.

10 Focus and f/# The speed of an opgcal system determines its sensigvity to focus changes (specifically defocus as a funcgon of displacement of the sensor focal plane relagve to the opgcs true focal plane). Fast systems have low f/# s. Slow systems have high f/# s. The size of the defocused image is the upstream or downstream displacement of the focal plane divided by the f/#.

11 Focusing an OpGcal System Most telescopes have a physical knob to turn or an electronic control of primary vs. secondary separagon to alter the locagon of the focal plane. In simplest terms focusing involves minimizing the FWHM or RMS size of the focal plane images of point sources. The image size minimum can be dictated by many factors Airy disc size / diffracgon limit System opgcs Camera pixel size Seeing

12 Focusing an OpGcal System Most telescopes have a physical knob to turn or an electronic control of primary vs. secondary separagon to alter the locagon of the focal plane. In simplest terms focusing involves minimizing the FWHM or RMS size of the focal plane images of point sources. The image size minimum can be dictated by many factors Airy disc size / diffracgon limit System opgcs Camera pixel size Seeing Extreme defocus shows you the telescope entrance pupil.

13 Seeking OpGmal Focus Because finding focus involves locagng the minimum of a funcgon, eyeballing focus can be frustragng and Gme consuming, especially with variable seeing. BeVer to take FWHM measurements vs. encoded focus value systemagcally and fit the minimum. Or find the wings, split the difference

14 Focus vs. Temperature Metals contract as temperature decreases. Since a metal structure establishes the primary/secondary separagon in a Cassegrain telescope or the objecgve/eyepiece separagon in a classical refractor opgmal focus is temperature dependent.

15 Tracking and Alignment A telescope cannot necessarily be ley on its own to track the stars The R.A. drive rate can be slightly off The gear train can be imperfect The polar axis can be misaligned

16 Guiding / Autoguiding With real Gme imagery, either from a sub-region of the science array or from a separate guiding array, tracking errors can be minimized with small guiding correcgons (buvon pushes) applied to the telescope drives.

17 Hand Paddles Manual control of telescopes occurs both at control room consoles, on computer screens and, in the dome, with a mulg-buvon hand paddle.

18 Hand Paddles Manual control of telescopes occurs both at control room consoles, on computer screens and, in the dome, with a mulg-buvon hand paddle.

19 Aperture Photometry A single pixel in a star s image contains flux from the star and background from the sky and potengally from unwanted celesgal sources.

20 Aperture Photometry Measure the sky in an annulus surrounding the star, hoping that the annulus median (not average!) flux is representagve of the background underlying the target.

21 Aperture Photometry Add up the flux in the (red) star aperture subtracgng off the esgmate of the sky flux from the (green) sky annulus.

22 Aperture Photometry Decisions, Decisions How big should the star aperture be? Too big and you get extra sky flux and associated Poisson nose (not to mengon contaminagng unwanted stars). Too small and you capture only a fracgon of the star s flux. Where should the inner and outer sky annulus radii be? Inner radius too Gght and the sky annulus will contain stellar flux biasing the background. Outer radius too large and the measured sky may be unrepresentagve of what underlies the target. What to reject and how? The sky annulus in the example contains a bright star. Use a cookie cuver to remove it from the accoungng? Reject its flux stagsgcally, for example using a median rather than an average. Reject pixels lying more than N standard deviagons above the mean in the sky annulus.

23 The encircled flux from the star increases with increasing primary aperture ungl the aperture captures most all of the flux. As long as the same aperture is used on the target and comparison stars the magnitude difference will be correct even if not all of the flux is in the aperture. In fact it may be preferable to use an aperture that that is not overly large to avoid contaminagon from background and other stars. Effects of Varying the Star Aperture

24 Effects of Establishing the Centroid Photometric results depend on the center of the extracgon aperture. Slight shiys will change, fracgonally, what pixels are in/out of the aperture. x cen = x y x y x * (Image[x, y] Background) (Image[x, y] Background) Chromey p. 311 Since since the aperture sum used to get the stellar (as well as background annulus) flux uses pixels and fracgonal pixels based on the centroid the results will change slightly with slight shiys in centroid.

25 Centroid Noise In high-precision photometry centroid noise may be the limigng factor. A robust centroiding algorithm is essengal. Consider a series of measurments with small sub-pixel centroid offsets.

26 PSF-fit Photometry In dense fields aperture photometry may become impracgcal. Inevitably aperture photometry also accommodates more background than desirable. Extract stellar flux in the leanest meanest footprint possible.

27 PSF-fit Photometry The peak counts for a star can be a good proxy for its flux. BeVer yet, the full stellar profile/shape provides informagon that can yield an opgmal esgmate of the total flux using least square fiong procedures where the centroid and peak are the free parameters.

28 PSF-fit Photometry The peak counts for a star can be a good proxy for its flux. Determine the stellar profile (PSF) from all of the stars on the frame. In the absence of significant opgcal distorgons all stars on an image have the same shape. Fit that shape to each star image. Fit the bright stars first and subtract, leaving a cleaner image on which to fit the fainter stars.