Observing with the Infrared Spectrograph

Transcription

1 Observing with the Infrared Spectrograph C. Grillmair, L. Armus GO Workshop November 2002

2 Outline 1) Meet the IRS IST 2) Basic IRS capabilities 3) Observing and readout modes 4) Data products and basic calibrations 5) Some specific data features 2

3 The IRS Instrument Support Team Current team members Lee Armus (IST lead) Carl Grillmair (Uplink cog. Sci.) Pat Morris (Downlink cog. Sci.) Jim Ingalls Sergio Fajardo-Acosta Phil Appleton Martin Burgdorf Jim Brauher Harry Teplitz Responsibilities coordination, interfaces, etc. IOC planning, AIRE, pipeline dev/test Calibration, pipeline dev/test Pipeline dev/test, IER dev, LMMS test support Pipeline dev/test, 2MASS PU target selection Pipeline dev/test, IRS manual editor IOC planning, pipeline dev/test Pipeline IA, pipeline dev/test IRS Peak-up, Peak-up Imaging, pipeline dev/test 3

4 The Infrared Spectrograph PI = Jim Houck, Cornell University Contractor = Ball Aerospace Key Features Uses 128x128 Boeing Si:As and Si:Sb arrays No moving parts Two, R~600 echelle modules ( , µm) Two, R~ longslit modules ( , µm) Peak-up imaging ( , µm) Acquisition of sources with poorly known positions. Peak-up on the science target or a nearby offset star whose relative position is accurately known. The IRS operates in staring or spectral mapping modes 4

5 The IRS Assembled 5

6 The IRS on the MIC Baseplate 6

7 Basic IRS Science The IRS will allow observers to perform spectroscopic observations of previously known sources (IRAS, ISO, 2MASS, etc.) and those discovered by SIRTF itself. extragalactic examples the physical conditions of the atomic/molecular gas in dusty galaxies via emission/absorption lines. the redshifts of optically obscured, distant galaxies. The IRS will enable spectroscopy at levels that are ~100 times more sensitive than those reached by ISO. 7

8 IRS Wavelength Coverage IRS High-Res IRS Low-Res LA - SB Mar 02

9 The IRS in the SIRTF Focal Plane 9

10 The 4 IRS Modules (view from above) Short-High Short-Low & Peakup Long-High Long-Low 10

11 IRS Slit Schematic nd Order ( µm) [1st Order ( µm)] st Order ( µm) Short-Low nd Order 1st Order ( µm) Long-Low ( µm) [1st Order ( µm)] Short-High 11.1 Long-High ( µm) ( µm) 11

12 IRS Peak-up Schematic Blue ( µm) Vignetted Field Red ( µm) Peak-Up Apertures Short-Low Array 128 x 128 pixels (1.8 arcsec /pxl) 1st Order spectrum µm 1st Order nd Order µm Red peakup Blue peakup 12

13 Basic IRS Capabilities Module Array Format (pxls) Pixel Size (arcsec) λ (µm) Resolution Short Low 128x Long Low 128x Short High 128x Long High 128x Sensitivity Levels (point source) Low Res: ~ µm (5σ, 500 seconds) High Res: ~3x10-18 W m µm (5σ, 500 seconds) Peakup: ~0.5 mjy See the real sensitivity plots on the web Saturation Limits (point & extended sources) Low Res: ~5 µm (8 seconds) or ~0.4 Jy arcsec -2 High Res: ~50 µm (8 seconds) or ~2.1 Jy arcsec -2 Peakup: ~0.5 Jy (4 seconds) or ~40 mjy arcsec -2 13

14 Outline 1) Meet the IRS IST 2) Basic IRS capabilities 3) Observing and readout modes 4) Data products and basic calibrations 5) Some specific data features 14

15 IRS Observing Modes 1. IRS Staring Mode The basic point and shoot mode. The slit, integration time, source type, source position(s) and peak-up option are specified by the observer. Peak-up options include use of the IRS, the PCRS, or no peak-up. The observer will usually perform a peak-up to accurately position the science target on the slit. Broad categories of photometric accuracy can be selected with IRS peak-up. For inertial targets, the IRS peak-up can be performed on the science target itself, or an offset star whose relative position is well known. Note: Since each sub-slit covers a separate wavelength regime, a full low or high resolution spectrum will require multiple SIRTF pointings. 15

16 Example of an IRS Staring Mode AOR Details: 1. IRS offset blue peakup target (moderate accuracy). 2. SH and SL selected. 3. Two cycles for SH, three cycles for SL. 4. Fixed, single science target (galaxy 1).

17 IRS Observing Modes (contd.) Peak-up Facts IRS peak-up can be done through one of two filters. These cover the µm (blue) and µm (red) bands. Point source IRS peak-up targets should be in the mjy range (blue) or the mjy range (red), but can be as faint as ~ 1 mjy (blue) or ~ 1-5 mjy (red). Extended source peak-up is available, but is less accurate**. For moving targets, peak-up can only be performed on the science target itself. IRS peak-up overheads can be substantial ( = sec) and are determined by source brightness and requested accuracy**. A visual peak-up using the PCRS is available for sources in the V ~ 7 10 mag. range. **see the SIRTF observers manual for full details on IRS peak-up 17

18 IRS Observing Modes (contd.) 2. IRS Spectral Mapping Mode The spacecraft is directed to perform a raster scan of specified dimensions parallel and perpendicular to the length of the slit. The slit, integration time, source type, source position(s) and peak-up option are specified by the observer. Cycles and Map Cycles are also specified. Note: Since the SIRTF roll angle is strongly constrained, spectral map position angles will only be selectable by constraining the observing window. (This is also true for staring mode) Note: Since the IRS slits are not aligned in the SIRTF focal plane, spectral maps using multiple slits will not be parallel on the sky. 18

19 Example of an IRS Spectral Mapping AOR Details: 1. No IRS peak-up. 2. Two full LL maps (Map Cycles = 2). 3. Offsets are 75.5 along the slit and 25.0 perpendicular to the slit. 4. There are 7 parallel steps and 5 perpendicular steps per map. 5. This is a sparse map, centered on the vignetted region of the LL slit (LL both).

20 All 35 steps in the map for April 24, 2003 Rows 1,7 only = first and last 5 steps. Rows 3,5 only LL 1 st order LL 2 nd order

21 IRS Readout Modes For spectroscopic observations, the arrays will always be read in what is called Raw Data Mode, where: The array is read non-destructively (4, 8, 16 or 32 times) at a limited number of intervals (1, 2, 4, 8, 16 or 32 seconds). Selection of an integration time sets the readout time and number of reads. Each of the reads (128 x 128 x 16 bits/pxl) is stored and transmitted to Earth, creating a 3-D data cube for each spectral ramp. for example, one ramp with 16 reads generates 0.52 Mbytes of data (128 x 128 x 16 bits/pxl x 16 reads) / (8 bits/byte) = 0.52 Mbytes Each of the four arrays can be turned on individually and simultaneously (reducing overheads for multi-slit observations), however, only one array can be read out at a time. 21

22 Outline 1) Meet the IRS IST 2) Basic IRS capabilities 3) Observing and readout modes 4) Data products and basic calibrations 5) Some specific data features 22

23 IRS Data Products 1. Raw Data Each observation (AOR) consists of number of Data Collection Events (DCEs). These DCEs, converted to FITS format, are the Raw Data Products. Each DCE consists of a set number of reads (samples) of the array at fixed time intervals without a reset between samples. example: a 256 second DCE can result in a 16 plane FITS data cube where each plane is a sample separated by 16 seconds of time. Each consecutive read has accumulated more photons than the previous read. This is the IRS Raw Data Mode, a.k.a. Sampling up the Ramp. 23

24 IRS Low Resolution Data Short Low Long Low 7.5µm 7.5µm 8.5µm Red peak-up 20.8 µm 20.0 µm 21.0 µm 5.3µm µm 14.2 µm Blue peak-up µm 14.2µm 7.5µm 40.0 µm 21.8 µm 24

25 IRS High Resolution Data Short High 10.0 µm Long High 37.0 µm 19.5 µm 19.3 µm 25

26 IRS Data Products (contd.) 2. Basic Calibrated Data = a FITS file made from each DCE which has all SIRTF-specific effects removed from the spectrum. A calibrated 2-D echellogram for the high-res modules. A calibrated long-slit 2-D spectrum for the low-res modules. Other cal files used to make the BCD will also be available. (e.g., the wavelength map, flat field image, illumination correction). Most observers will never need to redo the calibrations performed in generating the IRS Basic Calibrated Data product. 26

27 IRS Data Products (contd.) 3. Post-BCD Data = intended to convey the richness of the target spectrum and allow for first-look scientific analysis. For the short (5 pixel) IRS high resolution slits and the long (30 pixel) IRS low resolution slits, the post-bcd will be: A set of 2D, spatially rectified spectra, created by co-adding all observations obtained at the same position with a given slit. An extraction of all 1D spectra from the 2D rectified spectra. A final co-addition of all 1D spectra from the entire AOR for each slit. 27

28 IRS Data Product Schematic Raw data BCD Pipeline Processing Sample Time Post-BCD 1D spectrum Pipeline Processing Post-BCD 2D spectrum not shown f λ λ 28

29 Regular IRS Calibrations Calibrations to be performed during every IRS campaign will include: Spectro-photometric: Observations of A dwarfs and K giants and use of modelled and synthetic spectra. Flat fields: Observations of the zodiacal emission at high and low ecliptic latitudes and well-calibrated stellar sources. Internal stimulator lamps will allow monitoring of relative pixel-to-pixel variations with time. Wavelength: Observations of bright PNe and Be stars. Dark sky: Observations of the NEP and Elais-N1 field will be obtained (corresponding to all available integration times) and subtracted from matching science data from that campaign. Other calibrations (e.g. spectral leaks, scattered light) will be performed less regularly. It is expected that the IRS will be extremely stable, and that very high S/N spectro-photometric, flat fields, etc., will be built up over many campaigns. 29

30 Outline 1) Meet the IRS IST 2) Basic IRS capabilities 3) Observing and readout modes 4) Data products and basic calibrations 5) Some specific data features 30

31 IRS data features There are a number of data features that are being treated in the IRS pipeline and/or via specialized software tools. These are: 1) SL stray light (pipeline) 2) High-res order cross talk (pipeline) 3) Fringing (1D de-fringing tool) In addition, there has been a degradation of the response at long wavelengths in low resolution due to a de-lamination of the LL1 order sorting filter. Since it s initial discovery, there has been no subsequent decrease in sensitivity. This will be carefully assessed in IOC, and observing strategies / recommendations will be made to the community before proposals are due. 31

32 SL Stray Light & High-res crosstalk SL Stray Light Light from the peak-up arrays can spill over into SL 1 st and 2 nd order spectra. Exponential fits (based on lab data) are scaled to the inter-order areas to remove the stray peak-up light from the spectra. This is critical for proper SL flat-field generation. Cross Talk Light from closely-spaced echelle orders can overlap at low flux levels. Fits to (relatively) un-contaminated orders are scaled to all orders and a model of the cross-talk is created and subtracted. The largest corrections occur at the edges of the bluest orders. In most cases the correction is only a few %. In both cases the SSC will provide corrected and un-corrected science BCD s. All the flats will be corrected. 32

33 Peak-up stray light Un-corrected stray light from the peak-up arrays would affect our spectral flat fields 33

34 Exponential + Gaussian fits Spectral contamination is ~15% of the PU edge flux.

35 Peak-up stray light correction Before correction After correction Scaling Region 35

36 Order Cross Talk Cross talk = seen as a rise in the baseline level between the orders due to order overlap. The effect is most severe at the shorter wavelengths. 36

37 IRS Fringing Fringing Fringing is seen in both high-res modules at peak-to-peak amplitudes of 3-5% of the continuum (SH) and 6-10% of the continuum (LH). Periods of approximately 1.3 cm -1 (band-pass filter) and 3 cm -1 (detector) have been seen in ground test data. Unresolved fringes may also be present in the low-res modules. An interactive, 1D de-fringing tool will be available at launch. This tool performs (multiple) sine wave fitting on each SH and LH order. Fringe amplitudes and frequencies will be re-measured in IOC. 37

38 De-fringed spectrum Original spectrum Difference spectrum

39 Long-low Filter The performance of the long wavelength end of long-low (LL1) has degraded due to de-lamination of the order sorting filter. If there is no further degradation of the LL filter, the in-flight performance of LL1, as compared to estimates given in the current version of the SIRTF manual, will be characterized by: A loss of overall sensitivity of 30 50% Flat-fielding uncertainties that may preclude meeting the 5% relative spectro-photometric accuracy requirement in LL1. Possible short-wavelength light leaks that would require LL2 and SL observations to remove properly. Stay tuned for IOC/SV performance updates. 39

40 IRS Summary While there are clearly a number of optical and array effects that we need to monitor (and observers should be aware of), overall the IRS is versatile, well-designed, and simple to use. It will deliver: 1) Stable & well-calibrated spectra from ~5-40 microns 2) Very high sensitivity to point and extended sources (both inertial and moving targets). These observations should facilitate detailed studies of known dusty or distant sources, as well as open up whole new areas of Galactic and extragalactic study in the mid-infrared. 40