Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://www.stsci.edu/hst/training/events/InfraredInstrumentation/IR10Feb00.pdf Äàòà èçìåíåíèÿ: Tue Apr 23 15:20:07 2002 Äàòà èíäåêñèðîâàíèÿ: Sun Dec 23 09:18:52 2007 Êîäèðîâêà: Ïîèñêîâûå ñëîâà: ï ï ï ï ï ï ï ï ï

Infrared Instrumentation & Observing Techniques
The At mosph ere & Backgrou nd Sour ces
Don F. Figer 10 February 2000

Course Outline
v v v v v v

Introduction Atmosphere & Ba ckground Telescope & Optics Infrared Detectors Observ ing techn ique & reduction Questions & A nswers

1/27/00, S. Beckwi th 2/10/00, D. Figer 2/24/00, M. Robberto 3/9/00, B. Rauscher 3/23/00, M. Dickin son 4/6/00, all presen ters

http://www .stsci.edu/ssd/OTWG/

1


Aims for this lecture
v v v v v

Describe atm ospheric ef fects on IR observ ations Sum mar ize prim ary backgr ound sou rces Teach m ethods in estim ating sign al-to-n oise (S/N) Giv e S/N examp le problem s Giv e exampl es of ways to in crease S/N

Lecture Outline
v

The atm osphere
ü ü ü

absorption emi ssi on turbulenc e

v v v v

Sig nal-t o-noi se ratio (S/N) Backgrou nds S/N exam ple Mitig ation t echni ques

2


Atmospheric eff ects
v

Absorpti on
ü ü

reduced source flux dif fi cult cali brations in creased background nois e reduced integration tim es in creased object si ze (ðseei ngñ)

v

Em ission
ü ü

v v

Turb ulen ce
ü

All eff ects vary wit h wavelen gth , tim e, altitude, li ne-of sigh t

Atmospheric absorption
v v

Molecul es are the domi nant absorber s (H2O, CO2, Ox) Stron g fu nction of:
ü ü ü ü

wavel ength, atmospheri c waveban ds (JHKLM NQ) time , frequent cali bration (hour time scales ) alti tude, mountai n-top observi ng si tes li ne-of-si ght, lim ited target access and frequent cali bration

3


Atmospheric absorption versus
v

Sharp cutoffs
ü ü

defined primarily by H2O shape wavebands

v

v

High er transmi ssi on between l in es with hi gher resolution Can i ntroduce large cali bration errors for low resolution observ ations (MNRAS, 1994, 266, 497)

Altitude: 4200m Airmass: 1.0 H2O column: 1.2mm Resolving power 3000

"Thes e data, p roduc ed using the pro gram IRTRANS4, we re o btain ed fro m th e U KIR T world wid e w eb page s.ñ
http://www.jach.hawaii. edu/JACpublic/UKIRT/astronomy/calib/atmos-index.html

Atmospheric absorption versus - high res

Array defects

R = / 23,000
+ +

CO2 absorption lines

Ke ck II 10- m Fig er e t al. 2000, Ap J, acc ept ed

_

4


Atmospheric absorption versus altitude
v

Particle n umb er densiti es (n) for most absorbers fal l off rapidl y with i ncreasi ng alti tude.

I = I0 , e

-



, where is optical depth,
-x / x
0

ndx e
v v

dx

x0,H20 2 km, x0,CO2 7 km, x0,O3 15-30 km So, 95% of atmospheri c water vapor is b elow the alti tude of Mauna Kea .

Atmospheric absorption versus altitude

5


Atmospheric absorption versus airmass
v

The am oun t of absorbed radiati on depends upon the num ber of absorbers alon g the l ine of sigh t
AM=1 AM=2

Atmosphere

I = I 0, 10

- mag / 2.5

, mag = AM ,

where is atm. extinction coefficient.

Atmospheric emission
v v

Blackbody (therm al) Molecul ar (OH)

6


Thermal em ission
spectral radian ce, brigh tness, speci fi c in tensity : I = cos B(T) W m -2 Hz-1 sr-1
... emissivity (dimensionless)

Plan ck (blackb ody) fun ction: 2h3 1 B(T) = 2 c exp(h/kT) - 1 B(T) = 2hc2 1 5 exp(hc/kT) - 1

Peak in B or B: max (µm) = 3674 K T max ~10 µm f or room temp.

Blackbody Curves

7


Atmospheric emission: Blackbody
Total power on to a detector: P = A sky B(Tsky)
: sky: A: : : B(Tsky transmi ssi on of all optics x Q.E. emi ssi vi ty of sky telescope area soli d angle sub tended by f ocal pla ne aperture bandwi dth ): Planck fun ction
-10

At 10 µm, typi call y: ~ 0.2, ~ 0.1, A ~ 3x10 ~ 1.5 x 1013 Hz (10 µm f il ter), T ~ 270 K

m2 Sr

P ~ 10-9 W or ~ 4 x 1010 s-1

Atmospheric emission: OH lines
OH lines Photons s-1 m-2 asec-2 µm-1

Wavelength {µm}

8


Atmospheric emission: Molecular lines

Array defects

OH lines

R = / 23,000
+ +

Ke ck II 10- m Fig er e t al. 2000, Ap J, acc ept ed

Atmospheric emission versus time

9


Atmospheric Turbulence
v v

Static atm osphere bends l igh t (n = 1.000273 at 2 µm) Dyn ami c atmosph ere distorts the wavef ront

Orionis (ðBetelgueseñ)

Data from the W illiam Her sch el 4.2 metre Telesc ope in La Palm a. Wa veleng th 6 89.3 nm, Expos ure tim e 30 ms per fra me. The data were tak en w ith a simp le Spec kle Ca mera situa ted o n the GH RIL p latf orm of th e telesc ope. Th e da ta se ts wer e tak en w ith th e Spec kle Imag ing grou p a t Ca rdif f Univ ersity .

Atmospheric Turbulence
v

A di ff raction-li mi ted point spread function (PSF) has a f ull -width at half -max im um (FW HM) of:


v

FWHM

= 1.2

{m} {µm} {radians} = 0.25 {"} D{m} D{m} {µm} {"}, where r0 6 / 5 . r0{m}

In reali ty, atmospheric turbul ence sm ears the im age:


v v

FWHM

= 0.25

At Mauna Kea, r0=0.2 m at 0.5 µm. ðIsoplanati c patchñ is area on sky over wh ich ph ase is rel ativel y constant.

10


Atmospheric Turbulence
1.4O seei ng 0.5O seei ng no seei ng!

Lick 3-m
Fig er 1995 PhD Thes is

Keck I 10-m

Ser abyn, Shup e, & Fig er Nature 1998, 394, 448

HST/NICMOS 2.4-m
Fig er e t al. 1999 Ap J. 525, 750

Signal - definition
v

Signal is that part of the measurement which is contributed by the source.
S = F tQE {e - }, where h = atmteleinstr .
total

A



total

F: source flux tot al: transm issi on of atmosphere, telescope, optics t: in tegration tim e v

Must also apply aperture correction.

11


Noise - definition
v v v

Noise is that part of the measurement which is due to sources other than the object of interest. In sensitivity calculations, the ðnoiseñ is usually equal to the standard deviation. Random noise adds in quadrature.
N
total

=


i

Ni2 .

Noise - sources
v v v v v

Photon noise from source Noise from backgrounds Detector/electronics noise Systematic noise Data reduction noise

12


Noise - sources: Photon noise from source
v

The uncertainty in the source charge count is simply the square root of the collected charge.
N
source

= S=

total

A

F tQE {e - }. h

v

Note that if this were the only noise source, then S/N would scale as t1/2. (Also true whenever noise dominated by a steady photon source.)

Noise: Read noise vs. flux noise limited
4

3

slope=1/2 (flux dominated)

2
slope=1 (read noise limited)

1 0 1 2 LOG (t ime) 3 4

13


Noise - sources: Noise from background
v

Background is everything but the object of interest!
N
back

& =C

back

t=

total

A

F tQE {e - }. h

Background - sources
v

Atmosphere
ü ü

therm al mol ecular therm al scattering

v

Telescope
ü ü

v v

Zodiacal light Astronomical sources

14


Background - sources: Atmosphere
v

Thermal
& C
sky ,thermal

=

instr

tele A

sky B (Tsky )QE {e - s -1} h

v

OH
ü ü

The averag e OH li ne in tensity is approxi matel y 25,000 s-1 m -2 asec-2 µm-1. The conti nuu m between lin es is about 50 tim es lower than th is val ue (in the H band) .

Background - sources: Telescope - scattering
v

Mirrors
s Iscattered , where s is RMS deviation from a perfect surface.
2

v v

Baffle edges and walls Secondary support

15


Background - sources: Astronomical
v

Astronomical objects can be objects of interest or noise contributors, depending on the project.
ü

Sun lig ht, m oonl igh t ü Li ght scattered by sol ar system du st (ðzodiacalñ) ü Li ght em itted (th ermal ) by solar sy stem du st (ðzodiacalñ) ü Stars (especiall y in a crowded fiel d) ü Li ght em itted by i nterstell ar dust (ðcirru sñ)

Background - sources: Astronomical

16


Background - sources: COBE data

Background - sources: NGST
NGST Backgrounds
1.E+04 zod. lgt. pm_scat
- /s/pixel Signal [elec/sec/pix]

1.E+03

sm_scat 1.E+02 pm_therm sm_therm 1.E+01 dark readout 1.E+00 SPR_5 SPR_1000 1.E-01

1.E-02 0.1 1 Wavelength [µm ] 10 100

17


Noise - sources: Detector/electronics noise
v

The m easured sign al wil l depend u pon characteristi cs of the electr ical system
ü ü ü ü ü ü ü

reset noise (kTC) Johnson (resi stor) noise temperature-induced dri fts mi crophonics (change i n parasi tic capacitance) electric al pi ckup (especial ly vi a ground loops) all of these are usuall y l umped together as read noise dark current

v

Typi cal val ues for m odern-day system s approx. 10ós e- for NI R detectors and an order of m agni tude hig her f or MIR detectors.

Noise - classification in the frequency domain
v v v

ðWhiteñ noi se: equal power per u ni t frequen cy 1/f n oise: noi se power spectral density g oes as one over the fr equency o f the no ise. Pattern n oise: n oise power peaked at a discrete frequen cy - produces a fi xed pattern i n the i mag e.

18


Signal-to-noise ratio
S = N S


i

N

2 i

.

N N

total 2 back

=N =N

2 read

+N

2 dark

+N

2 source

+N
2 zodi

2 back

, where
2 other

2 sky ,OH

+N

2 sky ,thermal

+N

+N

.

Signal-to-noise ratio - example
v

v

Compare sensitivity (S/N) to a point source for Keck, a warm space telescope (HST-like), and NGST (ignore shot noise from source). Assume parameters in table, background fluxes from this lecture, 1000 second exposure, 3 e- read noise, 0.01 e-/s dark current, a filter bandwidth of 5 (/), and calculate for a range of wavelengths from 1-20 µm.

19


Near-infrared observing facilities

Facility HST SOFIA IRTF UKIRT NGST Keck

D (m) 2.4 2.5 3.0 3.7 8.0 10.0

s (") D.L. 1.0 0.5 0.5 D.L. 0.5

Tm (K) 290 230 273 273 40 273

e 0.0 0.1 0.1 0.1 0.0 0.1 5 5 2 0 3 0

Imager NICMOS FLITECAM NSFCAM IRCAM TBD NIRC

Spectrometer (NICMOS) FLITECAM CSHELL CGS4 TBD NIRSPEC

Signal-to-noise ratio - example
Relative Performan ce: Space vs. Ground
10000 HST-like 1000

(S/N) / (S/N )Keck

NG ST

100 10 1 0.1 0.01 0 5

Wavelength {µm}

10

15

20

25

20


Sensitivity of Future IR Facilities
5 Flux Limits in 10 4 seconds
1E-14 1E-15 SOFIA 1E-16 FIRS T F ( W/m2) 1E-17 1E-18 1E-19 1E-20 S T2010 1E-21 NGS T 1E-22 0 .1 1 10 W avelength (µ m) 100 1000 HS T WFC3 ALMA S IRTF

F AIR (8m)

Improving S/N
v

Atm ospheric ef fects
ü ü ü ü

Atmospheri c absorption: hi gher altitude OH emi ssi on: OH suppression i nstrumen ts Turbulen ce: adaptive optics Ulti mate ðf ix ñ is to go to space!

21


Improving S/N: multiple sampling
Noise for NIRSPEC Aladdin Array
80

Standard Deviation/root(2)

60
electrons

40

20
1/root(N)

0 0 10 20 30 Number of Fowler Samples 40

References
v v

v v

Al len 2000, ð Astrophysi cal Quan titiesñ Liv in gston & Wall ace 1991, ðAn A tlas of the Solar Spectrum in the Infrared f rom 1850 to 9000 1/cm (1.1 to 5.4 mi crons),ñ National Sol ar Observatory Techn ical Report #91-001 Maih ara et al. 1993, PASP, 105, 940 ðMethods of Experi men tal Phy sics , Vol. 12 - Part A: Astrophysi cs, Optical a nd Infrared,ñ ed. Carleton

22