Millimetre and Radio Astronomy Techniques for Star Forma:on Studies II

Transcription

1 Millimetre and Radio Astronomy Techniques for Star Forma:on Studies II John Conway Onsala Space Observatory, Sweden &Nordic ALMA ARC node

2 Today prac:cal details... For details see Introduc:on to Millimeter/Submillimeter astronomy by T. Wilson, March 2009, astro.ph , and textbook Tools of Radio Astronomy Wilson,Rohlfs and HuYemeister. Also a few viewgraphs adapted from, 2007 IRAM summer school especially talks of Thum and Pandre.

3 The APEX (sub)millimetre wavelength telescope

4 Antenna op:cs Onsala 20m IRAM 30m/APEX 12

5 Single pixel vs Camera obs Simplest case (e.g. Onsala 20m) radio telescope as light bucket dominated by radia:on coming from patch on sky of angular diameter θ= λ/d, the main lobe Hence in simplest form a single pixel device, though we can make images by poin:ng at different posi:ons and integra:ng but takes a long :me. Can build radio cameras with mul:ple receivers at primary or secondary focus. BUT spectral line sensi:ve systems expensive and difficult to build, max can be 3x3 or 7x7 pixels(!). Con:nuum only bolometers can have more pixels,but at cm wavelength number of pixels that can be crammed into focus limited, each pixel is many wavelengths in size.

6

7 Radio telescope as a Thermometer Last :me, strength of signal in Wm 2 Hz 1 if source more extended than antenna beam can be expressed conveniently as Temperature in K (the measured Antenna temperature ) using Rayleigh Jeans approxima:on. If source op:cally thick and in LTE this antenna temperature will equal kine:c temperature, T kin otherwise is (1 e τ )T ex Direct link to physical processes. Convenient to get output in Temperature units directly without calibra:ng Wm 2 Hz 1 coming into telescope aperture and conver:ng to Kelvin. We just compare powers coming out of receiver/amplifier electronics in milli WaYs when on our target source and when looking at loads of different temperature.

8 Complica:ons The receiver adds power (T RX, receiver temperature). At mm atmosphere adds power and also absorbs target sources. In fact telescope not just sensi:ve to small patch on sky θ= λ/d, but has side lobes, which pick up power from range of direc:ons including ground.

9 Beam switch to subtract atmosphere Incoming for 20m telescope with 1MHz BW W Beam-switch (needed only at mm wavelengths) Feed Amp Spectro meter V out (v) in milli-watts RX

10 Beam switch to subtract atmosphere Incoming for 20m telescope with 1MHz BW W Beam-switch (needed only at mm wavelengths) Feed Amp Spectro meter V out (v) in milli-watts RX

11 Single beam switching RX Adds receiver noise T RX and then multiples by gain G V right (ν)

12 Single beam switching RX Adds receiver noise T RX and then multiples by gain G V left (v)

13 Single beam switching V sig (v) = V right V left So get spectrum where receiver power removed and additive power effect of atmosphere also removed RX V left (ν) Gain G

14 Improved method Double beam switching V on = V right - V left with source in right beam path, form using fast mirror switching at 2Hz. Integrate for say 10sec 30sec. Then slew telescope so now source is in left beam path. Again form V off = V right V left with fast mirror switching at 2Hz, integrate again for say 10sec -30sec. Form Vsig = (V on V off )/2, which is proportional to received source power Advantage; compared to single beam switching, eliminates difference in gain between left and right paths. Slight disadvantage; lost slew time, but best choice for weak lines

15 Single and double beam switching efficiently remove addi*ve power effects of receiver and atmosphere, and ground (coming in from far side lobes). They do no remove mul*plica*ve (ayenua:on) effects of atmosphere, nor do they do the conversion from powers to temperature. To solve above problems, need to make calibra:on observa:ons, switching between blank sky and load temperature.

16 Calibra:on Spectra ( flats ) Ambient temperature load RX Gain G V right (v) = V amb (v)

17 Calibra:on Spectra Ambient temperature load RX Gain G V left (v) = V sky (v)

18 Calibra:on Spectra V cal (v) = V right (v) V left (v) or V cal (v) = V amb (v) V sky (v) Can be used together with V sig (v) to get opacity/gain corrected spectra in temperature units. RX Gain G V right (v) = V amb (v) Ambient temperature load

19 All quantities below are functions of velocity/frequency. All V quantities are output powers in mwatts, G is the electronics gain. With an ambient temperature absorber in signal path. When pointing on blank sky Where F eff is the forward gain of the telescope (approx 0.9), first term above is power contributed from atmosphere, second term pickup from the ground. Using that We obtain that the difference in power between blank sky and ambient load is

20 Chopper mirror calibra:on Where T A is the antenna temperature that would be measured if telescope was at top of atmosphere (i.e. before atmospheric attenuation), i.e the astronomical quantity of interest. From previous slide, substitute into above equation and rearrange... So to get opacity corrected calibrated data in K units, take the ratios of Δv sig spectra to Δv cal spectra multiply by T amb F eff This atmosphere opacity corrected T A versus frequency is what is delivered from the telescope, all ON-OFF, and flat field done automatically, no need to manually reduce this data (!!)

21 Final reduc:on steps Look for bad integra:ons Remove spectral baselines by fiong polynomial to line free parts. Francesco will explain more...

22 T ant (K) Final Result Taken with Onsala 20m telescope 17 days ago. Spectra of water maser emission at 1.3cm from disk surrounding supermassive black hole in NGC4258