Greenland Telescope: Single-dish science, instrumentation requirements Nimesh A Patel Satoki Matsushita Center for Astrophysics | Harvard & Smithsonian ASIAA

Nimesh Patel October 2018, Thule AFB 23 May 2019, EAO Futures meeting, Nanjing !1 Outline: • Current status • Single dish science goals • Summit Station • Future receivers development

Nimesh Patel October 2018, Thule AFB 23 May 2019, EAO Futures meeting, Nanjing !2 Completion of mechanical assembly of the Greenland Telescope 10 August 2017

!3 Thule Phase 1 Activities 2017-2019

• Antenna assembly (mechanical + electrical) completed at Thule Air Base, Greenland - September 2017 • Hydrogen maser installed in September 2017 • Servo tuning and tests in progress - October 2017 • Fringes demonstrated at Maunakea, with SMA + JCMT, using 230 GHz receivers and VLBI backend • 86 & 230 GHz receivers installed in late November 2017 • Pointing calibration in November 2017

• Joined the EHT run in April 2018

October 2017 First fringes on ALMA-Thule baseline, 28 January 2018

Mk4/DiFX fourfit 3.19 rev 2292 3C279.0166CV, 028-0843_b4, AG ALMA - THULE, fgroup B, pol XL Fringe quality 9

SNR 10.8 Int time 239.889 Amp 0.136 Phase 134.4 PFD 1.6e-17 Delays (us) SBD -0.000661 MBD 0.006959 Fringe rate (Hz) -0.000040 Ion TEC 0.000 Ref freq (MHz) 228221.3906 AP (sec) 0.205 Exp. e18j28 Exper # 3635 Yr:day 2018:028 Start 084300.00 Stop 084700.03 FRT 084500.00 Corr/FF/build 2018:079:230135 2018:080:102304 2018:080:065037 RA & Dec (J2000) 12h56m11.1666s Amp. and Phase vs. time for each freq., 7 segs, 195 APs / seg (39.94 sec / seg.), time ticks 60 sec -5˚47’21.525"

U Validity !5 L

A G

228221.39228279.98228338.58228397.17228455.77228514.36228572.95228631.55228690.14228748.73228807.33228865.92228924.52228983.11229041.70229100.30229158.89229217.48229276.08229334.67229393.27229451.86229510.45229569.05229627.64229686.23229744.83229803.42229862.02229920.61229979.20Freq (MHz) All 59.5 129.3 52.2 113.7 178.3 155.5 156.7 85.4 164.5 145.9 133.9 159.8 148.7 109.4 122.6 152.2 162.2 151.0 -158.7 162.0 135.5 126.4 134.4 137.8 123.9 133.3 116.4 98.1 139.7 52.7 95.7 Phase 134.4 0.3 0.1 0.1 0.1 0.2 0.2 0.1 0.1 0.2 0.2 0.1 0.2 0.2 0.2 0.1 0.2 0.2 0.2 0.1 0.2 0.2 0.1 0.1 0.2 0.3 0.1 0.2 0.2 0.2 0.0 0.1 Ampl. 0.1 893.0 527.3 478.9 415.5 53.8 781.4 278.3 111.5 309.4 497.0 894.5 735.7 128.2 348.1 773.5 690.3 568.2 808.7 449.6 86.1 322.2 383.1 703.2 250.8 464.5 300.5 713.9 458.8 777.4 896.6 477.0 Sbd box 464.8 U/L 1172/0 1172/0 1172/0 1172/0 1172/0 1172/0 1172/0 1172/0 1172/0 1172/0 1172/0 1172/0 1172/0 1172/0 1172/0 1172/0 1172/0 1172/0 1172/0 1172/0 1172/0 1172/0 1172/0 1172/0 1172/0 1172/0 1172/0 1172/0 1172/0 1172/0 1172/0 APs used A 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 PC freqs G 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 PC freqs A:G 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 PC phase A:G 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 0:0 Manl PC A 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 PC amp G 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 1000 B00UX B01UX B02UX B03UX B04UX B05UX B06UX B07UX B08UX B09UX B10UX B11UX B12UX B13UX B14UX B15UX B16UX B17UX B18UX B19UX B20UX B21UX B22UX B23UX B24UX B25UX B26UX B27UX B28UX B29UX B30UX Chan ids A Tracks B00UL B01UL B02UL B03UL B04UL B05UL B06UL B07UL B08UL B09UL B10UL B11UL B12UL B13UL B14UL B15UL B16UL B17UL B18UL B19UL B20UL B21UL B22UL B23UL B24UL B25UL B26UL B27UL B28UL B29UL B30UL Chan ids G Tracks Group delay (usec)(model) 1.53950718118E+04 Apriori delay (usec) 1.53950648530E+04 Resid mbdelay (usec) 6.95887E-03 +/- 2.8E-05 Sband delay (usec) 1.53950641917E+04 Apriori clock (usec) -2.1089119E+03 Resid sbdelay (usec) -6.61241E-04 +/- 8.8E-04 Phase delay (usec) 1.53950648546E+04 Apriori clockrate (us/s) -6.6459997E-06 Resid phdelay (usec) 1.63576E-06 +/- 1.3E-07 Delay rate (us/s) 5.37017872807E-02 Apriori rate (us/s) 5.37017874558E-02 Resid rate (us/s) -1.75172E-10 +/- 9.3E-10 Total phase (deg) 317.6 Apriori accel (us/s/s) -7.73193107321E-05 Resid phase (deg) 134.4 +/- 10.6 RMS Theor. Amplitude 0.136 +/- 0.013 Pcal mode: MANUAL, MANUAL PC period (AP’s) 5, 5 ph/seg (deg) 15.5 14.0 Search (4096X128) 0.136 Pcal rate: 0.000E+00, 0.000E+00 (us/s) sb window (us) -2.000 2.000 amp/seg (%) 14.8 24.4 Interp. 0.000 Bits/sample: 2x2 SampCntNorm: disabled mb window (us) -0.009 0.009 ph/frq (deg) 35.9 29.5 Inc. seg. avg. 0.137 Sample rate(MSamp/s): 116 dr window (ns/s) -0.011 0.011 amp/frq (%) 43.9 51.4 Inc. frq. avg. 0.140 Data rate(Mb/s): 7192 nlags: 464 t_cohere infinite ion window (TEC) 0.00 0.00 A: az 13.2 el 72.4 pa -167.9 G: az 175.1 el 7.5 pa -1.5 u,v (fr/asec) -1044.494 -30143.124 simultaneous interpolator Control file: cf_3601 Input file: /data-sc05/mike/e18j28/3635/028-0843_b4/AG..0166CV Output file: /data-sc05/mike/e18j28/3635/028-0843_b4/AG.B.1.0166CV EHT observing run April 2018 Keiichi Asada (ASIAA) Hiroaki Nishioka (ASIAA) ChenYu Yu (ASIAA) Nimesh Patel (CfA) !6 31 March - 9 April 2019, Global Millimeter-wave VLBI Array observations at 86 GHz

!7 Hydrogen maser house Control room, weather station

!8 Current Instrumentation

!9 Johnson Han (ASIAA)

2SB

!10 !11 Johnson Han (ASIAA)

!12 Two ROACH2 spectrometers, each with two IF inputs of 2.048 GHz b/w, 32768 channels CASPER ROACH2, dual 5 Gsps ADCs, Octal 10 Gbps Ethernet

!13 Summit Station

!14 Thule Air Base Position: 76.5° North, 68.7° West Altitude: 77 m Temperature: 5 C to – 25 C (average)

Summit Station Position: 72.5° North, 38.5° West Altitude: 3,200 m Temperature: - 10 C to - 50 C (average)

!15 View from level-2 platform Beginning of Ice sheet (traverse)

!16 Publications of the Astronomical Society of the Pacific, 129:025001 (11pp), 2017 February Matsushita et al.

Publications of the Astronomical Society of the Pacific, 129:025001 (11pp), 2017 February Matsushita et al.

Publications of the Astronomical Society of the Pacific, 129:025001 (11pp), 2017 February doi:10.1088/1538-3873/129/972/025001 © 2016. The Astronomical Society of the Pacific. All rights reserved. Printed in the U.S.A.

3.5 Year Monitoring of 225GHz Opacity at the Summit of Greenland Satoki Matsushita1, Keiichi Asada1, Pierre L. Martin-Cocher1, Ming-Tang Chen1, Paul T. P. Ho1, Makoto Inoue1, Patrick M. Koch1, Scott N. Paine2, and David D. Turner3 1 Academia Sinica Institute of Astronomy and Astrophysics, P.O. Box 23-141, Taipei 10617, Taiwan, Republic of China 2 Harvard-Smithsonian Center for Astrophysics, 60 Garden Street, MS-78, Cambridge, MA 02138, USA 3 Global Systems Division, Earth System Research Laboratory, National Oceanic and Atmospheric Administration, 325 Broadway, Boulder, CO, 80305, USA Figure 2. Time variationReceived 2016 plot September of 19; 225 accepted GHz 2016 opacity October 28; at published the summit 2016 December of 21 Greenland ice sheet. The measurement has been startedAbstract from 2011 August 17, which is Figure 3. Monthly quartile variation of 225 GHz opacity. Solid, dashed, and deWefined present as the the 3.5 day years 1monitoring in this results diagram. of 225 GHz opacity at the summit of the Greenland ice sheet (Greenland dotted lines are 25%, 50%, and 75% quartiles, respectively. The first and the Summit Camp) at an altitude of 3200 m using a tipping radiometer. We chose this site as our submillimeter telescope last months of this measurement, namely 2011 August and 2015 February, (Greenland Telescope) site, because conditions are expected to have low submillimeter opacity and because its fi location offers favorable baselines to existing submillimeter telescopes for global-scale Very Long Baseline have fewer data than the other months, so that the statistical signi cance is low. Interferometry. The site shows a clear seasonal variation with the average opacity lower by a factor of two during Summitwinter. The 25%, Camp. 50%, and It 75% is quartiles clear of the that 225 GHz there opacity is during a seasonal the winter months variation; of November through April are 0.046, 0.060, and 0.080, respectively. For the winter quartiles of 25% and 50%, the Greenland site is about Figure 1. 225 GHz tipping radiometer located on the roof of the Mobile Science Facitily (MSF) at the Greenland Summit Camp. opacity10%–30% worse in winter than the Atacama is low, Large but Millimeter high/submillimeter in summer. Array Figure(ALMA) or3 theshows South Pole sites. (A color version of this figure is available in the online journal.) theEstimated monthly atmospheric quartile transmission variations spectra in winter( seasonsolid, are similar dashed, to the ALMA and site dotted at lower frequencies lines (<450 GHz), which are transparent enough to perform astronomical observations almost all of the winter time with opacities are<0.5 25%,, but 10% 50%,–25% higher and opacities 75% at higher quartiles, frequencies respectively(>450 GHz) than those). December at the ALMA site. and This is due to atmospheric opacity for three months between late winter to binary format. The raw binary data are downloaded from the lower altitude of the Greenland site and the resulting higher line wing opacity from pressure-broadened saturated early spring, and the results are reported in Asada et al. (2012) Greenland to Taiwan regularly via internet. Marchwater lines tend in addition to to have higher dry the air bestcontinuum opacity absorption conditions, at higher frequencies. and Nevertheless, July tends half of the winter and Matsushita et al. (2013). After this measurement, we The tipping data have been reduced with the typical data totime have at the Greenland the worst. Summit Note Camp can that be used the for astronomicalfirst and observations the last at frequencies months between of this 450 GHz and moved the radiometer to the Greenland Summit Camp. The reduction method for tipping measurements (e.g., Matsuo 1000 GHz with opacities <1.2, and 10% of the time show >10% transmittance in the THz (1035 GHz, 1350 GHz, radiometer was installed on the roof of the Mobile Science et al. 1998). Since the MSF is a mobile facility on the snow, it measurementand 1500 GHz) windows.(i.e., Summer 2011 season is August good for observations and 2015 at frequencies February lower than) 380have GHz. One major Facility (MSF; Shupe et al. 2013), which was built by CPS in can shake due to wind or human activities inside the facility, so advantage of the Greenland Summit Camp site in winter is that there is no diurnal variation due to the polar night support of the ICECAPS project. The roof deck is at about 3 m that the radiometer tipping angle may be affected. In addition, above the snow surface (Figure 1). The measurement was the leveling accuracy for the radiometer is limited, so that the fewercondition, data and therefore than the the durations other of months,low-opacity conditions so the are statistical significantly longer signi thanfi atcance the ALMA site. started from 2011 August 17, and data were collected until radiometer may also have a small constant tilt. We assume that Opacities lower than 0.05 or 0.04 can continue for more than 100 hr. Such long stable opacity conditions do not occur is lower. In both diagrams, there is no significant annual 2015 February 12. We present here the data for these ∼3.5 the opacity is the same between the southern and northern as often even at the South Pole; it happens only for the opacity lower than 0.05. Since the opacity variation is directly years. skies, and the small difference in opacities derived from the difference.related to the sky temperature (background) variation, the Greenland Summit Camp is suitable for astronomical To measure the atmospheric opacity, we adopt the tipping southern and northern sky tippings is considered as the result of observations that need unusually stable sky background. method; we observe five angles from horizon (90=zenith, the tilt of MSF and/or the radiometer. Based on this Key words: atmospheric effects – site testing 42, 30, 24, and 19.2 , which corresponds to sec(z) of 1.0, assumption, we calculate the tilt angle and correct for it when fi 1.5, 2.0, 2.5, and 3.0) with 4 s integration at each angle. We deriving the opacity. In case the difference of the opacities Online material: color gure scan the mirror from south to north, namely observe the five between the southern and northern skies is large (i.e., when the 4.2. Cumulative Distribution and Histogram angles in both the southern and the northern sky (for the tilt angle is calculated to be larger than two degrees; this value

* measurements at Greenland). The blackbody target in the is also to allow some tolerance for the different opacities 1. Introduction observed size of Sgr A is obviously affected by the interstellarbottom of the radiometer is observed before and after the scan between the northern and southern sky), we judge this is due to We also made cumulative distribution plots and histograms l2 As various technologies for submillimeter (submm) wave scattering at 3 mm or longer wavelengths, following thefor the gain calibration. The total duration of a tipping real opacity differences, and we flag the data. observationsof the have measured advanced, the 225 prospects GHz for opacity very long whilescattering separating law (e.g., Shen the et al. 2005 seasons; Bower et al. 2006).measurement This is 75 s, and each tipping is performed every 10 baseline interferometry (VLBI) at submm wavelengths have effect, however, lessens at shorter wavelengths, and the sizeminutes. is Between the tipping measurements, the mirror is 4. Results and Discussions into winter and summer (Figure 4). Here we define winterl2 as pointed toward the zenith and the sky data are recorded every become a reality. Operations at shorter wavelengths improve observed to deviate from the scattering law at 1.3 mm 4.1. Time and Seasonal Variations 1 s. The output voltages, together with the mirror position, the proportionately the spatial resolution as compared with current (Doeleman et al. 2008). This has strongly motivated submm- between the beginning of November and the end of April (solid Figureblackbody 4. targetCumulative temperature, distribution and other monitoring plots data, and are histogramsFigure of2 225displays GHz the opacity measured in time variation of the centimeter and millimeter VLBI observations; we anticipate VLBI observations toward nearby SMBHs to resolve and recorded every 1 s into the hard disk of the host computer in 225 GHz opacity for the 3.5 year period at the Greenland that itline will be plots possible), to and resolve summer astronomical as sources May in throughimage emission October from their(dashed vicinities. line winter (solid lines) and summer (dashed lines). The vertical axis on the left- greater details, by a factor of 10. Upper limits of the intrinsic sizes of SMBHs havehand been side is for the cumulative distribution plots, and that on the right-hand plots). The quartiles for each seasonmeasured are so 0.046, far for Sgr A0.060,* (Doeleman and et al. 2008) and the The demand for better angular resolution is quite strong, side is for the histograms. Crosses on the cumulative3 distribution plots are the especially0.080 for the direct for imaging 25%, of nearby 50%, supermassive and 75%, black respectively,nucleus of M87 (Doeleman in winter, et al. 2012 and); upper limits to the holes (SMBHs); Sagittarius A* (Sgr A*), which is located at the sizes of both sources are about 40 marcsec. Theseopacity results quartiles of each season. The quartile for winter and summer are also center of0.089, our Galaxy 0.118, and therefore and the nearest 0.159 SMBH, for has summer been indicate(see that also much the longer crosses baselines and and/or higher frequencieslisted in the figure. imaged at various wavelengths usingfi the VLBI technique. The are needed to resolve and image the SMBHs. We therefore the values in the gure). It is obvious that the opacity in winter!17 is about half of that in summer at all the quartiles. The histograms show that the opacities1 of 0.04 and 0.10 are the opacities that occur most often in winter and summer, 1992 January and December (one year). Since both the ALMA respectively. and the South Pole sites are located in the southern hemisphere, We then compared the opacity quartiles using our 3.5 year we define winter as between the beginning of May through the period statistics with those at the ALMA and the South Pole end of October, and summer as November through April. The sites, which are well-established sites for submm observations. calculated quartiles for these three sites are listed in Table 1. The 225 GHz opacity data for the ALMA site have been For winter, the ALMA site is the best at the 25% quartile, obtained from Radford & Chamberlin (2000) and Radford South Pole is next, and Greenland Summit Camp is the worst, (2011), whose measurements have been made between 1995 but only by ∼0.005 (~12%) difference in opacity between each April and 2006 April (∼11 years), and that for the South Pole site. At the 50% quartile, the ALMA and South Pole sites are site from Chamberlin & Bally (1994, 1995) measured between almost the same, and the opacity is about 0.01 (about 25%)

4 !18 !19

structure designed to support the South Pole Telescope (SPT) [5]. Figure 6 is a computer model of the support structure. The goal of the spaceframe is to distribute the load of the telescope on the foundation and delay snow accumulation and drifts around the telescope. To better distribute the loads of the antenna, the support cone itself was redesigned with an adjustable 6-point interface to the spaceframe. The features of this new support cone are: possibility of leveling the antenna at this 6-point interface when needed, open bottom so that the cone and inside components are always at the temperature of the snow (-30 to -35°C), material is low temperature steel, support structures for encoder, and increased stiffness due to the full cone geometry connecting to 6 points instead of a truncated cone structure as shown on Figure 7.

Figure 5: Model of the GLT on the space frame partially embedded in the snow foundation (design of space frame and snow pad in progress). Isometric view showing all 4 side containers and receiver transfer container on the back (Left); sectional side view (Right)

Raffin P., 2014 SPIE paper

!20

Figure 6: Model of the spaceframe to support the GLT on the ice sheet

Single dish science

!21 Science goals: 1) M87 VLBI 2) Single-dish Submm and THz projects

R1-1 Publ. Astron. Soc. Japan (2016) 68 (1), R1 (1–41) doi: 10.1093/pasj/psv115 Advance Access Publication Date: 2015 December 14 Review

Review First-generation science cases for ground-based terahertz telescopes

1, 1 1 Hiroyuki HIRASHITA, ∗ Patrick M. KOCH, Satoki MATSUSHITA, Shigehisa TAKAKUWA,1 Masanori NAKAMURA,1 Keiichi ASADA,1 Hauyu Baobab LIU,1 Yuji U RATA,1,2 Ming-Jye WANG,1 Wei-Hao WANG,1 Satoko TAKAHASHI,3 ,4 Ya - We n T ANG,1 Hsian-Hong CHANG,1 Kuiyun HUANG,5 Oscar MORATA,1 Masaaki OTSUKA,1 Kai-Yang LIN,1 An-Li TSAI,2 Ye n - T i n g L IN,1 Sundar SRINIVASAN,1 Pierre MARTIN-COCHER,1 Hung-Yi PU,1 Francisca KEMPER,1 Nimesh PATEL,6 Paul GRIMES,6 Yau-De HUANG,1 Chih-Chiang HAN,1 Ye n - R u H UANG,1 Hiroaki NISHIOKA,1 Lupin Chun-Che LIN,1 Qizhou ZHANG,6 Eric KETO,6 Roberto BURGOS,6 Ming-Tang CHEN,1 Makoto INOUE,1 and Paul T. P.HO1,7

1Institute of Astronomy and Astrophysics, Academia Sinica, P.O. Box 23-141, Taipei 10617, Taiwan 2Institute of Astronomy, National Central University, Chung-Li 32054, Taiwan 3Joint ALMA Observatory, Alonso de Cordova 3108, Vitacura, Santiago, Chile 4National Astronomical Observatory of Japan, 2-21-1 Osawa, Mitaka, Tokyo 181-8588, Japan 5Department of Mathematics and Science, National Taiwan Normal University, Lin-kou District, New Taipei City 24449, Taiwan 6Harvard-Smithsonian Center for Astrophysics, Cambridge, MA 02138, USA 7East Asian Observatory, 660 N. A’ohoku Place, University Park, Hilo, HI 96720, USA

∗E-mail: [email protected]

Received 2015 August 3; Accepted 2015 October 28

Abstract Ground-based observations at terahertz (THz) frequencies are a newly explorable area of astronomy in the coming decades. We discuss science cases for a first-generation 10-m class THz telescope, focusing on the Greenland Telescope as an example of such a facility. We propose science cases and provide quantitative estimates for each case. The largest advantage of ground-based THz telescopes is their higher angular resolution ( 4 for a 10-m dish), as compared to space or airborne THz telescopes. Thus, high- ∼ ′′ resolution mapping is an important scientific argument. In particular, we can isolate zones of interest for Galactic and extragalactic star-forming regions. The THz windows are suitable for observations of high-excitation CO lines and [N II]205-µmlines,which are scientifically relevant tracers of star formation and stellar feedback. Those lines are the brightest lines in the THz windows, so they are suitable for the initiation of ground- based THz observations. THz polarization of star-forming regions can also be explored since it traces the dust population contributing to the THz spectral peak. For survey-type observations, we focus on “sub-THz” extragalactic surveys, the uniqueness of which is

C ⃝ The Author 2015. Published by Oxford University Press on behalf of the Astronomical Society of Japan. All rights reserved. For Permissions, please email: [email protected] R1-1 Publ. Astron. Soc. Japan (2016) 68 (1), R1 (1–41) doi: 10.1093/pasj/psv115 Advance Access Publication Date: 2015 December 14 Review

Review First-generation science cases for ground-based terahertz telescopes

1, 1 1 Hiroyuki HIRASHITA, ∗ Patrick M. KOCH, Satoki MATSUSHITA, Shigehisa TAKAKUWA,1 Masanori NAKAMURA,1 Keiichi ASADA,1 Hauyu Baobab LIU,1 Yuji U RATA,1,2 Ming-Jye WANG,1 Wei-Hao WANG,1 Satoko TAKAHASHI,3 ,4 Ya - We n T ANG,1 Hsian-Hong CHANG,1 Kuiyun HUANG,5 Oscar MORATA,1 Masaaki OTSUKA,1 Kai-Yang LIN,1 An-Li TSAI,2 Ye n - T i n g L IN,1 Sundar SRINIVASAN,1 Pierre MARTIN-COCHER,1 Hung-Yi PU,1 Francisca KEMPER,1 Nimesh PATEL,6 Paul GRIMES,6 Yau-De HUANG,1 Chih-Chiang HAN,1 Ye n - R u H UANG,1 Hiroaki NISHIOKA,1 Lupin Chun-Che LIN,1 Qizhou ZHANG,6 Eric KETO,6 Roberto BURGOS,6 Ming-Tang CHEN,1 Makoto INOUE,1 and Paul T. P.HO1,7

1Institute of Astronomy and Astrophysics, Academia Sinica, P.O. Box 23-141, Taipei 10617, Taiwan 2Institute of Astronomy, National Central University, Chung-Li 32054, Taiwan 3Joint ALMA Observatory, Alonso de Cordova 3108, Vitacura, Santiago, Chile 4National Astronomical Observatory of Japan, 2-21-1 Osawa, Mitaka, Tokyo 181-8588, Japan 5Department of Mathematics and Science, National Taiwan Normal University, Lin-kou District, New Taipei City 24449, Taiwan 6Harvard-Smithsonian Center for Astrophysics, Cambridge, MA 02138, USA 7East Asian Observatory, 660 N. A’ohoku Place, University Park, Hilo, HI 96720, USA

∗E-mail: [email protected]

Received 2015 August 3; Accepted 2015 October 28

Abstract Ground-based observations at terahertz (THz) frequencies are a newly explorable area of astronomy in the coming decades. We discuss science cases for a first-generation 10-m class THz telescope, focusing on the Greenland Telescope as an example of such a facility. We propose science cases and provide quantitative estimates for each case. The largest advantage of ground-based THz telescopes is their higher angular resolution ( 4 for a 10-m dish), as compared to space or airborne THz telescopes. Thus, high- ∼ ′′ resolution mapping is an important scientific argument. In particular, we can isolate zones of interest for Galactic and extragalactic star-forming regions. The THz windows are suitable for observations of high-excitation CO lines and [N II]205-µmlines,which are scientifically relevant tracers of star formation and stellar feedback. Those lines are the brightest lines in the THz windows, so they are suitable for the initiation of ground- based THz observations. THz polarization of star-forming regions can also be explored R1-2since it traces the dust population contributingPublications to the THz of spectral the Astronomical peak. For Society survey-type of Japan (2016), Vol. 68, No. 1 observations, we focus on “sub-THz” extragalactic surveys, the uniqueness of which is detecting galaxies at redshifts z 1–2, where the dust emission per comoving volume is C The Author 2015. Published by Oxford University Press∼ on behalf of the Astronomical Society of Japan. the⃝ largest in the history of the Universe. Finally we explore possibilities of flexible time All rights reserved. For Permissions, please email: [email protected] scheduling, which enables us to monitor active galactic nuclei, and to target gamma-ray burst afterglows. For these objects, THz and submillimeter wavelength ranges have not yet been explored. Key words: dust, extinction — galaxies: ISM — ISM: lines and bands — submillimeter: general — telescopes

!23

1Introduction et al. 2010), airborne telescopes [e.g., the Stratospheric Observatory for Infrared Astronomy (SOFIA); Young et al. Access to the terahertz (THz) frequency range or far- 2012], and balloon-borne telescopes is that it is possible infrared (FIR) wavelength range from the ground is mostly to operate large dishes with a high diffraction-limited limited by the absorption of water vapor in the Earth’s resolving capability. To clarify this advantage, it is con- atmosphere. Therefore, the THz region is one of the venient to have a specific telescope in mind in discussing remaining wavelength ranges unexplored from the ground. science cases. In this paper, we focus on the Greenland Tele- Space, balloon-borne, and airborne observations have so far scope (GLT). The GLT project is planning to deploy an Ata- been used to explore THz astronomy. There are only lim- cama Large Millimeter/submillimeter (ALMA)-prototype ited sites on Earth where the THz windows are accessible, 12-m antenna to the Summit Station in Greenland (3200 m and where ground-based THz astronomy is, indeed, pos- altitude) for the purpose of using it as part of submil- sible. Greenland (subsection 2.1), high-altitude (>5000 m) limeter (submm) very long baseline interferometry (VLBI) Chilean sites (Matsushita et al. 1999;Paineetal.2000; telescopes (Inoue et al. 2014a, see also section 2). Data of Matsushita & Matsuo 2003;Petersonetal.2003), and atmospheric transmission at the Summit Station have been Antarctica (Yang et al. 2010a;Tremblinetal.2011) accumulated over the past four years through our contin- are examples of suitable places for ground-based THz uous monitoring campaign (Martin-Cocher et al. 2014,see astronomy. also subsection 2.2). Our analysis indicates that site condi- Even in those locations suitable for THz observations, tions are suitable for a first-generation ground-based THz occurrences of excellent weather is still limited. Therefore, facility. Therefore, we mainly focus on the GLT in this observations need to be planned well in order to max- paper, noting that the THz science cases will likely be sim- imize the scientific output within the limited amount of ilar or even common for all the first-generation THz tele- observing time. Moreover, “first-generation” observations scopes except for their sky coverages. are especially important because they determine the direc- This paper also provides a first basis for ground-based tion of subsequent THz studies. There have already been THz observations for later generations of telescopes with some pioneering efforts of ground-based THz observations larger dishes, such as the initially proposed Cerro Cha- such as with the Receiver Lab Telescope (RLT: Marrone jnantor Atacama Telescope (CCAT)1 or any successor et al. 2004), the Atacama Pathfinder Experiment (APEX: project. With a diameter twice as large as the GLT, Wiedner et al. 2006), and the Atacama Submillimeter Tele- these telescopes will have even better sensitivity and res- scope Experiment (ASTE: Shiino et al. 2013). However, olution to further push the scientific results achieved by the attempts by these existing facilities have been difficult the GLT. and sparse due to challenging weather conditions. A ded- One of the major scientific advantages of the THz regime icated telescope at an excellent site—as proposed for the is that thermal dust continuum emission will be mea- Greenland Telescope—is, thus, paramount for successful sured around its peak in the spectral energy distribution THz observations. In practice, such first-generation obser- (SED). Moreover, the angular resolution (4 at 1.5 THz vations are associated with the development of THz detec- ′′ for the diffraction-limited primary beam) enables us to tors. Searching for targets that are relatively easy to observe spatially resolve the individual star-formation sites within but scientifically pioneering is of fundamental importance nearby molecular clouds, typically located within a dis- to maximize the scientific value of the instrumental develop- tance of 300 pc (sub-subsection 3.1.1). This high reso- ment. The first aim of this paper is, thus, to search for scien- ∼ lution is also an advantage for extragalactic observations tifically important and suitable targets for first-generation where star formation activities within nearby galaxies will THz science cases. The largest advantage of ground-based THz telescopes compared with space telescopes (e.g., Herschel; Pilbratt 1 http://www.ccatobservatory.org/ . ⟨ ⟩ • Diffuse ISM • Molecules • NII line at 205 µm • Dust continuum: star-forming regions • Polarization at THz frequencies • Bolometer surveys of high redshift galaxies • Flux monitoring of AGN sources • Gamma ray bursts • Spectral-line surveys • Water maser surveys

!24 • (A) Chemistry and evolution in the diffuse and dense ISM CO 13-12 line, NII line 1.5 THz heterodyne (multipixel) receiver

• (B) Collective effects of star-formation in extragalactic sources

650, 850 GHz ; Improve on Herschel resolution Bolometer arrays

• (C) Time variable submillimeter Universe

VLBI receivers

!25 Publications of the Astronomical Society of Japan (2016), Vol. 68, No. 1 R1-7

Ta b l e 2 . Representative terahertz lines.

Species Frequency (THz) Transition Excitation energy (K)

CO 1.037–1.497 (9–8)–(13–12) 248.87486–503.134028

HCO+ 1.070–1.337 (12–11)–(15–14) 333.77154–513.41458 HCN 1.0630–1.593 (12–11)–(18–17) 331.68253–726.88341

H2D+ 1.370 10, 1–00, 0 65.75626 3 3 N II 1.461 P1– P0 – CH 1.471 N 2, J 3/2–3/2, F 2 –2 96.31131 Publications of the Astronomical Society of Japan (2016),= = Vol. 68, No.= 1 + − R1-9 HD2+ 1.477 11, 1–00, 0 70.86548

Ta b l e 3 . Luminous protostellar sources for the THz line experiment. 3 THz observations on-source time of 5.8 min, they detected a CO (J 13– Name Lbol D α(J2000.0) δ(J2000.0) References∗ = 12) line toward Orion FIR 4, a cluster-forming region in 3.1 Important lines at THz (L ) (pc) (h m s) (◦′′′) ⊙ the Orion Giant Molecular Cloud. The main beam bright- In the THz regime thereL1448-mm are a number of interesting 4.4 250 but 03ness 25 temperature 38 87 30 and 44 05.4 the line width 1,2 of the THz CO line NGC1333 IRAS 2A 19.0 250 03 28 55.58 31 14 37.11 1,2 unexplored spectral lines, which should be useful for studies are 210 K and 5.4 km s− ,respectively.Ascompared SVS 13 32.5 250 03 29∼ 03.73 31∼ 16 03.80 2,3 of the ISM and star formation. In line observations, sky to the CO J 9–8 and 7–6 lines, there is no line wing NGC1333 IRAS 4A 4.2 250 03 29 10.50= 31 13 31.0 1,2 background subtraction is much easier than in continuum component with 10 km s 1 in the THz CO line. Multi- L1551 IRS 5 22 140 04 31 34.14 18! 08 05.1− 4,5 observations, and thus observations of intense THz lines transitional analysis of the “quiescent” component of the L1551 NE 4.2 140 04 31 44.47 18 08 32.2 5,6 should be the starting point of our THz astronomical L1157 5.8 325 20CO 39 lines 06.28 show that 68 02 the 15.8 gas temperature 1,7 and density traced experiment. Table 2 summarizes representative THz lines by the THz line are 380 70 K and (1.6 0.7) 105 References: (1) Jørgensen et al. (2007); (2) Enoch et al.3 (2009); (3) Chen, Launhardt,± and Henning ± × that have rest-frame wavelengths∗ that are accessible in the cm− , respectively. These results show that the THz CO THz atmospheric windows.(2009); These (4) Takakuwa lines etcan al. ( be2004 classified); (5) Froebrichline (2005 traces); (6) Takakuwa the very et hot al. (2012 molecular); (7) Shirley gas et in al. the vicinity of the (2000). into three categories; (i) very high-J (J is the rotational !26 protostars, without any contamination from the extended excitation state) molecular lines (CO, HCO+, and HCN); outflow component. The bulk of the outflows are appar- (ii) atomic fine-structure lines ([N II]); and (iii) pure rota- ently at lower temperatures. With the GLT, we aim at a THz CO lines (J 9–8, 10–9, and 13–12) are too low to be The CONDOR result above (Wiedner et al. 2006)shows tional lines of light= molecules (H2D+,HD2+, and CH). systematic survey of similar objects hosting highly excited detected even at the very high gas temperatures, while those that the THz CO intensity is 210 K toward Orion FIR CO lines. ∼ of lower-J CO lines (J 1–0; 3–2) are high. This is because 4(Lbol 50 L ). The lowest bolometric luminosity among 3.1.1 Tracers of star-forming= places Observations= of⊙ submm and THz HCN lines toward at the low gas density the higher J levels are not populated our target is 4.2 L (L1551 NE), and the expected THz Kawamura et al. (2002) observed several positions in the Sgr B2(M) by Herschel/the⊙ Heterodyne Instrument for the and the THz CO lines are too optically thin, as shown in CO intensity is 210 K 4.2 L /50 L 17.6 K ( 400 Jy Orion Molecular Cloud (OMC-1) by the ground-based Far-Infrared (HIFI) show× that the⊙ submm⊙ = HCN J 6–5∼ the lower left-hand panel. In contrast, at n 106 cm 3, within primary beam). Using the typical sensitivity= dis- 10-m Heinrich Hertz Telescope on Mount Graham,H2 = Ari-− and 7–6 lines exhibit “blue-skewed” profiles suggestive of zona.typical They of dense targeted cores the (dense-gas CO J condensations9–8 rotational in line molec- at infall,cussed while in the subsection HCN J 2.38–7 (see and also the table THz1), HCN we shouldJ be = = = 1.037ular clouds), THz. They all COdetected transitions some regions are thermalized with such a and high the 12–11able (1.06 to achieve THz) a lines 5σ exhibitdetection “red-skewed” of this source profiles with a sug- velocity 1 excitation along temperatures the ridge follow of OMC-1. the gas It kinetic has been temperature clarified gestiveresolution of expansion of 0.2 (Rolffs km s− etand al. an2010 on-source). The submm integration HCN time that(middle there right-hand are some panel). warm Consequently regions with kinetic the optical tempera- depths, linesof with 34 min. the blue-skewed Therefore, profiles the GLT originate can collect from gas a systematic with turesand thus130 the K brightness and hydrogen temperatures, number density become high106 evencm 3 for temperaturesample of hot100 regions K while in the the vicinity THz line of is protostars. emitted from ! ! − ∼ inthe the THz molecular lines. These cloud. results They show also that emphasized the THz the CO impor- lines are gas with higher excitation 330 K. These results suggest 6 3 ∼ tancean excellent of high tracer angular of resolutions dense (! 10 achievedcm− ) by and ground-based hot (! 100 K) that the lower-J HCN lines trace the outer infalling motion telescopesmolecular in gas. specifying the locations of the emission. toward the massive stars, whereas the THz line traces WithAccording the GLT, to radiation-hydrodynamic we aim at observing higher-excitation models of proto- CO the3.1.2 inner The expansion diffuse driven ISM: molecules by the radiation pressure and lines at a higher frequency, 1.5 THz, where a CO J 13–12 the stellar wind from the central massive stars. All of the star formation, such dense and high-temperature= molec- Some atomic or molecular lines in the diffuse medium (1.4969ular regions THz) should line is present. be present With at the the very latest high- stageJ molecular of proto- abovecan results be targeted imply at that THz high- frequencies.J THz molecular Fine-structure lines can be lines of lines,stellar “extremely collapse within hot” (! 300500 K) au molecular from the regions forming in central the uniquefundamental tracers to atoms probe and the very ions, inner which parts trace of protostars diffuse ( 30– vicinity of the forming protostars∼ can be traced. Moreover, and to study3 the gas motions there. ∼ star (Masunaga & Inutsuka 2000). For typical distances of 100 cm− ), transitional regions from ionized or atomic gas COprotostellar is the most sources strongly (d emitting300 pc), species the diameter in such ( a region,1000 au) toTo molecular quantify the gas physical in the conditions ISM, and traced thus byphoto-dissociation THz high-J enabling us to trace lower-mass∼ or more distant objects∼ than molecular lines, we performed statistical equilibrium calcu- corresponds to 3.′′3, which roughly matches the angular res- regions (PDRs), H II regions, and surfaces of molecular other species. Profile shapes of those lines are probes of gas lations of the CO lines based on the large velocity gra- olution of the GLT at 1.5 THz ( 4′′; subsection 2.3). With clouds. Among them [N II] (1.46 THz) can be observed motions in such regions. Wiedner∼ et al. (2006) built a 1.5- dient (LVG) model (Goldreich & Kwan 1974; Scoville & the GLT, thus, very hot molecular gas in the vicinity of the in the THz atmospheric windows. The THz line profiles THz heterodyne receiver, CO N+ Deuterium Observations Solomon 1974). For the calculations, the rotational energy forming protostars can be observed. The line profile shape provide kinematical information, which is a clue to the Reciever (CONDOR), installed it in the APEX, and per- levels, line frequencies, and the Einstein A coefficients of can then be used to trace the gas motions in such regions. cloud formation mechanism, or the disruption mechanism formed ground-based THz line observations. With a total CO are adopted from the Leiden Atomic and Molecular In table 3, we summarize possible target protostellar of clouds by the newly formed stars. objects for the first-generation THz line experiments. We There are several pure rotational lines of “chemically selected nearby (with distance D " 300 pc) and luminous basic,” light molecules in the THz region. Previous mm (! 4 L ) protostellar sources with ample previous (sub)mm molecular-line studies have shown that there are abundant ⊙ molecular-line studies. carbon-chain molecules toward dense cores without known Now we estimate the expected THz intensity of these protostellar sources, while toward dense cores with known sources. Considering that there is a linear correlation protostellar sources those carbon-chain molecules are defi- between the bolometric luminosities and the intensities cient (Suzuki et al. 1992). This result suggests that abun- of the submm CS (7–6) line toward protostellar sources dances of carbon-chain molecules can be used to trace evo- (Takakuwa & Kamazaki 2011), we simply assume that lutionary stages of dense cores toward star formation. CH there is also a linear correlation between the intensity of the is considered to be the starting point of carbon-chain chem-

THz CO line and the source bolometric luminosity (Lbol). istry (Leung et al. 1984), and observation of the CH line 5/16/2015 H2D+ observations give an age of at least one million years for a cloud core forming Sun-like stars : Nature : Nature Publishing Group distance of 120 pc (ref. 24).

+ The present observations provide the measurement of the ortho/para H2D ratio, and thus the corresponding ortho/para H2 ratio + + across the dense core (see Methods). The para­H2D spectrum observed with SOFIA and the ortho­H2D spectrum observed with + APEX are shown in Fig. 1, together with the model predictions (detailed below). We note that para­H2D shows a strong and narrow absorption profile against the far­infrared continuum emission caused by the central protostellar heating of the surrounding dust + grains, while ortho­H2D is observed in emission with a similar width. These observational facts are related to the cold temperatures + of the environment, where almost all para­H2D is in its rotational ground state (000) and is therefore observed in absorption (see the energy level diagram in Fig. 1). Owing to the nuclear spin conversion discussed in the Methods, the g5r/o16u/n20d1 (5111) and the fHir2sDt+ observations give an age of at least one million years for a cloud core forming Sun-like stars : Nature : Nature Publishing Group + excited rotational state (110) of ortho­H2D are populated even at the low temperature of the dense core. As a consequence, in natu+re.com Publications A­Z index Access provided to Harvard University by Harvard University Libraries Cart Login Register combination with the lower continuum brightness at larger wavelengths, the major contribution to the ortho­H2D signal observed at + 372 GHz is due to emission (see the energy level diagram in Fig. 1). In what follows, we estimate the amountAs rochf ipveara­HV2oDlum aen d516 Issue 7530 Letters Article + ortho­H2D in the line of sight causing the observed absorption and emission features by radiative transfer modelling. It turns out + that the ortho/para ratio of H2D is below 0.1 in the cool outer part of the dense core where the lines originatNeA. ThUiRs Eim |p LliEesT TaE vRery low ortho/para H2 ratio in this region. 日本語要約

+ Figure 1: Observed and modelled H2D spectra. + H2D observations give an age of at least one million years for a cloud core forming Sun­like stars

Sandra Brünken, Olli Sipilä, Edward T. Chambers, Jorma Harju, Paola Caselli, Oskar Asvany, Cornelia E. Honingh, Tomasz Kamiński, Karl M. Menten, Jürgen Stutzki & Stephan Schlemmer

Nature 516, 219–221 (11 December 2014) doi:10.1038/nature13924 Received 07 July 2014 Accepted 06 October 2014 Published online 17 November 2014 Print

The age of dense interstellar cloud cores, where stars and planets form, is a crucial parameter in star formation and difficult to measure. Some models predict rapid collapse1, 2, whereas others predict timescales of more than one million years (ref. 3). One possible approach to determining the age is through chemical changes as cloud contraction occurs, in

particular through indirect measurements of the ratio of the two spin isomers (ortho/para) of molecular hydrogen, H2, which decreases monotonically with age4, 5, 6. This has been done for the dense cloud core L183, for which the deuterium + 7 fractionation of diazenylium (N2H ) was used as a chemical clock to infer that the core has contracted rapidly (on a timescale of less than 700,000 years). Among astronomically observable molecules, the spin isomers of the deuterated 5/16/2015 H2D+ observat+ions give an age of a+t least one million years for a cloud core forming Sun-like stars : Nature : Nature Publishing Group trihydrogen cation, ortho­H2D and para­H2D , have the most direct chemical connections to H2 (refs 8, 9, 10, 11, 12) and their abundance ratio provides a chemical clock that is sensitive to greater cloud core ages. So far this ratio has not been + determined because para­H2D is very difficult to observe. The detection of its rotational ground­state line has only now become possible thanks to accurate measurements of its transition frequency in the laboratory13, and recent progress in 14, 15 + instrumentation technology . Here we report observations of ortho­ and para­H2D emission and absorption, respectively, from the dense cloud core hosting IRAS 16293­2422 A/B, a group of nascent solar­type stars (with ages of less than 100,000 years). Using the ortho/para ratio in conjunction with chemical models, we find that the dense core has + 7 been chemically processed for at least one million years. The apparent discrepancy with the earlier N2H work arises + because that chemical clock turns off sooner than the H2D clock, but both results imply that star­forming dense cores have ages of about one million years, rather than 100,000 years.

+ + Subject terms: Astronomy and astrophysics Interstellar medium a, The histograms show the ortho­H2D (top) and para­H2D (bottom) rotational ground­state lines as observed with APEX/FLASH and SOFIA/GREAT, respectively; the orange lines show the modelled line profiles. Intensities are given as antenna temperatures T* A and vLSR denotes the velocity with respect to the local standard of rest. b, Energy level diagram (in units of temperature, E/k, wMhaerine k is the Boltzmann constant) of the lowest rotational states of ortho­ and para­H D+. + 2 We detected the ground­state rotational transition of the para spin isomer of the deuterated trihydrogen cation (para­H2D ) at 1.370085 THz (ref. 13) (wavelength λ = 219 μm) towards IRAS 16293­2422 A/B using the German REceiver for Astronomy at Terahertz frequencies (GREAT14) onboard the Stratospheric Observatory For Infrared Astronomy (SOFIA15). This line has so far We model the observed lines using the previously derived dense core structure22 in conjunction with chemistry and radiative only been tentatively detected in absorption against the bright high­mass star­forming region Orion Irc2 by the Kuiper Airborne http://www.nature.com.ezp-prod1.hul.harvard.edu/nature/journal/v516/n7530/full/nature13924.html#methods 16 2/12 + Observatory . We also observed the ground­state line of ortho­H2D at 372.421 GHz (ref. 17) (λ = 0.8 mm) towards the same source using the Atacama Pathfinder EXperiment (APEX) 18 submillimetre telescope located in the Chilean Atacama desert at an altitude of 5,100 m. IRAS 16293­2422 A/B consists of a triple system of young (<100,000 years) solar­type protostars, comprising a close protobinary (A1/A2) and a third protostar (B) about 600 astronomical units (AU) away from these19, 20, surrounded by a massive envelope (a dense core of about two solar masses) with steeply decreasing (from the inside outward) temperature and density distributions21, 22. This dense core still bears the physical characteristics typical of starless cores on the verge of star formation (the so­called pre­stellar cores23), with the bulk of the material still at low temperatures (T < 20 K) and densities (number 6 −3 density of H 2 molecules n(H 2) < 10 cm ). The dense core is embedded in the dark cloud Lynds 1689N in Ophiuchus at a At kinetic temperatures T above ~12 K, the ortho/para H D+ ratio is completely determined in reactions with ortho­ and para­H , and it is http://www.nature.com.ezp-prod1.hul.harvard.edu/nature/journal/v516/n75230/full/nature13924.html#methods 2 1/12 closely tied to the evolution of the ortho/para H2 ratio. The shaded vertical region indicates the temperature range applicable to the dense core surrounding IRAS 16293­2422 A/B (at radial distances from the core centre of 3,000–6100 AU), while the horizontal shade indicates the + observed ortho/para H2D ratio. Together, these limits suggest a dense core age of at least one million years. The gas density, n(H 2) = 5 −3 10 cm , and the visual extinction, A V = 10 mag, are kept constant in this model.

+ + There!27fore, we have verified that the observed ortho/para H2D ratio is setting limits on the core age. The ortho/para H2D ratio + gives a more direct estimate of the ortho/para H2 ratio than the previously used deuterium fraction measurement of N2H (that is, + + + + N2D /N2H ; ref. 7), in particular for evolved regions with ortho/para H2 ratios of less than 0.01. Below this value, the N2D /N2H + + ratio loses correlation with the ortho/para H2 ratio (see Extended Data Figs 3 and 4). Therefore, at this point, the N2D /N2H + chemical clock stops while the clock based on the ortho/para H2D ratio keeps running. Our results indicate that the average −4 ortho/para H2 ratio is about 2 × 10 between radii of 3,000 AU and 6,100 AU (T = 13–16 K), which can be reproduced only at very late times of chemical evolution (see Extended Data Fig. 5). Our conservative analysis gives an age estimate of at least one million

years. The very low value of ortho/para H2 found in the core around IRAS16293­2422 is hardly possible to probe by any other + means, and we conclude that the timing set by the ortho/para H2D ratio is most relevant for constraining the duration of the dense cloud core phase in the course of star formation.

Methods Observational strategy + Although the rotational ground­state emission line of ortho­H2D has been observed towards several cold starless and low­mass 21, 28 + star­forming cores , para­H2D has previously been detected only tentatively, in absorption towards the Orion IRc2 region using 16 + the NASA Kuiper Airborne Observatory (KAO) . The ground­state rotational transition of para­H2D at 1.37 THz is extremely difficult to observe from the ground owing to poor atmospheric transmission and the resulting large system temperatures at terahertz frequencies. The transition frequency was not covered by the Herschel/HIFI bands29. The GREAT14 receiver onboard SOFIA15 now

http://www.nature.com.ezp-prod1.hul.harvard.edu/nature/journal/v516/n7530/full/nature13924.html#methods 4/12 Publications of the Astronomical Society of Japan (2016), Vol. 68, No. 1 R1-31 Time variable AGN sources Ta b l e 7 . Candidate AGNs for monitoring observations with 2. Submm VLBI with GLT at 345 GHz will have enough the GLT. sensitivity to detect the source without phased-up ALMA. Assuming that the blazar has a flat spectrum, it Name Alias Flux at 15 GHz z Optical ID [Jy] requires that the flux >0.3 Jy. 3. The object is accessible from the GLT site with zenith J0112 2244 S2 0109 22 0.265 BL Lac 0.48 + + angle <60◦. This gives a source declination larger J0319 4130 3C 84 0.0176 Galaxy 19.4 + than 12◦. J0721 7120 S5 0716 71 0.31 BL Lac 1.2 + + J0748 2400 PKS 0745 241 0.4092 QSO 1.15 + + J0854 2006 OJ 287 0.306 BL Lac 4.67 Measurement of the absolute flux density is expected to + J0958 6533 S4 0954 65 0.367 BL Lac 1.34 be limited by the calibration accuracy, which would typi- + + J1104 3812 Mrk 421 0.0308 BL Lac 0.33 cally be 10%. In order to achieve this accuracy robustly, + J1217 3007 ON 325 0.13 BL Lac 0.36 it is desirable to achieve a thermal noise level that is three + J1230 1223 M 87 0.0044 Galaxy 2.51 + times better than the calibration error. With this condition, J1653 3945 Mrk 501 0.0337 BL Lac 0.87 + around 4% degradation is expected to contribute to the J1719 1745 OT 129 0.137 BL Lac 0.58 + final accuracy due to the thermal noise. As a result, we need J1806 6949 3C 371 0.051 BL Lac 1.37 + only 30 min for the total on-source time at 345 GHz for all J1927 7358 4C 73.18 0.302 QSO 3.71 + + J2022 6136 OW 637 0.227 Galaxy 2.26 the sources in table 7. Taking into account the overheads, + J2143 1743 OX 169 0.2107 QSO 1.09 it will take one hour for each monitoring. Although the + J2202 4216 BL Lac 0.0686 BL Lac 4.52 observable sky from Greenland is limited, the same object + J2203 3145 4C 31.63 0.2947 QSO 2.6 can be monitored constantly over the winter season. Since + + time variability on a day–week timescale is expected, we !28 ∼ active in GeV bands during the last five years. Flux vari- propose to conduct this monitoring every three–four days. ability in this source from GHz frequencies to GeV energies is examined by Raiteri et al. (2012); there were at least three 5.2 GRBs with the GLT high states of the GeV emission, and multi-flaring peaks are associated with each event. The γ -ray and optical variability GRBs are among the most powerful explosions in the Uni- on a one-week timescale were tightly correlated, where the verse and are observationally characterized according to sampling is dense enough. In addition, there was also some intense short flashes mainly occurring in the high-energy intrinsic, but weak, correlation in light curves between the band (prompt emission) and long-lived afterglows seen γ -ray/optical and mm fluxes (taken by the SMA 1.3 mm from the X-ray to radio bands. The discovery of after- band). We suspect that the sampling rate at submm/mm glows was a watershed event in demonstrating the following wavelengths is still poor (> aweek).Intheshock-in-jet properties of GRBs: a cosmological origin for long–soft scenario, it is generally expected that the flare peak in the and short–hard GRBs (Metzger et al. 1997; Berger et al. submm/mm regime is followed with some time delay by the 2005; Fox et al. 2005); collimated narrow jets with energy 1051 erg for long–soft GRBs (e.g., Frail et al. 2001); asso- peak at longer wavelengths (i.e., mm to cm), as a conse- ∼ quence of the jet opacity, as demonstrated in radio light ciation with star-forming galaxies (e.g., Djorgovski et al. curves [e.g., PKS 1510–089 (FSRQ): Orienti et al. 2013. 1998); and a massive star origin for long–soft GRBs (Hjorth In conclusion, a dense monitoring of mm/submm fluxes et al. 2003;Staneketal.2003). Moreover, GRBs are now toward γ -ray bright blazars spanning less than a week is being exploited as probes of the high-redshift Universe, even the best strategy in order to test the shock-in-jet model and at z ! 8 (Tanvir et al. 2009). Because of their extremely high to explore the site and emission mechanism of the VHE luminosity, the highest-z events at the reionization epoch (z 8) have already been observed, and their discovery at z γ -ray flares in AGNs. ∼ > 10 is highly possible (e.g., Wanderman & Piran 2010). 5.1.5 Sample selection The standard fireball model (e.g., Mesz´ aros´ & Rees We started with the original sample of VLBI-detected Fermi 1997) predicts a double-shock system: a forward shock (FS) sources (Lister et al. 2011). From this sample we selected propagating in the external medium, and a reverse shock 17 sources, listed in table 7, with the following criteria. (RS) propagating into the ejecta. The RS emission uniquely probes the dominant composition (i.e., whether the com- 1. Submm VLBI with the GLT at 345 GHz will have position is dominated by baryons or magnetic fields) and enough linear resolution to resolve a blazar zone of properties (such as thickness and Lorentz factor) of the 0.01 pc. Practically, this condition limits the angular ejecta (e.g., Kobayashi 2000). These characteristics revealed 3 diameter distance to 10 Mpc (z < 0.37). by the RS can in turn constrain the unknown nature of !29 !30 Future instruments

!31 27TH INTERNATIONAL SYMPOSIUM ON SPACE TERAHERTZ TECHNOLOGY, NANJING, CHINA, APRIL 13-15, 2016

27TH INTERNATIONAL SYMPOSIUM ON SPACE TERAHERTZ TECHNOLOGY, NANJING, CHINA, APRIL 13-15, 2016 Development of 1.5 THz Cartridge-type Multi-pixel DevelopmentReceiver of 1.5 Based THz on Cartridge HEB Mixers-type Multi-pixel Yen-Ru Huang*, Chuang-Ping Chiu, Hsian-Hong Chang, Yen-Yu Chiang, and Ming-Jye Wang InstituteReceiver of Astronomy and Astrophysics, Based Academia on Sinica, HEB Taipei27TH 10617, INTERNATIONALMixers Taiwan SYMPOS IUM ON SPACE TERAHERTZ TECHNOLOGY, NANJING, CHINA , 13-15 APRIL, 2016 *Contact: [email protected], phone: +886-2-2366-5309 Yen-Ru Huang*, Chuang-Ping Chiu, Hsian -Hong Chang, Yen-Yu Chiang, and Ming-Jye Wang Institute of Astronomy and Astrophysics,of the quartz Academia and silicon Sinica, slab Taipeis are 10617,optimized Taiwan to 135.5 m and which can meet the ALMA specification, and the vacuum level of 5.7E-5 mbar at room temperature can be achieved. *Contact: [email protected] m respectively, phone which: +886 are-2 -depicted2366-5309 in Fig. 2(b). The temperatures of three stages during cooling procedure are shown in Fig. 4. The dashed line is the result of bare cartridge Abstract— A design of 2 × 2 NbN-based hot-electron-bolometer Greenland Telescope (GLT) [7]. An engineering model of body, and the solid line is that of the engineering model (HEB) mixer array receiver cartridge has been demonstrated cartridge which can be used for single-pixel and 2×2 array cartridge. The balanced temperatures of three stages are 2.8K, here by using multiple local oscillator (LO) beams. In our design, receiver has been constructed. The thermal analysis, cooling — 14.8K, and 79.5K for the bare cartridge, and 2.8K, 15.3K, and the 1.5Abstract THz LO A beam design is split of 2 into× 2 four NbN uniform-based hot sub-electron-beams with-bolometer a test ofGreenland cartridge , Telescopeand LO source (GLT) performance [7]. An will engineering be discussed. model of spacing(HEB) of 18 mixer mm by array using receiver a power cartridge distributor has, then been arrives demonstrated at a cartridge which can be used for single-pixel and 280.2K×2 array for the single-pixel engineering model cartridge respectively. The DC biases for cooled multipliers of the LO fourhere-pixel by silicon using multiple lens with local twin oscillator slot antenna (LO) (TSA)beams. through In our design, a receiver hasC ARTRIDGE been constructed DESIGN. C TheONCEPT thermal analysis, cooling largethe-area 1.5 beamTHz LO splitter. beam Anis split additional into four four uniform-pixel HDPE sub-beams lens iswith a test of cartridge, and LO source performance will be discussed.module are applied before cooling. The slightly increase of locatedspacing at 120 of mm18 mm in frontby using of the a powersilicon distributorlens to increase, then the arrives size at a The receiver cartridge comprises three cooled stages withtemperature at 15K and 110K stages is due to extra thermal operation temperatures of 4K, 15K, and 110K. To boost of beamfour- pixel waist silicon for fitting lens with to the twin aperture slot antenna parameter (TSA) of through sub- a CARTRIDGE DESIGN CONCEPT loading, such as cables, wires, and waveguides which are not reflectorlarge - ofarea GLT. beam Some splitter. cryogenic An additional tests of cartridge four-pixel ha HDPEve been lens isoutput power, cooled frequency multipliers are used in the LOincluded in the bare cartridge body. The receiver cartridge comprises three cooled stages with carriedlocated out at. In120 this mm article, in front w ofe the report silicon the lensdesign, to increase assembly, the sizemodule which is provided by VDI Inc. The fundamental thermalof beam analysis, waist and for some fitting testing to results the aperture of cartridge parameter. of subfrequency- operation of LO temperatures signal is generated of 4K, 15K, by a and synthesizer 110K. andTo boost reflector of GLT. Some cryogenic tests of cartridge have beenmultiplied output to power Q-band, cooled frequency frequency. Then multiplier the powers areis amplified used in the LO INTRODUCTION carried out. In this article, we report the design, assembly,Fig. 1and The modulefed schematic through which diagram the ofis vacuum the provided single--typepixel by and waveguideVDI four -pixel Inc. cartridge feedthroughThe fundamental Tthermalhe superconducting analysis, and someNbN testing based results hot-electron of cartridge-bolometer. design. (WR22) frequency at 300K of baseLO - signalplate tois reduce generated power by a los synthesizerses. The and (HEB) mixer is one of the commonest detectors for THz radio frequencymultiplied is further to Q- band multiplied frequency by . Thena cooled the powertwo-doubler is amplified INTRODUCTION astronomy due to its near quantum limit noise performance moduleand and fed a through cooled two the-tripler vacuum module-type to waveguide reach the feedthrough target and lowThe local superconducting oscillator (LO) NbNpower basedrequirement hot-electron [1], and-bolometer has frequency. (WR22) The at output 300K horn base of- plateLO module to reduce is fixed power at the losfocalses . The already(HEB) been mixer applie is done in severalof the common astronomicalest detectors observator fories THz [2 -radiopoint frequency of a 90 degree is further off-axis multiplied parabolic by mirror a cooled on the two 110K-doubler 4]. astronomyIn the past dueyears, to weits havenear sucquantumcessfully limit realized noise 1.performance4 THz plate module with an and effective a cooled focal two length-tripler of module 30 mm to to reach make the the target NbNand based low HEBlocal mixersoscillator with (LO) a receiver power noiserequirement temperature [1], andof hasRayleigh frequency. length The enough. output For horn 2 ×of2 LO array module receiver, is fixed a power at the focal aroundalready 2000K been and applie a bandwidthd in several of astronomical 3.5 GHz. Tohebservator receiveries [2distributor- point ofmodule a 90 is degree used tooff di-axisvide parabolic the LO beam mirror in to on four the 110K noise4] . temperatureIn the past canyears, be we further have improvedsuccessfully by realizedreplacing 1. 4a THzsub -beamsplate with with an a spacing effective of focal 18 mm. length After of that 30 the mm LO to and make the siliconNbN lens based with HEB anti mixers-reflection with acoating. receiver Innoise addition, temperature for ofradio Rayleigh frequency length (RF) enough.beams are For combined 2×2 array by receiver, a large -area a power astronomyaround receivers,2000K and multi a - bandwidthpixel receiver of 3.5is more GHz favourable. The receiverFig. 2beam (a) distributor Tsplitterhe four -pixelmade module power of mylar distributoris used films module. to with divide 13 (b) theThem thickness,calculated LO beam and into four becausenoise of temperature high mapping can speed. be furtherHowever, improved uniformly by injecting replacingtransmittance thena !32 subcouple and- beamsreflectance to the wit of HEB hpolarizing a spacing mixer beam adhered ofsplitters 18 mm.versus on the Afterthickness. backside that the ofFig. LOa 4 andThe temperatures of the three cooled stages versus time during cooling enoughsilicon LO lens power with to eachanti - pixelreflection is a criticalcoating. issue In in addition, a THz forsilicon radio lens frequency. The reflectance (RF) beamsof mylar are beam combined splitter by is a about largeprocedure -area. The dashed line is the result of the bare cartridge, and the solid line COOLING TEST OF CARTRIDGE is that of the engineering model cartridge. multiastronomy-pixel receiver. receivers, To achieve multi- pixelthe minimum receiver poweris more loss favourable and 20 %.beam In addition, splitter anmadeother of four mylar-pixel films HDPE with lens 13 ism located thickness, at and The vacuum and cooling test of cartridge has been uniformbecause beam of high profiles, mapping most speed.HEB arrayHowever, receivers uniformly apply injectingthe 120 mmthen in couple front of to the the silicon HEB lens mixer to increaseadhered the on size the of backside beam of a completed. The loading system with an ALMA-type testing polarizingenough beam LO power splitters to for each the pixel LO is beams a critical [5,6 ], issue including in a THzwaist silicon for matching lens. The the aperturereflectance parameter of mylar of beamthe sub splitter-reflector is To about understand the thermal loading after the synthesizer of Fouriermulti - phasepixel receiver. gratings, To wireachieve-grid the polarizers, minimum power multilayer losscryostat andof the is GLT shown antenna in Fig.. 3. There are two parallel slippery LO module is turned on, we mounted more sensors on tracks and20 a%. transportation In addition, an plateother on four the-pixel system, HDPE and lens the is located at structures,uniform etc. beam In profiles, our design, most simple HEB arraydielectric receivers slab beamapply the The design and assembly of the engineering modeldifferent positions to measure the temperature variation. Fig. 5 cartridge can120 be mm adjusted in front to of the the target silicon height lens byto increaseusing a handthe size of beam splitterspolarizing are used beam which splitters can for suppress the LO beams the variations [5,6], including of cartridge had been completed, and the testing of single-pixelshows the temperature evolution of each sensor when pulled stackerwaist. Tforhe matchingcartridge isthe mounted aperture on parameter the under -ofside the of sub -reflector transmittance and reflectance below 2.2% in the frequency design has been carried out first to check the performance. Figsynthesizer. of LO module is turned on and off. The two- Fourier phase gratings, wire-grid polarizers, multilayertransportation of the plate GLT, and antenna then be. slowly loaded into the testing range from 1.45 THz to 1.55 THz. 1 is the schematic diagram of the single-pixel and four-pixeldoubler module is heated from 108.8K to 152.1K as the structures, etc. In our design, simple dielectric slab cryostat beam alongThe the design slippery and tracks assembly. The Sumitomo of the RDK engineering-3ST model The receiver cartridges for the Atacama Large Millimeter receiver cartridge design. In these two engineering models, theamplified Q-band LO signals are turned on, and the splitters are used which can suppress the variationsthree of stage cartridge cryocooler had is been used, completed which has, andcooling the testi powerngs ofof single-pixel /submillimetertransmittance Array and reflectance(ALMA) belowhave many 2.2% inadvantages, the frequency arrangements design has of been cartridge carried are out identical, first to checkexcept the of performance. the mixertemperature Fig. of two-tripler module increases about 10K. including high cooling efficiency and good modularity which1.0 W at 4.4K for the 4K stage, 8 W at 18K for the 15K stage, However, the 4K plate and the mixer block (~ 3.4K) have a range from 1.45 THz to 1.55 THz. and 33block, W 1at is85the tKhe forLO schematic the coupling 110K diagramstage. module, of theand singlethe - numberpixel and of four IF -pixel can simplify the efforts on telescope site. In this paper, we temperature increase below 0.015K. We expect that the The receiver cartridges for the Atacama Large Millimeter channels. receiver Fig. cartridge 2(a) shows design the. In four these-pixel two engineeringpower distributor model s, the report the designs of single-pixel and 2×2 HEB mixer array performance HEB mixer won’t be affected during operation. /submillimeter Array (ALMA) have many advantages,module arrangements which consists of cartridge of TE and are TM identical, mode polarizing except of beam the mixer ALMA-type cartridge receiver at the frequency range of 1.45 - including high cooling efficiency and good modularity whichsplitters block, made the of LOquartz coupling and silicon module, slabs respectivelyand the number due to of IF 1.55 THz. The pixel number in our design can be further can simplify the efforts on telescope site. In this paper, wethe polarizationchannels. Fig. direction 2(a) of shows LO signal the fours. To-pixel achieve power uniform distributor extendedreport to the 9 designbased s onof currentlysingle-pixel available and 2 ×LO2 HEB power. mixer We array distribution patterns of the output sub-beams, the thicknesses target to deploy this multi-pixel THz receiver cartridge on the module which consists of TE and TM mode polarizing beam ALMA-type cartridge receiver at the frequency range of 1.45 - splitters made of quartz and silicon slabs respectively due to 1.55 THz. The pixel number in our design can be further the polarization direction of LO signals. To achieve uniform extended to 9 based on currently available LO power. We distribution patterns of the output sub-beams, the thicknesses target to deploy this multi-pixel THz receiver cartridge on the

Fig. 3 The loading system and testing cryostat for cartridge.

The vacuum sealing test of 300K base-plate shows a helium Fig. 5 The temperature distribution of the cold cartridge. leakage rate below 2.4E-6 mbar*L/s within 2000 seconds M4-1 26TH INTERNATIONAL SYMPOSIUM ON SPACE TERAHERTZ TECHNOLOGY, CAMBRIDGE, MA, 16-18 MARCH, 2015

components and resonators. Recently, we have begun readout line of the chip, each of which will address an measurements on a series of test devices intended to prove the individual resonator. The transmission gain and phase is then key technologies. In the course of the descriptions of each monitored to detect changes in the S-parameters of the component, we will present preliminary results from these resonators in response to illumination. The tones will be devices, and describe the planned measurement campaign in generated and processed at baseband frequency (0-500 MHz) detail. Finally, we will also discuss the development work and up- and down-converted to and from the desired readout being done on each part for the next generation of chips. frequency. All of this is done at room temperature, and the only cryogenic electronics required is the cabling to carry the II. SYSTEM OVERVIEW signal to and from the array and a low-noise HEMT amplifier. A detailed overview of the plan for the CAMELS prototype In the sections that follow we will discuss the plans for each instrument can be found in [5]. A system block diagram is of the components in detail and the present state of testing. 26TH INTERNATIONAL SYMPOSIUM ON SPACE TERAHERTZ TECHNshownOLOGY, in Fig. CAMBRIDGE, 1. At the basic MA level,, 16-18 an MARCH, IFBS consists 2015 of an M4-1 26TH INTERNATIONAL SYMPOSIUM ON SPACE TERAHERantenna toTZ couple TECHN radiationOLOGY, onto CAMBRIDGE,a transmission line MA; followed, 16-18 MARCH, 2015 III. OPTICAL DESIGNM4-1 by a bank of narrow bandpass filters to divide the signal into spectral channels; then a series of detectors to measure the Progress on the CAmbridgetotal integrated Emission power in each channel. Line Each spectrometer Surveyor M4-1pixel is integrated onto26TH aINTERNATIONAL single chip. KIDsSYMPOS willIUM be ON usedSPACE for TERAHER TZ TECHNOLOGY, CAMBRIDGE, MA, 16-18 MARCH, 2015 Progress on the CAmbridge power detection. These will be Emission a quarter-wave design, based Line Surveyor (CAMELS)componentson a length and of Niobiumresonators. Nitride Recently, (NbN) we and have SiO 2 begun microstrip readout line of the chip, each of which will address an measurementsshorted at oneon a series end aofnd test lightly devices capacitively intended to prove coupled the toindividual a resonator. The transmission gain and phase is then Christopher N. Thomas1,*, Ray Blundell2, D. Glowackakeyreadout1, David technologies line J. atGoldie. the In other the1, Paul course. Readout Grimes of the frequencies descriptions2, Eloy de in oftheLera each range Acedo 4monitored-6 1, Scott to Paine detect2, changes in the S-parameters of the Stafford Withingtoncomponent,GHz will(CAMELS)1 and be we used.Lingzhen will present The resonatorsZeng preliminary2 incorporate results from a sectio thesen ofresonators β- in response to illumination. The tones will be 1 devices,phase Ta andntalum describe (β- Ta) the , planned in which measurement the signal photons campaign are in able generated to and processed at baseband frequency (0-500 MHz) Cavendish1,* Laboratory, JJ2 Thomson Avenue,1 Cambridge, CB3 0HE1 , UK 2 1 2 Christopher N. Thomas , Ray Blundell , D.detail.break Glowacka Finally,Cooper wepairs , willDavid and also generate discussJ. Goldie quasiparticles,the development, Paul changingGrimes work theand, Eloy up- and de down Lera-converted Acedo to and, Scott from the Paine desired, readout 2Smithsonian Astrophysical Observatory, 60 Garden Street, Cambridge, MA 02138, USA Staffordbeingresonance done Withingtonon characteristic each part for the1. and nextThe generationLingzhen filters and of chips. opticalZeng 2 coupling frequency., All of this is done at room temperature, and the *Contact: [email protected] at > , 100phone GHz +,44 will 1223 be 764021 implemented in the sameonly cryogenic electronics required is the cabling to carry the 1 II. SYSTEM OVERVIEW signal to and from the array and a low-noise HEMT amplifier. Cavendish Laboratory,microstrip. JJ In Thomson operation, t heAvenue, spectrometer Cambridge, chips will be CB3 cooled 0HE , UK A detailed overview of the plan for the CAMELS prototype In the sections that follow we will discuss the plans for each 2 to 100mK in an Adiabatic Demagnetisation Refrigerator Smithsonian Astrophysicalinstrument Observatory can be found, 60 in [Garden5]. A system Street block, diagramCambridge, is of the MA components 02138 in, detaiUSAl and the present state of testing. *Contact: [email protected](ADR), in Fig.which 1. At is thecurrently basic level, being, an phone commissioned IFBS consists +44 1223of for an device 764021 antennatesting to at couple the SAO radiation. onto a transmission line; followed III. OPTICAL DESIGN Abstract — The aim of CAmbridge Emission Line Surveyorby a bank ofaddressing narrow bandpass issues filters such to divide as the flux signal- and into frequency calibration, (CAMELS) is to provide an operational demonstrationspectral of an channelsoperation; then a inseries varying of detectors background to measures and the the development of Integrated Filter Bank Spectrometer (IFBS) for mmtotal-wave integrated optimum power in observing each channel. strategies. Each spectrometer To do so, we will deploy a astro nomy. The prototype will observe from 103-114.7pixel GHz, is integrated onto a single chip. KIDs will be used for power detectionpathfinder. These will instrument be a quarter on-wave the design, Greenland based Telescope during its providingAbstract of— order Th e 500 aim channels of CAmbridge with a spectral Emission resolution Line of commissioning Surveyor addressing phase in 2016. issues such as flux- and frequency calibration, 3000. In this paper we discuss the design of the instrumenton aand length of Niobium Nitride (NbN) and SiO2 microstrip shorted at one end and lightly capacitively coupled to a ongoing(CAMELS) work towards is to its providerealisation. an Fabrication operational of a demonstrationfirst set of The of pathfinderan operation instrument in willvarying provide background four spectrometers and the development of Integrated Filter Bank Spectrometer (IFBS)readout for line at mmpixels, the -otherwave each. Readout providing frequencies 256 inspectral the range channels 4-6 with a resolution devices to verify the key technologies has recentlyGHz been will be used. The resonatorsoptimum incorporate observing a section of β - strategies. To do so, we will deploy a completed.astronomy. We Twillhe present prototype results will from observe a measurement from 103 -114.7R = ∆ν/νGHz, ≈ 3000 (a velocity resolution of ≈100 km/s). One phase Tantalum (β-Ta), in whichpathfinder the signal photons instrument are able to on the Greenland Telescope during its campaign to characterise resonator performance and describebreak Cooper pair pairs will and observegenerate quasiparticles, in the frequency changing range the 103-109.8 GHz (L) providing of order 500 channels with a spectral resolution of commissioning phase in 2016. our3000. planned In optical this tests.paper we discuss the design of theresonance instrument characteristicand the and ot. her The from filters 109.8 and - optical114.7 couplingGHz (H), . The mainFig. 2 scienceDetails of antennas for CAMELS chips. a) Proposed 4-probe horn operating at target > 100 is GHz 12,CO(1 will be-0) The implemented and 13 pathfinderCO(1 in- 0) the line same instrument emission coupling.from will galaxies b) provide Concept for testfour devices. spectrometer c) Photo of realised antenna on test ongoing work towards its realisation. Fabricationmicrostrip. of a Infirst operation, set of the spectrometer chips will be cooled device. I. INTRODUCTION in redshift range 0.05-0.13 (12CO) and 0.003-0.961 (13CO), devices to verify the key technologies hasto 100mK recently in an been Adiabatic pixels,Demagnetisatio eachn providingRefrigerator 256 spectral channels with a resolution Recent years have seen increased interest in filter bank which will allow survey work and mapping gasA. distribution Antenna s. completed. We will present results from(ADR), a measurementwhich is currently beingR = commissioned ∆ν/ν ≈ 3000 for device (a velocity resolution of ≈100 km/s). One spectrometers for mm, sub-mm and far-infrared astronomtestingy. at the TheSAO .L and H bandspair willtest will observe allow usin to the test frequency performanceThe final range in chips 103will- 109.8 use horn GHz antennas (L) for telescope Thecampaign driving forces to characterise have been progress resonator in superconductingperformance and very describe different observing regimes. The L band couplingis away, so from a transition is needed from circular waveguide to thinour-film planned circuit technologyoptical tests., allowing the fabrication of the the band edge and and the so backgroundother from loading 109.8 - is114.7 the low, IFBS GHz but chip. the (H) We. areThe investigating main science a scheme using four 12 13 rectangular waveguide probes suspended on a membrane, required filter banks, and the parallel development of Kinetic emission lines aretarget fainter is asCO(1 the galaxies-0) and are moreCO(1 distant.-0) line The emission from galaxies I. INTRODUCTION based12 on a design for the CLOVER [163] (Fig. 2a). This will Inductance Detectors (KIDs), which are near-ideal because of emission in thein H redshift band from range nearby 0.05 - gal0.13axiesallow ( shouldCO) polarisation and be -interlac 0.003ing-0.961 of two (filterCO)-banks, , allowing for the easeRecent with years which have large seen numbers increased of devices interest can bein filterbrighter; bank however which, the will background allow survey loading work relaxed from and filter the mapping spacing O2 while gas maintaining distribution frequencys. coverage. fabricatedspectrometers and read outfor[ 1]mm,. In particular,sub-mm the and compatibility far-infrared of astronomemissiony. line atThe the L window and Hedge bands (119 testGHz) will is muchTo allow avoid higher us the to membrane test performance processing step in during the early part of the test campaign, the test chips we have produced use the The two technologies driving forces has made have possible been the progress Integrated in Filter superconducting and subject to greater variability. Having two pixels Bank Spectrometer (IFBS), where antenna, filter bank and observing in eachvery banddifferent simultaneous observingly will regimes. allowa planar for antenna The sky Lcoupled band to is microstrip away . from This is a centre-fed detectorsthin-film are allcircuit integrated technology on the , sameallowing chip. the IFBSs fabrication are Fig.chopping 1 ofSystem the blockwithout diagtheram loss of band the of CAMELS observing edge instrument. and time so, and background forsingle comparison slot design, loading which isis illuminated low, but from the the underside of Fig. 2the Details chip of antennas through for CAMELS a window chips. a) in Proposed the holder. 4-probe horn Transmitted predictedrequired to be filter able banks,to offer andcomparable the parallel resolution development to grating ofof Kinetic the systematics emission in notionally lines are identicalfaintercoupling. as units. the b) Concept galaxies We for are test devices.are more c) Photo distant. of realised antenna The on test device. radiation is absorbed by the blackening on the holder. This and Fourier transform spectrometers in these wavebands, butThe KIDs will be readout in the canonical manner using a Inductance Detectors (KIDs), which are near-superpositionideal becauseinitially of microwave targetingof emission background tones transmitted in limited the along H performance, band the fromarrangement which nearby, foris shown gal inaxies Fig. 2b, should and a photo be of the realised at a smaller physical size (by exploiting the slow-wave effect example, would require an NEP of order 5A. x Antenna10-18 W/Hz 0.5 for the ease with which large numbers of devices can be !33 brighter; however, the background loading from the O2 or lumped element filters), wider instantaneous bandwidths each detector assuming observing conditionsThe atfinal Thule chips Airwill use horn antennas for telescope fabricated and read out[1]. In particular, the compatibility of emission line at the windowcoupling , edgeso a transition (119 isGHz) needed fromis much circular higherwaveguide to and without moving parts. An imaging array realised in this Base in Greenland. the two technologies has made possible the Integrated Filter and subject to greaterthe IFBS variability chip. We. are investigatingHaving a two scheme pixels using four manner, with moderate spectral resolution over a wide Though science drivers for an mm-waverectangular instrument waveguide are probes suspended on a membrane, bandwidth,Bank Spectrometer would be a transformative (IFBS), where technology antenna, for survey filter bankstrong, andthere observingare several in technological each bandbased simultaneous challenges on a design for in thely CLOVER will allow [6] (Fig . for2a). skyThis will astronomydetectors and mapping. are all integrated on the same chip. IFBSsdeveloping are KIDschopping for long without-wavelength loss operation. allowof observing polarisation The - interlacmosttime ing, and of two for filter comparison-banks, allowing for relaxed filter spacing while maintaining frequency coverage. Severalpredicted IFBS to projects be able, such to asoffer SuperSpec[ comparable2], DESHIMA[ resolution3] tosignificant grating is of reduced the systematics frequency separation in notionallyTo avoidbetween the membrane identical the processing units. step duringWe theare early andand Mi croSpec[Fourier4 transform], are targeting spectrometers sub-mm and in farthese-infrared wavebands readout, but signal initially and the opticaltargeting signal background. Thispart ofcomplicates the limitedtest campaign theperformance,, the test chips we which have produced, for use wavelengths. The CAmbridge Emission Line Surveyor selection of a material system that functionsa planar well antenna at coupledboth to microstrip. This is a centre-fed at a smaller physical size (by exploiting the slow-wave effect example, would require an NEP of order 5 x 10-18 W/Hz0.5 for (CAMELS) is a complementary project targeting mm- Fig. optical1 System block frequencies diagram of the CAMELS (high- instrument.loss required forsingle absorption) slot design, which and is illuminated from the underside of the chip through a window in the holder. Transmitted wavelengthsor lumped[5], andelement is a jointfilters effort), wider between instantaneous the Quantum bandwidthsreadout frequencies each (low detector-loss required). assuming observing conditions at Thule Air and without moving parts. An imaging arrayThe realised KIDs will in be thisreadout in the canonical manner using a radiation is absorbed by the blackening on the holder. This Sensors Group of the Cavendish Laboratory and the Harvardsuperposition ofIn microwave this paper Base tones w e transmittedin will Greenland describe along . the currentarrangement state is shown of the in Fig. 2b, and a photo of the realised Smithsonianmanner, Astrophysical with moderate Observatory spectral (SAO) resolution. It has two over development a wide workTh forough the pathfinderscience drivers instrument. for anWe mm will -wave instrument are mainbandwidth, aims: 1) to woulddemonstrate be athe transformative technologies necessary technology for a forbegin survey by giving strong, a system there level overvieware several of the instrument.technological challenges in mmastronomy-wave IFBS andand mapping.2) to show operationally that it can be This will be followeddeveloping by a detailed KIDs description for long of-wavelength the design of operation. The most used to make science-grade observations. The latter involves each of the components, including the filter banks, optical Several IFBS projects, such as SuperSpec[2], DESHIMA[3] significant is reduced frequency separation between the and MicroSpec[4], are targeting sub-mm and far-infrared readout signal and the optical signal. This complicates the wavelengths. The CAmbridge Emission Line Surveyor selection of a material system that functions well at both (CAMELS) is a complementary project targeting mm- optical frequencies (high-loss required for absorption) and wavelengths[5], and is a joint effort between the Quantum readout frequencies (low-loss required). Sensors Group of the Cavendish Laboratory and the Harvard In this paper we will describe the current state of the Smithsonian Astrophysical Observatory (SAO). It has two development work for the pathfinder instrument. We will main aims: 1) to demonstrate the technologies necessary for a begin by giving a system level overview of the instrument. mm-wave IFBS and 2) to show operationally that it can be This will be followed by a detailed description of the design of used to make science-grade observations. The latter involves each of the components, including the filter banks, optical Qanaaq high school students visit, 5 May 2019

!34 Mahalo!

!35