The Science Verification Process
Science Verification (SV) is the process by which we demonstrate that ALMA is capable of producing data of the quality required for scientific analysis, i.e. data good enough to be included in the peer-reviewed literature, and by which we fully test all observing modes expected to be available during Early Science (ES). This is achieved by making observations of a small number of selected astronomical objects. This work has been taking place since early 2011 (see Table 4 below). We took the very first SV data-sets with as few as seven 12-m antennas but the results were nevertheless encouraging. Today, all 66 antennas have been delivered - 54 12-m antennas and 12 7-m antennas in the Atacama Compact Array (ACA; Morita Array), which also comprises up to 4 of the 12-m antennas for total power observations. Currently, we typically observe with at least 34 12-m antennas in the 12-m ALMA array, up to 12 antennas in the 7-m ACA array, and at least two 12-m antennas for total power. This means of course that the observations that we are making now should all have good image quality and excellent sensitivity.
Verification and Demonstration of ALMA Capabilities
Until June 2012 SV focused on reproducing results that had already been obtained with other telescopes. Projects were chosen, via an informal “Call for Suggestions” from the astronomical community, based on the criteria of being the most suitable for testing specific Early Science (ES) observing modes and availability of existing comparable data, rather than those with the highest potential science impact. With the new capabilities that have become available from the Cycle 1 period onwards this approach became increasingly less feasible. Instead we now carry out a set of observations of a small number of specific objects chosen to demonstrate specific capabilities expected to be offered for future ES Cycles. As ALMA capabilities have now entered new parameter space, targets may not necessarily have existing directly comparable data taken with other telescopes. As has been the practice thus far, the reduced and calibrated datasets will be made available to the community for download as soon as the data have been successfully collected, reduced, and verified.
Collecting, reducing and releasing the data
SV observations are carried out as part of Extension and Optimization of Capabilities (EOC; formerly Commissioning and Science Verification) activities, which currently take place on an alternating schedule with ES observations. There are some planned breaks in the schedule, in particular a shutdown of operations during the month of February (the Altiplanic winter), and partial shutdowns for essential activities such as, for example, planned upgrades to the correlator or power systems. From 1 September to 30 November 2014 there was an uninterrupted period of EOC activity while we carried out the ALMA Long Baseline Campaign. Early Science observations re-started in December 2014.
The lists of targets below include an estimate of when it is expected that the observations will be made and/or when the data is anticipated to be released. The scheduling of SV observations is, however, necessarily flexible and the time needed to collect sufficient good-quality data and reduce and perform verification on it will vary considerably from target to target, so the dates given here are subject to change. Some data releases will happen more quickly than others.
The results of SV observations are made public via the ALMA Science Portal, with no waiting period, as soon as satisfactory observations and data reduction have been achieved. As of October 2012, each new release of data has been accompanied by an announcement on the ALMA Science Portal two weeks prior to release, as well as an announcement on the release date itself. Due to the nature of SV, the data released may not cover all the aspects listed below and, as already noted, some targets may not be observed at all. We also note that some data files released may be rather large so downloading and reducing them will take patience.
In order to reduce and/or analyze the ALMA data, the Common Astronomy Software Applications (CASA) package is used. Some targets are released with accompanying CASA Guides for the latest CASA version release (currently CASA 4.3). Feedback on the data and on the CASA software is always welcome.
SV data is reduced by staff members of the Joint ALMA Observatory (JAO) and the ALMA Regional Centers (ARCs), and the process is coordinated by the JAO.
Science Verification Targets
Five lists of SV targets are shown below. They are those which demonstrate:
- Proposed (2016) Band 5 SV targets
- The high angular resolution capability during the 2014 ALMA Long Baseline Campaign (Table 1)
- Cycle 2 capabilities (Table 2)
- Cycle 1 capabilities (Table 3)
- Capabilities prior to June 2012 (i.e. pre-Cycle 1) and publicly released by October 2012 (Table 4).
In these tables we list the targets, positions, observing bands and the currently anticipated timeline for the observations and/or data release. They are listed by the principal capability being demonstrated, but may suitable for verification of more than one capability. Click on the links for more details. Once released, a link to the dataset on the ALMA Science Portal will appear in the Status column.
Note that we did not expect to observe all the targets in Table 3. In most cases several possible objects were listed that were suitable for demonstrating a capability because we could not be sure which target would be available at the time when we were ready to make the observations. During the Cycle 1 period, commissioning of additional capabilities continued, so some observations of these objects will have included post-Cycle 1 capabilities such as longer baselines, more advanced observing modes and new receiver bands.
The targets that were listed at the time of the Cycle 0 Call for Proposals and were used to demonstrate the early (pre-Cycle 1) capabilities of ALMA and publicly released (up to October 2012) are shown in Table 4. They are organized by the main capability that was being verified (though obviously in many cases several capabilities were involved). All observations were carried out with the 12-m array. In some cases these targets were observed before 16 antennas were available and while many of the subsystems were still being tested, so they should not be construed to represent the quality of the data that can be expected from the system as it is today. They are provided here as a means for the user to become acquainted with the ALMA data structure, observing strategies and reduction techniques. Given that the data were taken during the construction phase, there may be more idiosyncrasies present than will be expected during full operations, so we ask the user to please review carefully the CASA guides provided with the datasets that represent unique observing modes or strategies.
The more recent SV targets (Table 1 / Table 2) are part of more focused campaigns than those that resulted in previous SV targets, and so the target lists are more concise. The addition of new planned SV targets is now announced via a news item on the ALMA Science Portal. In the specific case that targets of opportunity such as comets or transients are needed to test specific capabilities, data may be taken that is identical in purpose to SV but in this case there may be no opportunity to go through the normal process of announcing the targets in advance. However, the procedure for releasing target of opportunity SV data is identical to all other SV data. We will continue to confirm our ability to observe transient sources, so we do appreciate further suggestions for targets appropriate for those observations.
Capabilities demonstrated in 2016
The Band 5 receivers
- Arp 220: observations of the H2O line to demonstrate the ability to calibrate the water line across a broad frequency range in the wings of the atmospheric absorption at 183GHz.
- Sgr B2: spectral scan between 160 and 211 GHz to demonstrate the new science capabilities for molecular line surveys and chemistry using Band 5 and to demonstrate the ability of obtaining and properly calibrating spectral scans across the varying atmospheric absorption.
- U Her, W Hya, or VY CMa to demonstrate the ability for spectral line polarization observations of the water line in Band 5.
The listes SV targets are announced in ALMA Science Portal on 29 June New Targets for Science Verification: Band 5 news article.
The 2014 ALMA Long Baseline Campaign
The 2014 ALMA Long Baseline Campaign took place from 1 September to 30 November 2014. Its aim was to verify and demonstrate the high angular resolution capability of ALMA on baselines up to ~10 km, as usual by carrying out a set of SV observations specifically chosen to demonstrate the high angular resolution capability. The five SV targets are listed in Table 1. The scope of the campaign originally included demonstrating capabilities at frequencies only up to Band 6. However, due to the successful progress of the campaign we expanded the scope of the capabilities demonstrated to include Band 7 for two of the five targets. Details can be found in the News Items on the ALMA Science Portal.
We are currently completing the data reduction and verification of the Long Baseline SV targets. Once the verification and validation process is complete, the SV data and CASA guides/scripts will be publicly released on the ALMA Science Portal.
The release date of 17 February 2015 for four of the targets (see Table 1) was announced on the ALMA Science Portal on 2 February. The final installment of the data release will also be announced on the ALMA Science Portal two weeks in advance.
|ephemeris||6; continuum||12-m||Ephemeris target||
Released on 17 February 2015.
|Mira||02:19:20.79 -02:58:39.5||3 & 6; SiO lines & continuum||12-m||Released on 17 February 2015.|
|HL Tau||04:31:38.45 +18:13:59.0||3, 6 & 7; CO line & continuum||12-m||
Released on 17 February 2015.
|3C 138||05:21:09.9 +16:38:22||3 & 6; continuum||12-m||Polarization target||Observed. Data taking complete. Data reduction and verification underway. Data release anticipated for early March 2015.|
|SDP.81 (HATLAS J090311.6+003906)||09:03:11.61 +00:39:06.7||4, 6 & 7; CO line, H2O line, continuum||12-m||
Released on 17 February 2015.
Long Baseline Target - Juno
- Juno is an asteroid that is often used for ALMA calibrations.
- Band 6 spectral setup: four TDM (15.6 MHz channel width) spectral windows centred at 224.0, 226.0, 240.0 and 242.0 GHz.
- Capability demonstrated: 10km baselines in Band 6. The long baseline capability for ephemeris targets.
Long Baseline Target - Mira
- Mira (J2000 02:19:20.79 -02:58:39.5). Mira is a well-studied AGB star and is a prototypical Mira variable.
- Band 3 spectral setup: three TDM (15.6 MHz channel width) spectral windows centred at 100.21, 98.25 and 88.20 GHz. Four 61 kHz channel width spectral windows centred at 85.60, 85.76, 86.24 and 86.85 GHz.
- Band 6 spectral setup: one TDM (15.6 MHz channel width) spectral window centred at 229.6 GHz, four 0.122 MHz channel width spectral windows centred at 214.38, 214.09, 215.59 and 217.10 GHz, two 61 kHz channel width spectral windows centred at 232.69 and 231.90 GHz.
- Capability demonstrated: 10km baselines in Bands 3 & 6. Demonstration of similar-resolution (~0.06'' for a ~10km baseline configuration) imaging at two frequencies (Bands 3 and 6) in a mixed mode spectral setup (demonstrating both spectral line and continuum). With a proper motion of 240 mas/yr it will also provide an astrometric test.
- Example references: Ramstedt et al., 2014, arXiv:1410.1529; Cotton et al, 2008, A&A, 477, 853; Reid & Menten, 2007, ApJ, 671, 2068
Long Baseline Target - HL Tau
- HL Tau (J2000 04:31:38.45 +18:13:59.0). HL Tau is a young star with a circumstellar disk.
- Band 3 CO/CN spectral setup: two 61 kHz channel width spectral windows centred at 115.26 and 113.49 GHz and two TDM (15.6 MHz channel width) spectral windows centred at 102.91 and 101.10 GHz.
- Band 3 HCN/HCO+ spectral setup: two 61 kHz channel width spectral windows centred at 88.63 and 89.19 GHz and three TDM (15.6 MHz channel width) spectral windows centred at 90.8, 100.8 and 102.8 GHz.
- Band 6 spectral setup: four TDM (15.6 MHz channel width) spectral windows centred at 224.0 ,226.0, 240.0, 242.0 GHz.
- Band 7 spectral setup: four TDM (15.6 MHz channel width) spectral windows centred at 336.5, 338.4, 348.5, 350.5 GHz.
- Capability demonstrated: 10km baselines in Bands 3, 6 & 7 (continuum and spectral line). Demonstration of similar-resolution (~0.06'' for a ~10km baseline configuration) imaging at two frequencies (Bands 3 and 6). Higher frequency (Band 7) observations at long baselines.
- Example references: Greaves et al. 2008, MNRAS, 391, 74; Kwon et al. 2011, ApJ, 741.
Long Baseline Target - 3C 138
- 3C 138 (J2000 05:21:09.9 +16:38:22). 3C 138 is a strongly polarized (12%) quasar.
- Spectral setup: continuum spectral windows centred at the standard Band 3 & 6 frequencies.
- Capability demonstrated: 10km baselines in Bands 3 & 6. Demonstration of similar-resolution (~0.06'' for a ~10km baseline configuration) imaging at two frequencies (Bands 3 and 6) in a continuum spectral setup.
- Example references: Akujor et al, 1993, A&A, 274, 752.
Long Baseline Target - SDP.81
- SDP.81 / HATLAS J090311.6+003906 (J2000 09:03:11.61 +00:39:06.7) is a high-z (z~3), lensed, ultra-luminous starburst galaxy first detected in the Herschel-ATLAS Survey (Eales et al. 2010, PASP, 122, 499).
- Band 4 spectral setup: one 0.976 MHz channel width spectral window centred at 142.641 GHz to cover the redshifted CO J=5-4 line (576.267 GHz rest frequency) plus three TDM (15.6 MHz channel width) spectral windows for continuum centred at 144.6, 154.7 and 156.4 GHz.
- Band 6 spectral setup: one 0.976 or 1.95 MHz spectral window centred at 244.5 GHz to cover the redshifted H2O line (987.9 GHz rest frequency). Three TDM (15.6 MHz channel width) spectral windows centred at 228.0, 230.0 and 243.0 GHz, which includes the redshifted CO J=8-7 line (921. 8 GHz rest frequency).
- Band 7 spectral setup: three TDM (15.6 MHz channel width) spectral windows centred at 282.935, 294.935 and 296.935 GHz and one 1.95 MHz channel width spectral window centred at 284.872 GHz. This setup covers the redshifted CO J=10-9 line (1151.985 GHz rest frequency).
- Capability demonstrated: 10km baselines in Bands 4, 6 & 7. Demonstration of similar-resolution imaging at two frequencies (Bands 4 and 6) in a mixed mode spectral setup (demonstrating both spectral line and continuum.) Higher frequency (Band 7) observations at long baselines.
- Example references: Omont et al. 2013, A&A, 551, 115.
|3c286||13:31:08.3 +30:30:33.0||6, continuum||12-m||Polarization||Observed. Data reduction and verification underway. Data release will include a CASA guide.
|IRAS16293-2422||16:32:23 -24:28:36||4, CH3CN, CH3OH||12-m||Observed. Data reduction and verification underway.|
|NGC3256||10:27:51 -43:54:18||8, CO(4-3)||12-m||Observed. Data reduction and verification underway.|
|Comet C/2012 F6 (Lemmon)||ephemeris||6, HCN(3-2)||12-m||Ephemeris target with spectral line emission||Released|
|High Angular Resolution|
|Water masers||multiple||7,9||12-m||Long Baselines||VY CMa data released|
|G34.26+0.15||18:53:18.6 +01:14:58||9||12-m+ACA||Long Baselines||Not observed. We no longer plan to observe this target.|
|Mars||ephemeris||3,6,7||12-m+ACA+SD||Observed. Assessment of data quality and release potential under investigation.|
|Comet Garradd||ephemeris||3||12-m||Not observed. We no longer plan to observe this target.|
|CB54||07:04:20.9 -16:23:20||7||12-m||Software R9.1.1||Not observed. We no longer plan to observe this target.|
|Imaging extended structure|
|VV114/IC1623||01:07:47.2 -17:30:25||9||12-m+ACA||Observed. This target was observed but no data of suitable quality for release was obtained.|
|RXCJ1347-1145||13:47:31.2 -11:45:15||3||12-m+ACA||Initial observations carried out. Assessment of data quality and release potential under investigation.|
|Lambda Orionis||05:31:22 +12:05:00||3||12-m+ACA+SD||Software R9.1.1||Not observed. We no longer plan to observe this target.|
|HR3126/IC2220||07:56:50.9 -59:07:33||3||12-m+ACA+SD||Software R9.1.1||Not observed. We no longer plan to observe this target.|
|M16 (The Eagle Nebula)||18:18:52.7 -13:50:09||6||12-m+ACA+SD||Software R9.1.1||Initial observations carried out. Assessment of data quality and release potential under investigation.|
|NGC1512/10||04:03:54.3 -43:20:56||3||12-m+ACA+SD||Software R9.1.1||Observed. This target was observed but no data of suitable quality for release was obtained.|
|Fornax Cluster||multiple, ~RA 03h Dec -35||3||12-m+ACA||Software R9.1.1||Not observed. We no longer plan to observe this target.|
|Chamaeleon||multiple, ~RA 11h Dec -77||6||12-m||Software R9.1.1||Not observed. We no longer plan to observe this target.|
|Target||Coordinates (J2000)||Band||Transition||Corr Res||Cycle||Status||Reference||CASA guide?|
|Io Atmosphere||ephemeris||7||SO, SO2 (B7)||FDM||0 / 1||Observed||Moullet et al. 2010|
|Uranus, Neptune||ephemeris||6,7||CO,H2S,HCN, continuum||FDM||0 / 1||Observed (Uranus Band 7)||Hofstadter et al. 2009, Moullet et al (in prep)|
|Centaurus A||13:25:27 -43:01:08||6,7||CO(2-1), (3-2)||FDM||0 / 1||Released (band 6)||Espada et al (2009, 2010)|
|M100 / NGC 5247||multiple, 12-13h||3||CO(1-0)||TDM||0 / 1||Released (M100)||Helfer et al. 2003||See .|
|NGC 4038/9||12:01:53 -18:52:38||3,6,7||CO(1-0), (2-1), (3-2)||FDM||0 / 1||Released||Ueda et al (in prep), Wilson et al 2000||Band 7 CASA guide|
|High Spectral Resolution|
|TW Hydra||11:01:51 -34:42:17||3,6,7,9||CO, HCO+, HCN etc||FDM||0||Released (bands 3,6,7)||Qi et al. 2004, 2006, 2008; Hughes et al. 2011||Band 7 CASA guide|
|IRAS16293-2422||16:32:23 -24:28:36||6,9||multiple||FDM||0 / 1||Released||Bisschop et al. 2008, Yeh et al. 2008, Jorgensen et al. 2011||Band 9 CASA guide|
|High Resolution Spectral Survey|
|Orion (BN/KL and OMC1)||05:35:14 -05:22:23||3,6,7,9||spectral survey||FDM||0 / 1||Released||Beuther et al. (2004, 2005,2006) Wright et al. 1996, Blake et al. 1995|
|Low Spectral Resolution|
|NGC3256||10:27:51 -43:54:18||3||CO(1-0)||TDM||0 / 1||Released||Sakamoto, Ho & Peck, 2006||Band 3 CASA guide|
|HD163296||17:56:21 -21:57:22||6,7,9||multiple||FDM||0||Released (bands 3,6)||Hughes et al. (2008, 2011), Qi et al (2006, 2011)|
|Band 9 Imaging|
|Arp220||15:34:57 23:30:11||6,7,9||CO(2-1), (3-2), (6-5)||FDM||0 / 1||Observed||Sakamoto et al (1999, 2008, 2009), Matsushita et al 2009, Martin et al 2010|
|BR1202-0725||12:05:23 -07:42:32||7||[C II]||TDM||0 / 1||Released||Iono et al 2006|
|Sgr A*||17:45:40 -29:00:28||3,6||recomb. lines, 12CO(2-1)||FDM||0||Released (bands 3,6)||Kunneriath et al 2011, Zhao et al 2010|
|Not observed: these were on the original list but not observed (we no longer plan to observe them)|
|Lensed SMGs J1/J2||14:01:05 02:52:23||7||CO(3-2), cont.||TDM||0 / 1||Not observed||Frayer et al (1998, 1999), Weiss et al. 2009, Wardlow et al. 2010|
|NGC6334I||17:20:53 -35:46:58||3||CO, HCO+, HCN||FDM||0||Not observed||Hunter et al. 2006, Beuther et al. 2007|
|HH114mms||05:18:15 07:12:00||3||0||Not observed|
|R CrA Cloud Core||19:01:53 -36:57:21||?||HCO+(3-2)||FDM||0||Not observed||Groppi et al 2007, Chen and Arce, 2010|
|HD 107146||12:19:07 16:32:54||6,7||cont.||TDM||0||Not observed||Corder et al. 2009, Hughes et al. (in prep)|
 The observations were made in single-field interferometry or small mosaic modes, generally using dual polarization with 4 basebands of 1.875 GHz each with either high spectral resolution (FDM - ~1 MHz) or low spectral resolution (TDM - ~30 MHz). For some objects a higher resolution mode was used to match that of the comparison data.
 This target was observed but no data suitable for release was obtained.
 Additional, ACA and Single-Dish, observations and data reduction underway. A CASA guide describing the combination of ACA and SD data with the already released 12-m array data is in preparation. Data/guide release anticipated for January 2014.
Capabilities Demonstrated - Cycle 2
Demonstrate the Cycle 2 continuum polarization (TDM) capability, using the 12-m array, which is the only polarization observing mode offered for Cycle 2.
Capabilities Demonstrated - Cycle 1
Verify that coherence is maintained on ~1km baselines (compared to the ~400m baselines in Cycle 0) and that the calibration techniques, particularly phase correction, are working properly on these longer baselines. Bright compact sources are needed for this.
Demonstrate that the special steps required to observe and reduce the data on objects that move in RA and Dec work correctly in all cases, including both those objects that use the built-in ephemeris, e.g. planets and major moons, and those for which a special ephemeris has to be uploaded, e.g. comets. Also demonstrate that the Doppler corrections for such objects are made correctly. It is important that these are end-to-end observations as problems can arise in the data-capture and data-processing parts of the procedure. Dynamical selection of phase calibrators is also required since the objects move.
Cycle 1 capabilities include cases where the different basebands are used with different spectral modes (TDM/FDM) or different resolutions. The end-to-end process is more complicated than in Cycle 0, involving changes to the Observing Tool, the control of observations and the data reduction. In addition to making sure that the spectral setups are correct and that the data comes out of the final data reduction process with the correct frequency scales for the full range of possible combinations, we needed to demonstrate that the calibration processes made with different spectral resolutions are all applied correctly. An important additional capability is the use of spectral averaging which will make the data volume much smaller in many cases but again introduces many additional steps in the end-to-end observing process which need to be verified. We need to observe sources with a range of different spectral characteristics for this.
This is the most critical and complicated of the new capabilities for Cycle 1. We have to take well-matched data with the ACA and the 12m array and then combine these with the correct scaling and weighting into a single image (or rather spectral-image-cube). When single-dish measurements are included in the ACA observations we have to take and reduce those data in entirely different ways from the interferometric data and then bring the two data sets together. To establish that these processes are working correctly requires that we observe extended objects where we know exactly what the images should look like - e.g. planets - as well as spectral lines in objects with complex structure to show that we can image faint diffuse structures correctly even in the presence of bright compact features.
Band 9 observations will be particularly challenging because, in addition to the usual problems of getting good quality data and calibrating it at such high frequencies, the single-dish data requires a special observing technique to separate the signals from the two sidebands.
In Cycle 1 up to 15 sources could be included in an individual science goal. This can be done with the standard observing procedure so what needed to be verified was the end-to-end process including the generation of the scheduling blocks and data analysis.
For Cycle 1 we plan to make mosaics of up to 150 "pointings". To provide adequate UV coverage we need to get round all the points in about 30 minutes, and this has to include sufficient phase and amplitude calibrations. This requires a time per point of 10 seconds at the maximum, which in turn implies reducing the inter-subscan time to no more than 4 seconds, which will give us ~50% of time on source, i.e. ~29% loss of sensitivity to overheads, as well as reducing the latency of the calibration measurements. We needed to demonstrate that the software enhancements to do this are all working correctly and that the end-to-end process including the creation of the grid of pointings and the reduction of the large images can be done in a practical manner.