In the first 5 cycles of ALMA operations, data for more than 1800 projects have been calibrated and manually imaged for quality assurance purposes before being delivered to the PI and being added to the ASA. However, imaging is a time-consuming process and therefore for each project the quality assessors only imaged small portions of the data to verify that the PI requirements were achieved. As a consequence, images generated with manual imaging are highly inhomogeneous and incomplete. The recently developed imaging pipeline has improved this rate, but its usage as Quality Assurance (QA) tool has been implemented only during cycle 4 and applied only since then.
The Additional Representative Images for Legacy (ARI-L, Massardi et al. 2021) in the ALMA Science Archive is an ALMA Development project approved by JAO in December 2018 for execution and based on the results of a feasibility study approved by ESO and reported in Massardi et al. (2019). The project aim is to increase the legacy value of the ASA by bringing the reduction level of ALMA data from Cycles 2-4 close to the level of the more recent Cycles processed with the ALMA Imaging Pipeline.
Currently, an archive miner can choose to download for each MOUS only the products and/or the raw data. To complement or re-produce the product images, once all the data are downloaded, calibration scripts must be run with the same CASA version that the analyst used to generate them. This might become an issue for datasets of the earliest cycles. Nevertheless, the archive is considered a precious resource and is broadly used (Fig. 1 and Stoehr, 2017), so that more than 30% of the publications with ALMA data exploited the ASA resources.
The ARI-L complete set of imaging products will be highly relevant for all science-cases and enhance the possibilities of exploitation of archival data. Indeed, images for all the datasets will:
facilitate the archive access and the data usage for science purposes also to non-expert data-miners,
provide a homogeneous view of all data to compare datasets and to do a more conscious download selection,
allow to more profitably link the archive to several tools of visualization and analysis (e.g. VO, CARTA, ADMIT, KAFE, ...)
allow to generate previews similar to those planned for the following cycles.
The project aimed to produce and ingest into the ASA a uniform set of full data cubes and continuum images covering at least 70% of the data from Cycles 2-4.
The project was officially expected to finish in June 2022. The main goal was reached in January 2022 ahead of schedule and, on the consideration of its success, a six-months no-cost extension was accorded to increase the fraction of delivered datasets to even higher rates.
ARI-L produced and ingested into the ASA a uniform set of full data cubes and continuum images covering at 86% of the data from Cycles 2-4 processable with the ALMA pipeline. These cubes complement the more limited QA2-generated image products which cover only a small fraction (< 10%) of the observed data for those cycles.
The project was leaded by the Italian node of the European ALMA Regional Centre (ARC) in collaboration with staff from the ESO and UK European ARC nodes.
The ARI-L images were generated, certified for quality and ingested into the archive as external products. The workflow and rocessing and quality assurance procedures are described in the ARI-L project website and papers (Massardi et al. 2021, and Massardi et al. 2022).
ARI-L image, README and Weblog files can be queried in the ALMA Science Archive through the “ari_l” label in the “Collection” column, visualised in a dedicated preview and downloaded as single file among the "Externally delivered products" in the Request Handler system.
The calibrated Measurement Sets for all the ARI-L delivered MOUS are stored at the Italian Astronomical Archive (INAF-IA2) facilities and are publicly available.They can be requested via e-mail to help-desk@alma.inaf.it .
Comprehensive reviews of science cases that could benefit from similar images in the archive are in the ARI study report (Massardi et et al., 2019), in the KAFE paper (Burkutean et al., 2018), in the ALMACAL papers (Bonato et al., 2018) and in the ARI-L collaboration papers (Massardi et al. 2021).
For more information on the ARI-L activities please see the ARI-L documents.
(Left panel) Fraction of the ALMA publications that make use of either Science Verification data (green line) and PI-only (orange line). The fraction complementary to all data (blue line) exploits archived non-proprietary data. (Right panel) Cumulative data flow into the archive (green line) and out of the archive (blue line) in TB. (Stoehr et al. 2017)
Snapshots of the ASA query interface showing (top) a search for projects in Cycle 4 (i.e., project code 2016*), with the “ari_l” flags in the “Collections” column, and (bottom) the download interface listing the ARI-L products as “External”.
Note that single ARI-L images can be accessed with the remote CARTA viewer available for all the ASA images.
Snapshots of the ASA query interface showing the interactive previews for an MOUS in the project 2015.1.01151.S comparing the CH3OH line in the galaxy NGC 4945 in the QA2 image (panel A) and in the ARI-L image (panel B)
The spectral line forest as published by the PI (top panel, Beltrán et al. , 2014 ), for the high-mass star-forming region G35.03+0.35, compared with the QA2 products available in the ALMA archive (bottom panel).
To illustrate the improvement in the archive user’s experience provided by the ARI-L efforts, we show below the line forest that could be obtained from the ARI-L products in a comparable frequency range for a high-mass star-forming region.
Light curve for the blazar PKS J0635-7516 in ALMA band 3 and 6. This source is an ALMA calibrator and its image is not available to date in most of the products for the hundreds of MOUS in which it has been observed. This image was generated with the Key-analysis Automated FITS-images Explorer (KAFE, Burkutean et al. 2018) tool loading FITS images for hundreds of ALMA projects including the source as calibrator Liuzzo & et al. , 2019 ; Bonato et al. , 2018 .
Zeroth moments of the brightness distribution (corresponding to integrated brightness) of the CO(5-4) line for the dusty obscured galaxy XID42 (D’Amato et al. 2020) obtained exploiting the QA2 (top left) and the rebinned ARI-L (center) products for MOUS uid://A001/X340/X2c of the project 2015.1.01205.S. For comparison we present also the difference between the zeroth moments (top right), and the comparison of the line spectra.
Images and spectra ingested into the ASA for the project 2016.1.00875.S MOUS uid://A001/X88f/X11a (a spectral scan of a region of the Central Molecular Zone of the Milky Way). The first two images show the QA2 and ARI-L products for the same channel, which is indicated by the green arrow in the two spectra shown below. The image on the right shows the structure associated with the bright spectral feature indicated by the red arrow in the ARI-L cube spectrum; this is absent in the QA2 cubes.