I discuss in this post some ideas for how Web 2.0 can be used in professional astronomy, and the example I will used is that of building collaborative astronomical image mosaics from a collection of input images with the Montage mosaic engine.
One simple idea that is already being used is a wiki/blog for discussion of how to use the Montage engine, successes and failures that users have, and user feedback that can be used to improve Montage. The interactive website Astrobetter offers astronomers the opportunity to share tools and expertise, and image mosaic creation is one of the topics that has been discussed on this site. The comments have in fact led directly to improvements in the organization and content of the Montage web site. They are also leading Montage to deploy a dedicated user forum for users to share innovative uses of the Montage toolkit, workarounds to problems and identify defects.
Scientists in many fields have generally not taken to Web 2.0 as a collaborative tool and as a means of sharing data. A research report by the Research Information Network, a British policy unit, is to our knowledge the only detailed study of why this is the case in the fields of social, physical and biological sciences. The primary barrier is apparently a lack of understanding of how to get started and what benefits will accrue to scientists (See also my blogpost “25 Things for Researchers and Social Media!”. Despite this lack of take-up, there are many science and outreach projects that can be done effectively with Web 2.0. Some examples are described below; they illustrate the unrealized potential of Web 2.0 in astronomy, a discipline increasingly dominated by large, multi-institutional (and, often, multi-national) research consortia, which in theory should make it an ideal domain for Web 2.0.
One project would be building a collection of image mosaics produced by many users. This could be done at many levels. For example, the Montage project could host mosaics produced by others. This seems particularly useful in the case of images that are produced for outreach. Simple outreach mosaics (or science mosaics) could also be placed on Flickr. Or users could post their own mosaics on servers that they choose. In any of these cases, searching or browsing of these images also need to be considered. If all images were on one site, they would be easy to browse. If they are on multiple sites, a registry (or a federation of registries) is needed. Additionally, image tags are needed for searching. FITS files already have metadata incorporated, but the standardization of this metadata for images is could be handled by a standard that is under development for this purpose. This tagging could be manual or automated. The registry could also have an RSS feed or use twitter to allow interested groups to see new mosaics that had been registered.
Web 2.0 also potentially could allow collaborative building of atlases from surveys. For example, using the virtual data concept (Foster et al. 2002), once an atlas is defined (meaning a definition of a set of image plates from one or more survey bands that would be useful to a community of users, including a projection of the plates, and the boundaries of the individual plates), a library could be built with pages for each plate. However, each plate would not be built until a user wanted it. A request for a plate that had not yet been built would launch a run of Montage. This could either be done for public data, with the atlas being publicly accessible, or it could be done only within a collaboration, with only the collaborators having access to the atlas initially. This could be done by Montage or others, or it could be done as a service that would provide overlays over Google Sky, for example. Having such atlases available might be helpful to researchers who want to compare new data against some data of known quality and provenance.
Using the mashup concept, Montage components could be turned into services that could be exposed to users with inputs and outputs named to make them clearly understandable, perhaps as Google gadgets. This would require the allow users to compose services as they choose, including mashing them up with other non-Montage services, though the cost of running these services would have to borne by someone, perhaps Montage or an infrastructure such as a TeraGrid science gateway or a hosted commercial service such as Amazon EC2 or Windows Azure.
By the early 2000s, many scientists were designing unique user interfaces and tools to access compute and data resources. The TeraGrid Science Gateways program (Wilkins-Diehr et al. 2008) started in late 2004 to try to encourage this to continue, with common tools being developed for the scientists and communities who were building the gateways, and TeraGrid resources being used for the compute and storage parts. This was originally done using mostly Web 1.0 technologies, but currently, the TeraGrid Science Gateways program is investigating how to integrate Web 2.0 technologies as well.
Foster, I., J. Voeckler, M. Wilde, and Y. Zhao. Chimera: A Virtual Data System for Representing, Querying, and Automating Data Derivation. Proceedings of 14th International Conference on Scientific and Statistical Database Management (SSDBM’02), 2002.
Wilkins-Diehr, N., D. Gannon, G. Klimeck, S. Oster, and S. Pamidighantam. TeraGrid science gateways and their impact on science. Computer, 41(11):32-41, 2008.
This post is excerpted from the paper Collaborative Astronomical Image Mosaics by Daniel D. Katz, G. Bruce Berriman, and Robert G. Mann. To be published in “Reshaping Research and Development using Web 2.0-based technologies.” Mark Baker, ed. Nova Science Publishers, Inc. (2011). I wish to thank my collaborators for permission to post material in this blog.