Next: Correcting Some DRAO Synthesis Telescope Wide-Field Imaging Problems
Previous: Reduction of Bidimensional Spectral Data Obtained with the Integral Field Spectrographs of the 6-m Telescope
Table of Contents --- Search ---
PS reprint
P. T. Wallace
Starlink Project, Rutherford Appleton Laboratory, Chilton
Didcot, Oxfordshire OX11 0QX, United Kingdom
New techniques such as VLBI, observations from space and lunar laser ranging have brought about rapid advances in fundamental astronomy. The existing mechanisms for agreeing and promulgating revised constants and computational procedures were never intended to cope with the present flow of results. Fortunately, new methods of information exchange, in particular through global computer networks, are becoming available, and might offer more responsive alternatives to traditional channels such as IAU Bulletins and paper journals.
A related problem is the simultaneous need for (a) an interlocking and self-consistent set of canonical constants and procedures and (b) the current best estimates of individual constants and procedures such as planetary masses and precession/nutation models.
Against this background, the Subgroup on Standard Procedures (SGSP) of the IAU Working Group on Astronomical Standards (IAU/WGAS) was in 1991 asked to prepare a report on standard procedures needed in fundamental astronomy, which (a) should have a maximum degree of compatibility with the International Earth Rotation Service (IERS) standards, (b) should include the implementations of procedures in the form of tested software and/or test cases and (c) should be available both in written and machine-readable form (IAU 1991). After preliminary discussions, the Subgroup issued a questionnaire (IAU 1992), followed by a campaign to compare five independently-developed implementations of the IAU precession formula. The WGAS concluded that it was both desirable and feasible to establish a mechanism to provide such procedures to the astronomical community, named it ``SOFA'' (Standards Of Fundamental Astronomy) and made detailed recommendations to the IAU describing how the SOFA service should be managed and operated. This led to a formal IAU resolution (IAU 1994) which recommended that:
The service will emanate from the IAU/SOFA Distribution Center, the location of which has yet to be determined, and will be operated according to advice from three panels:
The present status of the scheme is that membership of the Reviewing Board is being decided, and preparations are being made to allow potential sites for the SOFA Distribution Center to bid for this role.
Although the Distribution Center will be able to supply the SOFA products in hardcopy form or on magnetic/optical media (the recipient paying handling and shipping costs), the principal and preferred mechanism will be via the Internet. At least three methods of obtaining information electronically will be supported:
Overall, the emphasis will be on automatic operation, with minimal load on the staff operating the SOFA service.
The intention is for SOFA to offer a group of relatively small standard procedures, rather than any large, all-embracing software packages. Each standard procedure will correspond to a specific computation in fundamental astronomy, and will be canonical in nature, uncolored by purely practical issues such as efficiency, user-interfaces or input/output. Making full use of machine precision will, however, be a consideration.
The great care with which the procedures will have to be designed and evaluated means that rapid progress cannot be expected, and it is likely to take several years to provide even this basic set:
Although one of SOFA's key roles will be to provide access to canonical procedures, many users of the service will be looking for practical software that they can use directly or build into application programs. Indeed, there is a large audience for whom the fine distinction between a canonical procedure and a merely accurate and reliable one is of no importance, as long as whatever is supplied can be used without undue difficulty.
In general, there are two types of software which SOFA needs to address: (i) individual functions (subprograms, subroutines, etc.) and (ii) executable programs offering multiple facilities. Items of Type (i) form the basis of SOFA, but it is less clear that it is appropriate to include items of Type (ii), especially early on. Whatever software is provided, users will expect the following characteristics (in no particular order):
IAU 1991, Resolution A4, Recommendation VIII, Item 3
IAU 1992, WGAS Circular 51.2.1
IAU 1994, Resolution B11
Wallace, P. T. 1994, in Astronomical Data Analysis Software and Systems III, ASP Conf. Ser., Vol. 61, eds. D. R. Crabtree, R. J. Hanisch, & J. Barnes (San Francisco, ASP), p. , 481