In Section 1.4 we stated that astrometry must be developed within an extragalactic reference frame to microarcsecond accuracies. The objective of the present chapter is to provide the theoretical and practical background of this basic concept.

International Celestial Reference System (ICRS)

A reference system is the underlying theoretical concept for the construction of a reference frame. In an ideal kinematic reference system it is assumed that the Universe does not rotate. The theoretical background was presented in Section 5.4.1. The reference system requires the identification of a physical system and its characteristics, or parameters, which are determined from observations and that can be used to define the reference system. In 1991 the International Astronomical Union agreed, in principle, to change to a fundamental reference system based on distant extragalactic radio sources, in place of nearby bright optical stars (IAU, 1992; IAU, 1998; IAU, 2001). The distances of extragalactic radio sources are so large that motions of selected objects, and changes in their source structure, should not contribute to apparent temporal positional changes greater than a few microarcseconds. Thus, positions of these objects should be able to define a quasi-inertial reference frame that is purely kinematic. A Working Group was established to determine a catalog of sources to define this frame that is now called the ICRF.

Astrometry is positional astronomy. It encompasses all that is necessary to provide the positions and motions of celestial bodies. This includes observational techniques, instrumentation, processing and analysis of observational data, positions and motions of the bodies, reference frames, and the resulting astronomical phenomena.

The practical side of astrometry is complemented by a number of theoretical aspects, which relate the observations to laws of physics and to the distribution of matter, or celestial bodies, in space. Among the most important are celestial mechanics, optics, theory of time and space references (particularly with regards to general relativity), astrophysics, and statistical inference theory. These scientific domains all contribute to the reduction procedures which transform the observed raw data acquired by the instruments into quantities that are useable for the physical interpretation of the observed phenomena. The goals of this book are to present the theoretical bases of astrometry and the main features of the reduction procedures, as well as to give examples of their application.

Astrometry is fundamental to, and the basis for, all other fields of astronomy. At minimal accuracy levels the pointing of telescopes depends on astrometry. The cycle of days, the calendar, religious cycles and holidays are based on astrometry. Navigation and guidance systems are based on astrometry, previously for nautical purposes and now primarily for space navigation.

Astronomy and astrophysics are also strongly dependent on astrometry.

The apparent direction in the sky at which a celestial object appears is not the actual direction from which the light was emitted. What is observed is the tangent direction of the light when it reaches the observer. For reasons that will be discussed in this chapter, the light path is not rectilinear and several corrections describing the effects of bending, or shifts in direction, are to be applied to the direction from which the light is observed to determine the actual direction of the emission. We shall not deal here with the various transformations undergone by the light within the observing instrument; they are particular to each case. Some examples are given in Chapter 14. We shall consider only the direction from which the light came when it entered into the instrument. One has to consider the atmospheric refraction, the shift in direction due to the combination of the speed of light with the motion of the observer, called aberration, and the bending of light in the presence of gravitational fields. The latter has been already presented in Section 5.4.2, but will be revisited in Section 6.4. Similarly, the geodesic precession and nutation are to be considered when relating the positions from a moving reference frame of fixed orientation to a fixed reference frame of the same orientation (see Section 7.5).

For many years, the theory of relativity was ignored for astrometry because the effects were much smaller than the accuracies being achieved. For the motions of bodies of the Solar System, Newtonian theory was adequate, provided that some relativistic corrections were introduced. Actually, there was a long period of questioning whether the theory of relativity was correct or not, and thus relativistic corrections were introduced in a manner to determine whether observational data could then be better represented.

However, over the past 30 years, there has been a very significant improvement in the accuracies of observations, all confirming the conclusions of general relativity. So its introduction became not only acceptable, but necessary. In 1976, the International Astronomical Union introduced relativistic concepts of time and the transformations between various time scales and reference systems. In 1991, it extended them to reference frames and to astrometric quantities. Now, with plans for microarcsecond astrometry and with time standards approaching accuraries of one part in 10-16, and better in the future, it is necessary to base all astrometry, reference systems, ephemerides, and observational reduction procedures on consistent relativistic grounds. This means that relativity must be accepted in its entirety, and that concepts, as well as practical problems, must be approached from a relativistic point of view. In 2000, the IAU enforced this approach further by extending the 1991 models for future requirements in such a way that they become valid to accuracies several orders better than those currently achieved in 2000.

Until 1998 the reference frame was defined in terms of the Solar System and the dynamics induced by the motions of Solar System bodies and their shapes. The two reference planes were the equator of the Earth and the ecliptic, the mean plane of the Earth's orbit. The intersection of these two planes, the equinox, was the fiducial point. Since both the equator and the ecliptic move owing to solar and lunar gravitational forces on the shape of the Earth and perturbations by the planets on the Earth's motion, the equinox moves with time. Therefore, each dynamical reference frame had to be defined for a specific epoch and observations and predictions were transformed to and from standard epochs to the chosen times of observations.

We have presented, in Chapter 7, the new adopted International Celestial Reference System (ICRS) based on the extragalactic radio sources. Now, the fundamental reference frame is fixed for all times. The IAU introduced in 2000 two space fixed systems. The Barycentric Celestial Reference System (BCRS) and the Geocentric Celestial Reference System (GCRS) have been defined in terms of metric tensors and the generalized Lorentz transformation between them, which contains the acceleration of the geocenter and the gravitational potential. The ICRS is to be understood as defining the orientation of the axes of both these systems for each of the origins. The International Celestial Reference Frame, (ICRF), determined from VLBI observations is the realization of the ICRS and can similarly be geocentric and barycentric.

As a rule, an astrometric observation does not directly provide the quantity sought, but rather some function of it and of a number of other parameters. The procedure that is used to obtain it is called reduction of observations. Generally, one needs to use several observations taken at different times and conditions, and treat them together. Astrometric data reduction includes taking into account a variety of effects that are described in the following chapters. But it also consists of a mathematical treatment of the data, for which one generally assumes some statistical properties. Among other things, one must be aware of the fact that a numerical result has no physical significance, and is not useful, unless there is an indication of how good the numbers are. The results must be supplemented by another number, called uncertainty, which provides an interval (sometimes called error bars) within which there is a stated probability that it includes the true value. Sometimes, when several quantities are computed simultaneously, and are mutually dependent or correlated, other numbers called correlation coefficients must be determined.

This chapter presents succinctly the most commonly used statistical tools to perform the data reduction. There are many textbooks in statistics or data treatment to which one should refer in order to go deeper in the understanding of the procedures or get more information.

Astrometry is undergoing fundamental changes. The celestial reference frame based on bright optical stars is being replaced by the extragalactic reference frame observed at radio wavelengths by Very Long Baseline Interferometry (VLBI); Hipparcos has proven the capabilities of astrometry from space; photographic plates are being replaced by charge coupled devices (CCDs) as the detectors of choice; optical interferometers are replacing transit circles and astrolabes; accuracies of tenths of arcseconds are being replaced by milli- and microarcseconds; the dynamical reference frame is being replaced by a kinematic reference frame; Global Positioning System observations are changing the determinations of Earth Orientation Parameters; and the theory of general relativity is required as the basis of astrometry.

This book is designed to provide the fundamentals for the new astrometry at the milli- and microarcsecond accuracies. The intent is to start from basic principles, without approximations, and develop the fundamentals of astrometry at microarcsecond accuracies. It is based on the general theory of relativity and the concepts introduced by the International Astronomical Union (IAU) in the past decade. The book provides the definitions and applications of the International Celestial Reference System (ICRS) to astronomy and astrometry. It is also designed to provide the philosophy and concepts of astrometry for the future, the principles behind the algorithms, the reasons for astrometry and its relationships with astronomy, geophysics, planetary sciences, astrophysics, cosmology, and celestial mechanics.

The previous chapters have provided the bases for the applications to observations. This chapter will provide examples of the reduction and analysis of observational data. Software for coordinate conversions and individual effects, such as precession, nutation, proper motions, aberration, refraction, etc., is available as NOVAS at the US Naval Observatory website. A software program that performs on-line the full reduction of a CCD plate, called PRIAM, is available at the IMCCE web site of the Paris Observatory.

Observing differences, classical and new systems

In Chapters 7, 8, and 10, we introduced fixed and moving reference systems. At present, we are in a transition period, when old and new systems may be used, although it is to be expected that the new system will progressively remain the only one used, because they have important advantages. However, it is still necessary to be acquainted with both.

Referring positions to the fixed reference frame at epoch J2000.0, namely the International Celestial Reference Frame (ICRF), instead of the reference frame and positions referred to the equinox and the ecliptic and based on the FK5, improves the accuracies of the positions. The ephemerides are now on the ICRF. Astrometric observations made with respect to ICRF positions only require the corrections for proper motions of the reference stars and appropriate differential corrections to reduce the magnitudes of the plate solution parameters (Section 14.5).

What characterizes the new observational techniques is, of course, not only that they appeared recently – say in the past couple of decades – but also that they strive to reach one or two orders of magnitude better accuracy than the classical techniques. The unit for describing the new astrometric capabilities is no longer one tenth of a second of arc, but one thousandth of a second of arc, a subunit that we shall designate throughout this book by the abbreviation mas (milliarcsecond). Plans are underway to reach microarcsecond (μas) accuracies. Such a gain in precision impacts directly on all aspects of astronomy and, particularly, on the reduction procedures. One may divide these new techniques into three major groups: interferometry, timemeasuring techniques, and space astrometry.

New detectors

Many astrometric techniques take advantage of the development of much more sensitive new detectors. Among them, one must mention the following.

• The CCD detector. The CCD (for charge coupled device) has become a major tool in astronomy since the 1970s (Monet, 1988). It consists of a semiconductor device, where a photoelectric effect takes place when light reaches it, producing an electronic image. This image is preserved by arrays of small positive electrodes, which attract photoelectrons and keep them in a similar array of potential wells, providing the possibility of long exposure times by adding photoelectrons in the same pack (Kovalevsky, 2002). […]

In contrast with celestial reference frames, which have existed since the time when catalogs of stars over the whole sky became available, a global reference system for the positions on the Earth did not exist until direct geodesic links could be performed between regions separated by oceans and, more generally, by geographically or politically impassable barriers. So there were a number of local, geodetic coordinate systems, called datums, to which the positions of terrestrial sites were referred. They were given under the form of parameters defining the shape and the size of a reference ellipsoid, as well as its orientation with respect to some conventional features such as a mean pole and a zero meridian. The ellipsoid parameters were determined to best fit the local geoid (equipotential surface corresponding to the mean ocean level), and attached to the Earth by conventional coordinates of an initial point.

Introduction

At the beginning of the space age, positional and, later, laser or Doppler observations of satellites were used to link the individual datums and to place them in a unique terrestrial coordinate system. However, locally, countries continued (and many still do) to use their own datums for surveying and legal objectives. But for scientific purposes, this was a much too complex system, and a global terrestrial reference system had to be developed.