The International Earth Rotation Service (IERS) was established in 1987 by the International Astronomical Union (IAU) and the International Union of Geodesy and Geophysics (IUGG), and it began operation on 1 January 1988. The primary objectives of the IERS are to serve the astronomical, geodetic and geophysical communities by providing the following:

1. • The International Celestial Reference System (ICRS) and its realization, the International Celestial Reference Frame (ICRF).

2. • The International Terrestrial Reference System (ITRS) and its realization, the International Terrestrial Reference Frame (ITRF).

3. • Earth orientation parameters required to study Earth orientation variations and to transform between the ICRF and the ITRF.

4. • Geophysical data to interpret time/space variations of the ITRF with respect to the ICRF, i.e., of the Earth orientation parameters, and to model such variations.

5. • Standards, constants and models (i.e., conventions) encouraging international adherence.

This presentation primarily covers the first three IERS functions from the operational point of view.

Inclusion of Polar Motion and UT1 data (corrections) is essential to achieving the full accuracy in NASA (and other) spacecraft navigation. Tracking station locations are routinely determined to the meter-level and must be available inertially (on the rotating Earth) to equal accuracy a priori in real-time. This implies knowledge of UT1 to 2.5ms and polar motion to lm (and for practical use for the first time accurately predicted and disseminated to the navigation computer programs). This was essential from 1965 onward as we prepared for Apollo, for Apollo itself, and all subsequent missions. This paper presents the history behind the author’s “discovery” circa 1965 at JPL that UT1 and polar motion were not then included in NASA’s Orbit Determination models and programs; how this omission was being “observed” in station location determinations (and errors) from spacecraft radio tracking data; how it was rectified; the essential nature of these data in precision spacecraft navigation including Apollo; and how this became undoubtedly the highest and most critical application of these data in history. There followed widespread recognition of the many observers present and past who tirelessly and often with little visibility outside of professional astronomy made the necessary observations and calculations. As an historical paper, there is room for a few interesting anecdotes and personalities in NASA, as well as the late and much missed Dr. Markowitz then head of the USNO time service, and his successor Dr. Klock, with both of whom the author had the honor, professional benefit, and great personal pleasure of working during this time.

After one hundred years of intensive study, some aspects of polar motion remain poorly understood. This motion of the Earth’s axis of rotation with respect to its mantle and crust has been observed and studied since the creation of the International Latitude Service. Research has shown that the motion of the pole is quite complicated and distinguished by components over all time scales. Observationally, the secular, Chandler and annual components of polar motion are well documented, and various geophysical processes have been suggested as likely causes. Other components such as the approximately thirty-year motion, the high-frequency motion and the daily/sub-daily motion remain as subjects of extensive research.

Polar motion and Universal Time have been regularly determined since 1899 and 1956, respectively, at a number of observatories all over the world. Before the International Earth Rotation Service (IERS) was established in 1988, the classical astrometry instruments such as visual zenith-telescopes, PZTs, transit instruments, astrolabes etc. were used. The survey of all these instruments and the methods of observation used is described. The values of instantaneous latitude and UT0-UTC made at a set of selected observatories and based on individual star observations have been collected at the Astronomical Institute in Prague during the past years. They were recalculated using the most recent astronomical standards and the Hipparcos Catalogue, and used to determine the Earth orientation parameters (polar motion, celestial pole offsets and Universal Time). The most recent solution, based on about 4.5 million observations with 47 different instruments at 33 observatories, is described and the results of polar motion presented.

VLBI has been used for high precision EOP measurements since 1980. The typical formal error for current routine monitoring of polar motion is ~80 µas. A program for continuous EOP observations using VLBI is being incrementally implemented. There are unexplained differences between EOP values derived from simultaneously observing networks.

Results of Earth orientation parameter (EOP) determinations from NEOS-A and CORE programs reduced using the package OCCAM version 3.5 at the Institute of Applied Astronomy of Russian Academy of Science (IAA) are presented. Results of EOP determination using different reference stations, relativistic models, the IERS Standards (1992) and IERS Conventions (1996) are also presented.

The Satellite Laser Ranging (SLR) technique has been used to determine Earth Rotation Parameters (ERP) for over twenty years. Most of results contributed to the International Earth Rotation Service (IERS) are based on analysis of observations of Lageos 1 & 2 satellites collected by the global tracking network of about 40 stations. Now five analysis centers submit operational (with 2–15 days delay) solutions and about ten analysis centers contribute yearly final (up to 23 years) ERP series. Some statistics related to SLR observations and analysis are presented and analyzed. Possible problems in SLR observations and analysis and ways of its solution are discussed.

Since January 1995 the International GPS Service (IGS) has been combining and analyzing daily polar motion (PM) series, produced and submitted by seven IGS analysis centers (ACs) for the IGS Final orbit/clock combinations. Since June 30, 1996 the IGS Earth Rotation Parameter (ERP) series that accompany the IGS combined orbits, also include combined PM rates. Furthermore, since March 1997, the IGS LOD (Length of Day) solutions are based on separate combinations of AC LOD solutions calibrated and weighted according to the IERS Bulletin A definite values. Similar to AC orbit solutions, the PM solutions have improved considerably since 1995, so that currently the IGS combined and the best AC PM solutions are at or below the 0.1 mas precision level, although PM biases may exceed .1 mas. Comparisons of AC ERP and PM rate solutions with the IGS Final combined ERP series revealed signals with 7 and 14-day periods for some AC solutions.

During 1998, the IGS Final and the best AC PM rate solutions compared with Atmospheric Angular Momentum (AAM) at 0.3 mas/day (rms) with an average correlation of about 0.8 and 0.6 for the PM x and PM y rate components. The correlation varied considerably with time and frequency, though significant correlation already started from 6-day periods and reaching maxima within 10 to 50 day period bands. Most of the remaining signal in the PM rate solutions could likely be accounted for by Ocean Angular Momentum (OAM) as seen from the comparisons of combined OAM and AAM with the IGS PM series during 1995 and early 1996 when also OAM data were made available. During this period the IGS PM rates agreed with the combined OAM + AAM series with 0.3 and 0.2 mas/day (rms) for the PM x and y components and with an average correlation of about 0.8 for both PM components.

Since 1978 at the Calern Observatory-France and from 1996 at the Observatório Nacional (O.N.)-Brasil solar observations have been carried out, aiming primarily to study the solar diameter. The ensemble of these two series is noteworthy by the length of Calern’s, which spans about two solar cycles, and by the high density of O.N.’s, with typically 20 observations per day, on each side of the meridian, all year around. The precision of a single measure is 015 for the northern hemisphere observations and 02 for the southern hemisphere observations.

The characteristics of the combined series enabled detailed searches for anomalous trends (e.g., from atmospheric origin) and systematic bias (e.g., from instrumental origin). The analysis show the series robust for either. Thus there is a possiblity to investigate the variable time of transit of the solar disc across the almucantars of observation.

The apparent solar time, defined by the actual diurnal motion of the Sun, shows variations because of the varying rate of motion of the Sun in hour angle. That is, due to the inequalities in the annual motion along the ecliptic and to the inclination of the ecliptic to the equator.

In this work we detail the series characteristics and discuss the determination of the hour angle of the true Sun, as exemplified by the equation of time. The accuracy of the determination varies with the solar declination and the length of the observed apparent orbit. Typical precisions follow those for the semi-diameter determinations.

The TOCAMM (TOrino CAgliari Measuring Machine) project undertaken jointly between Torino and Cagliari Astronomical Observatories aimed to convert the old measuring machine ASCORECORD into an automatic and impersonal one. This program is intended to contribute to the link of the HIPPARCOS Catalogue to the ICRS through the determination of precise position of optical counterparts of 80 extragalactic radiosources taken from the IERS list and to investigate the astrometric accuracy of the Guide Star Catalog (version 1 and 2). The calibration test phase, carried out first at the Astronomical Observatory of Torino and after at Cagliari Observatory, where the machine has been now installed, indicate that the available positional accuracy is about 0.5 microns in both x and y coordinates.

Two long time series of polar motion were analysed with respect to a linear drift, decadal variations, Chandler wobble and annual wobble: the C01 series published by the International Earth Rotation Service (IERS) and the pole series which J. Vondrák, obtained by re-analysis of the classical astronomical observations using the HIPPARCOS reference frame (1899.7–1992.0). By a least-squares fit the linear drift of the pole, usually called ‘secular polar motion,’ was determined to 3.31 milliarcseconds/year (mas/yr) toward 76.1° West longitude. For this fit the a priori correlations within each pair of pole coordinates were taken into account, and the weighting function was calculated by estimation of empirical variance components. The decadal variations of the pole path were determined by Fourier analysis. Using a sliding window analysis, the variability of the periods, the amplitudes and the phases of the Chandler wobble and annual wobble was investigated. The variances of the results and the number of iterations needed to get a convergence in the nonlinear approach show that the new time series by Vondrák is more homogeneous and consistent than the IERS C01 series.

A continuous composite series of polar motion components extending from 1846 until now called EOP (IERS) C01 is available at the Earth Orientation Section of the Central Bureau of the IERS. This series is the basis of the IERS system. It relies on different series derived from optical astrometry until 1972 and geodetic techniques since. It is given at 0.1 year intervals (1846–1889) and 0.05 year intervals (1890-now). Its accuracy has dramatically improved from 100 mas in 1846 to about 0.2 mas at present.

Now the IERS combined solutions involve mainly the contributions of VLBI, GPS and SLR techniques. It is regularly recomputed to take advantage of the improvement of the various recent individual contributions and of the refinement of the analyses procedures.

The objective of this paper is to describe this long-term polar motion series and to present the evolution and the state of the art of the multi-technique EOP combined solutions and the predictions regularly computed at the IERS/CB.

We report anomalies in the roughness of the path of the pole of rotation at the time of the Bolivia earthquake of June 9, 1994 (Mω = 8.2) and of the Kurile Islands earthquake of October, 4, 1994 (Mω = 8.3). This finding reinforces the results of previous studies, which have indicated changes in the polar motion curvature at the time of major earthquakes.

Several decades ago Markowitz was the first to report the existence of fluctuations in the motion of the Earth’s axis with a period of about twenty-four years. This empirical term was often considered not real but only an artifact due to local effects. In this paper long-term variations and the relationship between the 30-yr Markowitz wobble and changes in the Earth’s rotational speed are briefly discussed.

Apart from the “shaking” near the epicenter that is the earthquake, a seismic event creates a permanent field of dislocation in the entire Earth. This redistribution of mass changes (slightly) the Earth’s inertia tensor; and the Earth’s rotation will change in accordance with the conservation of angular momentum. Similar to this seismic excitation of Earth rotation variations, the same mass redistribution causes (slight) changes in the Earth’s gravitational field expressible in terms of changes in the Stokes coefficients of its harmonic expansion. In this paper, we give a historical background of the subject and discuss the related physics. We then compute the geodynamic effects caused by earthquakes using Chao and Gross’ (1987) formulas based on Gilbert’s (1970) normal-mode summation scheme. The effects are computed using the centroid moment tensor (CMT) solutions for 15,814 major earthquakes from Jan., 1977, through Feb., 1999, as provided in the Harvard CMT catalog. The computational results update those of Chao and Gross (1987) and Chao et al. (1996), further strengthening their findings and conclusions: (i) the strong tendency for earthquakes to make the Earth rounder and more compact (however slightly) continues; (ii) so does the trend in the seismic “nudging” of the rotation pole toward the general direction of ~ 140°E, roughly opposite to that of the observed polar drift, but two orders of magnitude smaller in drift speed.

The long series (1899.9–1979.0) of latitude observations made at Carloforte in the framework of the International Latitude Service is analyzed to determine a secular trend. The value of the linear rate obtained from the classical astrometric results agrees fairly well with the value derived from recent SLR and GPS observations carried out at the Cagliari Observatory.

To explain the observed effects in the Earth’s polar motion, a mechanism of the relative motion of the lower mantle and upper mantle with a boundary at 670 km of depth is proposed. According to the new approach, the Earth’s layers (including separate plates) are considered as nonspherical, heterogeneous celestial bodies, interacting with each other, with the Moon and the Sun and executing a wide spectrum of relative motions in different timescales. The small displacements of the centers of masses of the lower and upper mantles and their relative rotations have here a primary importance. These displacements display themselves at various time scales (from a few months to millions of years), and their manifestations are readily detected in the regularities of the distribution of geological structures as well as in many geodynamical processes. Important regularities of the ordered positions of the plate centers, of their triple junctions, hot spots, systems of fractures and cracks, geographic structures, fields of fossils, etc., are observed as consequences of certain displacements and inclined rotations (Barkin, 1999). At geological time intervals, the slow motion of the layers causes mutually correlated variations of the processes of rifting, spreading, subduction, regressions and transgressions of the sea, of the plate motion, formation and breakdown of super continents, etc. The motions and the accompanying tectonic mass redistribution cause variations of the components of the Earth’s inertia tensor and geopotential, which lead to variations of its diurnal rotation and polar motion. Explanation of the main properties of the perturbed Chandler polar motion has been done.

The history of observations of the Chandler motion is presented in brief and the importance of the ninety years of the ILS data is stressed. Essential features and different empirical models of the Chandler motion are discussed.

The 14-month Chandler period is associated with the free nutation of the Earth about its spin axis. The observed value of the Chandler period comes usually from the analysis of astronomical series of polar motion data as well as from superconducting gravimeter measurements. At the observation level a periodicity of about 420–440 days also was noticed in microseismic activities. Recently we found evidence of a signal with period similar to the Chandler one in the Sardinia rainfall time series.