By Yuri ILYASOV, Dr. Sc. (Tech.), head of Department, Lebedev Physical Institute (FIAN), RAS

In 1982 a team of the RAS FIAN scientists and their colleagues from the All- Russia Scientific Research Institute of Physico-Technical and Radio-Technical Measurements, RF GOSSTANDART (State Standards Committee) received an author's license for their truly unprecedented breakthrough. They introduced (with a priority from 1979) what was called a pulsar time scale, and in 1991 -a method of synchronization of remotely spaced time standards by pulsar signals. It took the international scientific community nearly ten years to recognize the priority of Russian physicists and follow them in this field. And at the end of the 20th century the US Naval Observatory (USNO) identified the three most promising time standards: a mercury one with ion traps (USA), a cesium one with laser cooling (France) and a pulsar application (Russia).

And it all started in August 1967 when researchers of the Mullard Radio Astronomical Observatory, Cambridge (Britain), analyzing recordings of observations of discrete space radio sources, came across a brief interference. At first sight this looked like some accidental snag, but later on it began to appear every night. The head of the project, Prof. Antony Hewish (later a Nobel Prize winner) decided to keep an eye on the "unpleasant" disturbance in order to determine its nature and, in the final analysis, if possible, get rid of it somehow.

As a result, it turned out that the "stray" signal arrives regularly, but every time this happens 4 minutes earlier that before. This fact attested to its extraterrestrial origin because this is exactly the time interval at which the appearance of celestial bodies at given points is daily accelerated. And that means that what we were dealing with was not an interference, but a signal from some unknown space source. A "hunt" was started after it and on November 28, 1967 at right ascension 19.19 hours, declination +21 ^o from the celestial equator, radio astronomers registered the first evident pulses (arriving at periodical intervals of 1.33 s) from the space "stranger" which was codenamed CP= 1919+21 (Cam-bridge Pulsar). Shortly after astrophysicists identified three of its "brethren" (0834+06, 0950+08,1133+16) and it became clear that what the scientists were up against was a completely regular phenomenon. These objects were called pulsars for their ability to generate radiomagnetic emissions in a pulsed manner, and November 28, 1967 was recognized as the day of their discovery. Since then pulsars have been the object of close

Pages. 11

observations and studies of many radioastronomical observatories across the world, including our own, located at the town of Pushchino (Moscow Region) which is equipped with the biggest radio telescope DKR-1000 and a unique large phased array LPhA FIAN.

The main question facing the scientists concerned the nature of the new celestial sources of radio emissions. The original suggestion was put forward by the Cambridge team "pioneers". In their opinion these could be white dwarfs which, under certain conditions, could enter a regime of periodical changes of their size in which their diameter is dwindling in proportion to attenuating inner energy. In the process of this shrinkage the energy of nuclear processes within the star is boosted as a result of which it swells again. The period of such variations can be of the order of 1 s and more.

But the two pulsars, discovered in 1968 (one by Australian astronomers in constellation Vela with a pulse period of 0.089, and the other one by American observers in the Crab Nebula with an even shorter period of only 0.033 s), "buried" once and for all the idea of white dwarfs, which had they been "performing" in this regime, would have simply disintegrated.

A different explanation of the nature of this phenomenon was suggested by our compatriot-the first director of the Pushchino Observatory of FIAN, Prof. Viktor Vitkevich, Dr. Sc. (Phys. & Match.). His idea was that a rapidly spinning celestial object possesses what he called a localized area of emission which hit at every spin the "line of sight" of the observer telescope. This model was reminiscent of a lighthouse beacon. Capable of such rapid spins can be some very compact space objects tens of kilometers in diameter only. Such parameters are possessed only by neutron stars, but what remained unclear was where they obtain the energy required for such powerful radio emissions.

At this point we should make a small diversion. After the discovery of the neutron in 1932 by the British scientist (later Nobel Prize winner) James Chadwick, our compatriot, Acad. Lev Landau, who also won a Nobel Prize, suggested the possibility of existence of what were called neutron macro-droplets, produced at super-high pressures. In 1934, the German astronomers Walter Baade and Fritz Zwicky (both worked in the United States) predicted the formation of neutron stars as fragments left by a blast of a supernova which had "shed" its envelope and shrunk in the process of a collapse down to a "sphere" about 20 km in diameter and with a mass within 1+2 solar masses. With the momentum* remaining unchanged this kind of a "gyro" should accelerate up to high angular velocities with a period of rotation of about 1 s and less.

Another stage in unravelling the mystery of pulsars was an attempt made in 1964 by then a young talented scientist, and now Member of the Russian Academy, Nikolai Kardashev, to explain the nature of the Crab Nebula produced by an explosion of a supernova in 1054 (recorded by Chinese chronicles) by a powerful magnetic field inside. In the scientist's view this formation should have been less than 100 km in diameter, spin with a period of less than 1 s and have a magnetic field of about 10 ¹² Gs.

* Momentum-measure of mechanical movement for a material body-the product of its mass and linear velocity. - Ed.

Pages. 12

This fitted in a remarkable way the parameters of a pulsar which was later (in 1968) discovered in the Crab Nebula. Finally, in 1967 the Italian physicist Franco Pacini presented the first model of radio emission of a rapidly spinning neutron star with a strong magnetic field.

All-round studies of pulsars (both old ones and newly discovered) were launched in the early 1970s with building their models and development of a theory of their generation of radio emission and studies of their general and individual characteristics. With this aim in view scientists of the FIAN Pushchino Radio- Astronomical Observatory installed in 1973 a new telescope in the form of a dipole phased array, consisting of 16,384 vibrators (pick up pulsar signals) on an area of some 8 hectares (200 m from east to west and 400 m from north to south) which received the name of Large Phased Array (LPhA) Scanning (Phased). Later on its technical parameters and remarkable potential will be discussed in greater details. And now let me just say that the telescope operates in the meter wavelength range.

Like the one in Cambridge, our instrument picks up the pulsar signal when it crosses the local meridian, but the LPhA is much more sensitive and has greater noise immunity than the Cambridge pulsar array. And it registers signals during 4 - 15 min, which is twice as long as its Western analogs.

In the building of the new radio telescope it was necessary to study the stability of the period of rotation of various pulsars. It has now been established that the period of succession of the received pulses of them is the same-like in unusually precise watches. But what will their "behavior" be like at long operational intervals (more than a month), what will be the likely peculiarities of the new time scale if, of course, it is recognized by the scientific community? What methods of observations can best be used? And we continued seeking answers to these and related questions.

Years of studies of neutron stars conducted with the help of the LPhA FIAN instrument have confirmed the high stability of their spin period. Also observed and measured has been their secular linear deceleration which occurs because of the slowly dwindling mechanical energy of their rotation which is spent on emissions. In young stars this deceleration per 1 s amounts to about 10 ^-13 , and for old ones- up to 10 ²¹ , that is several fractions of picoseconds (10 ^-12 s) after a year. This process is of a highly "determined" nature and can be easily taken into account if pulsars are regarded as space clocks.

In the course of our observations we arrived at a novel idea - can stable pulsars be used for forming a new fundamental time scale? We translated this idea into an application for an authorship certificate which, as had been mentioned before, we received in 1982.

So, what are the advantages of our proposal? By the time of its submission the international scientific community had accepted as the standard one ephemeris second, as 1/31556925, 9747th part of a tropical year for 1900 for January 0 at 12 h ephemerical time* (by a decision of the General Assembly of the International Astronomical Union-IAU, 1995, and a resolution of the International Bureau of Measures and Weights, 1956). Also established in 1967 was the atomic second, equal to 9.192.631.770 periods of oscillations as emitted by the isotope of cesium Cs ¹³³ (13th General Conference on Weights and Measures). All the national centers of time and frequencies in the world obey these standards and build their own scales accordingly. However, every individual country has its own national standards (of calibration) which, naturally enough, is causing a serious problem of the unification of measurements and comparison of national scales which is the responsibility of the International Bureau of Weights and Measures in Paris. Comparative studies indicate that in different countries the quality of standards (today these are hydrogen masers) is not the same. As a result the relative accuracy of reproduction of the nominal frequency of the quantum standard* amounts approximately to 10 ^-12 - 10 ^-13 , which means that within a year clocks corrected by these standards will show a time difference of several microseconds- something simply inadmissible in certain fields of technology. And the obvious conclusion is that we simply have to have a common keeper of time intervals for us all.

Pulsars offer a nearly ideal solution of this problem. They are one and the same for all users, they are located without the Solar system, accessible for observations from any point of the globe and are "long-lasting" (their "service life" is measured in millions of years). And their period of rotation is also convenient - of about one second or even fractions thereof. If a "group scale keeper" is formed of more than two identical stars, then, as one can easily see, this

Recording of radio emission from 0329+54 pulsar.

* Ephemerical time-independent variable in equations of movement of celestial bodies. - Ed.

* Quantum frequency standards-devices for accurate measurements of oscillations frequency, based on frequency changes of quantum transitions (SHF and optical spectra) of atoms, ions and molecules from one state into another. Used in navigation and time service as standards. - Ed.

Pages. 13

Promising time standards: a - ephemerical second; b - atom second; c - pulsar time scale.

scale becomes autonomous and needs no comparisons with any other.

But translating this idea into reality proved to be no simple matter. First, because of the spin of the Earth around its axis and its orbital rotation around the Sun the distance from the observation point to a pulsar, and, consequently, the time of arrival (TOA) of its pulse, are permanently changing. So, it would be logical to find a point in space which is free from the above "defects". For example, the center of masses of the Solar system. Given the name of the barricenter, it would have been the most appropriate position for the purpose. This would have also been promoted by the fact that the time of arrival (TOA) of a pulse from a pulsar to the center and to the ground radio telescopes can be easily intertied by three mathematical ratios which represent a reliable algorithm of conducting a pulsar scale.

Another problem we faced consisted in the difficulty of registration of the incoming pulsar signals. Their spectral flux density amounts to only several milli-Jynski (mJy)*. And let us note for comparison that the flux density of a 10 Wt transmitter set up on the Moon would have been million of times greater. And, finally, an exact measurement of the TOA of a pulse would be obstructed by both internal and external interferences, or noises in the specialist terminology.

All of the aforesaid attests to the fact that: for a unit for keeping a pulsar time scale, or base, it is necessary to have a large radio telescope, better a foil steer-able one which could accumulate pulse signals over a long time tracking for object. In order to make its antenna more sensitive it is necessary to use a broad frequency band for the reception. One will also have to have devices capable of "relating" the data of pulse arrival to a certain metrological base-in other words-a high-precision system of local clocks comparison on the radio telescope. In our case this will be comparison with the GOSETALON (Federal Standard) scale. This process is implemented with the help of a receiver of time signals of the RF GOSSTAN-DART (State Standards Committee).

And now let us describe in more precise terms the radio telescope check-up complex. This includes, above all, the antenna, receiver equipment, time service and a computer. These demands are folly met by the AS-102/64 receiver complex built in Pushchino on the base of the aforesaid LPhA FIAN and operating in the meter wavelengths band. Since 1976 it has been used for observations and measurements of the time of signals arrival from several chosen stable (frequencies of rotation of 1 Hz) pulsars. Those of them - 0834+06 and 1919+21-turned out to be the most convenient. For more than 20 years of their monitoring the exact TOA deviated from the predicted (precalculated) by not more than several tenths of microseconds, which is within the limits of chance variations of measurements. And that means that pulsars are "performing" as the best top-class clocks with their relative error margin over 20 years being not more than 10 ^-13 - 10 ^-14 . Such accuracy is simply impossible now for quantum standards over such long time intervals.

And no matter how convenient for observations were normal or recycled pulsars, there are also different ones in nature, rotating much faster. The existing instruments do not fit them because in the meter wavelengths range short pulses are considerably "stretched out" while passing through interstellar space consisting of rarified plasma clouds. According to calculations the most fitting for observations of such "fast"

* A unit of measurement in radioastronomy of the power of incoming emissions in the frequency band of IHz, on an area of 1 m ² , i.e. one Jan=10 ²⁶ Wt/Hz * m ² . - Ed .

Pages. 14

Pages. 15

Residuals of TOA of pulsar 1937+21.

neutron stars is the decimeter wavelength range. Following the discovery of millisecond pulsars (1983) foreign specialists also came out in favor of our concept of establishing a new pulsar scale.

To put these ideas into practice, FIAN experts started work in 1991 on a new unit for the decimeter wavelength band. They took as the basis the TNA-1500 antenna system (one of the largest in the world) designed and built by the Special Research Bureau of the Moscow Power Engineering Institute. It has a fully steerable dish 64 m in diameter. The multichannel receiver was designed by our specialists in conjunction with our counterparts from the Scientific Research Institute of Radiophysics (Nizhni Novgorod), and the precision time service-in cooperation with the GOSSTANDART Institute of Metrology of Time and Space. Our specialists also upgraded the algorithms of pre-calculations of the TOA of pulse and synchronization of the start of observations by the equipment.

From 1993 to 1996 the whole set of the instruments and equipment was being tested at the Space Communications Station of the Special Research Bureau of the Moscow Power Engineering Institute at Medvezhye Ozero (Moscow Region). Later it was moved to the town of Kalyazin (Tver Region) where the second TNA-1500 telescope was constructed. This was done because the level of interferences there is much lower than around Moscow. That latter consideration has been chiefly responsible for the fact that observations of the fastest millisecond pulsars В 1937+21 (period of rotation- 1.56 ms) started back in 1996 and is being continued to this day. The data obtained over nearly 6 years of monitoring attest to the really high stability of the chosen "reference" neutron star. This conclusion is born out by divergence in the pre- calculated and the registered TOA of pulse (so-called residuals) within the root mean square (RMS) value of 2 - 3 ms. And that means that using pulsar signals one can predict an event on a time scale one year in advance and even more with an error of just a few microseconds.

The pulsar time scale suggested by our specialists and being adopted in this country now is winning more and more supporters. Today regular pulsars monitoring is conducted in practically every major radio astronomical center around the world-a fair measure of the tangible contribution of Russian scientists to the progress of fundamental metrology of time.

Illustrations provided by the author

Permanent link to this publication: