Memorie della Societá Astronomica Italiana
Vol. 51 (June 1980), 247–261



1Inst. of Astronomy, N. Copernicus Univ., Toruñ, Poland
20sservatorio Astronomico di Trieste, Trieste, Italy

(Received 25 May 1980)


Methods and problems of solar radio data recordings at different observing stations are given.
Improvements and standardization of data are suggested.


Vengono presentati i metodi ed i problemi di registrazione dell'attivitá radio solare in vari Osservatori. Viene data un'indicazione per migliorare e rendere piú standardizzata tale attivitá.


In May 1976 a memorandum accompanied by a questionnaire and a letter concerning problems of single-frequency solar observations at metric wavelengths was distributed to 30 observatories. In doing this we were encouraged by prof. H. Tanaka. It was hoped that this action, when well responded to, could be the first step towards an international development in this kind of observations.

We did receive 16 replies and comments, some of which contain very interesting ideas. A review of the collected material that is presented in this paper seems to show the existence of a real need of further improvements in metric wavelength solar observations, regarding both data and instrumentation problems. Some valuable points remarked by J. van Nieuwkoop {1978) may be an outline for further efforts in this field.


The description of the instrumentation used for single-frequency observations by the various stations is provided in form of Tables. Some spectrographs were also included, since they allow for single-frequency data recording: the Dwingeloo spectrograph is a complex of 60 adjacent single-frequency receivers, the Zürich-Beilen spectrographic data can be digitized at given frequencies (up to 500) and the Nançay spectrograph presents similar possibilities (120 channels with 1 MHz bandwidth and 100 channels with 0.2 MHz bandwidth). Most of the detailed entries in the Tables come from the replies to the questionnaire, others from the literature.

The Tables are in general self-explanatory. The four-letter abbreviations of station names are those used by WDCs. The columns of Table I headed "gain" and "HPBW" give respectively the antenna gain and half-power beamwidth (in degrees): we are confident that these are the measured values, even in those replies in which this is not specified. In Table II, "dynamics" is the ratio between the highest and lowest sensitivity thresholds. The "noise" column gives the receiver noise figure, and that of "sensitivity" gives the least detectable signal in solar flux units (1 s.f.u. = 10–22 W·m–2·Hz–1). In Table III, the "speed" column contains recorder tape speed in cm·h–l or the sampling fre¹uency in Hz, "dynamics" represents the recorder ratio of greatest and lowest signals. The rightmost column gives an estimated accuracy figure in percent. If two figures in this column are separed by a comma, the first one regards the mean flux density measurement accuracy, the second one the distinctive event data accuracy.

Table  I

Antenna Systems

ABST 221 16 × dipole azimuthal 19.5 l2×15 BLEN* 100-1023 7 m paraboloid azimuthal - 30-3 BORD 930 7.5 m paraboloid equatorial 34 2°25'×1°21' CRAC§ 406 7 m paraboloid equatorial - 8.6 CRAC§ 810 7 m paraboloid equatorial - 3.4 CULG 43 corner reflect. adj. in δ - 60 CULG 80 13 m paraboloid equatorial - 20 CULG 160 13 m paraboloid equatorial - 10 DWIN 509-533 25 m paraboloid azimuthal - - DWIN 574-614 25 m paraboloid azimuthal - - DWIN 655-666 25 m paraboloid azimuthal - - HARS 228 7.5 m paraboloid equatorial - - HIHA 100.5 10 m paraboloid equatorial 13.4 20 HIRA 200 10 m paraholoid equatorial 18.8 10 HIRA 500 6 m paraboloid equatorial 25.8 6 H0VN 242 7.5 m paraboloid azimuthal 20 11 NANC 169 16×4m paraboloid equatorial 11 30 interferometer NANC 169 ditto + 2×10 m equatorial 16 5 paraboloid NANC 150-469 6 m paraboloid equatorial 18 14-30 ONDR 260 7.5 m paraboloid azimuthal 23 12 ONDR 536 7.5 m paraboloid azimuthal 29 7 ONDR 808 7.5 m paraboloid azimuthal 33 5 TYKW 1000 3 m paraboloid equatorial 25.2 7.5 TORN 127 2 × gain standard adj. in δ 7.7 65 10λ interferom. TRST 237 10 m paraboloid equatorial 27 9.5 TRST 327 10 m paraboloid equatorial 28 8.4 TRST 408 10 m paraboloid equatorial 30 6.3 TRST 408 2×8 dipoles adj. in δ 38 24x26 100λ interferom.
* From September 1979 - formerly DURN § proposed abbreviation for Cracow - formerly CRA

Table  II


ABST 221 total power 8 70 - 6 - - BLEN 100-1023 digital spect- 1-3 60 1440 K .0005 10 film rograph BORD 930 total power 30 - 2000 K 1.5 - - CRAC 406 total power 2.8 20 4.5 dB 1.0 0.5 audio CRAC 810 total power 2.1 20 4.5 dB 1.0 1 audio CULG 43,80,160 total power 1 40 2.5 dB - - audio DWIN 509-533 spectropolarim. 2.7 30 - 0.01 - film DWIN 574-614 spectropolarim. 0.9 30 - 0.01 - film DWIN 655-666 spectropolarim. 2.7 30 - 0.01 - film HARS 228 total power 1 - 6.0 dB - - audio HIRA 100.5 polarimeter 0.1 30 300 K few 1 - HIRA 200 polarimeter 0.25 50 270 K 1 0.5 audio HIRA 500 polarimeter 1 50 450 K 0.5 0.2 audio HOVN 242 total power 0.56 22 2200 K 2 1 audio NANC 169 total power 1 35 150 K 0.1 - - NANC 169 correlation 0.6 51 150 K 0.005 - - NANC 150-469 total power 1-0.2 50 1500 K - - LED ONDR 260 total power 0.8 60 348 K 1 - - ONDR 536 total power 0.75 60 580 K 1 - - ONDR 808 total power 0.6 60 1450 K 1 - - TYKW 1000 polarimeter 10 >30 800 K 0.5 - - TORN 127 total power 0.23 28-40 160 K 2 & 5 0.3 audio TRST 237,327,408 polarimeter 2.5 90 3.0 dB .5&.01 1 audio TRST 408 total power 0.5 16 6.0 dB 0.1 1 visual

Table  III

Recording and Observations

ABST 221 chart 24 cm/h 100:1 flux dens. 2.5,1 BLEN 100-1023 film, digital - - flux dens. - tape (events) & circ.pol BORD 930 chart 30 cm/h 64:1 flux dens. 10,10-25 CRAC 406,810 chart 12 cm/h 200:1 flux dens. 5,20 CULG 43,80,160 3-channel chart 2 Hz - flux dens. - & magnetic tape DWIN 509...666 film,digital ta- 120-540 - flux dens. 5 tape pe,2-3-ch.chart cm/h (film) & circ.pol. 10 chart HARS 228 chart - - flux dens. 10 HIRA 100.5 2×2-ch. chart 36 cm/h - flux dens. 40 & circ.pol. HIRA 200 2×2-ch. chart 36 cm/h 10^5:1 ditto 25 HIRA 500 2×2-ch. chart & 36 cm/h 10^4:1 ditto 20 digital tape H0VN 242 chart 6 cm/h 100:1 flux dens. 15,30 NANC 169 chart 120 cm/h - flux dens. 10-20 NANC 169 film, 5 Hz (film) - flux dens. 5(fl.) digital tape 25 Hz (tape) & circ.pol 10(pol.) NANC 150-469 film, digital 45 cm/h(film) - flux dens. - tape, 2 charts 1-100 Hz(tape) ONDR 260,536,808 chart 120 cm/h - flux dens. 10 TYKW 1000 2-ch. chart,di- l2 cm/h(ch.) 6:1 sum & diff. - 4 Hz (tape) of 2 c.pol. TORN 127 2 charts 16 cm/h 200:1 flux dens. 10,30 TRST 237,408 2-ch. chart, 20 cm/h 1600:1 sum & diff. 10 analogue tape (chart) of 2 c.pol. TRST 327 analogue tape 8600 cm/h - ditto 10 TRST 408 2-channel chart 72 cm/h 100:1 flux dens. 15


Nearly all the comments about the problems mentioned in the memorandum are presented hereafter. The year following the name of an author in quotations will be omitted when we refer to a reply to the memorandum and/or the questionnaire.

Among problems of general interest it was pointed out that the construction of composite dynamic spectra, e.g. of type IV bursts, is difficult because of the small number of frequencies available, the different calibration systems, the different time constants, and the subjectiveness of choosing spectral contours, making a reference to wide-band (if low-dynamic) spectrograms necessary (Slottje).

1. Absolute Calibration

Reference sources such as Cas A or Cyg A are rather weak for calibration purposes; also the secular decrease of Cas A flux is not very well determined. Generally it may be seen that the stations that do not use absolute calibration sources are forced to this by low antenna gain (Beilen, Bordeaux, Hiraiso, Nançay spectrograph, Ondøejov, Toyokawa). At Hiraiso they tried to use a reference antenna of measurable gain; the outputs of the antenna to be measured and of the reference one are applied to the R and L port of the receiver and are compared (Yamashita). In Toyokawa a large horn is used as reference antenna (Tanaka). In Bordeaux the calibration is carried out every week by means of a hot and a cold noise source (Poumeyrol).

The following stations take Cas A as reference source: Dwingeloo (every 3 months), Abastumani (once a year), Hoeven (every 2 months), Nançay (169 MHz interferometer), Toruñ (monthly) and Trieste (237 and 327 MHz polarimeter, 408 MHz interferometer, every 2 months).

Cyg A is used by: Nançay (169 MHz radioheliograph and interferometer), Harestua (comparing Cyg A region with a cold region at α = 9h40m and δ = 41°) and Trieste (408 MHz interferometer). Note that the Trieste 408 MHz polarimetric channel is calibrated on the interferometer.

At Culgoora the quiet fluxes at 80 and 160 MHz are periodically measured with interferometric techniques by comparison with sources of known flux (Vir A, Tau A, etc.) to determine the effective area of the radiometer aerial. The matching of the 80/160 MHz aerial feed is measured every day as a check on the stability of the effective area (Nelson).

In order to avoid difficulties connected with the absolute calibration and to reduce the inhomogeneity of data reported by various observatories it would be useful in the opinion of Kharadze and his colleagues to organize simultaneous observations of the Sun, at least for those stations where this is geographically possible. Through this cooperation more accurate results could be found, to be shared with other stations in order to overcome instrumental differences. In June 1979 a workshop was held in Nançay to discuss the results of such a joint observation day (May 12, 1979 was chosen) in which five observatories took part. The proceedings of this meeting are summarized in Elgaröy et al. (1980), from which Figure 1 is taken; considering that generally the data are given with an accuracy of about 10%, the fit must be accepted as rather good. In the same meeting it was proposed to repeat periodically this kind of joint observations; the observation of a "smooth" type IV burst would be useful as a test of the extrapolation to higher flux densities.


Figure 1. — The solar background spectrum for May 12, 1979 (after Elgaröy et al. 1980). The dashed curve represents the Quiet Sun as given by Smerd (1964). Station names are those used by the WDCs.

Suggestions were also made as to the agreement on one absolute flux density scale. The researchers taking part in the Nançay meeting have agreed to accept the formula by Baars et al. (1977), slightly modified as follows. The flux density of Cas A is expressed by:

S(f,t) = So exp[a(t – to)] fαo + b(t – to)

where to = 1980.0, a = –0.0097 yr–1, b = 0.00126 yr–1 and
for 0.2 ≤ f <  30 GHz, αo = –0.770, So = 0.2720 sfu while
for .02 < f < 0.2 GHz, αo = –0.681, So = 0.3140 sfu and S is in sfu, f in GHz.

The flux density of Cyg A is given by:

log S(f) = 4.695 + 0.085·log f – 0.178·log2f

were S is in 10–26 W·m–2·Hz–1, f in MHz.

2. Galactic Background

The galactic flux at 80 and 160 MHz of Culgoora has been measured around the ecliptic in absence of the Sun and is subtracted as required. This procedure needs the effective area of the aerial to remain constant (Nelson). Basically the same method is used in Hoeven (Klaassen) and has been used in Trieste before 1969.

In the opinion of Yamashita the above mentioned method cannot be regarded as decisive. A similar idea is expressed also by Slottje who thinks that interferometry is needed when the Quiet Sun is to be measured (of course the problem is less critical when the bursts are concerned as they appear above the background). Presently in Torun interferometric technique is used routinely and so this is not a real problem; this technique is even more developed for the Nançay instruments (the 169 MHz interferometer having a resolving power of 3.5', the 169 MHz radioheliograph 1.15').

3. Ground Reflections

Some hours in the morning and in the evening are lost in Harestua because of ground reflections: this is particularly important in observatories where snow or ground-frost are often present (Hoeven, Beilen), but it is noted also in other stations. In Hiraiso the observing site is near the seashore, at a height of about 25 m: interference patterns are shown in the observations at sunrise, but it is hoped that a shielding skirt around the antenna edge or ground shaping works (banks, trenches) may be effective in preventing this (Yamashita).

In Toruñ, broad beam antennas are used and this leads to a continuous influence of ground-reflected radiation on observations. The effect is present all the year as a function of the altitude of the Sun. It is hoped that after a few years of observations the required factors may be established to correct the previously derived data.

4. Methods Used against Man-Made Interference

Metric observations may be strongly affected by interference of both natural and artificial sources. To distinguish between solar events and interference in Hiraiso they first notice the coincidence of the supposed event and occurrences that appear in observations at other frequencies, including ionospheric absorption measurements. A narrow-band solar radio emission may usually be checked by measuring its circular polarization. It may clearly differ from the radiation of an artificial source (Yamashita, 1974).

Further, it is not difficult to find out whether a phenomenon is of solar origin if high-time resolution recordings are used: on these, an interference is clearly visible because of its sudden beginning and end. Also checking the sky in a neighboring re.gion may be useful. Raffaelli (1975) presented a system for 408 MHz observations at San Miguel (not included in the Tables) consisting of two antennas one of which is directional and the other omnidirectional. An input of the radiometer is switched alternatively from one aerial to the other with a frequency of 1 kHz. A solar event, unlike interference, is observed by both antennas with the same shape.

It is also possible to distinguish between solar emission and interference by using a noise discriminator — an electronic circuit or simply human ears. Ears are a sensitive detector of interference, once trained, but are not suitable for routine observations. An electronic noise discriminator is a combination of two RC circuits of different time constants. When connected to the same detector these two circuits show different responses and their ratio is constant for random noise while it varies in other cases (Yamashita, 1974).

5. Methods Used to Cover a Wide Dynamic Range

A proper dynamic range is achieved in Culgoora by using logarithmic amplifiers. The logarithmic gain is high for Quiet Sun and weak bursts but decreases for strong bursts. The changeover is gradual and is set at about 7 dB above Quiet Sun plus background level (Nelson). Also the Nançay spectrographs employ these amplifiers with very good reliability (Daigne, Bougeret).

In Bordeaux they use an automatic divider of the output signal by 4, 16 or 64 (Poumeyrol). In the 200 MHz receiving system of Hiraiso the input signal is divided in two levels, namely the Quiet Sun (low) and the Active Sun (high) levels. The high level has a dynamic range of about 30 dB, while the low level is limited to 20 dB or less for reasons of linearity. This method, however, has some deficiencies (smooth connection between the channels). A quite new design has been put into operation for 500 MHz observations. It is a receiver that keeps the input level of the linear detector and its driving stage rather constant, with some definite clearance, with the aid of attenuators inserted automatically at the input when large signals are present; at the output a microprocessor is employed. Logarithmically compressed flux (10 to 100,000 sfu) is recorded beside linear-scale flux (0 to 400 sfu) and poiarization degree (–1 to +1) (Yamashita).

At the Trieste Observatory, paper strip recorders (with a speed of 20 cm·h–l) are automatically attenuated by a factor of 2 when the recorded signal reaches end-of-scale level. There is the possibility of attenuating the signal up to 16 times. When the output signal exceeds a given threshold, an attenuation of the antenna signal (in steps of 10 dB) is automatically triggered.

In Ondøejov two paper strip recordes per frequency are used; the ratio of sensitivity between them is 1:10 (Tlamicha).

6. Data Reduction

Deficiencies of data reduction processing are noted by Slottje: handling consumes much time and the results (e.g. composite dynamic spectra) are often disappointing. Besides, burst classification (see e.g. Tanaka, 1975) still depends strongly on the observer; Klaassen suggests that observers agree on using a manual in which several examples of burst profiles in different frequency ranges would be illustrated. He also suggests to report the reduced data in accordance with the accuracy of measurements; since this is generally about 10%, the data should be given with 2 significant digits only.

Finally one of us (KMB) pays attention to the lack of consistent criteria for variability index evaluation. He suggests a slight modification of the Manual by Tanaka (1975) that would take into account the dependence of noise storm burst height on the integrating time constant of a system (Borkowski, 1977).

From the above reported material and suggestions, a fair proposal to obtain a much-needed standardization of data recording is the following. For mean flux density measurements (hourly data) a paper strip recorder having a speed of 20 cm·h–1 and a time constant of 1 sec would look sufficient. When distinctive events are considered, as 0.1 sec (or less) lasting burst structures do exist, a time constant of 0.01 sec is needed and the recorder speed must be proportionately higher, while logarithmic amplifiers should be employed. Such a "high rate" recording system should of course be used only during active periods, a magnetic tape support being preferable to a paper strip recorder.


The material collected here seems convincing enough about the need of considerable wark to be done in improving and standardizing both observing techniques and data reduction methods. New frequencies of observation are necessary for a better covering of the metric spectrum, too. Most of the problems, however, are purely technical. A future improvement in this respect may be just a wider information suggesting the best designs to overcome each particular problem. Essentially, as J. van Nieuwkoop rightly notes, we must know and improve our accuracy, both as regards deviations from an absolute calibration in the full amplitude range, and the stability as a function of temperature and so on. He suggests to define a specific aim, and to compare our instruments with it; to use the system block-diagram, to choose carefully the type and quality of essential components, to discuss and decide between observers and equipment, manufacturers about the ways of obtaining the best results even by modifications in hardware, data handling or calibration procedures (van Nieuwkoop, 1978). This is in our opinion a very good way of starting.

The most unpleasant difficulty concerns the absolute calibration; though it should not be expected that this problem will be solved quickly and easily, the main factors responsible for this may be singled out and the best actions to overcome them could be decided. One of these is the fixing of an absolute flux scale for the overall metric band, to be respected by every observatory. This, as well as other problems outlined in the preceding paragraphs, may be relatively easily solved by international cooperation. Something very promising in this direction has been done during the above mentioned Nançay meeting, though among few stations and without the support of a major international agency; all these problems would in fact require an official organization to be effectively solved, such as a working group formed by URSI, CESRA or IAU participants, in which experienced scientists would lead the action.


This report was possible thanks to many individuals who are being or have been actively working in the field of metric wavelength solar observations and who kindly shared their experience and information with us. We are grateful to A.O. Benz (Zürich-Beilen) , J.-L. Bougeret (Meudon-Nançay), G. Daigne (Meudon-Nançay), E.K. Kharadze (Abastumani), M.A. Klaassen (Hoeven), U. Koren (Trieste), A. Maxwell (Fort Davis), C. Mercier (Meudon-Nançay), G.J. Nelson (Culgoora), F. Poumeyrol (Bordeaux), C. Slottje (Dwingeloo), O.P. Sveen (Harestua), H. Tanaka (Toyokawa), A. Tlamicha (Ondøejov), F. Yamashita (Hiraiso) and S. Ziêba {Cracow). Special thanks are due to Dr. J. van Nieuwkoop for his inspiring comments. We are also very much indebted to Dr. J. Hanasz who critically read a preliminary version of this paper.


Baars, J.W.M., Genzel, R., Pauliny-Toth, I.I.K. and Witzel, A.: 1977, Astron. Astroph. 61, 99.

Borkowski, K .M.: 1977, Physica Solariterr. 4, 13.

Elgaröy, Ö., Slottje, C., Tlamicha, A., Urbarz, H., Zanelli, C., Zlobec, P., Bougeret, J.-L., Kerdraon, A. and de la Noë, J.: 1980, Astron. Astroph. Suppl. Ser., (in press) [44 (1981), 165].

van Nieuwkoop, J.: 1978, private communication.

Raffaelli, J.C.: 1975, private communication.

Smerd, S.F.: 1964, Ann. of the IGY  34, 331.

Tanaka, H.: 1975, Instruction Manual for Monthly Report, Toyokawa.

Yamashita, F.: 1974, private communication.