ITU-R P 371-8-1999 Choice of Indices for Long-Term Ionospheric Predictions《长期电离层预测的指数选择》.pdf

1、STD*ITU-R RECMN P-371-8-ENGL 1999 = 4855232 053b377 807 W Rec. ITU-R P.371-8 1 RECOMMENDATION ITU-R P.371-8 CHOICE OF INDICES FOR LONG=TERM IONOSPHERIC PREDICTIONS (1963-1970-1974-1978-1982-1986-1990-1995-1999) The iTU Radiocommunication Assembly, recommends 1 that the 12-month running mean sunspot

2、number R12, or alternatively the 12-month running mean value of , the 2800 MHz solar radio noise flux (is. 12), be adopted as the preferred index to be used for predicting monthly median values of foF2 and M(3000)F2 over all timescales: substantially equivalent results should be obtainable by the us

3、e of either of these indices; 2 over dl timescales: that 12 be adopted as the preferred index to be used for predicting monthly median values of foE and foF1 3 that predicted values of these indices should be determined by means of the modified McNish-Lincoln procedure (see Annex 1) using latest ava

4、ilable measured monthly index values for the present solar cycle and the average of past cycle values for future cycles; 4 the same index may be adopted for all such charactenstics with little loss of accuracy; that where propagation predictions require simultaneous use of values of different ionosp

5、heric characteristics, 5 ionospheric predictions may not be sufficiently accurate. that caution be shown in the use of the recommended indices at high magnetic latitudes, where the resulting ANNEX1 1 Introduction The concept of indices for long-term ionospheric predictions relies on the. assumption

6、that the important characteristics of the ionosphere, such as the critical frequencies of the various layers and the MLTF factor M(3000)F2, depend in a systematic way on certain measurable quantities concerned with solar radiation. It should however be noted that the correlation between these indice

7、s and the actual ionospheric characteristics does not necessarily imply a causal relationship, but rather an indication of associated phenomena. Changes of solar activity in general contain the components: - - - a fairly regular component with a period of about 11 years, which represents the weil kn

8、own cycle of solar activity; a component that has a quasi-period of about a year or a little less; and eratic fluctuations with periods of less than a month. COPYRIGHT International Telecommunications Union/ITU RadiocommunicationsLicensed by Information Handling Services2 STD=ITU-R RECMN P-371-8-ENG

9、L 1999 = 4855232 O536378 743 E Rec. ITU-R P.371-8 2 Sunspot numbers For studies of the min component of the solar cycle, the 12-month running mean sunspot number R12 is used because the resultant smoothing considerably reduces complicated rapidly-varying components, but does not obscure the slowly-

10、varying component. The definition of RIZ is: in which Rk is the mean of the daily sunspot numbers for a single month k, and R12 is the smoothed index for the month represented by k = n. The two main disadvantages in the use of Rl2 are: - - Nonetheless, R12 appears to be the most useful parameter for

11、 long-term studies and predictions concerning the F2 layer. the most recent available value is necessarily centred on a month at least 6 months earlier than the present time; it cannot be used to predict the shorter-term variation in solar activity. 3 Index Consistent and reasonably long series of O

12、bservations of the solar radio noise flux at about 10 cm wavelength have been made by Canadian, Japanese and other laboratories. The monthly mean, , of the daily values from Canada expressed in units of Wnr2 HT1, should be regarded as the reference data for this index. is more closely correlated wit

13、h E-layer critical frequency than are noise flux values at other wavelengths. As the solar flux observations are only available from 1947, the sunspot numbers remain one of the longest senes of observations of a natural phenomenon. Therefore, the continued collection and recording of sunspot observa

14、tions is encouraged. 4 Other indices Over past years, a substantial number of different indices have been considered to try to represent long-term changes of the different ionospheric characteristics, but of these the ITU-R recommends the indices R12 and ml2 for ionospheric predictions. 5 Correlatio

15、n between 12 and R12 The recommended relationship between R12 and mI2, also indicated in Fig. 1, is: 2 = 63.7 + 0.728 Ri2 + 8.9 x 10“ RI, 6 The prediction of indices There is as yet no method whereby it is possible to predict accurately indices for the next sunspot cycle, or, more generally, for a c

16、ycle which has not yet begun. Indices that have been calculated by using harmonic analysis, or by using empirical and statistical laws applied to observations over some earlier and even recent cycles, have not proved useful in predicting those for a new cycle. After a sunspot minimum has been observ

17、ed, future development of the cycle can be extrapolated to a certain extent, although the deviations have been observed to be rather extreme. COPYRIGHT International Telecommunications Union/ITU RadiocommunicationsLicensed by Information Handling ServicesSTDeITU-R RECMN P.371-8-ENGL 1999 4855232 053

18、b379 b8T 9 Rec. ITU-R P.371-8 3 In the United States of America, R12 is predicted using an improvement of the McNish-Lincoln objective method. First a mean cycle is computed from all past values of R12 starting from the sunspot minimum of each cycle and continuing eleven years thereafter. For predic

19、tion of a value in the current cycle, the fit approximation is the value of the mean cycle at the stated time after minimum. This estimate is improved by adding a correction proportional to the departure of the last observed value for the current cycle from the mean cycle. With the current computer

20、programs, a new prediction for each month of the remainder of the cycle can be made as soon as a new observed value becomes available. The statistical uncertainty of the prediction is fairly small for the first few months after the last observed value, but becomes large for predictions 12 months or

21、more in advance. As soon as a minimum is identified, new correction factors can be computed by including the observed values for the preceding cycle, for application to the new cycle. FIGURE 1 Relationship between R12and Q 12 260 240 220 200 180 e 160 2 140 lu) 100 80 60 0 20 40 60 80 100 120 140 16

22、0 180 200 RI 2 Predictions of R12 for one year ahead are also carried out by the Sunspot Index Data Centre (SIDC) in Brussels. An example of their predictions, for solar cycle 22, is shown in Fig. 2 and can be compared with the observed smoothed values. Predictions of 12, based on the McNish-Lincoln

23、 method, are carried out by the Radiocommunication Bureau (BR). Measured and predicted values of R and and their 12-month running mean values (Ri2 and 12) are published by the BR in the monthly Circular of Basic Indices for Ionospheric Propagation (and are also posted on the world wide website of th

24、e ITU). The SIDC also makes available, via electronic mail, the measured and predicted values of R, with access through the file transfer protocol anonymous procedure. COPYRIGHT International Telecommunications Union/ITU RadiocommunicationsLicensed by Information Handling Services4 Rec. ITU-R P.371-8 FIGURE 2 An exampie of predicted and observed sunspot numbers, R, (cycle 22) 1 986 I 9x11 1990 1992 Yew 1994 1996 1998 - Observed smoothed wammo Predicted smoothed one year ahead COPYRIGHT International Telecommunications Union/ITU RadiocommunicationsLicensed by Information Handling Services