Although much attention has been focussed on the recent events on 50MHz involving contacts between North America and Europe, there is evidence to show that such events have occurred every summer since 1980 and quite possibly before that. In fact not only has propagation to North America been evident, but South America has also been represented, mainly in the form of a regular series of reports of reception of the FY7THF beacon in French Guiana.

The existence of a beacon is extremely helpful in looking at propagation statistics.. If we consider the dates and times of FY7THF reception first, we find the following pattern:

Table 1. FY7THF reception in Europe
F layer	Postulated Es
November - December 1979	June, July 1980
October 1980 - January 1981	June, July 1981
April 1981	May - July 1982
October 1981 - February 1982	May - June 1983
October - December 1982	June 1985

There is good evidence that the events on the left of Table 1 were F layer, since they coincided with high solar flux and sunspot activity. What then of the events on the right? fy7thf reception in Europe 1979-1985

Figure 1 shows the distribution of times of openings consolidated into half hour periods. The open boxes correspond to the openings in the October to February months, whilst the filled boxes are for the May to July summer months.

The F2 reception events correspond to reports from UK amateurs on 59 days during the four winters involved. The Es reception events correspond to reports from UK, France and Sweden on 24 days. It can be seen immediately that a quite different time of day is favoured by the summer events, which is another indication that a different mechanism is responsible.

Map showing boundaries of Es zones and percentage of the daytime hours during the June solstice months (6am to 6pm, May to August) during which fEs exceeded 5MHz for the period 1948 to 1954 (after Smith, 1957)

temoral zones of Es

Correlation of transatlantic events with solar and geomagnetic indices

In an attempt to establish the cause of the 50MHz events, the sunspot numbers and geomagnetic indices have been examined. Considering first the relation with sunspot numbers, Table 5 shows the sunspot numbers (Provisional Brussels numbers except 1983 and 1984 which are definitive, and 1985 which are provisional Boulder numbers) for days on which FY7THF was heard.

Table 5 Summertime observations of FY7THF
Date	Time start	Time end	Heard by	Sunspot No	Ap
27/06/80	2045	2122	G4BPY	185	3
28/06/80	1730	1936	G4BPY	188	3
29/06/80	1447	1449	G4BPY	149	5
11/07/80	2055	2227	SM6PU	87	10
11/07/80	2032	2034	G4BPY
13/07/80	2001	2010	G3WBQ	105	9
13/07/80	2032	2110	G3WBQ
13/07/80	2200	2235	G3WBQ
14/07/80	2002	2025	G3WBQ	128	9
14/07/80	2302	0010	G3WBQ
14/07/80	2030	2040	G4BPY
05/06/81	1850	2004	G4BPY	55	6
13/07/81	1810	1815	G3WBQ	153	9
13/07/81	1921	1923	G3WBQ
13/07/81	1940	1949	G3WBQ
13/07/81	2029	2034	G3WBQ
25/05/82	1900	1935	SM6PU	88	8
05/06/82	2140	2230	G4BPY	111	6
17/06/82	2029	2031	G3TCT	136	6
17/06/82	2100	2125	G4GLT
18/06/82	1724	1937	G4GLT	134	7
06/07/82	2022	2113	G4GLT	32	15
07/07/82	1920	2033	G4GLT	33	21
08/07/82	2017	2019	G4GLT	49	15
08/07/82	2052	2054	G4GLT
09/07/82	1850	2153	G4GLT	61	10
20/07/82	1900	1940	G4GLT	180	24
16/05/83	1600	1700	F0FDB	99	9
06/06/83	1918	1922	G4GLT	85	11
07/06/83	1030	1032	G4JCC	104	6
21/06/83	1757	1905	G4GLT	117	20
11/06/85	1909	1935	G4BPY	88	13
11/06/85	1917	2046	G4GLT
16/06/85	1308	1330	G4BPY	50	4
16/06/85	1720	1840	G3WBQ
21/06/85	1800	1930	G3TCU	12	10

It can be seen that sunspot numbers vary between 12 and 188. Applying statistical tests to the events in June and July 1982, it can be shown that there is no significant difference between the numbers on Es days and a random selection. (The method used is the Kolmogorov-Smirnov goodness of fit test for completely specified distributions - see Elementary Statistical Tables, H R Neave. George Allen And Unwin Ltd 1981).

We turn next to the relation with the planetary geomagnetic index, Ap, the average daily amplitude, prepared by the University of Gottingen.

Cumulative distribution of Ap

Fig 8 shows the distribution of Ap values for each June and July between 1980 and 1984. On the same figure is shown the distribution of Ap for days on which transatlantic Es occurred (both W/VE and FY7). We can see that nearly 70% of Es days have Ap less than 10, compared with only 45% of all days in June and July 1980-84. Therefore days with low geomagnetic activity are best for transatlantic Es. Applying the Kolmogorov-Smirnov test as before we can compare the distributions. The results show that the distribution for all Es days (W, VE, FY7) is significantly different from a random selection at 99% confidence.

Comparing the distribution of Ap for the "FY7" days only, there is a significant difference at 90% confidence.

Conclusion

Transatlantic propagation at 50MHz in the summer is not sunspot related, nor is it transequatorial. The correlation with low geomagnetic index indicates that auroral-E is not involved. The most likely explanation is multi-hop Es.

“Transatlantic propagation by Sporadic E at 50MHz” - RadCom, July 1986

This article analysed the incidence of transatlantic openings and showed through a statistical analysis that summer time openings to the UK were not sunspot-related, nor transequatorial, but that multi-hop Es was the most likely explanation. The following text is an abridged version of the original article.