144 MHz Long Distance Radio Propagation
from Western Europe into the Atlantic Ocean
Analysis of tropospheric inversion layers and the atmospheric refraction index
along radiowave propagation paths exceeding 3700 kilometers
Volker Grassmann, DF5AI
September 6, 2005 - updated December 12, 2005
Tropospheric radio links between England/Ireland and the Canary Islands have been reported several times in recent years and G4ASR even considers 144 MHz contacts between the British and the Canary Isles an every year experience (see, e.g.,  and the references cited therein). In August 2005, however, VHF radio amateurs have reported record-breaking results in tropospheric dxing which will be briefly summarized in this document. To explain the meteorological origin of this extraordinary dx QSOs, we will analyse upper air sounding data and will calculate the vertical distribution of the atmospheric refractive index along the radio propagation paths. Between the northwest of Spain and the Canary Islands, intense inversion layers are found between 1000m and 2000m altitude indicating significant discontinuities in air temperature and, in particular, in air humidity. Thus, significant discontinuities are also present in the vertical distribution of the atmospheric refractive index which has apparently enabled ducting of radiowaves and, in consequence, long distance 144 MHz communication links exceeding path lenghts of more than 3700km.
144 MHz dx band openings in June and July 2005
Considering the many dx QSOs between Spain and Portugal on one hand and the Azores, Madeira and the Canary Islands on the other hand, it is almost impossible to provide a detailed picture of the dx scenarios in June, July and August 2005. It appears that dx distances between, say 1000 and 2000 kilometers, and even more, were an every day experience to many radio amateurs, see the "tropo reports" in . As an example, we will briefly summarize the dx results from Joachim (CT1HZE) located in IM57NH at the south coast of Portugal. Among his QSOs to the Azores (around 1380 km), Joachim has in particular established contacts to Mauretania, i.e. he heard the 5T5SN radio beacon (which is operational since May 2005) and he worked the beacon operator 5T5SN (June 5) representing the very first 144 MHz contact between Portugal and Mauretania (2245 km). Between July 9 and 19, Joachim had several contacts even to the Cap Verde Islands, i.e. to D44TD corresponding to a distance of 2681 km, . This band openings have supported dx QSOs on VHF, UHF and SHF too. On 432 MHz, Joachim worked Fred (CU8AO) on the Azores (June 21, 1953 km) and, again, 5T5SN (July 6) and the latter one represents the Portuguese terrestrial tropo record on 70 cm wavelength. The very first SHF contact between the main land of Portugal and the Azores was established by Joachim and Fred on July 6 (1296 MHz, 1953 km) which represents another Portuguese tropo record, .
Very long dx QSOs in August 2005
Remarkably, 144 MHz radio stations from England and Ireland came into play too. On July 2, Tim (G4LOH), located in IO70JC in southwest England, worked a number of radio stations on the Canary Islands corresponding to distances around 2600 km . On July 16, he also received signals from the Azores (2344 km), i.e. from the CU8DUB beacon.
Note: The above mentioned radio beacons, i.e. 5T5SN and CU8DUB, are part of the Dubus Beacon Project initiated by the Dubus magazine. This remarkable project deploys 50 and 144 MHz radio beacons around the Atlantic Ocean and has activated the CU8DUB transmitter just a couple of days before it was received by G4LOH. The Dubus team around Joachim (DL8HCZ/CT1HZE), Dithmar (DF7KF), Nicholas (5T5SN), Fred (CU8AO), and many others, currently plans to implement another VHF beacon in Tunesia and has already put the Bermuda beacon (VP9DUB) into the air. The project also plans VHF radio beacons in the South Atlantic region, e.g. on Ascension Island.
The big bang came in the morning of August 7: Tim, again, worked several EA8 stations and finally managed two QSOs with Alex, RW1ZC/mm, travelling on a vessel close to the west African sea coast . The distance of 3493 km landmarked a new 144 MHz dx record in the IARU Region 1 . On August 15, Charles (EI5FK) came into play also working RW1ZC/mm in four independent QSOs . Alex's ship has changed its position though and the brand new IARU Region 1 record has been extended even further to 3751 km , see the red line in figure 1. Two weeks later, i.e. on August 29, it was Tim again managing another QSO with Alex corresponding to 3444 km .
The refractive index N
Radio amateurs are aware of long distance VHF propagation caused by tropospheric inversion layers and it appears plausible to keep an eye on inversion layers when interpreting the above mentioned dx QSOs. In this paper, however, we will not discuss neither the creation nor the various types of atmospheric inversion layers (interested readers may refer to basic information on this subject, e.g., in , and to a more detailed discussion, e.g., in  and ).
Tropospheric inversion layers may cause reflection and also ducting of radiowaves resulting in long and very long radio commincation paths . All this phenomena may be interpreted by analyzing the refractive index N which controls the radiowaves' ray path in the atmosphere. Calculating the refractive index N, we obtain a summation of two terms, i.e. N = Nd+Nw where Nd and Nw denote the so-called dry and the wet term, respectively. The dry term is mainly controlled by air temperature and air pressure, the wet term considers the air humidity in particular (see the below formulas). In practice, the dry term contributes about eighty percent to the actual value of N, i.e. the wet term only contributes around twenty percent . However, Nw is typically more important than Nd in tropospheric radio propagation because the wet term is a much more variable than the dry term .
Thus, inversion layers relevant in VHF radio propagation may be identified by discontinuities in the vertical profile of the dry and wet component of the refractive index. In the following, both quantities are initially represented by the air temperture and by the dewpoint which are both available in meteorological data. The dewpoint denotes the temperature where the air's water vapour content begins to condensate corresponding to 100% relative humidity, see, e.g.,  and . In the following, we will refer to the dewpoint because it is a convenient quantity which is measured in Celsius, i.e. the air temperature and the dewpoint may be displayed in the same temperature profile. In the temperature diagrams, dry air is indicated by a more or less large difference between air temperature and the dewpoint, in wet air, on the other hand, the difference is rather small. We will finally compare the vertical profiles of air temperature and dewpoint with the calculated profiles of the atmospheric refractive index.
Upper air sounding data
VHF radio amateurs often refer to weather maps to understand (or to predict) actual dx conditions in tropospheric radio propagation. In the European sector, for example, the University of Cologne distributes actual weather patterns which are considered in particular useful for VHF radio amateurs, see . Weather maps can provide infomation on the horizontal distribution of, for example, air temperature and pressure (at ground level or, alternatively, in higher altitudes) but cannot provide detailed information on the vertical structure of the troposphere.
In this paper, we refer to the upper air sounding data which is distributed in the internet by the University of Wyoming, see . The data provides a large variety of detailed information, for example, on air temperature, air pressure, dewpoint and other parameters between ground and more than 30.000m altitude. This information is available for a large number of places around the world (airports) typically at 00 and 12 UT. By using the internet portal, the user may select various graphical presentations or, alternatively, may access ASCII tables summarizing all data in easy-to-handle data columns. All this information is useful in radio propagation studies and has been applied, for example, in the analysis of the May 20, 2003 dx opening between central Europe and the Canary Islands, see . Readers interested in the backgrounds of upper air sounding may refer to the internet tutorial given by the University of Columbia, see .
Vertical profiles of air temperature and dewpoint
Figure 2 displays the dx QSO between G4LOH in southwest England and RW1ZC/mm close to the west African sea coast on August 7, 2005. The diagrams display the vertical profile of air temperature (red) and dewpoint (blue) along the radio propagation paths, i.e. at Camborne (southwest England), La Coruna (northwest Spain), Funchal (Madeira), Guimar (Canary Islands), Sal (Cap Verde Islands) and Dakar (Senegal), respectively. The meteorological data was taken from the University of Wyoming web site (see  and the blue infobox) and has been post-processed by using a spread-sheet computer program. The left column displays sounding data from midnight, 00 UT, the right column displays the 12 UT data, i.e. this column is most representative with respect to the dx QSOs between 0927 and 1149 UT. The QSO data was taken from G4LOH's personal web site  and the map display is adopted from the BeamFinder analysis software . By clicking on figure 2, the user may download the large version of the graphics.
Tracing the propagation path from England to west Africa, a number of remarkable features attract our interest. At G4LOH's location, we find around 80% relative air humidity between ground and 1000m altitude (see the station Cambore at 12 UT and note the small difference between the red and the blue line). At 1200m, however, the dewpoint (blue line) drops sharply from -0.4C to -11C accompanied by a slight increase of 1 or 2 degree in air temperature (the numerical values are taken from the original data table which is not shown here). Between 1815m to 1900m altitude, the dewpoint returns to air temperature by increasing sharply from -16C to more than +1C reaching +5C at 2100m. Thus, between 1200m and 2000m, we are facing a layer of relatively dry air (16% air humidity), the ambient troposphere shows more than 70% relative humidity though.
Near Spain (see the station La Coruna at 12 UT), we find dry air in all height intervals which is indicated by the relatively wide horizontal spacing between the red and blue line (refering to the data table, the relative humidity decreases from 50% at ground level to 13% at 2000m altitude). Thus, the radiowaves have travelled in dry atmosphere and no significant discontinuity becomes visible in the profile. This is however not true when consideriing Madeira and the Canary Islands, see the stations Funchal and Guimar at 12 UT. Here, we find intense inversion layers around 1800m (Funchal) and 1000m (Guimar), respectively, indicated by a significant increase in air temperature and an even more significant decrease in dewpoint. Further south, those inversion layers however disappear, see the station Dakar at the southern end of the chain of observatories (the positive peak around 8000m may be ignored because it is interpreted an inaccuracy in the sounding data).
In figure 3 and 4, the same type of analysis is shown for the dx scenario on August 15 and August 29, respectively. Without discussing the details, we also notice sharp inversion layers between 1000m and 2000m altitude which appear to concentrate in the sea area from northwest Spain down to the Canary Islands. At lower latitudes, the inversion intensity is much smaller (see the station Sal) and disappers altogether in the Dakar data (which might have something to do with the fact that inversion layers over sea terrain cannot be observed by the continental observatory at Dakar). At higher latitudes, i.e. in England and Ireland, wet ground air develops into dry air at higher altitudes without showing distinct layer-type features though.
Discussing the refractive index N
We have found strong indication that all this dx QSOs were enabled by atmospheric inversion layers between 1000m and 2000m altitude. Interpreting this results in terms of the vertical gradient in the refractive index, VHF radio propagation was mainly influenced by variations in the wet component Nw of the refractive index N, i.e. air temperature inversion has played a role too but was less important here. This is also documented by figure 5 which displays the dry and wet term and also the total refractive index N by using the data from Funchal on August 7, 12 UT. Note the variability of the total index N which is almost exclusively controlled by the variability of Nw althougfh its magnitude is much smaller than the magnitude of the dry term.
The refractive index N should be actually referred to as the modified refractive index which is commonly used in literature because it provides values which are more convenient in practical applications than the orginal physical index n. By using
the refractive index of, say n = 1.0003 now reads N = 300 which, by the way, corresponds to typical values at low altitudes below 1000m. The total refractive index N may be calculated by using the formula (see, e.g., ):
where T denotes the air temperature in Kelvin, p the air pressure in hPa, f the relative humidity (percent value divided by 100). The saturated water vapour pressure es (which is also measured in hPa) is calculated by using MAGNUS' empirical formula (see, e.g., , )
where t denotes the air temperature measured in Celsius. With this mathematical tools and with the Wyoming sounding data, radio amateurs may calculate vertical profiles of the refractive index N in tropospheric radio propagation similar to figure 5.
Considering the so-called standard atmosphere, the refractive index is 325 N-units at ground level which decreases by 44 N-units per height-kilometer . This is also visible in figure 5 when focusing on altitudes not higher than 1200m. Within a small interval at higher altitude, the refractive index however drops sharply by 70 N-units which indeed represents an intense inversion layer. Thus, the height profiles in air temperature and dewpoint (see figure 2 to 4) indeed indicate major discontinuities in the refractive index which have certainly affected radiowave propagation between England/Ireland and the west coast of Africa. We may speculate that all this inversion layers were caused by warm air moving over the cold sea water probably supported by calm weather because no turbulence has destroyed the layer formation and its apparent persistence during the day (this assumption is supported by all observatories which indicate low wind speeds of less than 10 knots at ground level).
Speculations on the radiowave propagation paths
Having identified intense inversion layers between 1000m and 2000m altitude, we however cannot yet explain the true radiowave propagation paths. To explain path lengths of more than 3000km, tropospheric ducting appears an plausible interpretation though. Elevated ducts may guide the radiowaves along a wavelike propagation path through the troposphere and ground ducts, on the other hand, may reflect radiowaves between an inversion layer and ground - both scenarios may support long distance propagation on VHF, UHF and SHF .
Figure 6 and 7 display the vertical profile of the refractive index in the north and in the south of Funchal, i.e. at Camborne and Guimar, respectively (August 7, 2005, 12 UT). At Camborne (figure 6) the refractive index N shows a sharp decrease around 1200m altitude but also a sharp increase just below 2000m. A similar feature, much smaller though, is also visible around 3700m altitude. Thus, we find a layer of relatively wet air (see A in figure 6) embedded within two layers of relatively dry air at, say 1900m and 3700m, respectively, which altogether may act, perhaps, as an elevated wave guide. It is however believed that this feature did not play a major role in the August 7 scenario because of two reasons: 1) the radiowaves were emitted at ground level (note that G4LOH is located close to the Camborne observatory), i.e. due to the first discontinuity at 1200m the radiowaves could probably not access any duct in higher altitudes, 2) the refractive index shows another interesting feature at ground level which appears more significant and which is also visible in the Guimar data (see B in figure 6 and also in figure 7). This feature appears to indicate a low altitude inversion layer that has formed just above the sea surface.
We may speculate that this meteorological feature has supported the generation of a waveguide between sea level and 1200m altitude enabling long distance communication from England down to the west coast of Africa. It appears remarkable, that the same feature is visible in two places separated by 2600 km, i.e. at Camborne and also at Guimar, it is not visible at Funchal though. Nevertheless, this conclusion must be considered speculative and cannot be verified with scientific conviction. It is also important to note that the sounding data cannot resolve the vertical fine structure of the refractive index, i.e. variations of ten meter scale lengths, or so, cannot be identified at all, those features may however also play an important role in tropospheric ducting of VHF radiowaves . On the other hand we may argue, that small scale variations in the local refractive index can hardly represent constant features over distances of more than 3700 km, i.e. small scale features must be considered less important in the interpretation of the August 7, 2005 scenario.
New awareness for dx links between Europe and west Africa
The dx scenario from August 2005 may be interpreted by intense inversion layers between 1000m and 2000m altitude, open questions however remain in the interpretation of the radiowaves' true propagation path through the troposphere. And there is another open question which appears even more important: To the author's knowledge, intense inversion layers in the sea area northwest and west of Africa must be considered an usual rather than an extraordinary phenomenon (note, for example, the vertical profiles in figure 4.5 in the interpretation of the 2003 dx opening between central Europe and the Canary Islands  which displays inversion layers similar to the above figures 2 to 4). We therefore need to raise the question why this type of extraordinary dx openings did not occur more often in the past - perhaps, it did but it was not observed because of a lack of VHF amateur radio stations in western Africa.
In the past, VHF radio propagation between Europe and west Africa could not be considered a subject within the scope of radio amateurs but we may assume that the dx openings in summer 2005 have changed the radio amateurs' awareness considerably. This dx openings have been stimulated by two major achievements in amateur radio: 1) maritime mobile dx operation has significantly increased in recent years, i.e. we may now study ocean radiowave propagation paths which provide dx opportunities not available on the continents; 2) the Dubus Beacon Project has deployed VHF radio beacons in strategic geographical positions supporting permanent dx observations to islands in the Atlantic Ocean and to places on the African continent. From this perspective, the August 2005 dx QSOs may have an importance beyond our present understanding. In ten or twenty years or so, we will possibly consider the summer 2005 dx scenario the door opening event to new experiences in terrestrial VHF long distance radiowave propagation in amateur radio.
Vision or speculation?
In fact, we cannot exclude similar dx openings in the future, e.g. between coastal radio stations in western Europe and, say the Cape Verde Islands extending the IARU region 1 record in tropospheric dxing beyond 4000 km, perhaps (this speculation has been raised by other radio amateurs too, see, e.g., ). To fire your imagination: in 1996, an Australian 144 MHz radio beacon has been observed on Reunion Island (Indian Ocean) corresponding to a distance of more than 6000 km, see  and the references cited therein. Can we have similar dx openings in the Atlantic Ocean too?
Surprisingly, we may find two Brazilian islands in an appropriate geographical position, i.e. the St. Peter and Paul Archipelago (around 6000km) and also the island Fernando de Noronha (around 6500km measured from the southwest of England), see figure 8 (both sites are listed, by the way, by the Brazilian Commission of Geological and Paleobiological Sites ). In both cases, the dx path would cross the same geographical areas which were discussed in this document, i.e. regions where intense inversion layers may be considered no unusual phenomenon. The author has discussed the opportunity of VHF dxpeditions in this part of the world with two experts, i.e. with Flavio (PY2ZX), who has visited both islands in recent years, and with Joachim (CT1HZE), head of the above mentioned Dubus Beacon Project: they are both sceptical , . In fact, the St. Peter and Paul islands are rather small without any significant infrastructure except of a geological research station (see the photo assembly). This is indeed a difficult place for ham radio operation although dxpeditions, see, e.g.,  and also local scientists  put this islands in the air already. A better deal is the island of Fernando de Noronha, perhaps (see the site report in ). Volker Ermert, University of Cologne, has supported our analysis by providing sounding data measured at Fernando de Noronha on August 15 and 29, 2005. Figure 9 displays the August 29 data at 12 UT indicating some inversion around 2000m altitude although its intensity is much smaller compared to figure 5 to 7. The data does not justify any speculation about the dx openings' true extend into lower latitudes on August 29 but, nevertheless, it appears remarkable that signatures of tropospheric inversion layers may be found along the northern half and also at the southern end of a 6500km path from England to the central part of the Atlantic Ocean.
I am grateful to V. Ermert, University of Cologne, Germany, for providing valuable comments and for providing the sounding data of the station Fernando de Noronha. Special thanks to T. Campos, University of Natal, Brazil, for providing detailed information on the St. Peter and Paul Archipelago and to S. Erasmi and C. Kollatschny, both at University of Göttingen, Germany, for establishing the contact. I am grateful to F. Archangelo, PY2ZX, and J. Kraft, CT1HZE, for providing information on ham radio opportunities on the Brazilian islands in the Atlantic Ocean. I also wish to thank the University of Wyoming for granting free internet access to the sounding data and its archives.