Contacter l'auteur / Contact the author

Recherche dans ce site / Search in this site







Software review

A point-to-point transmission forecast set for a power of 100 W in August 2004. Only the year and month are used (CCIR coefficients) with a SSN of 85. The path is set between Brussels and San Francisco using a Yagi at the transmitter side and a dipole at the receive site with a takeoff angle of at least 3. Working in SSB, the SNR is set to 48%. Remain to run the simulation and display the result in a chart.

VOACAP propagation analysis and prediction program (II)

Main menu and settings

After have launched VOACAP that loads quite rapidly, a large screen pops up as shown at right. No embellishment, but an input screen with push buttons where all data are gathered at a glance.

This kind of interface is also shared by the other programs provided with VOACAP like ICEPAC. However, the number of fields and submenus available depends on the model used.

The main menu includes five dialogs : File, Run, View, Save as and Help.

"File" menu permits to save all settings in a .VOA extension file in the subdirectory \saved\default; "Run" menu permits to start the calculations for your specified conditions and create five different output files (Circuit, Graph, Batch, Distance and Time); "View" menu permits to display charts for various output parameters supported by the model. "Save as" menu permits to save either your transmit, receive or the whole circuit using a generic name, and the "Help" opens a short Windows HLP file explaining what represent input fields and their specifications. But this help is neither very verbose nor extended. However, it should read at least once.

As shown at the bottom of the screen in the "Input Help:" message box, a red commment is displayed. Each input field is indeed associated to a contextual help that activates when you move your mouse over the field. IMHO this comment replaces advantageously the information available in the Help menu.

This main input screen is your interface with the point-to-point model. That means that all circuit data must be set first in this screen and associated subscreens before expecting to launch the calculation or display a forecast. In some working conditions, you will get an error message (yes, in a red box !) if you don't respect this procedure.

What is a circuit ? A "circuit" is defined as a complete communication system including users on both ends of the communication channel, both antenna systems, the transmitter, the receiver, the working frequencies, and more.


Some 30 propagation methods or set of algorithms (functions) are provided to establish a forecast, among which the MUF-FOT-Es, MUF-LUF-FOT and FOT-MUF-HPF to name some well-known acronyms. Without more detail, it is better to select the default "Method 20" that takes advantage of all the system performance. But in some working conditions it is advisable to select the appropriate method.

For example, if you work with the longest paths over 10,000 km, "Method 21" uses special functions that are worth to be tested. In between, for target locations ranging between 7 and 10,000 km "Method 30" takes advantage of statistical functions to smooth irregularities between both models. At last if you only need to know MUF and FOT and display the result in a chart, select "Method 8" or "Method 9". However not all have been cross-checked with VOACAP results, and according to George Lane (2001), among them there are Method 16 and... Method 9.


They are ionospheric parameters that take into account additional effects like the presence of oceans or the day of the month. The idea came after engineers corrected predictions using CCIR coefficients applying Transmission Loss tables to determine the variability of the MUF. Substracting data taken between 1958-63 to the current ones, they noted huge difference that were compiled in these both coefficient tables.

The default is "CCIR/Oslo", but they cannot be interpolated when using the date-of-the-month. They lack also of data when the path goes over the oceans, a deficiency of CCIR coefficients. When the date-of-the-day is used in the field "Month" ("Groups" group) URSI 88 (Australian) coefficients must be selected (their name come after 1988, the date of release by Australian researchers). They provide a linear fit between the SSN calculated during the quiet solar cycle (minimum) and the paroxysm of its activity (maximum). And as we told, data over the oceans are also considered.

However, under specific conditions (for example requesting a special ionospheric map without recalibrating the prediction against measurement) URSI 88 coefficients introduce errors in VOACAP calculations, mainly when the foF2 data do not match the epoch at which other critical parameters are measured. This problem was "solved" in using only the month as input. Indeed, VOACAP was initially designed for experimenting with prediction algorithms, but the experimentation was never finished due to lack of funding. Take a look at the list of available prediction methods and parameters, and you will understand what I mean... So excepted a special reason, keep the default CCIR/Oslo coefficients and do never enter the day-of-the-month because you have all chance to ruin the correlation defined for the specified prediction.


The Groups subset provides two field groups, a range of months/days and the associated smoothed sunspot number (SSN), to not confuse with the sunspot number (see below).

VOACAP takes into account a 12-month average of the SSN. The "Month" field can be filled manually or automatically pressing on button "Seasons" (1,4,7,10) or "All MONTHS" (1 to 12). It also accepts a decimal integer which format is "MM.dd" ( For example June 30 must be encoded as "06.30". But as we explained just above, in this case you must select the URSI 88 (Australian) coefficients, and this is not recommended.

The "Groups" subset, used to set the month and SSN values. Do never enter the day-of-the-month or the daily sunspot number or expect errors in forecasts.

In some applications using a simplified interface, the fact to enter the day of the month, forces the application to automatically switch to URSI 88 (Australian) coefficients without warning.

In all cases, using the day-of-the-month, expect to get some "unattended" errors in forecasts. Thus avoid to use it. But there is another reason to never request a forecast for a specified day of the month with VOACAP : this is SSN.

First, we must know that the average solar flux index (SFI) and the smoothed sunspot number (SSN) are correlated by the next relation :

SFI = 63.7 + 0.727 SSN + 8.95x10-4 SSN2

where SSN refers to the F10.7-derived SSN using the Penticton Radio Observatory noon value.

For example, a SFI of 85 is equivalent to a SSN of 28. 

If you lost this formula, use the small G4ILO's HFProp program that converts automatically these values.

Now, if you check on the Internet you will find different SSN values. How to get the correct SSN ? We must follow George Lane's advice, one of the main VOACAP developer who recommends to use the SSN listed at the National Geophysical Data Center at NGDC FTP website (for users who cannot access that FTP site, here is the GIF equivalent content) because other SSNs - like the temporary values maintained at SEC - are supposed to give erroneous results. For the small story, only the NGDC's SSN has been used to calibrate predictions from VOACAP, hence George's recommendation. To well understand this problem, I warmly suggest you to read on this site the next page dealing with the various SSNs, why they exist, then to come back to resume this review.

To read : The short history of the Smoothed Sunspot Number

If during periods of low sun activity the different SSN values (e.g. SSN, SSNf, SSNe, etc) don't differ much from one site to another (e.g. NGDC vs. SEC) or in relation to the real-time values, during the period of sun paroxysm, these averages can be very different from the daily real values. So, whatever the improvement made on the VOACAP engine of your application, if you want to estimate the confidence of your propagation forecast, even if the circuit required reliability is set to the highest value (SNRxx of 90% or higher), always run the simulation with values up to 20% higher and lower than the SSN to see what happens. In the field I have noted that your DX coverage can vary from 10 to 25% !

At last, as you noted, VOACAP doesn't provide any connexion to the Internet to update automatically these fields or to use the system date. Hopefully other interfaces solved this problem to name WinCAP Wizard that calls GeoAlert-Extreme Wizard or DXAtlas that calls IonoProbe to retrieve the monthly update.

VOACAP predictions for a circuit between Belgium and Brazil scheduled in September with 100 W PEP output, a takeoff angle of at least 3, a S/N reliability (SNR) of 50 dB in SSB and a circuit required reliablity (SNRxx) of 90%. For left to right results for an SSN of 122 in 2000 during the maximum of solar cycle 23, an SSN of 27 in 2004 and an SSN of only 7 during the minimum that occured end 2006. In 2000 (left), at 2200 UTC on 14.131 MHz we had a signal strength in Brasilia of -134 dBW or about S4. Another graph predicted a S/N ratio of 24 dB, thus good signals. In a few years we observed a gradual closing of upper bands, beginning with the 10 and 12m bands during the first two years. By 2004 (center), late in the evening the 15 and 17m bands were harder to work or never open to DX in SSB. On 20m at 2200 UTC the signal strength was already close to -160 dBW, in the noise, and the S/N dropped to -1 dB ! End 2006 (right) VOACAP predicted till worse conditions with a signal strength of -182 dBW, and a SNR of... -40 dB ! Not that long distance QSO were impossible to work, even though, but propagation conditions were deeply affected by the low solar flux that did not more ionized ionospheric layers as it did some years earlier. We had to rely on lower bands like the 40m to recover a signal strength equivalent to the one of 2000 but with a lower SNR close to 15 dB at best, but often lower than 10 dB.

As you have understood, the SSN or the solar flux at 10.7 cm (SFI) is the main factor to consider when calculating the propagation for a band. Its influence is as important as the one of your antenna takeoff angle. In addition, if the sun activity (radiation pressure) is not without effects on the ionization of ionospheric layers, it also affects the temperature and pressure of the high atmosphere of the Earth.

This second effect mainly affects the lowest band propagation, a factor too rarely considered in propagation programs that ignore the weather conditions as well as other factors (e.g. Appleton's magneto-ionic theory), and VOACAP doesn't go against the rule.

For example, there is no input in VOACAP for the geomagnetic field, A- or K-indices despite their utility to detect a possible geomagnetic disturbance or a storm. This can be a problem if you want to work via the north pole, during auroral activities, or on the top band, to establish short-term forecasts, or when working with vertical antennas at low latitudes. Here also, more recent applications developed new methods to bypass this problem, using for example real-time ionosondes (e.g. DXAtlas).


This group contains coordinates (longitude/latitude in two digits) and the name of the transmitter location. These data are extracted from several text files sorted by cities, continent, state, DX entity or any other custom selection.

The "Transmitter" group.

You can add as many location files as you want. By default they are saved in the \geocity, \geonatio and \geostate subdirectories.

The standard format is : 


Example extracted from "Europe.geo" :


As you note coordinates are recorded in degrees and minutes. City names or countries do not need to be sorted. This is VOACAP that converts coordinates in decimal values and sort the location file before displaying data on screen.


This group works exactly the same way as the Transmitter and requires no special comment. Other settings related to the noise  level at the receive location are taken into account in the "System " group that will be reviewed on the next page.

Both Transmitter and Receiver settings can be saved in generic files respectively called "transmit.def" and "receive.def" located in the \userdb subdirectory that lists all locations specified in the circuit(s). All are text files that can be modified by the user as long as the format is strictly maintained.


Clicking on the button you can select either a short or a long path (long great circle via the antipode) for the prediction. All depends on your target location, the time of the day and your antenna gain. For VOACAP a long path means also any path longer than 10,000 km and in this case it also uses a different algorithm, "Methode 21",  to calculate propagation.


Whatever frequencies listed in these fields, if you request a graph the propagation calculations are always performed for integer frequencies between 2 and 30 MHz. These range of frequencies are only used to mark the scale on the output graphs. 

In addition they can be used to select the default frequencies or a set of two user-definable frequency lists.

Next chapter

System group

Page 1 - 2 - 3 - 4 -

Back to:


Copyright & FAQ