What can we expect from a HF propagation model ?

Setting of a VOACAP model (IV)

To be complete, here are all parameters (as of the 2004 release) that must be properly set in a propagation program like WinCAP Wizard 3 using the VOACAP model, two propagation analysis tools that are reviewed in the linked pages.

The transmit terminal (TX)

- Transmit power and losses : use the input power delivered to the antenna less losses in transmission lines or antenna couplers (e.g. checked on a external SWR-meter).

- Power variation with frequency : if the power difference exceed a few dB, it can be compensated by adjusting the antenna gain. In some antenna designs the radiator can be moved away or bring closer to the parasitic elements. However this change affects also the F/B and F/S ratios.

- Select the right antenna for each band : useless if you use a multi-band design. However most advanced programs take into account several antenna types and apply them to a range of bands. In addition, with a directional array, don't forget to set the Azimuth field in the right direction (although you can always work a station using the side or the rear lobe of a beam !)

- Select the correct S/N ratio (aka Service type) : SNR gives you a measure of the communication circuit quality; it defines whether a band is open or closed for a specified reliability level. Being given that it is an important source of error, it requires some more explanations.

SNR is a monthly median signal-to-noise ratio value expressed in dB. It is different in SSB and CW, this latter mode allowing much more noise on bands and poorer working conditions. This parameter is often split in several fields : S/N (SNR), S/N reliability (SNR-Rel) and S/N required reliability (SNRxx). The reliability is the percent of time that the S/N ratio exceeds the required reliability. The reliability is defined as the fraction of days that successful communication may be expected at a given date at the specified time and frequency. It represents thus the expected average performance during undisturbed days of the month. Based on field experience, this is generally considered to happen when the geomagnetic index Ap ≤ 27 and the K-index ≤ 4. By default SNR-Rel = 90%, the industry standard. Consequently the required reliability (SNR-xx) is the S/N ratio exceeded on 90% of the days. However, amateurs content of a SNR-Rel = 50%.

The required SNR reliability doesn't appear out of the blue. It depends on the operation mode bandwidth as follows :

SNR = 10 + 10 log (bandwidth in Hz)

For example, a 3 kHz bandwidth SSB signal gives a SNR of 45. A 500 Hz CW tone is close to 37. It is thus useless to require an SNR higher than 50%, excepting if you want to work in "Hi-Fi" conditions or in modes offering a larger bandwidth.

SNR charts are very instructive once the coverage area from the station has been understood. In a VOACAP model for example "point-to-point circuit graphs" (a propagation estimation along two stations including all their properties) showing the S/NR and SNR-Rel vs. the time and band can be animated through their ranges to better understand when the bands will be open for a particular circuit.

Now, a tiebreaker : we know that the VOACAP model can display many variables. A common question is whether the signal or the SNR level must be used to measure the circuit quality ?

Intuitively we have a tendency to check signal predictions, what most freeware or low cost propagation programs used by default, because most if not all modern receivers have a built-in S-meter, and it is common practice to compare relative receive signal strenght between stations.

However this value assumes that incoming signals are transferred to the receiver without any loss. When the sky wave is reflected by grounds of different dielectric constant and permeability (land, rock, sea water, etc) or penetrate in a noisy environment, it is far to be the case. You can also loss more than 50% of power in using a system badly tuned. So, to avoid taking into account these variations, it is more reliable to use the signal and noise levels (dB>μV) to calculate SNR predictions because they are treated as a ratio that remains unchanged by individual station factors.

The receive terminal (RX)

At the other end of the circuit it is often the big unknown, the terra incognita of your call. This section is often ignored or amateurs set default values as it is often difficult to specify the working conditions of a remote station that sometimes you even don't suspect the existence or the working conditions !

In theory here are the parameters to take into account :

- The remote transmitter power

- The remote antenna specifications (model, gain and bearing)

- The man-made noise level at the remote site (or at least try to know is he works in the country or in a large city where QRM is expected to be stronger)

Tips : If you have not the least information about your remote station, use an isotropic antenna (vertical) with a 3 to 6 dB gain and select the most appropriate type of area (residential, rural, etc). But as soon as you work that station, try to get the specifications of his/her antenna and the type of city in which he or she lives to improve the forecast.

Here also if you have not the least information about the remote antenna system, specify an isotropic antenna with a 3 to 6 dB gain.

Reception Area predictions

Also named Area coverage, this parameter permits to check if your signal can be heard from the specified receive location in estimating the noise level at the remote site.

Area coverage

It shows transmission or reception areas, the latter displaying the result in a chart. 

To get an accurate forecast, and especially to know the Most Reliable propagation Mode (MRM) it is important to match the antenna radiation pattern to the elevation angle set in the simulation. In this way you can determine whether or not your antenna system performs at best

The area coverage permits to compare the effects of using different antenna designs (e.g. dipole, vertical, beam, etc). In addition the effect of the ground can be estimated (signal losses over lands, better propagation over seas, etc).

When displaying the reception area it is easy to show the gain of an antenna and see how far goes its radiation pattern (e.g. Yagi vs. vertival).

MUF, LUF and FOT charts

To answer to our first two questions, when we wonder what band will be open at what time to reach such a DX country, a part of the answer can be found in calculating what Maximum Usable Frequency is generally open 50% of the time; this median value is called the MUF. Remember that 50% means a 50-50 chance to work in the specified conditions. That means also that sometimes you will only have 1% of chance to work that DX, at another occasion 100% of chance to work it, but in the average 15 days per month the higher frequency will be open as predicted. It's a good news, but don't go too fast...

The MUF is a statistical value associated to a determined degree of reliability, and thus, it doesn't tell all the story. As all median value, it gives however a good overview of the range of frequencies over which we expect some DX openings. But what's the matter below and above the MUF or using a different reliability ?

We have first to know that the MUF strongly depends on the ionization level of the F-layer, and is used to define the uppermost frequency that is reflected by the F-layer at a distance of 3000 km from the transmitter as shows the graph displayed at right.

Frequencies well below the MUF are affected by the low ionospheric layers at daytime that have a tendency to lift the Lower Usable Frequency, aka LUF. The D-layer shows a strong absorption while the E-layer reflect often shortwaves more than expected. So, on the 40-m band for example, by 11 AM and 3 PM local time in summertime, we can experiment deep QSB due to respectively an increasing and decreasing of the D-layer density and its raising/descent to higher/lower frequency. This QSB affects transmissions with a fading up to 9 S-point during some seconds to some tens of seconds repeating during tens of minutes !  On the contrary, above the MUF your chance to make contacts are almost null. Due to their high frequency, sky waves are not more reflected by the F-layers (F2 or F) and escape into space.

Using sky waves it is thus impossible to work on frequencies too away below or above the MUF.

At left, a first way to display the MUF. This map was prepared for 3000 km radio signal paths by the Solar terrestrial dispatch (spacew) and shows in addition the gray line, the auroral oval and the sun position. Click on the image to get a real-time update from DX.QSL.NET. At right, two other views, more traditional, displaying the MUF and LUF as a function of time. HFProp uses either an iso-contour map showing critical frequencies on the world map for a specified time and circuit (here from ON to FY) or it displays the MUF and LUF curves in a chart vs.time.

The Highest Possible Frequency, HPF, is the upper usable limit exceeded 10% of the time, or 3 days per month, or say in other words, in exceptional conditions. During the 90% of time we should use the Frequency of Optimal Transmission (name after the French original words), aka FOT. It is defined as the statistical frequency during which the MUF can be exceeded of 85% (what writes also FOT = 0.85 x MUF). It is thus lower than the MUF as is reliability is higher. This range of frequencies spreading between the FOT and the HPF is 4 MHz wide or larger, sometimes so wide that it includes two ham bands. To know the probability to use such exceptional conditions, there is only paramater to check : the "required reliability" of the signal-to-noise ratio (SNRxx or SN-Rel) for the specified circuit.

At last, the MUF predicts ionospheric conditions but without taking into account other variables (noise, antenna gain, etc). As it, the MUF chart can thus be misleading and give disappointing results. It should be used with the SNR predictions to get an accurate measurement of the circuit quality.

Beacon predictions

HF beacons from the NCDXF/IARU international network transmit without interruption in high speed CW (22 wpm) on various HF frequencies. They can help to "feel" the propagation and see openings towards each of the 18 countries from which they transmit.

However, used without more information, the estimation is not very accurate and you need to take into account additional data like the SSN (smoothed sunspot number), MUF, required reliability and receive antenna properties to name some important factors.

Summary chart

As its title states this chart provides a general overview of the circuit from one end to another. It was devised by George Lane, the VOACAP sponsor, at the time he was working with the U.S. Army HF communications.

This chart is similar to a MUF chart but includes all system variables so that you can get at a glance the system integrity at any frequency or time of the day. It is very useful because performance of your station and your chances for making contacts in good conditions depend on the correct adjustment of all variables of the system. Therefore a "good" propagation prediction program, I mean an accurate model and complete should be able to help you optimize your working conditions and provide you the expected success in DXing.

Next chapter

Running predictions

Page 1 - 2 - 3 - 4 - 5 - 6 -