5MHz Forecasting – Project
On this site there is available both the outcome of the project and an article that introduces and explains the basics. However, this does not tell the whole story; space and other constraints are limiting factors.
The whole thing started as an idea, which in essence is quite simple. That is, to use the 5MHz database’s recordings to forecast future events. How do you do that, well again it’s quite simple, gather all the recordings together that are similar to those that are envisaged and statistically work out the answer. The heart of the problem whilst said out loud can easily be grasped. However, how is this going to be achieved with nearly 1.5M entries in the database?
This is where Roger Pettett, G7TKI came in. Luckily enough we are both members of the same local Radio Amateur Club, C.A.R.S. At our first meeting during a reorganisation of unwanted items around Chelmsford, we discovered a convergence of interests. Roger liked my idea and saw the possible potential, his background in computing technologies and experience in manipulating vast databases were the catalyst to get the idea moving.
Once the various parameters had been worked out and the tolerances that they would be used within had been established, G7TKI went away and designed the program. It is written in PERL, a big plus because at least I understand the rudimentary basics of this language. It was not long before we had a working system, it only remained then to fine tune it. Fine tuning such a program was one of the initial components. It fell upon me to perform this task; I used the 5MHz beacon chain itself to verify each forecast. Of course I checked each element of the output against the database, not straightforward with so many database entries. There are only three beacons but their geographical distribution enables a good approximation of all U.K. latitudes.
This frequency gives excellent U.K. coverage if NVIS techniques are employed. The main deficiency of this is at night when the foF2 (F2 Critical Frequency) drops below the frequency. Also occasionally during the lower parts of a solar cycle, it is possible that the foF2 will never reach this frequency, but this is relatively rare.
Any comments/suggestions e-mail – firstname.lastname@example.org