North Queens TV White Space Trials
Our long term test 22 is producing some unexpected results in mid winter we hadn't expected much variation in signal strength, we are saw over 10dBm, the strongest correlation seems to be with temperature as shown in the graph below.
Although snow accumulation in the tree tops does has a negative affect we rarely see it as heavy snow fall is usually accompanied by high winds in our area. When high winds are not accompanied by snow the does appear to be a slight positive correlation with wind speed when temperatures are below freezing.
We have now run this test from January until late June and the findings in the graph above hold true.
The lowest temperature we experienced was -17C and the highest +30C the was essentially no variation in signal strength from +5C to +30C
We can only speculate as to the reason(s) why, a large part of the direct signal path is though the tree tops so variation in signal absorption in this area could have a significant affect, once the temperature dropped below freezing water and some of the water vapour would turn to ice, some of which would fall to the ground especially if a wind was present.
The effect seems most noticeable between zero and -10C and seems to diminish after that although we have less data below -10C at the moment.
It's possible that the microclimate that exists in and just above the tree tops is having a noticeable effect on this type of signal path and perhaps a signal path with 70% of the fresnel zone in free space would not exhibit this behaviour, it would be interesting to compare 2 links of similar length, but unfortunately would be very hard to do in our terrain.
We are hoping to have access to a higher tower in the future which will make such a comparison easier.
We are seeing minimal variation in heavy rain and fog less than 4dbm.
Link signal strength and climate data January to late June xslx file download