51 minutes ago
Hello,
I have a question regarding the algorithms used to select and involve station signals in lightning location calculations.
My station 2891 is located in a relatively remote region, but it should nevertheless play an important role in detecting lightning near the edge of the network coverage.
During the last season, I noticed that my station is involved in lightning calculations much less frequently for strikes located 0–1000 km to the west of my station. This is exactly the direction where, theoretically, the station should be particularly useful for closing the triangulation geometry.
For some reason, the servers often do not seem to use my station for these calculations, even when, from my perspective, the received signals are of very good quality and the distance is not large enough for the signal to become problematic or overloaded.
I have attached several images illustrating this. I marked in red the areas where the servers use my station significantly less often, compared with the areas marked in green.
In the red areas, I can see my station participating in the calculation for at best about half of the lightning strikes, and often considerably less than that.
At the same time, my station participates in the calculation of distant lightning strikes over Turkey, Iran, and the Caucasus mountains almost all the time.
I have also noticed that almost all stations around Moscow, as well as newer stations located within approximately 300 km of my station, do not seem to have this problem. Their signals are apparently accepted by the servers for lightning detection to the west without any significant reduction in participation.
I have tried to reproduce the settings used by several Moscow-area station operators, including similar antenna parameters. I also experimented with different GPS parameters and tried using the same thresholds and filters. However, none of these changes significantly affected the situation.
This makes me wonder whether there is some aspect of the server-side station selection algorithm that I am missing.
It seems somewhat unusual to me that the station is regularly used for much more distant lightning in the south, including Turkey, Iran and the Caucasus, while being used much less frequently for relatively nearby lightning to the west.
I would appreciate any explanation of how this selection works and whether there is anything I can adjust on my station to improve its participation in western calculations.
Thank you!
I have a question regarding the algorithms used to select and involve station signals in lightning location calculations.
My station 2891 is located in a relatively remote region, but it should nevertheless play an important role in detecting lightning near the edge of the network coverage.
During the last season, I noticed that my station is involved in lightning calculations much less frequently for strikes located 0–1000 km to the west of my station. This is exactly the direction where, theoretically, the station should be particularly useful for closing the triangulation geometry.
For some reason, the servers often do not seem to use my station for these calculations, even when, from my perspective, the received signals are of very good quality and the distance is not large enough for the signal to become problematic or overloaded.
I have attached several images illustrating this. I marked in red the areas where the servers use my station significantly less often, compared with the areas marked in green.
In the red areas, I can see my station participating in the calculation for at best about half of the lightning strikes, and often considerably less than that.
At the same time, my station participates in the calculation of distant lightning strikes over Turkey, Iran, and the Caucasus mountains almost all the time.
I have also noticed that almost all stations around Moscow, as well as newer stations located within approximately 300 km of my station, do not seem to have this problem. Their signals are apparently accepted by the servers for lightning detection to the west without any significant reduction in participation.
I have tried to reproduce the settings used by several Moscow-area station operators, including similar antenna parameters. I also experimented with different GPS parameters and tried using the same thresholds and filters. However, none of these changes significantly affected the situation.
This makes me wonder whether there is some aspect of the server-side station selection algorithm that I am missing.
It seems somewhat unusual to me that the station is regularly used for much more distant lightning in the south, including Turkey, Iran and the Caucasus, while being used much less frequently for relatively nearby lightning to the west.
I would appreciate any explanation of how this selection works and whether there is anything I can adjust on my station to improve its participation in western calculations.
Thank you!