Efficiency

[cutty]

Hi All
I realise that storms are moving phenomena but the movement is relatively slow and my thought was that there might be a simple way of tuning a detector. Well tuned detectors must benefit the network as a whole. Not everyone would have the time to analyse a lot of historical data and, like me, many have probably not learnt enough to use it effectively.

Maybe something along the lines I suggest would be a sort of 'Quick Start Guide' for those with limited time and/or knowledge run in parallel with the 'proper tools' if the server capacity would allow it. Brian has pointed me to the Strikes versus Signals graphic which gives a guide as to what I am doing but not what I should be achieving in comparison to stations which should be showing similar results.

As recently we have has storm clusters distant from each other I have been able to improve my performance somewhat by watching other stations fairly close to me. What I am suggesting is a formalisation of that.

I see the point of Egon's grading system but the spread of the system must mean there are some who 'don't have the time or knowledge/lost interest' who be be stuck in group one but might be motivated by a simple method of improving.
Regards
Alan

Now, think a moment... We don't 'tune' these systems or antennas.. we do optimize
Decide whether your want a high effectivity / efficiency number. That would indicate 'station' performance.
My suggestion does NOT relate to 'network' figures... only what an operator has decided.
With today's network density, except for those fringe areas, where stations might elect to go 'group TWO' if they can stay 'efficient'...
why would 'high signals' low stroke' stations not desire to be more accurate. efficient, effective, within a 'smaller radius'?
My systems normally are operating with 'radius' of 30-1200 KM --- and the only thing that knocks my numbers down, as a station,
is my dadgum sporadic noise makers that don't trigger interference mode.  In my location with my environment, for me to
send 60000 signals trying to hit 'South America' as I once could, years ago, is foolish, and does NOTHING for network except
send a bunch of signals that won't be used for locating anyway...
So as far as my 'station / location' 'quality' my station can NEVER perform as it used to, or is capable of if some local parameters
were changed... So, I simply optimize down for better overall performance, since the network does NOT work any better with
all my extra signals, even if they're strokes, greater than about 1200km.

[mwaters]

I don't know if this has been mentioned here yet, but a station's efficiency, ratios, etc. can vary from very high to very low depending on where the storms are in relation to it. Perhaps that should be taken into account?

[mwaters]

Hi All
I realise that storms are moving phenomena but the movement is relatively slow and my thought was that there might be a simple way of tuning a detector. Well tuned detectors must benefit the network as a whole. Not everyone would have the time to analyse a lot of historical data and, like me, many have probably not learnt enough to use it effectively.

Maybe something along the lines I suggest would be a sort of 'Quick Start Guide' for those with limited time and/or knowledge run in parallel with the 'proper tools' if the server capacity would allow it. Brian has pointed me to the Strikes versus Signals graphic which gives a guide as to what I am doing but not what I should be achieving in comparison to stations which should be showing similar results.

As recently we have has storm clusters distant from each other I have been able to improve my performance somewhat by watching other stations fairly close to me. What I am suggesting is a formalisation of that.

I see the point of Egon's grading system but the spread of the system must mean there are some who 'don't have the time or knowledge/lost interest' who be be stuck in group one but might be motivated by a simple method of improving.
Regards
Alan

Yes. There is a thread "Should we have a FAQ?" but it hasn't had any activity for weeks.

[cutty]

I don't know if this has been mentioned here yet, but a station's efficiency, ratios, etc. can vary from very high to very low depending on where the storms are in relation to it. Perhaps that should be taken into account?

Doesn't that proposed suggestion accomplish that very thing?  It DOES NOT look at 'network'... only the station's activity.

[Alanpenwith]

Hi
Sorry if I used the wrong term. By tuning I meant setting the station up to the best advantage of the network. 
So IF it were possible to pick a storm and see the results of stations likely to agree with my attempt to optimise to 1200km it gives me something to work on. As I see it I don't need or want to compare my attempts with stations seeing 5000km strikes.
It is more or less what I have been doing and it seems to be working. I have watched the live maps and my signals chart to see if stations around me are reacting similarly. It works after a fashion, it's fairly quick and easy - my stats have improved so my idea was only to see if the principal had any mileage. 
Sometimes a person seeing things with a different perspective can spark a new train of thinking - or maybe not !
Regards
Alan

[pasense]

@Cutty:
first problem with the proposed figures is that they do not discriminate between real sferics and noise. The signal sent category contains both and we would first like to reduce noise


second problem is what mwaters wrote: the position of storms both affects the signals sent and the strokes detected, but in different ways. For my station at the moment very often the storms are in a favorable position and I get sometimes very high participation in the detected strokes, so the stations performance looks good. When the storms are in a less favorable position relative to my station, the performance looks lousy. So you could judge the station either as "excellent" or as "lousy", depending where the storms are. 


Third problem is that many, many stations send signals of real sferics that are much distorted by noise. That their signal can be used anyway looks like a miracle to me.


Another serious problem has been spelled out clearly by Alanpenwith: we need tools that show us quickly how a station is performing without looking at lots and lots of actual signals from our and other stations. I do not think that this can be achieved with only these two numbers proposed.


I propose to make a list of examples for good sferic signals and also for interference noise, and for clean signals and noisy signals, together with hints how to improve the situation. People could then browse this catalog and check how their station is performing.


In addition waterfall plots would help, histograms of signal strength of sent and used signal, etc.

[readbueno]

Pasense, In the small storms that we have had in the North Atlantic for the last 48 hours or so I have been actively monitoring eight stations in the area of interest.
The "signal to stroke ratio" for these stations varies from on average, only over this time period, from 2 signals with 1 stroke, to 8200 signals for three strokes:
i.e. In the last hour and taking the figures hourly your station's figures are 9,725 signals sent 76 stroke recorded. This give the basic ratio of about 130:1 for your station
My figures over the same period are 135 signals sent for 63 strokes recorded this ratio is 2.14:1.
I have only included stations that were recording strokes in this storm, Their squares were lighting up on the map.
They were all Blue stations running ferrite antennas and E field of reasonable size (100 mm).
There is a big disparity between the figures that somehow needs to be explained.
These are stations in reasonable geographical area, using the same equipment, so how come the big difference in ratios, between signal sent and strokes recorded, i.e. 2:1 to 2600:1?
This was based on average, over a short amount of time, using a small sample, but these are not the worst nor the best stations on the graphs or in the lists either.
If this sort of measurement was taken for all stations, which the server could probably do, much easier than one person just doing the sums on a calculator, we would have a better picture about the state of the network in any given period and geographical area, in relation to the amount of lightning activity and ambient noise.
With this information we could then see where our individual station rates are over a different period of time and the affect of any adjustments that we make.
Taking short "snapshots" and saying, "Yes, today I am near the top or near the bottom of the list." is very volatile and changes from minute to minute and is maybe not the best way of assessing either the stations performance, regardless of what it is called, nor probably is it a good way to establish the "relative health" of the network.
I have asked similar questions to you about What is an acceptable signal, and about various forms of noise and interference, in these forums, and I have read the forums in Wx Blitzortung as well, I have responded to the "Should we have a FAQ" forum etc., etc.
Yes, it would be nice if someone with the skills could make new tools or continue in the development of the tools that we do have.
This also takes time and we have to be clear what we want the tools to do.
Basically, we must each take responsibility for our own stations, within the guidelines established, or to be established, and make the best possible adjustments and positioning of our stations as we can.
Not everyone is in an ideal situation in regards to this and we should not expect perfection, however, even given the average stations limitations and the "time" commitments of the station owners, we should be able to do better than what the graphs and the quality of some of the signals in the archives are showing.

I have made several suggestions, both in these forums and in conversation with several individual stations.
They amount basically to this:-
1. Reduce your gains to a much lower level, yes, you may lose a few distant strokes, but your overall efficiency will be much higher.
2. Increase the thresholds to a good level, this will stop many forms of annoying interference triggering the station, and "local lightning is always going to exceed the threshold anyway.
3. Make use of the station itself as a tool to find your local sources of noise.
This can be greatly aided with a Long Wave AM radio, tuned to a reasonably unpopulated frequency and just walking around your house and your locality noting what the prominent sources of noise and interference are.
At least some of these will be in your control and you should be able to deal with these.
Some may be in the public domain, Noisy streetlights and power lines, etc, these should be reported to the appropriate body.
Some other may not be able to be fixed, although a polite letter or patient conversation may mitigate some of these sources caused by your neighbours.
Some noise and interference is not reducible at source Traffic noise, factories, Military installations etc, and you will have to resort to filters on the control board itself. or using different gains and threshold at times of high noise.
Even in these situations you can still record a lot of lightning, using the nulls on your antenna or shielding, but you will have to live within limitations that others may not have.
If you try just these three things you may be pleasantly surprised!

We are still pioneering and like all pioneers to a certain extent we are making the map as we go along, but at least a good path has already been established and although it is not plug'nplay (Where's the fun in that? You might just as well watch the public screens!) it is not too much of a challenge either.
We still have a long way to go, and a lot of things to do, and I am sure that we will all grow as individual and the network will be even better than it is.
Enjoy, do your best and learn from each other, we all have different skills, and sharing in this activity is a good thing by itself.
Highest regards,

Brian.

[pasense]

Readbueno wrote
..............................................................................................................
I have made several suggestions, both in these forums and in conversation with several individual stations.
They amount basically to this:-
1. Reduce your gains to a much lower level, yes, you may lose a few distant strokes, but your overall efficiency will be much higher.
2. Increase the thresholds to a good level, this will stop many forms of annoying interference triggering the station, and "local lightning is always going to exceed the threshold anyway.
3. Make use of the station itself as a tool to find your local sources of noise. 
-------------------------------------------------------------------------------------------------------------
to 1 taking a look at http://wwlln.net/n showing the thunderstorms active on earth, you may notice that there is a lot of action within reach of the Canary Islands, especially towards the west and south. Attenuation over sea is low, so the sferics can be recorded well here. You should note the the station is also receiving sferics from the Americas and contributing to the location there. In South America and Africa there are simply not enough stations yet so many real signals the station here receives are not yet used by the network.
2.) I have no indication whatsoever of local noise sources. There simple is no "annoying interference" triggering the station, the signals are almost completely from sferics. 
3.)There are no local sources of noise of relevance. The only one I have not really managed to trace concerns only the E field, hence the low amplification on this channel. 


The bottom line is that I chose to run an "island station" aiming at coverage of the Atlantic around and from the figure below you can see that it works: up to a distance of 1000 km the efficiency is 100%, gradually dropping to 80% at 4000km. Then the American continent comes into play and the recognition rate drops.
You chose to run a station optimized for local recognition and that's what makes the difference.
A quality measure of an island station is the S/N ration, where N means real noise and not sferics from distant strokes, and of course the location accuracy. The efficiency and efficacy measure are of little value here.
Cutty has already proposed to make this distinction between isolated stations and integrated stations.

[readbueno]

Pasense, That is exactly what I am saying, your station is not one of the less productive stations and seems to be within the better half of the graph.
The stations that my advice is mainly aimed at are the stations at the bottom right hand corner of the graph some of whom are sending tens or even hundreds of thousands of signals per hour and only recording very few strokes.
I am also part of the Region three network and report many strokes there as well, the graph that you chose to show is at the height of the lightning season here.
maybe you would compare that with today's graph, after all we don't wish to be accused of comparing Apples and Oranges.
You appear to have no local activity recorded at all?
Maybe there is not any lightning to ever be recorded in the sunny Canary Islands unlike in rainy Galicia?

Brian. 

[pasense]

readbueno,
since I started the station here there was no thunderstorm on the Atlantic nearer than about 1000 km, and some scattered thunderstorms over North Africa, so there was simply no local lightning at all. This may change when we get lows from the west, with lightning and the sorely missed rains, but at the moment nothing like this is in the forecasts.