The Left Scale:

left nominal
The color bars are a Frankfort Weather addition to the server produced graphs. They indicate the nominal performance regions expected of the three Frankfort Weather Blitzortung systems. This is a percentage / ratio scale, from 0 to 100% explained below. It does NOT measure the actual 'signal count' expressed in real numbers on the right hand scale.

The below explanations are by the developer, Dr Egon Wanke.
Dr. Wanke mentions 'adjust threshold ' several times. Normally Frankfort Weather will avoid threshold adjustments, preferring to leave them at the base ADC reference of 100mv. We may quickly adjust the gains of the analog amplifier stages in an affected channel, rather than the digital portion, due to the volatile noise environment at the Frankfort Weather location. For best overall performance, we avoid 'Automatic' modes, operating with 'manual' settings. It IS possible to achieve very fine discrimination using Gains / Threshold, 'Ignore Signals Below Threshold Percentage ' and other filters. Unfortunately, our local EM environment is too unstable for such precision.

The White Curve:

The white curve shows the absolute number of signals sent over the past two hours. The scale is on the right side of the table. The value is normally between 100 and 10000. With high values (> 8000) there is usually a fault, see also the explanation of the red background lines. In the event of interference, set the gain lower or set the threshold somewhat higher (150mV)

The Yellow Curve:

The yellow curve shows the relative number of signals sent that were accepted after simple checks. Here it is checked whether the signal is actually a pulse. The yellow curve should mostly be over 80%. Otherwise, the signal / noise ratio is either too small or the signals have no pulse shape or are superimposed too strongly with other signals. If the yellow curve is low and the number of signals is high, then usually only noise is received. Set the gain somewhat weaker or set the threshold somewhat higher (150mV). Reposition your antenna to a different location to compare the different locations.

The Green Curve:

The green curve shows the relative number of accepted signals that could be assigned to the localized lightning strikes. This should be at least 10%, the more the better. If the thunderstorm is very far away, high values cannot be reached here. However, if the green curve is always zero, then the received signals are exclusively noise with a pulse shape and not flash signals. Please check your antenna system and location.

The Blue Curve:

The blue curve shows the relative number of assigned signals that was actually used for the calculation. Only a suitable selection of the assigned signals is used for the actual calculation. If the blue curve shows high values, your station is extremely important at the moment for calculating the lightning strikes of the current thunderstorm. In general, the blue curve will always show something when the green curve shows something. If the blue curve shows nothing, even though the green curve has high values, then this is not a problem. In these cases, you just have to wait until the thunderstorm is in a reception group that you cover better than the other stations.

The Vertical Lines

The vertical lines indicate error situations. The most common errors are bad GPS (blue) and interference (red). If the GPS is poor, you should check the GPS antenna. In the case of interference situations, you should observe whether you appeared regularly at certain times. As a result, the sources of interference can often be localized, e.g. Christmas lights after dusk. Note that you can display the graphic up to the minute grid.

NOTES

Another basic thing: Do not operate your detector permanently in automatic mode, but always in manual mode in the long term. In order to find the correct amplifier setting for trouble-free operation, you have to watch your station for a long time, during thunderstorms and without thunderstorms. Leave the threshold in the range 100 to 150 mV. It is more appropriate to work with a lower threshold (100mV) and a lower gain.

Single numerous indices about the efficiency or effectiveness of the detector have a high entertainment value (after all, our hobby should also be fun), but are not well suited to adjust the station. I will show you examples of well-tuned stations and stations that are not so well tuned

. — Dr.Egon Wanke, April 2020