We had a couple of storms go through our area yesterday. With the first round of storms, the Tempest reported almost 2600 lightning strikes within a period of about 1.5 hours. With the second round, the Tempest reported a little over 1900 strikes within a period of about 1 hour.
Does that strike count seem reasonable? Iād be surprised if we actually had that many lightning strikes within measurable distance in that amount of time.
Iām not sure where you are but Iām in Michigan and we did have a significant mesoscale convective system go through with a LOT of lighting the other night. My lightning count was similar. I think the total lightning count for this type of system traversing the state would be plausible but the tempest is only supposed to be measuring out to what, 25 miles? Thereās just no way IMO.
I will say that when lightning has occurred, the Tempest has been spot on. Iām not getting phantoms because Iām triple checking with 3 other resources I use. To me, the total # is irrelevant as long as itās giving me a heads up notification that itās actually occurring. It really did itās job recently when I was totally unaware of an approaching storm when I had the kids outside and blue skies at that moment.
Tempest is scaled to detect in a circle of 40 miles ā¦ and sometimes even more if the nose level is low.
I think that is 40km, not 40 miles.
true, the official specs show 40 km though I just had one coming with 47 km here and Weather34 seems to catch even way over the limit mentioned.
Guess Franklin gives a āsureā range. But indeed officially it is 40 km
thanks for pointing it out
Thanks everyone for the information.
We had another storm system go through last night and the Tempest measured over 3400 strikes within a time span of about 1.5 hours.
Iām not sure how to verify these numbers are correct. According to the Tempest map, there are a few Air/Sky stations that are somewhat close but none closer than 15 miles or so. Those systems registered a significantly lower strike count than mine for the same time period.
Iām probably grasping at straws here, but my house is next to an easement where high voltage transmission lines run. Could those possibly be playing havoc with the lightning sensor?
I would think if the transmission lines were causing problems, youād see strikes even in good weather. Were you closer to the storms than the other stations you mentioned? If so, that could account for your higher strike count.
yes that was my initial reaction as well, but then I figured that perhaps it is also pretty moist. I donāt know the exact conditions, but there might leak some current through sparks on the electric wires.
@bchelbig 3400 strikes is a lot, but remember it covers a detection area of 1250 square km. For a big storm 3400 strikes might be true in such an area, but it has to be a really, really big storm.
The AIR+SKY stations donāt have the same hardware + backend smarts that the Tempest does. To give you an ides of the difference Iāve attached two screenshots. The first one is of the Tempest with the approximate time I enabled the lightning setting on the AIR, the second pic is from the AIR. Both stations are at my house and within 50 feet of each other.
Again, only compare from the red indicator line on.
i think they both use the same lightning detector. However the parameters defining the sensitivity of the detector can be changed on the tempest. In your case it almost looks as something is wrong with you air unit.
@dsj does the air use the same smarts to add extra lightning that wasnāt detected by the unit? i would think so, but does it?
Maybe I misunderstood but I thought this statement by David meant that Tempest had things that the AIR didnāt for various reasons.
Here is a quote from https://community.tempest.earth/t/lightning-reports-tempest/6262/25?u=gizmoev
And later in the same thread:
As for my AIR being defective, remember that I had lightning turned off until the red indicator line. It didnāt record vary many strikes. We have so few lightning storms here and I got tired of the false positives that I turned it off. I have not had those issues with the field test Tempest units.
I think those statements by David were very general statements not super specific for the tempest (even though he mentioned it). As an example, the raincheck functionality, first introduced for the sky unit, would fit those statements. I donāt think that adding missed strikes back in later via the server, would be specific for the tempest. That is just a software thing. The thing that is specific, is that the tempest can change itās threshold for detecting strikes.
The answer is the Tempest does have things that the AIR doesnāt, both on the device itself (firmware) and in the āback-end smartsā. Yes, in theory we could improve AIRās reporting of lightning using the same approach taken with Tempest. We may decide to tackle that in the future, but we are not doing that yet. As a practical result, lightning reported by Tempest will generally be better than AIR.
PS: My statement that āTempest is more than just hardwareā IS a general statement that includes more than just the Tempest hardware device. AIR & SKY are more than just hardware too and, in fact, are part of the overall Tempest System!
PPS: It is a bit of a semantic challenge to distinguish the āTempest Hardwareā separately from the āTempest Systemā. We recognize that we probably need a better name for the piece of hardware that is currently referred to as āthe Tempestā but we donāt have one yet. For now, the piece of hardware that you install at your house is āthe Tempestā and everything else (including the piece of hardware) is the āTempest System.ā
Tempest Station ?
Tempest Sensor Suite ?
(anything that gets us the chance for more Yeti mugs!!! 
4 posts were split to a new topic: Go wacko inventing a better name for the tempest hardware
Those are real, validated strike counts. Lightning strikes the Earth about 100 times each second on average. Thatās about 8,000,000 strikes per day on average.
Letās look at some radar and lightning archives from that June 3rd afternoon Derecho event that raced across PA and through NJ. Not only was it super cool but actually a super rare storm event you were right in the middle of. With a lightning activity radius of about 40km, itās no wonder your Tempest reported so many strikes.
Hi David,
iām feeling a bit disappointed. implementing false lightning suppression for the tempest, but not for the Air even though it is just a software thing. Will Sky/Air users be left behind? Usually it isnāt the best thing for software development to create a separate branch for newer hardware. If one does, it almost always implies that no development or improvement will happen to the other branch. Too bad.
I did a quick analysis of the lighting strikes today, comparing the Air and Tempest. The data, while not one for one identical was similar enough I could match several entries. Iām impressed with the direction WeatherFlow is heading.
Thanks for the information. Itās good to know that the system is functioning properly and providing accurate measurements. This discussion has definitely been a learning experience for me.
BTW, Iām loving my Tempest! Keep up the good work!
I came to the forum to ask about some of the strike count date I have received this week. Last week we had some storms come through and the sound was over 2000 and that seems correct and all seemed well. But then on July 6th we had a major storm come through New Jersey and there was a huge number of lightning strikes rain and hail. I found when watching the sound live that day that the court kept returning to 0 and then would count up a few strikes then return to 0. I thought this was odd but wanted to see what it said in the history the next day. That show 0 strikes. When I look at the graph of the lightning from the time of the storm, it is almost solid blue. Anyone experience similar issues?
