Accuracy

[kev_rm]

I have a question, and I don't mean to be antagonistic.  I am a sailor. I have been sailing around the globe for many years and I try to use all of the available information to me to inform my decision making.  In the tropics,  particularly near the equator, lightning is a very real risk.  I have had three strikes within 200m of me, and a direct strike is on the order of $50,000USD and 1 year of time to repair.  I know four friend boats that have been hit.  

My question is this - and please if you have something to contribute, talk about independent observations to calculated results.  Not about "here's the math for the calculated result"

How is  accuracy measured with this platform?  In simple terms - if there is a set of data that represents real actual strikes  or cloud-to-cloud - position and time that we've used to validate?

[mendip_defender]

As a rule we don't compare ourselves to the professional piad for services as we are just a group of hobbiests.

I reffer you to a post made some time ago,
https://forum.blitzortung.org/showthread...8#pid18468

[GeezerD]

Hello Sailor Kev!

A sailboat is pretty much just a well-grounded lightning rod in the flattest possible place, rigged up so it'll float, right? What kind of lightning protection do you use onboard?

If you are talking about detecting lightning strokes from a sailboat, you have some relatively unique needs that are not likely well served by Blitzortung. Blitzortung is a detection network, connected by low-latency internet. I think you would be better served by a single station receiver and detector on your boat, and even there your unique needs, such as the mobile nature of  your antenna, will require special support compared to a stationary installation.

Is it even better over the lifetime of the hull to have a good air terminal, grounding conductor and connections, and grounding plate? Like, in California, it's rather more economical to bolt your house to your foundations than it is to buy the state earthquake insurance they offer, and I can give you four reasons why...

That said ..

Coverage in the tropics, pretty much anywhere south of the Tropic of Cancer and into the southern hemisphere is very poor. Blitzortung is primarily a northern hemisphere, temperate region sort of thing, because of the location of the entirely volunteer detectors and the non-commercial nature of the entire endeavor, as mendip_defender mentioned.

Anecdotally, Blitzortung accuracy is pretty good. I was able to use Blitzortung dynamic maps around my small town in Colorado as a proxy for seeing the flash of lightning. After that, I was able to count seconds to estimate rough distance, and verify both distance and rough bearing when I heard thunder. It was fun.

Compared to other lightning detection and location services, blitzortung has some characteristics that you would have to be aware of. Unlike Blitzortung, other services may detect strokes other than Ground to Cloud or Cloud to Ground. When you compare Blitzortung detections to other services, you are likely to see about twice as many strokes in the other services as you see in Blitzortung, because of that. Again anecdotally, I've done this is California for strokes in the Diablo Range and east toward the Sierras, and of the strokes that were reported by both Blizortung and the "other" service accuracy was still good.

D

[MDuperrier]

Coverage in the tropics, pretty much anywhere south of the Tropic of Cancer and into the southern hemisphere is very poor.

100% right, unfortunately. Trying to improve the network and detection ratio in Africa for exemple is a real challenge.

I would not rely on Blitzortung (yet    ) for safe sailing

   Marc

[kev_rm]

I guess I'm not understanding very well - Perhaps I should better define what I mean by accuracy:

Let's say, In objective reality, there are five lightening strokes each at a particular place on the planet in the period of five minutes, in the area of coverage by detectors

How many are detected?

Of the ones that are detected, relative to the actual location of the stroke, what is the positional error? 

Now you say, hey this is volunteer, this is a hobby, this is not a paid service - don't get me wrong, I really really appreciate it! 

But - I think if there is not any actual experimentation, I think that should be disclosed and understood - gospel vs. science. Kind of fundamental.

[cutty]

There are various places throughout the main blitzortung documentation where the phrases 'experimental' 'hobby' 'don't rely on', 'do not use data for...' etc are used.
If you are unable to relate to the 'hobbyist' 'experimental' nature of this project, then you're surfing the wrong site. This is CONSTANTLY revising computations, assignments, etc. Chasing the discharge point (Ground Point) of a C-G stroke is partly math, partly luck. There is NO way any of us are going to claim a "percentage of strokes detected" and report it. Likewise we MIGHT state that our goal is locate a stroke ground point within 1000 meters, and that would be a ball park estimate.  We might refer to it as 'deviation', which can be from meters to miles (km).  Likewise a 'multiple stroke' discharge may happen so rapidly that we sense it as a single stroke. I've PERSONALLY documented stroke within meters, and in a few cases DEAD ON the actual ground point.... others mis-located by kilometers.  So, on some of the map displays you can see a 'max deviation' by mousing over a new strike... that will be displayed in nsec, most likely.  That can be translated by converting to distance. Note that a lightning impulse DOES NOT travel at the speed of light, however.... maybe 'half' that rapid speed, depending. So you can say any 'deviation' or 'accuracy' or 'efficiency' is based on a 'half-fast' calculation. Pun intended. 

https://docs.lightningmaps.org/general/l...detection/

[mendip_defender]

This might better explain Accuracy and Precision,
https://sciencenotes.org/what-is-the-dif...precision/