Probability of Lightning Strike

[mechanicalchad]

I have a facility that appears to have damage that could be from a lightning strike. I have data from the local weather office that shows 11 lightning strikes in the area giving both latitude and longitude. None of them are exactly on my location, but a couple of them are quite close. I would like to determine the probability that the location provided in the data report could be my location.

According to an IEEE paper, “[t]he random errors in the direction measurements plus any residual site errors in the direction measurements plus any residual site errors have a standard deviation of about 0.9º. (IEEE, Cummins et al, “The U.S. National Lightning Detection NetworkTM and Applications of Cloud-to-Ground Lightning Data by Electric Power Utilities”, November 1998).

So now that I have the standard deviation can I predict the probability that the measured lightning strike could actually be the one at my location? Can I safely assume a normal distribution? Should I figure out the probabilities for each latitude and longitude coordinate separately and then multiply them out?

Any help is appreciated! Probabilities was never one of my strong suits.

Thanks,
Chad

 

[GregLocock]

If they are quoting standard deviations they have already assumed normality. No, don't separate lat and long, use distance, by pythagoras.

Incidentally do you realise how enormous 0.9 degrees is?

Cheers

Greg Locock


New here? Try reading these, they might help FAQ731-376

http://eng-tips.com/market.cfm?

 

[mechanicalchad]

Thanks Greg!

I figure 0.9 degrees is about 100 km which does seem rather large, and does seem to call all of the location data into question. However, upon further review, that was just one measuring device, and they have an algorithm that narrows that accuracy down. However, they didn't give the algorithms standard deviation which is too bad.