NASA Pinpoints Lightning The Old-Fashioned Way

ke4roh writes: "As a child, I would watch a lightning flash and count the seconds until I heard the first clap of thunder. Get three kids counting in different places, and you could figure out where that cloud-to-ground strike was by coordinating their counts. That's the premise behind NASA's latest lightning detector, according to a press release. It uses a radio to detect the strike and four microphones spaced about 20 feet (7 m) apart. The neat part is its accuracy - about 15 feet (5 m) within a 1 mile (1.6 km) radius. The information should help them determine if lightning may have damaged sensitive launchpad equipment."

Re:fp

[uncoveror]

They are also working on a new launch vehicle [uncoveror.com] that won't be so sensitive to lightning.

neat.

[Hadlock]

neat. so i'm guessing these are directional microphones, in a semicircle, pointed twords the launch pad. sounds like you could build your own "lightning dectector" with an old 486, and a couple of ISA sound cards... just run a script waiting for an abnormality in the TV/FM tuner in the box, and then start recording... could be an interesting project.

How it works

[ke4roh]

Actually, the microphones don't have to be directional. First, the antenna picks up the radio noise of the lightning strike, then some time (measured) elapses, and the first microphone receives the signal. At that moment, you know the lightning struck on a circle of appropriate radius around that one microphone.* Wait a fraction of a second, and another microphone hears it. You can then draw another circle around the second microphone, and it overlaps the first in only two places. Wait until the third microphone picks it up, draw your third circle, and they will only overlap in one place. Add another microphone for increased accuracy, and you're done. The microphones could be in any configuration, but the farther apart they are, they more accurate the results can be.

* Try this at home: Count the time until lightning arrives. It's about 5 seconds per mile (3 seconds per kilometer), so divide the number of seconds by 5 and you get the lightning's distance in miles (by 3 for km). If you know the distance to the lightning (without the direction), you know that the lightning struck somewhere on a circle with that radius and you at the center.

[BTW: For more unit conversions than you can shake a stick at, visit Russ Rowlett's Units of Measure site [unc.edu] which helped me check the numbers above.]

Why pick up audio when RF is available?

[martyb]

I was pondering this very thing a few days ago when a lightning storm passed by. I turned on my AM radio and discovered there was quite a range of frequencies that would pick up when there was a flash of lightning. So, my thoughts were to use a number of AM radio receivers, arranged equidistant around a circle, each sending their signals to a processor located at the center of the circle.

Now, it appears that the NASA folks are using microphones to pick up the sound of the thunder. Thunder is just the noise made by the lightning. Why not pick up the RF field, itself? I would think there would be less distortion and a clearer signal. Instead of 4 microphones, use 4 AM radios. Otherwise the technology would be the same in measuring the path differences (which was so clearly explained in the parent post).

Now, I'm aware the RF signals can also experience distortion (multi-path, etc.) so I'm not claiming that AM receivers would be a panacea, but I would think that it would be more precise than anything that could be done with microphones.

The preceding is based on what I'd learned in physics classes back in college... Is there anyone here who is more versed in RF signal processing who would care to comment?

Re:Why pick up audio when RF is available?

[morcheeba]

answer:
1. sound is much slower than radio waves, so the electronics can be slower and simpler (and therefore cheaper). To "slow down" a radio signal, you'd need a large diameter ring of sensors -- something that may not fit on top of a launch tower
2. Lightning is a one-shot deal; it's far easier for electronics to measure delay as a phase angle, which would require a continous waveform.

fun fact: different frequencies travel at different speeds. Thats why far-away thunder "rolls" for a period of time, while near lighting cracks. Theoretically, you don't need the RF component to determine distance. (Fibre optic guys hate this natural phenonemnum because it distorts what they send.)

Re:Why pick up audio when RF is available?

[martyb]

answer: 1. sound is much slower than radio waves, so the electronics can be slower and simpler (and therefore cheaper). To "slow down" a radio signal, you'd need a large diameter ring of sensors -- something that may not fit on top of a launch tower

Thanks for the feedback! It prompted me to do some calculations and further investigation. Apparently, there are TWO different SOLLO systems: SOLLO 1 [nasa.gov] and SOLLO 2 [nasa.gov]. (Note: There's an error in the HTML for the SOLLO2 page where a less-than-symbol can be mistaken as the start of an HTML element.) Here's the description for SOLLO 2:

The basic concept of SOLLO 2 is the estimation of the distance of a lightning strike from the difference between the time of arrival of the visible flash and the audible thunder. SOLLO 2 locates lightning strikes to within a meter radius. The single receiving station includes an electric-field sensor at the center of a horizontal circle of 1- or 2-meter radius, four microphones placed on the circle at 90-degree intervals, and a fifth microphone 1 or 2 meters above the center. A nearby lightning strike causes the electric-field sensor to put out a pulse that is used to start the timer and to trigger the digitization and recording of microphone outputs as functions of time. The differences between times of the arrival of thunder at the five microphones range up to a few milliseconds. These differences are determined with[in] 10 s, from real-time digital cross-correlation among the microphone outputs. The direction from which the thunder came (thus the direction to the lightning strike) is computed by finding the set of direction cosines of the sound-propagation vector that yields the least squares best fit to the time-of-arrival differences, given the known speed of sound.

So, to answer my own question of why not use just RF? The speed of light is approximately 900,000 times the speed of sound; let's just call it one million (10^6) to make the calculations easier. Then, the differences between times of the arrival of thunder at the five AM receivers would range up to a few nanoseconds and the differences would need to be determined within 10 picoseconds. Ouch! I'm no electrical engineer, but that looks like terahertz to me. Getting it back down to gighertz would require expanding the baseline by a factor of 1000. That implies arranging the RF receivers around the perimiter of a circle 2 km in radius. SOLLO 2 would fit in the back of a pickup truck; mine wouldn't be quite so portable. =)

Re:Why pick up audio when RF is available?

[morcheeba]

cool, thanks for the research!

Re:Why pick up audio when RF is available?

[ke4roh]

Would the RF signal be that clear? Consider that for timing purposes, you just listen for the first crack of thunder, and likewise listen for just the first bit of static from the lightning. But an electrical storm is full of little bits of static discharge, and you'd be interested in its location on the ground - would that also be the first bit of signal received? Only when the lightning travels ground to cloud, otherwise the signal from the cloud end would show up first. Since lightning travels slower than light speed, it's hard to tell where on the ground the strike was by observing the first bit of static. Lightning travels faster than sound, so you can listen for the first sound, and it will have come from close to the ground. By using the EM radiation together with sound, you avoid all the complexities of the lightning's path effecting the EM radiation or the sound.

NASA "Pinpoints"...

[Wrexen]

its accuracy - about 15 feet (5 m) within a 1 mile (1.6 km) radius.

I don't know what kind of pins you're using, but they're probably not very effective at that size...

Re:NASA "Pinpoints"...

[quintessent]

Well, that depends what you're doing with them. They might be useful for crushing buildings, for instance.