How does a stormscope/strikefinder actually work?

[Dave Jacobowitz]

Okay, this question is just about curiosity and the remote possibility
of an interesting DIY project. I'm an EE, so have the background do
understand (but working in the computer field, may not have the recent
experience to do so. ;)

Just how do lightning detection systems work?

It seems to me that the E/M discharge from lightning is essentially
noise, and so would be rather wide-band. So one could easily detect
that a static discharge had occured by looking for a sudden
simultaneous burst of energy in two or more rather disparate frequency
ranges.

But finding where the discharge was... that seems harder. Clearly, one
filter out one band of frequency and use direction antennas and a
little math to figure out the azimuth to the strike.

So now we have a way to detect a strike and work out its angle
relative to the receiver.

But how do you get distance? All I can think of is having multiple
antennas on the aircraft, separate by some known distance, and using
simply time-domain analysis to convert the relative time of flight to
each of the antennas into a position. This would seem to require at
least three antennas to work, and also would require an uncommonly
precise way of measuring time considering that you can't get antennas
very far away from each other on a light aircraft.

So, how do these devices _actually_work? What frequency ranges do they
work in? How many antennas do they have? How do they determine
distance?

thanks,
Dave Jacobowitz

-- jacobowitz73 --at-- yahoo --dot-- com

[Paul Tomblin]

In a previous article, (Dave Jacobowitz) said:
>But how do you get distance? All I can think of is having multiple

I'm pretty sure they guess the distance based on the strength. So a very
strong strike will show as closer than it really is, and a weak one will
show further away.


--
Paul Tomblin, PP-ASEL-IA _|_ Rochester Flying Club web page:
____/___\____ http://www.rochesterflyingclub.com/
___________[o0o]___________

[Peter Duniho]

"Paul Tomblin" > wrote in message
...
> In a previous article, (Dave Jacobowitz) said:
> >But how do you get distance? All I can think of is having multiple
>
> I'm pretty sure they guess the distance based on the strength. So a very
> strong strike will show as closer than it really is, and a weak one will
> show further away.

This is basically true, and the 'spherics receivers have been known to
suffer from that problem. However, my understanding is that part of what
makes the receivers so successful is additional filtering or logic done on
the processing that helps correct for those errors.

Frankly, I suspect the original poster has a better background for
understanding exactly what kind of signal analysis could be used than
anything I could post. But I'm guessing that things like differences in
signal propogation for different frequencies and/or amplitudes provides some
sort of way to compare signal ratios to help refine the distance estimate.

Pete

[David Rind]

Dave Jacobowitz wrote:
> Okay, this question is just about curiosity and the remote possibility
> of an interesting DIY project. I'm an EE, so have the background do
> understand (but working in the computer field, may not have the recent
> experience to do so. ;)
>
> Just how do lightning detection systems work?
>
> It seems to me that the E/M discharge from lightning is essentially
> noise, and so would be rather wide-band. So one could easily detect
> that a static discharge had occured by looking for a sudden
> simultaneous burst of energy in two or more rather disparate frequency
> ranges.
>
> But finding where the discharge was... that seems harder. Clearly, one
> filter out one band of frequency and use direction antennas and a
> little math to figure out the azimuth to the strike.
>
> So now we have a way to detect a strike and work out its angle
> relative to the receiver.
>
> But how do you get distance? All I can think of is having multiple
> antennas on the aircraft, separate by some known distance, and using
> simply time-domain analysis to convert the relative time of flight to
> each of the antennas into a position. This would seem to require at
> least three antennas to work, and also would require an uncommonly
> precise way of measuring time considering that you can't get antennas
> very far away from each other on a light aircraft.
>
> So, how do these devices _actually_work? What frequency ranges do they
> work in? How many antennas do they have? How do they determine
> distance?
>
> thanks,
> Dave Jacobowitz
>
> -- jacobowitz73 --at-- yahoo --dot-- com

You can find a number of websites that can provide a better description
than I can, but the basic concept is that:

1) The intensity of most EM bursts from lightning over a range of
frequencies is such that the strength at a given frequency is
proportional to the strength at other frequencies.

2) Some frequencies suffer very little atmospheric absorption and so
give an unabsorbed measure of the strength of the lightning.

3) Some frequencies are significantly absorbed by the atmosphere.

4) By comparing the unabsorbed frequencies and the absorbed frequencies,
you can make a reasonable guess as to how much atmosphere the EM burst
traversed getting to the receiver, and so can predict how far away the
lightning was.

--
David Rind

[Nathan Young]

On Wed, 14 Jul 2004 22:55:06 +0000 (UTC),
(Paul Tomblin) wrote:

>In a previous article, (Dave Jacobowitz) said:
>>But how do you get distance? All I can think of is having multiple
>
>I'm pretty sure they guess the distance based on the strength. So a very
>strong strike will show as closer than it really is, and a weak one will
>show further away.

This is true, and although it is an 'error' in the actual distance, at
least the error is in the pilots' favor. Keeping us further away from
the stronger storms...

[Dean Wilkinson]

Paul is correct. Distance is determined by the relative strength of
the EMP (electro-magnetic pulse) from the strike, based on
statistically averaged historical lighting EMP emissions data. Most
stikes get placed close enough to the proper distance on the screen to
provide useful data. Direction is determined using directional
antennas.

Dean Wilkinson
B.S.E.E.

(Paul Tomblin) wrote in message >...
> In a previous article, (Dave Jacobowitz) said:
> >But how do you get distance? All I can think of is having multiple
>
> I'm pretty sure they guess the distance based on the strength. So a very
> strong strike will show as closer than it really is, and a weak one will
> show further away.

[Dave Jacobowitz]

Ah! Guessing. I had not thought about that approach. It's interesting
that they can do this estimation and that for the most part it works
pretty well. I mean, at least in the strikes seem to clump up.

I do believe that ground-base strike measuring equipment used for
forecasting and storm tracking can better isolate the position of
lightning strikes -- but in that case, they *do* have the luxury of
spreading out their sensors.

thanks!
-- dave j

(Paul Tomblin) wrote in message >...
> In a previous article, (Dave Jacobowitz) said:
> >But how do you get distance? All I can think of is having multiple
>
> I'm pretty sure they guess the distance based on the strength. So a very
> strong strike will show as closer than it really is, and a weak one will
> show further away.

[Dave Jacobowitz]

Guys,

Thanks for the responses. I was thinking only in terms of the the most
basic first principles of radio propagation, and now I see that that's
probably not a reasonable approach.

Making reasonable assumptions about the relative signal power in
different frequency segments of a lightning strike, and then knowing
something about the attenuative properties of moist air to those
frequency segments, could clearly work, and I don't doubt that's what
airborne lightning detection does.

Of course, now you're in a situation that requires one to know
something about lightning and something about air, but I guess that's
life!

Now I'm really curious to see a spectrogram of several lightning
strikes to see what's predictable about them!

thanks again,
-- dave j, PP-ASEL, no lightning detection on board :(



David Rind > wrote in message >...
> You can find a number of websites that can provide a better description
> than I can, but the basic concept is that:
>
> 1) The intensity of most EM bursts from lightning over a range of
> frequencies is such that the strength at a given frequency is
> proportional to the strength at other frequencies.
>
> 2) Some frequencies suffer very little atmospheric absorption and so
> give an unabsorbed measure of the strength of the lightning.
>
> 3) Some frequencies are significantly absorbed by the atmosphere.
>
> 4) By comparing the unabsorbed frequencies and the absorbed frequencies,
> you can make a reasonable guess as to how much atmosphere the EM burst
> traversed getting to the receiver, and so can predict how far away the
> lightning was.

[Dave Jacobowitz]

Guys,

Thanks for the responses. I was thinking only in terms of the the most
basic first principles of radio propagation, and now I see that that's
probably not a reasonable approach.

Making reasonable assumptions about the relative signal power in
different frequency segments of a lightning strike, and then knowing
something about the attenuative properties of moist air to those
frequency segments, could clearly work, and I don't doubt that's what
airborne lightning detection does.

Of course, now you're in a situation that requires one to know
something about lightning and something about air, but I guess that's
life!

Now I'm really curious to see a spectrogram of several lightning
strikes to see what's predictable about them!

thanks again,
-- dave j, PP-ASEL, no lightning detection on board :(



David Rind > wrote in message >...
> You can find a number of websites that can provide a better description
> than I can, but the basic concept is that:
>
> 1) The intensity of most EM bursts from lightning over a range of
> frequencies is such that the strength at a given frequency is
> proportional to the strength at other frequencies.
>
> 2) Some frequencies suffer very little atmospheric absorption and so
> give an unabsorbed measure of the strength of the lightning.
>
> 3) Some frequencies are significantly absorbed by the atmosphere.
>
> 4) By comparing the unabsorbed frequencies and the absorbed frequencies,
> you can make a reasonable guess as to how much atmosphere the EM burst
> traversed getting to the receiver, and so can predict how far away the
> lightning was.

[Jim Weir]

There are two theories about how to determine distance, and there are two
lighning detector companies in the market. Each of them strongly propounds
their theory to be the best.

In either case, the direction is done with an electronic version of the old ADF
goniometer using an e-field sense antenna and an h-field loop antenna.
Combining those two with an appropriate phase shifter gives you a cardioid
pattern with a sharp null. Phase shift until you are in the null and this gives
you heading relative to the loop antenna.

The question is where to look in the spectrum for the noise. In some tests I
did as a very young engineer, we found that the lightning spectrum peaked around
50 kHz. and one of the systems on the market looks very closely around this
frequency. Their algorithms have done a very good job on predicting range by
signal strength averaged over many strikes. Quite accurately.

The other company says that 50 kHz. gives the maximum amount of energy, but that
looking in a rather broad bandwidth gives more accurate results. The actual
number is a trade secret, but my suspicion from the components involved is that
they look in a noise bandwith of a few hundred Hz. but sweep the range from 50
kHz. to somewhere in the 3 MHz. range. As a function of WHAT they hear and
comparing one frequency to another for the same strike, they predict distance.
Quite accurately.

Howzat?

Jim



(Dave Jacobowitz)
shared these priceless pearls of wisdom:




Jim Weir (A&P/IA, CFI, & other good alphabet soup)
VP Eng RST Pres. Cyberchapter EAA Tech. Counselor
http://www.rst-engr.com

[James M. Knox]

Jim Weir > wrote in
:

> Howzat?

As usual, pretty durn good.

From my conversations with a couple of the BFG guys that worked the "series
I to series II" transition, the first real trick was to accurately
characterize the lightning RF data. In several parts of the country they
put up triangulation systems (ground based). This allowed them to
accurately pinpoint any strikes in area. They then outfitted a couple of
planes and flew them around the outskirts of any storms that moved through.

The result from the planes was the recorded RF broad-spectrum energy
patterns from strikes at known distances. It was from this that they found
that they could do a good estimate of distance by essentially matching the
basic pattern to one of several basic shapes (this is partly how they
exclude cloud-to-cloud) and then comparing the difference between the
"paradigm" shape and what was actually measured.

The rest was just a lot of high-speed (for the time) DSP.

jmk

[G.R. Patterson III]

"James M. Knox" wrote:
>
> It was from this that they found
> that they could do a good estimate of distance by essentially matching the
> basic pattern to one of several basic shapes (this is partly how they
> exclude cloud-to-cloud) and then comparing the difference between the
> "paradigm" shape and what was actually measured.

Dumb question time. Why exclude cloud-to-cloud? Wouldn't that type of lightning also
indicate conditions one would wish to avoid?

George Patterson
In Idaho, tossing a rattlesnake into a crowded room is felony assault.
In Tennessee, it's evangelism.

[William W. Plummer]

G.R. Patterson III wrote:
>
> "James M. Knox" wrote:
>
>>It was from this that they found
>>that they could do a good estimate of distance by essentially matching the
>>basic pattern to one of several basic shapes (this is partly how they
>>exclude cloud-to-cloud) and then comparing the difference between the
>>"paradigm" shape and what was actually measured.
>
>
> Dumb question time. Why exclude cloud-to-cloud? Wouldn't that type of lightning also
> indicate conditions one would wish to avoid?
>
Good question. I think the idea that a lightning bolt is an impulse
that excites all frequencies equally at the source. So it shouldn't
matter if it is cloud-cloud or cloud-ground. And, how does a receive
know which is which anyway?

[Andrew Gideon]

Dave Jacobowitz wrote:

> I do believe that ground-base strike measuring equipment used for
> forecasting and storm tracking can better isolate the position of
> lightning strikes -- but in that case, they *do* have the luxury of
> spreading out their sensors.

Now *that* would be a terrific use for mode-S: cooperative weather analysis.

- Andrew

[Andrew Gideon]

William W. Plummer wrote:

>> Dumb question time. Why exclude cloud-to-cloud? Wouldn't that type of
>> lightning also indicate conditions one would wish to avoid?
>>
> Good question. I think the idea that a lightning bolt is an impulse
> that excites all frequencies equally at the source. So it shouldn't
> matter if it is cloud-cloud or cloud-ground. And, how does a receive
> know which is which anyway?

I'm stuck on the question "why would I care?" I mean, if I want to avoid a
t-storm, I don't particularly care whether the discharges are to the ground
or within the atmosphere.

This brings me to my main concern about this type of device, at least as I
understand it. Static discharge occurs after the storm is already worth
avoiding. If I'm in the clouds, this seems terribly likely to permit a
storm to suddenly appear much too close.

Solutions based upon drop size/density (ie. RADAR) would appear to be more
useful in that regard.

So...is it really safe to fly in the clouds with naught but spherics for
weather?

BTW, since my club's aircraft are all carrying strikefinders, I'd appreciate
any references to descriptions of how best to leverage these in IFR flight.

- Andrew

[Peter Duniho]

"Andrew Gideon" > wrote in message
online.com...
> I'm stuck on the question "why would I care?" I mean, if I want to avoid
a
> t-storm, I don't particularly care whether the discharges are to the
ground
> or within the atmosphere.

I've yet to see any reference that indicates that the devices actually do
filter out cloud-to-cloud lightning. So far, all we've got to go on is a
Usenet post.

> This brings me to my main concern about this type of device, at least as I
> understand it. Static discharge occurs after the storm is already worth
> avoiding. If I'm in the clouds, this seems terribly likely to permit a
> storm to suddenly appear much too close.

First, I have no idea why you say that "static discharge occurs after the
storm is already worth avoiding". Lightning is a very good indicator of
*active* thunderstorms, exactly the sorts of storms you'd want to avoid.

Secondly, storms can appear quickly true, but it's not like one's going to
engulf you in an instant.

> Solutions based upon drop size/density (ie. RADAR) would appear to be more
> useful in that regard.

They are sometimes useful, sometimes not. Consider, for example, that it is
entirely possible to have heavy rain, perfectly safe to fly in without a
thunderstorm. Also consider that radar suffers from attenuation (heavy rain
hiding even heavier rain farther away), while lightning detection does not.

> So...is it really safe to fly in the clouds with naught but spherics for
> weather?

I haven't had a chance to use them myself, so I can't answer that question
first-hand. However, those who ought to know say that lightning is actually
a much better predictor of thunderstorm strength, and of whether what's out
there is a thunderstorm at all (of course) than rainfall is. Lightning
detectors sure seem to be the standard equipment preferred in places like
Florida, and they seem popular elsewhere at all, for those frequently
dealing with thunderstorms (embedded or otherwise).

> BTW, since my club's aircraft are all carrying strikefinders, I'd
appreciate > any references to descriptions of how best to leverage these in
IFR flight.

Turn it on. Stay away from the strikes. I understand your hesitance to
just cruise right on into developing thunderstorms without knowing ahead of
time how well the Strikefinder works. But I'd have to say you're at least
in a better position than most of us to report on how well they work. :)

If I had a plane available to rent with a Strikefinder installed and lived
somewhere that isolated thunderstorms happened with any frequency, I would
take that opportunity to go out flying when one or more isolated
thunderstorms are around, and see what the Strikefinder says. You don't
need to get very close at all for the Strikefinder to tell you what it sees.
Compare the information from the Strikefinder against that from the Nexrad
radar information (available for pretty much everywhere in the US to anyone
with an Internet connection), and see for yourself whether you think the
Strikefinder does a reasonable job of highlighting the dangerous storms.

Pete

[Teacherjh]

>>
First, I have no idea why you say that "static discharge occurs after the
storm is already worth avoiding". Lightning is a very good indicator of
*active* thunderstorms, exactly the sorts of storms you'd want to avoid.
<<

I think he (or she?) means that the storm becomes worth avoiding =before=
static discharge happens. So, there is a window of development of storms that
the stormscope does not cover. I don't know whether this is true or not, but I
think that is what was being said.

There are certainly "almost thunderstorms" that it's best to fly around.

Jose




--
(for Email, make the obvious changes in my address)

[Peter Duniho]

"Teacherjh" > wrote in message
...
> I think he (or she?) means that the storm becomes worth avoiding =before=
> static discharge happens. So, there is a window of development of storms
that
> the stormscope does not cover. I don't know whether this is true or not,
but I
> think that is what was being said.

A developing storm won't necessarily show rain either. Falling rain occurs
during the mature and dissipating stages of thunderstorm development, and
while water droplets may be present during the developing stage (being
lifted by the updraft), there may not necessarily be enough to show up on
radar as a significant storm.

I agree that his statement might have meant what you said, rather than what
I thought it meant. But I still don't see how it would imply 'spherics
devices are inferior for detecting thunderstorms. Also, while I'm not
positive, if I recall correctly lightning is present in any thunderstorm
where turbulence and strong updrafts are present, regardless of the stage of
development.

In other words, lightning is a very good indicator of what kinds of
thunderstorms should be avoided.

> There are certainly "almost thunderstorms" that it's best to fly around.

No doubt...but I haven't seen anything that would suggest a Stormscope or
Strikefinder wouldn't identify those storms.

Pete

[James M. Knox]

"G.R. Patterson III" > wrote in
:

>
> Dumb question time. Why exclude cloud-to-cloud? Wouldn't that type of
> lightning also indicate conditions one would wish to avoid?

The official answer is "where would you plot it???"

Some of the newer models have a feature that allows you to temporarily
disable the suppression feature. This was in response to pilot requests
(pilots who got tired of seeing a lot of lightning and NOTHING showing up
on the screen).

jmk

[Peter Duniho]

"James M. Knox" > wrote in message
2...
> > Dumb question time. Why exclude cloud-to-cloud? Wouldn't that type of
> > lightning also indicate conditions one would wish to avoid?
>
> The official answer is "where would you plot it???"

What do you mean? You'd plot it where it happens, just as with
cloud-to-ground strikes.

How could that possibly be an "official" answer?

Pete

[Andrew Gideon]

Peter Duniho wrote:


> A developing storm won't necessarily show rain either. Falling rain
> occurs during the mature and dissipating stages of thunderstorm
> development, and while water droplets may be present during the developing
> stage (being lifted by the updraft), there may not necessarily be enough
> to show up on radar as a significant storm.

This I didn't know.

> I agree that his statement might have meant what you said, rather than
> what
> I thought it meant.

Frankly, I'm not sure how your interpretations of what I wrote are
different. I think you've both a pretty good grasp of my concern.

> But I still don't see how it would imply 'spherics
> devices are inferior for detecting thunderstorms.

I didn't write this. I'm trying to find out just how well they work as
avoidance tools, true, but that doesn't imply that I don't like the device.
I don't know enough yet.

> Also, while I'm not
> positive, if I recall correctly lightning is present in any thunderstorm
> where turbulence and strong updrafts are present, regardless of the stage
> of development.

From my reading - light yet, I admit - it sounded like lightening occurred
only after the developing stage is well along. It takes time for the
difference in potential to grow enough that discharges occur, as I've
understand what I read.

Did I misunderstand?

[...]
>
>> There are certainly "almost thunderstorms" that it's best to fly around.
>
> No doubt...but I haven't seen anything that would suggest a Stormscope or
> Strikefinder wouldn't identify those storms.

Wouldn't a strikefinder, by definition, not see a storm that wasn't yet a
thunderstorm?

I liked your idea, BTW, about trying the strikefinder experimentally. I was
up in poor-but-VFR conditions this weekend, and I was trying to do
something of that sort. Isolated t-storms were predicted, so it seemed a
good opportunity.

Unfortunately, the visibility was sufficiently poor and there were enough
clouds around that I couldn't really see much at a distance. There were
strikes showing, but I couldn't match them with anything visually.

Also, nothing was clustering. There were regions with indicated strikes,
but of no major density. If I had to guess, I'd say that my understanding
about the time it takes for discharges to occur is wrong, and this is a
demonstration of "near t-storms" appearing on the strikefinder. Alas, this
really is just a guess.

I've more reading to do, in the meantime.

- Andrew

[Andrew Gideon]

James M. Knox wrote:

>> Dumb question time. Why exclude cloud-to-cloud? Wouldn't that type of
>> lightning also indicate conditions one would wish to avoid?
>
> The official answer is "where would you plot it???"

Azimuth: Where the strike is strongest. Distance: Based upon strength.

Aren't these how strikefinders handle cloud-to-ground strikes?

> Some of the newer models have a feature that allows you to temporarily
> disable the suppression feature. This was in response to pilot requests
> (pilots who got tired of seeing a lot of lightning and NOTHING showing up
> on the screen).

Yes, well, that would bug me too <laugh>!

- Andrew

[Michael]

Andrew Gideon > wrote
> From my reading - light yet, I admit - it sounded like lightening occurred
> only after the developing stage is well along.

Depends on what you mean by well along. In fact, on a day with strong
thermal activity, you can actually detect the static discharges from
thermals. They don't fit the models developed for lightning strikes,
and often a thermal that is close will show up as a strike much
farther away.

> It takes time for the
> difference in potential to grow enough that discharges occur, as I've
> understand what I read.
>
> Did I misunderstand?

No, that part is right. But realize that when the activity is strong,
it doesn't take that much time. Basically, by the time there is
enough liquid water for RADAR to see it, there are static discharges
strong enough for spherics.

> Wouldn't a strikefinder, by definition, not see a storm that wasn't yet a
> thunderstorm?

Not really. Static discharges need not be lightning to be detectable.

> Also, nothing was clustering. There were regions with indicated strikes,
> but of no major density. If I had to guess, I'd say that my understanding
> about the time it takes for discharges to occur is wrong, and this is a
> demonstration of "near t-storms" appearing on the strikefinder.

Exactly.

A spherics device won't keep you dry (since it can't see
non-convective rain) but it will keep you out of severe turbulence.
In fact, I've yet to encounter anything worse than occsional light
turbulence while relying solely on the spherics. Moderate turbulence
in clouds is something I've only ever experienced when allowing a
controller with RADAR to vector me through an area I would not have
entered without RADAR assistance.

Michael

[James M. Knox]

"Peter Duniho" > wrote in news:10fnu8jacpfvs32
@corp.supernews.com:

>> The official answer is "where would you plot it???"
>
> What do you mean? You'd plot it where it happens, just as with
> cloud-to-ground strikes.
>
> How could that possibly be an "official" answer?

I would have to look, but I think it was in either the manual or a brochure
(FAQ) that came with my Stormscope.

[Andrew Gideon]

Michael wrote:

> Depends on what you mean by well along. In fact, on a day with strong
> thermal activity, you can actually detect the static discharges from
> thermals. They don't fit the models developed for lightning strikes,
> and often a thermal that is close will show up as a strike much
> farther away.

Ah ha! This is new to me too.

>> It takes time for the
>> difference in potential to grow enough that discharges occur, as I've
>> understand what I read.
>>
>> Did I misunderstand?
>
> No, that part is right. But realize that when the activity is strong,
> it doesn't take that much time. Basically, by the time there is
> enough liquid water for RADAR to see it, there are static discharges
> strong enough for spherics.

Hmm. Okay.


[...]

>> Also, nothing was clustering. There were regions with indicated strikes,
>> but of no major density. If I had to guess, I'd say that my
>> understanding about the time it takes for discharges to occur is wrong,
>> and this is a demonstration of "near t-storms" appearing on the
>> strikefinder.
>
> Exactly.

So what was I seeing? Static discharges from thermals that could grow into
t-storms?

> A spherics device won't keep you dry (since it can't see
> non-convective rain) but it will keep you out of severe turbulence.
> In fact, I've yet to encounter anything worse than occsional light
> turbulence while relying solely on the spherics.

In this case, what are you avoiding? Clusters? Any "strikes"? What's your
threshold for "not that way"?

BTW, when you write "RADAR" in making your comparisons here, are you
referring strictly to ATC RADAR? Or are you including airborn and/or
down/uploaded NEXRAD

- Andrew

[Michael]

Andrew Gideon > wrote
> So what was I seeing? Static discharges from thermals that could grow into
> t-storms?

Yes, pretty much.

> In this case, what are you avoiding? Clusters? Any "strikes"? What's your
> threshold for "not that way"?

Depends on how much turbulence I'm willing to tolerate. If I avoid
all strikes, I never get anything worse than some light chop and
thermal activity. If I avoid only clusters of strikes, and ignore
individual dots, I've accepted some increase in risk, and also
accepted that I'm going to have light turbulence and some moderate
chop at times.

If I'm going for a weak spot in a line - an area with relatively
little clustering - I know I'm in for it. I won't do this kind of
penetration without some sort of RADAR service - either decent ATC
RADAR or following a RADAR-equipped light aircraft, preferably both.
At that point, I slow to Va-10, ask for a block altitude, and strap
in. At that point, I KNOW I'm in a convective area - I'm just
counting on RADAR to take me around the worst of it. I don't do this
when I can avoid it - it's kind of like flying single engine low IFR.
If everything works it will be OK, but do it long enough...

As a last resort, you can always switch to the short range mode and
steer away from any dots that how up in front of you. It works most
of the time, but realize that if you're doing that, you've already
screwed up.

> BTW, when you write "RADAR" in making your comparisons here, are you
> referring strictly to ATC RADAR? Or are you including airborn and/or
> down/uploaded NEXRAD

Actually, NEXRAD is what ATC has these days, and yes, I'm talking
about airborne RADAR as well. By the time there is enough water being
suspended for airborne RADAR to see it - meaning distinguish it from
the surrounding non-convective light to moderate precip - it will show
up on a Stormscope in an obvious way.

See, the real challenge of using spherics (or RADAR) is not avoiding
weather. That's easy. Simply don't go anywhre you see strikes (or
returns). Unfortunately, that doesn't get you where you are going.
The challenge is to ignore the light/moderate turbulence (or precip)
but avoid the cells. That always requires some amount of judgment and
interpretation. How much clustering is acceptable? Depends on lots
of factors. It's never perfect. Unless you always turn away from the
first indication, you are eventually going to penetrate a cell - the
odds will eventually catch up with you. Of course the same is true of
flying single engine IFR.

Michael