[Prev][Next][Index][Thread]
Re: (meteorobs) More on LM estimates and sources of error in ZHRs
Lew wrote,
>Note also that though LM estimates ARE highly subject to error, it would not be
>strictly correct to call them "subjective" as Rob does. Nor would I expect
>shower association to be quite as significant an error source as Rob suggests,
>at least under conditions where many meteors per hour were being recorded and
>they were predominantly from ONE shower...
>
>However, his point is very strong regarding meteor magnitude estimates: these
>are highly subjective, and their accuracy probably varies from hour to hour and
>from sky region to region for a given observer. That is why making significant
>use of magnitude distributions in ZHR calculations seems like an inherently
>problematic - but also a probably necessary - endeavor?
I agree with Lew's comments but was a little surprised at all the comments
above he attributed to me. These appear to be from the message that I
have included below.
As it happens, I did refer to "subjective" assessment of lim mag, but that
was an unfortunate shorthand and Lew is correct in pulling me up on that.
The point I was intending to make was that lim mag estimates will involve
error (= uncertainty) and this contributes to the overall error. Marco
and Peter Bus also pointed out to me that the limiting magnitude estimates
do no necessarily relate to the observed rates; one observer could record
less stars but see more meteors.
Lew's other point with regard to incorrect assignment of meteors to a
shower was on the assumption that I was referring periods of high rates,
in which case I mostly agree. However I was writing in general. A formal
error can be calculated from the number of shower meteors seen and is
based on these meteors being a true sample of the shower. However,
mistakes in shower assignment add to the overall error and for minor
showers could be very significant. There are two aspects to this;
the chance alignment of sporadics or other shower meteors with the
radiant of another shower or simple mistakes. There are lots of ways this
could happen from incorrect knowledge of radiant position through not
correcting for radiant drift and zenithal attraction, an inappropriate
radiant diameter, meteor seen in periphery etc. There is also the problem
with VERY high rates in the inability to correctly count all the meteors
that can be seen.
I hope this doesn't sound like being negative. Visual meteor observations
are an important method of gauging shower activity, but a realistic error
analysis (uncertainty) that may give substantially larger errors than are
often suggested, should not be seen as downgrading their value. This
enhances their value through better representing the reality.
Cheers, Rob
>From rmn@aaocbn.aaodot gov.au Wed Dec 22 02:31:36 1999
>Date: Sat, 18 Dec 1999 06:37:31 +1100 (EST)
>From: Rob McNaught <rmn@aaocbn.aaodot gov.au>
>To: meteorobs@jovian.com
>Subject: Re: (meteorobs) ZHR of meteor showers II
>A ZHR of 270.5 implies an accuracy that is basically impossible in visual
>observation. It wouldn't surprise me that this should more correctly
>read
>300 +/- 200. One can derive a formal error based on the sqr root of the
>number of shower meteors seen, but this is the absolute minimum for what
>the true error is and assumes that there are no errors in assigning
>shower
>association. Every manipulation of the data, in correcting for the lim
>mag, radiant elevation, perception coefficient, field obscuration by
>cloud
>etc intruduces additional error. But ultimately, the biggest error is
>the
>observer's subjective assignment of values to the observed quantities;
>shower association, meteor magnitude and limiting magnitude.
>
>Only the significant figures should be quoted in the final result and
>there is no value in claiming greater accuracy than the observations
>allow. A figure of 300 +/- 200 should not be seen as bad observation or
>as of no value, but as (possibly) a more valid representation of the
>available data. When the observations are combined with the data from
>many others, the overall error can drop dramatically, and then the
>intercomparison of individual observations with this combined result can
>indicate the magnitude of the error in the individually derived rates
>more reliably.
>
>Note that "error" is just as important a calculated value as the ZHR
>value
>it is related to. It is not "incorrectness", and nobody should feel
>afraid of quoting a large error when this what the data implies. The
>example of the error I use above is purely hypothetical and is not
>actually calculated from the observed data. Too often there are
>arguments
>about what the TRUE ZHR is for some shower maximum when the problem is
>probably more one of the error values
>
>Cheers, Rob
To UNSUBSCRIBE from the 'meteorobs' email list, use the Web form at:
http://www.tiacdot net/users/lewkaren/meteorobs/subscribe.html
References: