[Prev][Next][Index][Thread]
(meteorobs) Asher-McNaught model in doubt... (fwd)
written by Mark Kidger in a reply to a query
whole text here:
http://www.meteors.com/cgibin/cometlinear/discussion_board_messages/30045.html
the last paragraph:
>The whole situation is very interesting. As it stands, the Asher-McNaught
>model is facing failure or, at least, being seen to lack the precision
>that it appeared to offer. This may make the predictions of a big storm in
>2001 far less convincing. More likely though, we'll see a repeat of the
>Perseid activity curve where new data changes the results significantly;
>that, at least, is my gut feeling. Certainly, I feel unhappy about the
>discrepancies that are being seen with the model and would prefer to doubt
>the data first!
Here is my reply to this
Mark Kidger raises several valid points but I feel his overall assessment
is unduly harsh. There are several issues involved here, some
interconnected, that limit any conclusions as clear cut as Mark suggests.
Given that there are two aspects to the theory, the timing from the dust
trail theory and the dust trail density model, I shall deal with them
separately.
Data from 1866, 1867, 1869, 1966 1969 and 1999 all have the predicted time
of maximum within about 5 minutes of that observed. However it has to be
noted that these encounters are all within the limited geometry of the
nominal dust trail core to -0.0007 AU (sunward) of the core. It is
reasonable to expect encounters significantly outside this region to be
further from the nominal time. Initial results suggest that the 2-rev
encounter at -0.0012 AU is reasonably predicted, with the Ondrejov radar
giving peaks ~7 minutes after and ~20 minutes after the nominal predictions.
With the 4-rev trail encounter the next night, Ondrejov observed a maximum
some 8 minutes early. Also with the 8-rev trail, Ondrejov noted a maximum
some 8 minutes early. It is interesting that both the 4 and 8-rev
encounters were at a geometry of +0.0008 AU. Okay, so not quite the 5 minute
accuracy, but there is another issue here that becomes relevant for the more
distant trails. The evidence is that the dust trails are strongly flattened
sheets, but in reality they are probably highly elliptical in cross-section.
The skewed passage of the dust trail as it moves outwards from perhilion
crossing the Earth's orbit, means that the Earth encounters the peak later
if passing sunward of the dust trail core and earlier if on the far side.
This is exactly the form of difference encountered in 2000; 2-rev at
-0.0012 AU 7 to 20 mins late, 4-rev and 8-rev at +0.0008 AU ~8 minutes
early. This effect is discussed by David and I in our paper "Variation of
Leonid maximum times with location of observer" Meteoritics & Planetary
Science 34, 975-978 (1999) (see section Close Approach to the Orbit). I hope
I have not been selective in choosing the radar data here but these are the
only reliable times of maxima I have access to, the visual data being of
variable quality due to the variable observing conditions. (Note that in the
above analysis I applied a topocentric correction to the observed times at
Ondrejov by subtracting 2 minutes. The true correction will differ, but will
likely be close to this.)
The next question is in the ZHR predictions. In this, I'll come straight out
and say that I believe the analysis of Lyttinen is the most complete. David
and I concentrated on fitting the storm data and were aware of the under
predictions when we extrapolated outside the storm region. However I believe
we can incorporate the 2000 data along with 1969, and probably other data,
to represent these lower intensity encounters. This would probably produce
results that are quantitatively similar to those of Lyttinen. As it stands,
our very simple model is qualitatively similar to Lyttinen. The encounters
in 2000 are clearly unreasonable extrapolations beyond the our fit, and I
personally am delighted with how our model stood up. The 4 and 8-rev trails
were comparable intensity (factor of 2 is fine by me considering the
probable disruption of the 8-rev trail) and stronger than the 2-rev trail.
Remember, our fit used an (arbitrary) double Gaussian and a similar function
with shallower slopes will result in an improved fit to the 2000 (and 1969)
data.
So overall, I am more than happy with the 2000 data, but recognise that an
improved dust trail density model is required to fit non-storm encounters.
However as it stands, I believe the storm encounters we predict for 2001
and 2002 will occur within about 5 minutes of our predictions and about a
factor of two in the rates. If being out by a factor of 4 in ZHR in 2000 is
considered poor, it shows how far expectations for Leonid predictions have
come in the last year!
Cheers, Rob
To UNSUBSCRIBE from the 'meteorobs' email list, use the Web form at:
http://www.tiacdot net/users/lewkaren/meteorobs/subscribe.html
Follow-Ups: