[Prev][Next][Index][Thread]

Re: (meteorobs) Leonids 2000-2006



Given that I'm very short of time at the moment, my response will be seen
as most unsatisfactory, but I shall endevour to make a more detailed
response in a couple of weeks or so.

Given that dust trails are clearly present in meteor streams, and that
their geometry can only be determined with rigorous dynamical calculation
it is impossible to predict the ZHR of a dust trail encounter without
looking at the detailed geometry of the encounter.

Examination of the Leonid dust trail encounters makes it clear that
missing the core of a young dust trail by about 20 Earth radii will
result in rates lowered by about a factor of 100.

Each year the Earth goes through the region of the dust trails gives
a lottery of which dust trail is encountered and how close.  Each
section of the many dust trails that could be encountered in any
particular year, has a different age, density and mass distribution.
To predict the ZHR using the dust trail theory required calculation
of these quantities, but the most important part to be considered are

1) What is the exact density profile of a dust trail?
   Theoretical approaches are very model dependent in this, so we
   just fitted Gaussians knowing that these were just a poor
   approximation to some of the factors involved (and we specifically
   stated this in the original publications).

2) Knowledge of historical Leonid storm activity.
   For this we went with the independent and detailed analysis of
   Peter Brown.  The historical record will contain many uncertainties
   but Brown's analysis is the most uniform of all available.
   Despite being unhappy with some of the values Brown derived (and he
   specifically noted that most of his results were lower limits),
   it would be unreasonable to pick and choose the data.

Given this, our simple semi-empirical model fitted all the historical data
to the double Gaussian with only about 20% error.  I don't try to
hide from the fact that our 1999 predictions were too low by a factor
of ~10.  However, by rejecting only the 1833 ZHR, we can fit all the
other storms (1866, 1867, 1869, 1966, 1999) with much the same accuracy.
Only 1867 is discrepant by a factor of 3 and this fit makes Brown's 1833
ZHR too low by a factor of 2 (meaning 1833 and 1966 would have similar
ZHRs by our model).

To reiterate:

The Leonids contain many young dust trails

The geometry of the encounter with a dust trails requires detailed
dynamical calculation

The profile of a dust trail must be known to allow the ZHR to be
calculated.

Our semi-empirical approach requires historical data to define the
profile, but the data is limited.

Following 1999, one solution to our poor ZHR prediction was to reject
the 1833 storm ZHR (or to double it).

The new fit, including 1999, makes almost no change to our 2001 and 2002
predictions, as the width of our Gaussian fit has increased whereas
the peak intensity has remained much as before.

My critique of other methods of storm prediction (including Ferrin's)
was sent to this list in early November.  The only other approach than
dust trail modelling, that I see could work is the detailed dynamical
modelling of the Leonid stream as a whole by Peter Brown and colleagues.
The only current limitation with his study appears to be a lack of
computer power.  If the number of dust particles he integrates were to
be increased by about a factor of 1,000 it would be possible to look
at the integrated flux in a bin size much closer to the Earth rather
than the 0.002 AU he is currently using.  As stated above, dust trails
vary in flux by a factor of ~100 over ~20 Earth radii (~0.0009 AU)
in their radial profile, so a much smaller bin size than 0.0020 AU is
required to prevent contamination by outlying dust trails that do not
intersect the Earth.  Given this extra computer power, I believe
Brown's approach will be more valuable, in that it would not only
include the dust trails (present by default without specific
modelling of their profiles) but also the background.  In other words,
Brown's work will give a meaningful prediction in ANY year.  The dust
trail model looks only at storms and outbursts, which is a topic of
specific interest, but cannot say anything about the background.

Cheers, Rob

Robert H. McNaught
rmn@aaocbn.aaodot gov.au

To UNSUBSCRIBE from the 'meteorobs' email list, use the Web form at:
http://www.tiacdot net/users/lewkaren/meteorobs/subscribe.html

References: