[meteorite-list] Leonid Meteor Shower Forecasts: 'It Looks Like We All Were Wrong'
From: Ron Baalke <baalke_at_meteoritecentral.com>
Date: Thu Apr 22 09:47:14 2004
Leonid Meteor Shower Forecasts: 'It Looks Like We All Were Wrong'
By Robert Roy Britt
27 November 2001
Preliminary shooting star tallies and interviews with the four groups who
predicted this year's Leonid meteor shower reveal that while strides were
made in the young science of meteor forecasting, nobody got it right in
"Right now it looks like we all were wrong, in various degrees -- perhaps me
worse than others," said Bill Cooke, a NASA scientist whose forecast for a
peak in Hawaii was most unlike the other three.
Nonetheless, experts say meteor forecasting, if not perfect, is at least
coming of age. The other three forecasts were not entirely off base, and
without question the 2001 version of the annual event was a storm, as
expected by all four research groups.
Scientists classify a meteor storm as one in which the hourly rate of
shooting stars, calculated over a 15- to 20-minute period, exceeds 1,000.
Some forecasts had predicted such a rate for North America and elsewhere,
but there was no agreement on if or where such a storm would take place.
The Leonid meteor shower is created by dust left in space by comet
Tempel-Tuttle, which lays down a separate trail each 33.2 years as it orbits
the Sun. Forecasting the shower involves figuring out which trails Earth
will pass through and how dense they will be.
At Mt. Lemmon, Arizona, a group of experienced meteor observers, using a
high-tech counting technique, gave this report: "We observed an above 'storm
strength' activity level (> 1000 m/hr ZHR) from about 1000-1130 UTC."
Translation: They saw more than 1,000 meteors per hour just before dawn on
Sunday, Nov. 18.
Robert Lunsford of the American Meteor Society and six other people used
remote "smart-mice" to feed their Mt. Lemmon observations into a PC operated
by James Richardson, using software developed by Morris Jones to perform
real-time hourly-rate calculations from multiple observers. The work was
part of a NASA-sponsored research effort at the Ames Research Center.
What these avid and experienced observers saw was, of course, similar to
what many amateurs and first-time observers witnessed -- a glorious storm of
shooting stars that won't be repeated for nearly a century.
A group of international observers northeast of Beijing, China, counted a
peak hourly rate of 2,400 shooting stars. Other skywatchers in Asia and
Australia reported rates of nearly 3,000 per hour. In one of the most widely
watched forecasts, however, researchers David Asher and Rob McNaught had
predicted an hourly rate for the region of 8,000.
Other scientists used video cameras and radio receivers to record the event.
Final and official numbers for peak hourly rates, called ZHR, may not come
for weeks or months, after all the observations are sorted out and
But some general conclusions can be drawn.
"The rates for the Western Hemisphere were higher than expected while those
for the Western Pacific were lower," said Lunsford. "The shower also
produced better activity for a longer period of time than expected. At least
the timing was close enough that most people were able to view the event."
Many observers reported a long peak that began after or lasted past the
Lunsford and Rainer Arlt, an astronomer at Astrophysikalisches Institut
Potsdam in Germany, both said that a forecast group led by Finnish scientist
Esko Lyytinen was most accurate at predicting the storm. They had called for
a peak of 2,000 per hour for North America.
In general, all the forecast groups seemed to do better at predicting the
timing of peak activity rather than the number of shooting stars that would
fill the sky during the peak, several scientists said.
"ZHR predictions are still hard to achieve," Arlt said. The difficulty is
not directly the result of bad prediction models, he said, but instead is
due to less than perfect data on past storms. Accounts of previous meteor
showers are often drawn from newspaper articles or amateur tallies.
Arlt said forecasts for 2002, when another meteor storm is expected, will
now need to be scrutinized.
The 2002 shower will be accompanied by a full Moon, which will drown out
most of the fainter meteors. Still, many avid meteor observers will plan
trips to favorable locations in attempts to recapture the magic of 2001.
SPACE.com asked a leader from each forecast group how he thought their work
stacked up against the others this year. Below are the replies, edited for
brevity and clarity.
[ZHR = peak hourly rate; UT=Universal Time; rev.=revolution (7 rev. means 7
orbits ago for comet Tempel-Tuttle, which leaves a new trail of debris every
33.2 years. These revolutions are also discussed as the year in which they
occurred, as in "1699" being a dust trail left in that year.)]
David Asher (of Armagh Observatory with Rob McNaught):
I haven't had time to assess this. At present my complete guess is that
Lyytinen et al. may be the 'winners' but that models that followed the 'dust
trail technique' pioneered by Reznikov et al. were all broadly correct, as
Rob McNaught and I (among others) were 100 percent sure they would be, even
though we didn't know which would be the 'winning' model among them.
The shower was further confirmation that meteor storm prediction is
essentially solved, albeit many people should be able to do exciting new
research to try to refine the details.
Esko Lyytinen (with Markku Nissinen, Tom Van Flandern):
According to early data, the rates, especially at the Pacific area, seem to
be a little below the predicted. The times predicted were quite good, at
least enough for choosing the observing location.
What is especially interesting is the timing of the 7 rev. outburst (visible
in Americas). Our last modeling with nongravitational effects put it about
half-an-hour later than the previous simple trail models. This seems to be
about the observed (it may have occurred even a bit later). I think that
this gives quite strong support to our model. Even beforehand I was very
interested to see the timing of this outburst peak.
Peter Jenniskens (of NASA's Ames Research Center):
Our peak rate from FISTA was about ZHR = 1200/hr peaking at 10:40 UT, while
our ground station at Mt. Lemmon had about 1500/hr. They also recorded a
narrow spike up to 2600/hr around 11 UT, the cause of which is not clear
yet. My personal estimate from the ground was around 1500/hr. Peak ZHR from
our ground station in Australia was about 1900/hr at about 18:15 UT (1866
peak). A modeling would be needed to determine the peak rate and time of the
It is clear that a comparison of the video records is needed to make sure
these numbers are well calibrated. If true, they definitely would support my
prediction that the 1767 dust trail was closer to Earth's orbit and the 1866
dust trail further away than earlier predictions by Asher and McNaught as
well as Lyytinen.
Esko Lyytinen's model did good regarding the peak times, especially where it
concerns more subtle effects from gravitational perturbations on the shape
of the dust trail near Earth's orbit. Observations have surpassed the
sophistication of the model by Asher and McNaught. Brown and Cook are wrong.
Bill Cooke (of NASA's Marshall Space Flight Center, with Peter Brown):
If we judge the accuracy of the forecasts from the very preliminary numbers
we have now using the criteria of intensity and timing, then I can state the
* Asher/McNaught called it to within 20 percent or so
* Brown/Cooke also had it right to within 20 percent
* Lyytinen/Van Flandern factor of 2 too high
* Jenniskens factor of 4 too high.
* Asher/McNaught off by about 30-40 minutes (too early)
* Brown/Cooke no peak at this time
* Lyytinen /Van Flandern - off by about 30-40 minutes (too early)
* Jenniskens - off by about 30-40 minutes (too early)
The 1799 peak predicted by Brown/Cooke apparently did not materialize, but
it has been suggested that the error in the timing of the first peak was due
to influence from 1799 material, which was not a significant influence in
forecasts other than Brown/Cooke.
* Asher/McNaught , Lyytinen/Van Flandern 2.5 times too high
* Brown/Cooke more than a factor of 3 too low
* Jenniskens too high by 33 percent
* Asher/McNaught, Lyytinen/Van Flandern just slightly early
* Brown/Cooke about 30 minutes early
* Jenniskens about 30 minutes early
This is the way things look now. However, these numbers are very preliminary
and are almost certain to change (The IMO revised their 1999 numbers at
least 3 times, and that was for only one peak).
Another item of interest is that the data we collected on the night of the
17th may show a small peak from 1932 and 1965 trails; this was not predicted
in any forecast.
Received on Tue 27 Nov 2001 11:53:11 AM PST