[Contents] [Index] [Search] [Home] [Up] [Previous] [Next]

Millstone Hill UHF Meteor Observations: Preliminary Results

Philip J. Erickson
Atmospheric Sciences Group
MIT Haystack Observatory
Westford, MA 01886

November 19, 1998


The Millstone Hill 440 MHz UHF backscatter radar system has recently run experiments in a high-time resolution, raw data mode designed to obtain amplitude and range distribution information on both sporadic meteors and the Leonid meteor shower. This page presents preliminary results of those observations.

Observation Mode

The Millstone raw data observation mode uses the 220 foot zenith antenna for maximum sensitivity to meteor head echoes. These echoes are often less than 0.1 second duration as the fast-moving meteors fly through the vertically directed beam. A 13 baud Barker code provides a power profile with 1.2 km resolution every 5.904 msec, which should mean at least 2 to 4 hits on even the very fast moving Leonid meteors (mean velocity of approximately 71 km/s).

Initial goals are to determine the altitude extent and amplitude distribution of observed meteor echoes. Future plans include analyzing decorrelations of the Barker code power profiles for potential determinations of Doppler shifts.

You can find an online description of the experimental mode itself here .

More information about the VAMPIRE raw data taking system is available here .

Observation Dates and Plans

We have made the following recent observations:

Preliminary Results

Due to the large volume of data collected, only a small representative sample of the full observations are presented here. (The experiment generates 4 GB of raw data in a 4 hour run at current signal raster settings.)

Detection Criteria

Following the work of Pellinen-Wannberg and Wannberg (1994) [1] and Wannberg et al (1996) [2] with the EISCAT VHF and UHF radar systems, we applied a simple criterion to detect meteor echoes. After Barker decoding, the first range gate with a power value greater than 3 times above the standard deviation was flagged as a "meteor echo event". Statistics were then collected on those events.

A single IPP could, of course, have echoes from more than one meteor, but this preliminary analysis simply records the location of the first large echo. Note also that we plot meteor echo events here, which should NOT be confused with actual meteor counts. A single meteor can produce 5 to 15 "meteor echo events".

Sporadic Meteor Altitude Distributions: 10 November 1998

The figure below shows a histogram of the distribution of observed UHF sporadic meteor echoes as a function of altitude. Data from 1233 to 1311 UT on November 10, 1998 is shown. The histogram gap at 82 km is deliberate, due to persistent clutter echoes, and results below 70 km are also likely affected by ground clutter.

10 Nov 1998  Sporadic Meteor

The results are strikingly similar to Pellinen-Wannberg and Wannberg's observations of meteor echoes. The sharp altitude cutoff in the events observed above approximately 100 to 105 km is a geophysical effect, which Pellinen-Wannberg and Wannberg call the "meteor height ceiling". The meteor height ceiling varies with radar transmitter frequency, and is caused by ablating ionization which piles up in front of the meteor, creating an overdense plasma backscatter return which appears suddenly at and below the ceiling height.

We also note that this experimental mode differs significantly from the EISCAT observations, in that the latter used 2 second integrations for meteor detections. Since we have much more data available, we can collect statistics at a much faster rate.

Leonid Meteor Altitude Distributions: 17 November 1998

This section is currently under revision, due to large changes in our detection and processing algorithms. More information will be posted as soon as it is available.

Post-Leonid Meteor Altitude Distributions: 18 November 1998

The figure below shows the echo count histogram for the November 18, 1998 period immediately after the Leonid shower. The y axis scale is back to the same level (0 to 50) as the 10 Nov 1998 sporadic meteor plot. Once again, data is plotted from nearly the same local time period (1230 to 1300 UT) as the previous two periods.

18 Nov 1998  Post-Leonid Meteor Histogram

In terms of overall echo count rates, conditions are back to sporadic meteor levels. There is some hint of a persistent tail in the altitude distribution of echoes, but during the 30 minute period shown, statistics are sufficiently low that conclusions such as this are somewhat premature.

Time Variations of One Meteor Echo Sequence: 18 November 1998

Finally, the figure below presents a sequence of 10 panels which demonstrate the echo signature of an individual meteor passing through the vertically directed zenith antenna beam. The panels plot echo power as a function of altitude for individual radar IPPs from 12:37:24.583 UT onward; time progresses down the left column and then down the right. The constant-power clutter echo at approximately 81 km is the same one omitted from histogram statistics above.

18 Nov 1998  Meteor Echo Sequence

(Click here for an expanded version of this plot.)

The meteor echo is the double-humped signature appearing from approximately 85 to 100 km. This plot demonstrates the value of recording individual radar IPPs. The meteor velocity is fast enough that it only takes approximately 40 milliseconds or 10 IPPs to pass through the antenna beam. The power profile which results has the same characteristic shape as that seen by Wannberg et al (1996) [2]. Although the meteor echo itself is very compact in range, the echo signature is spread out over more than 15 km, and is the result of a high Doppler shift causing decorrelation of the transmitted Barker pulse code. We plan to analyze these signatures, as Wannberg et al (1996) [2] have done, to extract Doppler shift statistics.


The Millstone operations staff, and in particular Chris Farrell, Ching Lue, and Glenn Campbell, have been instrumental in overcoming technical obstacles and making possible these observations.


[1] Pellinen-Wannberg, A. and G. Wannberg: Meteor observations with the
European incoherent scatter UHF radar, J. Geophys. Res, 99, 11,379-11,390,

[2] Wannberg, G., A. Pellinen-Wannberg, and A. Westman: An ambiguity
function based method for analysis of Doppler decompressed radar signals
applied to EISCAT measurements of oblique UHF/VHF meteor echoes, Radio
Sci., 31, 497-518, 1996.

Philip Erickson <pje@haystack.mit.edu>
Last modified: Wed Jan 27 08:13:34 1999
[Contents] [Index] [Search] [Home] [Up] [Previous] [Next]