data mining for meteor echoes using high power uhf … lincoln lab meteor processor pulse/threshold...

● Introduction● Lincoln Lab Meteor

Processor● Pulse/Threshold ● August 7, 2003

Results

Kristen FrazierCornell University

Mentor: Phil Erickson

August 7, 2003

Data Mining for Meteor Echoes Using High Power UHF Radar

● Introduction● Lincoln Lab Meteor

Processor● Pulse/Threshold ● August 7, 2003

Results

Kristen FrazierCornell University

Mentor: Phil Erickson

August 7, 2003

Local Sunrise

Orbit

IntroductionMicrometeors

- 10-5 – 10-4 m2

SporadicDetected at 90-140 km<Microsecond Duration

Micrometeor

Incoherent Scatter Radars

+ +

+ +

+ +

+ +

+ +

__

_ _

__

_ _

__

_ _

__

_ _

__

_ _

+ +

+ +

__

_ _

3) IONIZED PLASMA5

) Alti

tude

Ran

ge (k

m)

90 km

140 km

1) MICROMETEOR

2) ABLATED ATOMS

4) 'HEAD ECHO' DETECTED

Translation Program“nrs_reformat”

RAW DATA Meteor ProcessorNASA Meteor Project

-Millstone Hill Collaboration-A. Coster, S. Close

Meteor Processor Program-D. Durand

Raw Data Sample

Single Meteor Streak

Single Meteor Streak

Color Indicates AmplitudeRadar Clutter From White Mountains

1

2

Pulse Length: 40 microseconds

Raw Data Sample

Double Meteor Streak

VS.

VS.

Single

Pulse Wave Form Results

VS.

2.00

3.00Best Results: 2.00 – 3.25Manually ChangedGate Index = 15

Threshold Results

239 Meteors Logged68 Visibly Detectable30 Manually Tagged

Conclusions

• Why Double/Single Streak Phenomenon? • Why is There Such a Difference in the Log vs. Plots• Would Different Thresholding Give Better Results?

Questions:

Peak Range Rate: 50 – 60 km/secPeak Mean Altitude: 106-108 kmDuration: 50-100 millisecondsPeak Correlation Coefficient: -1

Statistics:

239 Meteors Logged

68 Visibly Detected

30 Manually Tagged

Acknowledgments