Data Mining for Meteor Echoes Using High Power UHF … Lincoln Lab Meteor Processor Pulse/Threshold...
Transcript of 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