technical interchange meeting spring 2008: status and accomplishments
TRANSCRIPT
Technical Interchange Meeting
Spring 2008: Status and Accomplishments
TASKS
• Task 1. Support Deployment of SZ-2 RV Ambiguity Mitigation Algorithms– To support the ROC in upgrading the SZ-2 RV
Ambiguity Mitigation algorithm – To discuss any anomalies in the SZ-2
algorithm encountered by the ROC at operational sites
TASKS
• Task 2. Support implementation of Staggered PRT – Determine operational scanning strategy and if any
additional PRFs are required – Update AEL to allow any PRT ratio – Analyze other PRT ratios– Support implementation of the Clutter Spectrum Filter– Support validation and verification of Staggered PRT
Clutter Spectrum Filter
• Task 3. Spectrum width improvements – Provide an algorithm that includes the spectrum width
improvements
TASKS
• Task 4. Spectral Processing – Provide information on NSSL’s spectral
processing• Spectral processing for dual polarization variables
(Bachmann and Zrnic paper, and Bachmann’s PhD) • Ground clutter identification using Polarimetric Spectral
densities (Melnikov and Zrnic paper Radar Conf. Australia and report to be put on NSSL’s WEB)
Polarimetric Spectral Density of ZDR
Polarimetric Spectral Analysis
• Spectral Density of Differential Reflectivity SDR (vk) = |Ah(vk)|2 /|Av(vk)|2
• Spectral Density of Differential Phase
SDP(vk) = arg{Ah*(vk)Av(vk)}
• Spectral Density of Cross Correlation Coefficient Sρhv(vk) = Running sum product of three spectral coefficients
-30 -20 -10 0 10 20 30-10
-5
0
5
10
15
20
Z DR (
dB)
-30 -20 -10 0 10 20 300
0.2
0.4
0.6
0.8
1
hv
-30 -20 -10 0 10 20 30
-100
0
100
Velocity (m s-1)
(
o )
-20 0 20
0
20
40
Velocity, m s-1
Pow
er,
dB
@ 30 km
-20 0 20
0
20
40
Velocity, m s-1
Pow
er,
dB
@ 40 km
-20 0 20
0
20
40
Velocity, m s-1
Pow
er,
dB
from 30 to 35 km
-20 0 20
0
20
40
Velocity, m s-1
Pow
er,
dB
from 40 to 45 km
H
V
H
V
H
V
H
V
Spectral density of ZDR
Spectral density of ρhv
Spectral densities of Powersi.e., Doppler Spectra
Densities are medians from 20 rangelocations, 30 to 35 km
Histograms of PolVariables from
weather and clutter
Histogramsof pol varobtained from full spectra and from 3 linescentered on 0Doppler
Cold Season Warm Season
Adaptive Clutter Recognition Criterion
-Compute the polarimetric variables, ZDR, ρhv, andδ from the spectra at and near zero Doppler
-For SNR>3 dB:
Declare clutter
If {(ZDR < -2 dB) OR (ZDR > 5 dB) OR (|ρhv|<0.8) OR (δ > 20o)}
Otherwise it is not clutter
Adaptive Generation of Clutter Map using Dual Polarization Spectra of ground clutter and weather at Horizontal and Vertical polarization
Notch filter Filtered spectra for the decision
AP cases
a) Sep 2007c) Oct 2007
HHistogramsSep 2007 case
Clutter identified with the algorithm(red) in a field of weather echoes (aqua marine)
Distribution of Clutter to Noise Ratio in the H and V channels
GMAPapplied independentlyeverywhereto H and Vchannels
Adaptive GMAP applied equally toboth channelsaccording tothe spectralrecognition(Serbianoven bakedbread) algacting on the H channel
Probability of Detection
Performance
• Simulations– Add time series data from clutter and weather
and apply the classification criterion
• Probability of false alarm ~ 5%
• Probability of detection ~ 90 %
• Addition of coherency threshold (NCAR) could improve the performance?
Polarimetric Spectral Analysis Separates Insects from BirdsBachmann and Zrnic 2007Bachmann’s PhD (NSSL,WEB)
10 20 30 40 50 60 70 80 90 100 110
-60
-40
-20
0
20
Range, km
Pow
er,
dB
0.5o
6.0o
Radial @ 180 - power spectral density
10 20 30 40 50 60 70 80 90 100 110
-60
-40
-20
0
20
Range, km
Powe
r, dB
0.5o
6.0o
Range, km
Pow
er, d
B
Power along the radial
0 20 40 60 80 100Range, km
35
0
-35
Vel
ocity
, m s
–1
Power spectral density field Power, dB
-20 0 20
-60
-40
-20
0
H-channel spectra @ range 30km
Velocity, m s-1
Pow
er, d
B
-20 0 20
-60
-40
-20
0
V-channel spectra @ range 30km
Velocity, m s-1
Pow
er, d
B
Spectral density of , Zdr, Average spectral densities for ranges from 30 to 70 km for each radial of PPI to get Spectral VADs
Vel
ocity
, m
s-1
exp.21,el. 0.5, range from 30 km to 70 km
50 100 150 200 250 300 350
35
0
-35-35-30-25
-20-15
Vel
ocity
, m
s-1
50 100 150 200 250 300 350
35
0
-35
0.6
0.7
0.8
0.9
Vel
ocity
, m
s-1
50 100 150 200 250 300 350
35
0
-35
0
5
10
Vel
ocity
, m
s-1
50 100 150 200 250 300 350
35
0
-35
-100
0
100
Azimuth, degree
N NE E SE S SW W NW N
Insects
Birds
Power, dB
Zdr, dB
, degree
Polarimetric Sea Clutter Algorithm (Ryzhkov et al. paper)
• This is Spring bonus
• Fuzzy logic algorithm uses: SD(P), SD(ΦDP), LDR, ZDR, ρhv, and V
• Tested on SPOL data from 2001 (Washington coast)
Sca
tterg
ram
s