Aucun titre de diapositive - PowerPoint PPT Presentation

1 / 16
About This Presentation
Title:

Aucun titre de diapositive

Description:

Franck Garestier (1), Kostas ... (1) ONERA, BA 701, Ecole de l 'Air, 13661 Salon AIR, France ... I.N.R.A. test site, Avignon, France. INRA test site ... – PowerPoint PPT presentation

Number of Views:94
Avg rating:3.0/5.0
Slides: 17
Provided by: gares
Category:

less

Transcript and Presenter's Notes

Title: Aucun titre de diapositive


1
Analysis of forest
parameters and agricultural field structure from
high resolution PolInSAR X band data Franck
Garestier (1), Kostas Papathanassiou (2), Irena
Hajnsek (2), Pascale Dubois-Fernandez (1)
Xavier Dupuis (1) (1) ONERA, BA 701, Ecole de
l Air, 13661 Salon AIR, France (2) DLR,
Microwave and Radar Institute, 8230 Wessling,
Germany
2
Plan of the presentation
  • Presentation of the RITAS data
  • Ground truth campaigns (during and after
    acquisition)
  • Pine forest
  • Random Volume over Ground model
  • Assumptions and correction of the data
  • Mean extinction coefficient estimation
  • Wheat type fields
  • Signature of oriented volume observation
  • Conclusion

3
Presentation of the data
  • Campaign Radar Imagerie Thématique
    Agricole et Sols
  • Team
    INRA, CETP, BRGM, CEMAGREF and ONERA
  • Date acquisition in March 2002 by
    R.A.M.S.E.S. system
  • Data X band
    mono-pass PolInSAR
  • Size of the resolution cell 0.9 m x
    0.9 m
  • Swath 800 m
  • Mean incidence angle 30

4
Ground truth campaigns
INRA test site orchards specie, height,
biomass, maturity, water content soils
roughness, hygrometry
during the
acquisition The entire data vegetation
specie, height, tree density, maturity
buildings height after the
acquisition
optic
radar
I.N.R.A. test site, Avignon, France
Montfavet
5
  • Presentation of the RITAS data
  • Ground truth campaigns (during and after
    acquisition)
  • Pine forest
  • Random Volume over Ground model
  • Assumptions and correction of the data
  • Mean extinction coefficient estimation
  • Wheat type fields
  • Signature of oriented volume observation
  • Conclusion

6
Pine forest Random Volume over Ground
model Cloude Papathanassiou
2001
18 m
Representation of Pauli basis polarimetric
HHVV HV HH-VV
Polarimetric DEM HH HV VV
7
Pine forest interferometric coherence
Interferometric coherence for a
polarization straight line in the complex
plane Ground to volume ratio with
and the coherency matrixes of canopy and
soil Interferometric coherence of the volume
only contribution depends on and

Im
Re
Complex plane
8
Pine forest limitation of the data
antennas
High rate switch dysfunction during the
flight problem in V emission
decorrelation on the whole images
Decorrelation on VH and VV channel (small loss
of coherence for VV)
9
Pine forest dual (lite) pol
Im
The slight loss of coherence in VV
channel should induce a bias in parameter
extraction overestimation of the
forest height use of only two
polarisations HH and HV
degradation of the estimation (compared to
Quad-pol mode) interferometric coherence
associated to HV
HH HV VV
Re
Complex plane
10
Pine forest canopy decorrelation
Im
range axis
resolution cell
gHV
gHH
Re
Complex plane
Interferometric phase of the ground remains
unchanged
The roughness of the canopy induces a scalar
decorrelation of decorrelation coefficient
only applied on the volume contribution
Radial correction ( 0,07)
Impossible to extract and
simultaneously
Assumption on to estimate forest height
11
Pine forest coherences and heights
slight underestimation of the forest height
6 m vertical separation (9)
0

1
Coherence of HH and HV channels
Mono-pass acquisition high coherences
0.9 lt lt 0.95 (Coherence observed
on bare surfaces 0.98)
Phase standard deviation -
phase estimation accuracy 3 -
forest structure
12
Pine forest volume height
18 m height pine forest without underlying
topography
2 1,8 1,6 1,4 1,2 1 0,8 0,6 0,4
(dB/m)
0 10
20
30
0 10
20
30
Height (m) histogram of the estimated ground level
Forest height (m) histograms for different
extinction values
Height retrieved with high extinction
Dual-pol mode reduces the accuracy of
estimation
13
Pine forest extinction estimation
18 m height pine forest without underlying
topography

Mean forest height (m)
Mean forest height error (m)

Mean extinction coefficient (dB/m)
Mean extinction coefficient error (dB/m)
Mean extinction coefficient of a sparse pine
forest at X band
A large interval of mean extinction
coefficient gives a satisfying inversion. Accurac
y of 0.5 dB/m
14
  • Presentation of the RITAS data
  • Ground truth campaigns (during and after
    acquisition)
  • Pine forest
  • Random Volume over Ground model
  • Assumptions and correction of the data
  • Mean extinction coefficient estimation
  • Wheat type fields
  • Signature of oriented volume observation
  • Conclusion

15
Wheat type fields orientation effect
Bare soil surface
HH HV VV
INRA 8 wheat orchards
Representation of Pauli basis polarimetric
HHVV HV HH-VV
16
Conclusion
Pine forest Phase centers corresponding to
HH and HV separated wave
penetration in the canopy of a sparse pine forest
at X band determination of a high mean
extinction coefficient using RVoG model
forest height estimation low sensitivity to mean
extinction coefficient (HV phase center
close to the top of the canopy) Wheat type
fields observation of an orientation effect
due to the vertical structure
Write a Comment
User Comments (0)
About PowerShow.com