As of 29 March 2001
Round tables will be scheduled after each session held at the workshop. The seed questions below were prepared by chairman and co-chairman of each session so that we can hold more fruitful meeting.
Mon. 2 April 2001
Chair : Wolfgang-Martin Boerner
Co-chair/Report : Francesco Mattia
(From Richard Bamler)
1. What will the partially polarimetric ASAR data give us in terms of thematic interpretation?
2. What could be the application of polarimetric interferometry with ASAR?
(From Kazuo Ouchi)
3. I understand the basic techniques of InSAR/DInSAR have been well established. What are the key problems left in this area and possible solutions/approaches? How realistic is the InSAR to be utilized in practical applications, that is, actually to be used on a routine basis?
(From Wolfgang-Martin Boerner): See questions and queries put forth in working group report of CEOS-WGCV-MRS/SAR, Toulouse 1999 on Round Table Discussions for gPolarimetry/Interferometry/Polarimetric SAR interferometry (10 pages) - forwarded to list of participantsh. A selected set of extracted questions is also provided:
4. Are the fundamental studies on the gTheory of Radar Polarimetry & Interferometryh completed? What is still missing as regards the combined polarimetric radar/SAR interferometry theory?
5. Are the existing closed-form optimization procedures in POL-SAR, IN-SAR and POL-IN-SAR image analysis complete? What is still required?
6. What are the new results that have been obtained regarding the use of gcorruptedh Partial (mainly amplitude-only) versus Fully (well-calibrated complex scattering matrix) polarimetric SAR image feature interpretation? How will incomplete radar polarimetric SAR image information impact the future utility of space-borne SAR imaging?
7. Why do we require - under all circumstances - well calibrated polarimetric scattering matrix SAR image data takes? How do we go about measuring and preserving absolute gpolarimetric phase informationh required especially for POL-D (RP)-IN-SAR Imaging?
8. Is it more advantageous to use multi-look SAR imaging data take formatting or to use fully polarimetric SLC (Single Look Complex) SAR image data takes formatting? What does the proper polarimetric Lee-Wishart approach accomplish?
9. How do the DEM recovery procedures developed from non-polarimetric IN-SAR, from POL-SAR and from POL-IN-SAR compare? Why is it truly essential to use complete polarimetric POL-IN-SAR image data takes for determining unambiguous DEMs, i.e., separating canopy vegetation cover from soil surface, etc.?
10. Why does POL-IN-SAR imaging play a major role in proper biomass estimation and why do we need to make use of the entire HF/VHF to EHF bands within about 1 MHz to 90GHz? Why does the CEOS not request the acquisition of pertinent frequency band utilization?
11. Why has there not been a more open-minded approach adopted for the fusion of EWB Remote Sensing information, covering the HF-VHF-UHF-EHF to FIR-IR-NIR-VIS-UV spectral regions which includes EWB-D-POL-IN/TOMO-SAR as well as Hyper-spectral POL-SAR imagery?
12. Why was not sufficient attention paid to the development of multi-parameter vector-scattering inverse and modeling techniques? Would it be possible to establish another gInternational CEOS Working Group on EWB-D-POL-IN-SAR Modelingh? Would this not become an essential top priority? This is probably the best way, maybe the only way, to maintain an open and fair assessment of results, while simultaneously providing a forum to motivate and stimulate new work in this important area; would it not?
(From Ridha Touzi)
13. Using the alternating polarization mode of ENVISAT, what will be the best combination HH&VV, HH&HV, VV&VH, HV&VH as a function of the application considered.
14. Muti-look JPL compressed format introduces significant error (cf Touzi poster) for targets which are not azimuthally symmetrical, and point targets. How could this influence the results published on such targets. Is such format suitable for future satellite?
(From Eric Pottier)
15. Concerning partial polarimetry: No more chance to apply a fully polarimetric optimisation of the coherence of interferometric data ! What is the relevance of a two dimensionnal optimisation process ? This is only possible when phase is still available.
16. With partial polarimetry: no more chance to derive a good accuracy of the surface slopes using only one single image (no interferometry but just polarimetry). This can be useful to solve the problem of phase unwrapping.
17. What is the impact of partial polarimetry on unsupervised segmentation of POL-IN SAR images (note: Jong Sen has already adressed this problem in case of supervised segmentation).
18. Can FULL polarimetric and/or multifrequency remote sensing be seen as a way to provide enough information to invert more and more complex direct models ?
Chair : Fabio Rocca
Co-chair/Report : Jean-Claude Souyris
(From Richard Bamler)
1. How would a "dream" SAR satellite configuration (wavelength, polarization, orbit, additional instruments, timing, ...) look like for
a) topography reconstruction
b) deformation measurement?
2. How can other instruments of ENVISAT (e.g. MERIS for water vapor) help us improve SAR interferometry?
Tue. 3 April 2001
Geometry & Radiometry/SAR processing Session
Chair : Richard Bamler
Co-chair/Report : T.B.D
Chair : Robert Hawkins
Co-chair/Report : T.B.D
(From Robert Hawkins)
1. What are the lessons we have learned from the program initiatives in gcalibrationh from JERS, ERS, SIR-C, and RADARSAT-1?
2. How could these lessons be best utilized in new programs such as ASAR, ALOS, and RADARSAT-2?
3. What is the best approach to ensure users have confidence in the calibration of their imagery?
4. How best can calibration information and updates be forwarded to users?
5. Have we made any progress in the potential useful use of distributed targets for radiometric calibration?
6. We all assume that around orbit calibration variations are small - is this correct?
Instrument design and new concept Session
Chair : Jean-Claude Souyris
Co-chair/Report : T.B.D
Wed. 4 April 2001
Ocean applications Session
Chair : Benjamin Holt
Co-chair/Report : Hiroshi Kawamura
Ship detection/Sea ice/Soil applications Session
Chair : Kazuo Ouchi
C-chair/Report : Scott Hensley
(From Kazuo Ouchi)
1. Can the estimation of sea ice age and thickness be made purely from the sequential SAR data only? I should think that some prior geophysical knowledge must be required. If so, what are they? How versatile is the technique? Can the same technique be applied to other sea ice rather than Arctic one? (CEOS-SAR01-011)
2. Measurement of soil moisture from SAR data is a very complex problem. It involves the backscatter contributions from (1) soil moisture, (2) vegetation cover and (3) surface roughness. Unless (1) is singled out, in a quantitative manner, from other two contributions, one cannot make reliable measurements, no matter what one does in the post-processing stages. How does the NN approaches this problem? I understand that the IEM applies to surface scattering. Does the model takes into account the contributions of (2) and (3) ? Is so, how can it be done, and what are the relative contributions, and what is the accuracy with reference to the in-situ measurements? (CEOS-SAR01-026)
3. I am a joint-author for this paper. However, the following discussions may be of some interest. (1) So far, the phenomenon has not been observed in other SAR data. Why? (2) How much effect does it cause for ISAR processing? From the motion tracking of a particular "point", can the motion be estimated? (CEOS-SAR01-026)
Forest application Session
Chair : Thuy Le Toan
Co-chair/Report : Shaun Quegan
(From Thuy Le Toan)
1. Do we have a good understanding of the need to have temporal maps of forest structure and biomass in applications (carbon budget at regional and global scales, forest production function, forest ecological and environment functions..)
2. Do we have a sufficient understanding of the requirements in terms of accuracy, temporal repeat coverage, spatial resolution for a given type of application?
3. Are the published retrieval algorithms fully validated against relevant ground?
(From Thuy Le Toan)
1. Do the SAR community knows sufficiently what are the information products needed in scientific and commercial applications? (Or do we just look at what is possible to retrieve from SAR data).
2. Do we have a sufficient understanding of the requirements on the information product level on accuracy, temporal sampling, spatial resolution ?
3. Are the retrieval algorithms mature enough for applications?
3.1. Are the research and application demonstration results sufficiently validated (with relevant ground data)?
3.2. Are the current algorithms sufficently robust for global or regional applications?
3.2.1. Are the data-sets used for validation sufficient (diversity of conditions) ?
3.2.2. Is the development of underlying basic theory sufficient?
3.2.3. Is the development of software tool sufficient to provide a thematic map which can be used by non-radar users ? (geometric and radiometric correction, filtering and inversion etc..)
4. Using current and forthcoming spaceborne data (ERS, Radarsat, ENVISAT, ALOS), what joint effort needs to be done to bring research works to effective or operational use.
What are the applications to be prioritize?
5. Do the state-of-the art in research and application demonstrations using theory and airborne (and experimental spaceborne) data, sufficiently mature to provide the radar system parameters optimum for a given application.
6. For the applications under consideration, can we quantify the advantages to have more advance systems (e.g. multi-frequency polarimetric interferometric) compared to multitemporal ERS, Radarsat, ENVISAT, ALOS ?
Chair : Masanobu Shimada
Co-chair/Report : Gordon Keyte
(From Gordon Keyte)
1. What sensitivity should be sought in order to obtain accurate cross polarimetric measurements at L band?
2. Is it possible to use the Amazon rain forest for L band SAR cross track antenna measurements? (Since there will be significant penetration, the radiometric characteristics at different incidence angles will vary),
3. Should active calibration transponders be set up to calibrate the cross track pattern and how many transponders are required?
4. Will ionospheric disturbances affect the accurate recording of phases between polarimetric channels at L band? How do we calibrate for these effects?
Thu. 5 April 2001
Chair : Yves-Louis Desnos
Co-chair/Report : Manfred Zink
(From Manfred Zink)
1. Further recommendations regarding the collection of strategic data sets (SDS). SDS will be acquired as sort of a background mission in addition to the individual requests. The following recommendations have been at the last workshop:
Wide Swath: global coverage every six months for interferometry, coverage of coastal zones, Antarctica
INSAR: areas outside SRTM coverage to support ERS tandem data
2. Is it desirable to start creating a global high-resolution radar backscatter database?
3. What is the potential of the new ASAR modes and products for science and operational applications?
- Near Real Time stripline products (Image Mode, Alternating Polarization, Medium Resolution, Wide Swath, Global Monitoring Mode)
- Alternating Polarization products HH/VV or HH/HV or VV/VH (PRI,GEC,SLC,Medium Resolution)
- ASAR Global Monitoring Mode(1km resolution, HH or VV, products per orbit, global coverage in 6 days)
- Selectable Polarization and Incidence angle High Resolution Products in Image mode
4. What is the potential of the Newly developed ASAR Wind Wave algorithm and the new level 2 product? Should it be extended to Alternating Polarization?
(From Kostas Papathanassiou)
5. Passive (receive-only) radar (radiometer) sensors on microsatellites (small and cheap) following an (expensive) active SAR mother sattelite - as for example proposed by CNES with the Cartweel configuration - are an exiting new idea for revolutionising SAR remote sensing. They upgrade significantly the performance of a space-borne mission and offer manifold SAR data. Are there any thoughts about a polarimetric SAR microsatellite configuration considered with any of the planned sattelittes launching in the next years (ENVISAT, ALOS, RADARSAT 2) ?
6. A passive microsatelitte configuration operating in a dual polarised receive-only mode allows the bistatic measurement of the scattering matrix even if the active SAR sensor (mother) is only dual polarised. This can be an innovative concept for a (not too expensive) future fully-polarimetric (bistatic) interferometric space-borne SAR mission. Are there any thoughts for upgrading the pure performance of the ASAR system with a dual polarized receiver micro-satellite configuration???
(From Jong-Sen Lee)
7. ENVISAT can collect dual polarizations but incoherent in phase. What is the impact on information extraction (such as, geophysical parameters, land-use classification, etc.) using amplitudes without their phase difference information?
8. The cross-polarization (HV or VH) has been ignored by most researchers in ocean sensing. We have found (Lee et al., IEEE J. Oceanic Engineering, Oct. e98), however, HV is more important in detecting Gulf Stream front at certain imaging geometry than co-polarizations. This is due to HV is sensitive to azimuth surface tilt. ENVISAT will have cross-polarization and dual polarization modes. What can cross-polarizations improve current ocean remote sensing techniques?
Copyright by NASDA/EORC 2001
Oral : 58 presentations
Poster : 17 posters