Dear Colleague,
The organization committee has now compiled your replies to the workshop in Moelle. Your suggestions have been most valuable and we urge you to continue this interaction with us. After reading this information letter you have another chance to interact with us and to influence the final format of the meeting. We should remember that part of the purpose of this workshop is educational, part is to provide focus, and hopefully part is to get a good idea on how to proceed in attacking the fundamental stumbling block to obtaining computational algorithms. Hopefully, some good collaborative efforts will develop from this workshop.
The current participant list and some information about the accommodation are appended to the tentative schedule. Please read!
THE TENTATIVE SCHEDULE we suggest is:
(1) The workshop will consist of 4-5 sessions: Thursday PM, Friday AM and PM, and Saturday AM and (possibly) PM.
(2) The first 1-2 sessions will involve a critical review of the 1D results for time-domain layer-stripping methods. We will attempt to determine just what results, techniques, and concepts from the 1D work can be carried over to the multi-D problem. This will hopefully help the group to begin to define and appreciate the fundamental problems in the multi-D time-domain case. Further, this part of the workshop will include a review of just what people currently do in multi-D computational inversion (optimization, approximations, combinations of the two). This will be important to provide some context and for further down the road where we will ultimately want to combine exact algorithms with optimization methods to account for realistic limited and noisy data. Hopefully, this part of the workshop will get everything established and everybody rolling.
(3) The remainder of the workshop will focus on the computational regularization issues. This will include such topics as which formulation (data) to choose (wave field (Green functions), scattering operator kernel, Dirichlet-to-Neumann operator kernel), how to computationally apply the square root and related operators, and, the clearly related issue of how to regularize the ill-posed problem.
We now ask for your immediate input on the following two issues:
A) What comments do you have on the basic breakdown of the workshop? Do you have any suggestions for improving the basic plan?
B) Assuming that the basic plan remains intact, we want suggestions on how to organize the 2-3 sessions on the computational regularization issues? We are looking for an effective way to break up, present, and discuss this material.
Best regards,
Gerhard Kristensson