sdi (seismic data inversion)

---> sdi
Abbreviations
SDI  Strategic Defense Initiative 


trying to understand processing and inversion
trying to unite      processing and inversion

classic formulation: inverse theory
finding the different elements

many images and applications early on. 
start as a collection and then structure.
try replacing code by Java as soon as possible 

book    selfcontained
chapter selfcontained; can be isolated easily

each book, chapter selfcontained
each chapter a problem set
contains: paper, Cakefile, README, LOG, [Fig], [subdirs]
each subdir a figure
Data, Src can be shared:
top-level src  dir which creates library
top-level data dir
((top-level data dir which creates data library (simple preprocessing) ?))
all general (multi used) software as classes or subroutines in Library
all general (multi used) data in library
start by copying what necessairy form Jon's book
everything small, or dynamic: home dir: (paper, code, cakefile, header file)
everything bulky,static but non-reproducible on matt (Fig, Data)
everything bulky, reproducible on (room?) matt: (executables,lib)
everything bulky, temporairy on scratch: (intermediate data files)
general  cake rules /usr/local/lib/cake/..
document cake rules /sdi/Cake/sdirules    (generalize for all books)
local    cake rules /sdi/chapter/Cakefile
personal cake rules ~/lib/cake 

design process: 
	first draft paper as outline (doubles as outline and as presentation)
	then create figures (as slides possibly)
	then finish writing paper (fill in the words)

topocs:
	multiscale interpolation problem for seismic data
	self-similar regressions
	DMO operator and predictionaire filter
	Simple application: statics, trace balancing
	Learning about the Covariance matrix by PEF

data sets: 
	gulf of mexico from Jon's BEI
	mobil (AVO and decent dtructure, 2D, well-logs)
	oesberg (3D, little structure, AVO, well-logs)
	husky (structure)
	salt 3D (structure, full 3D)
	TDF examples (gallilee, seabeam, seasat (!))
	Some easy synthetics: templates



unification of: 
numerical analysis (e.g. multigrid, eigenvector and inversion stuff)
wave propagation ( and other differential equation stuff)
statistics (especially baysian)
time series processing (especially prediction error theory)
standard geophysical data processing methods
other cool applications ... 

topics:
	multiscale interpolation problem for seismic data
	self-similar regressions
	DMO operator and predictionaire filter
	Simple application: statics, trace balancing
	Learning about the Covariance matrix by PEF


write Acknowledgements
write Introduction

presentation: 
reproducible research
on the web 
object oriented (java)
numerical analysis framework

