Thin bed inversion using non-linear techniques (& other things) May 2008.

POWERPOINT SHOWS -

Click here for an extensive report by David Paige on a project to map the thickness of a coal seam (see results at the left). Then give me your thoughts at dpaige1@sbcglobal.net

here for a later show that continues the coherent noise theme, this time on 2D data.  This show explores the possibility that an entire 2D interpretation may be based on noise caused by a shallow coal seam. Again, this is a long PowerPoint, so give it time. Your comments would be appreciated.

here for the latest normal site that describes my newest plans for areas like the North Sea, with emphasis on how AVO results can be improved (or beaten). If you only look at one, choose here (you can go back to the others). One reason is my current thinking on the fact that seismic tuning controls most everything. Of course you might have come from there.
For a PDF report on the coal seam project CLICK HERE
Coherent noise, wavelet development, tuning effect on AVO theory, interpretation fallacies & non-linear inversion are topics covered inside the main show. To fully understand one subject, one must be familiar with the others. Seismic results depend on an intricate set of intertwined factors, and I do my best to integrate them.  I use many PowerPoint shows to illustrate my points. The goal of noise removal is better inversion, and I keep trying. A major coal bed inversion project is inclluded.

For many visitors the discussions cover more than they want to know, but there are a lot of pictures. 

for the full discussion  Click here            dpaige1@sbcglobal.net

While the cost of preliminary analysis is not trivial, it is a small price to pay. It does require access to the raw field data however.  Once the noise is identified, rational decisions on reprocessing can be made. 

If you plan to bookmark, do it here. It gets complicated inside. 

If you don't go on, a favor please. Since it will have cost me from $.41 to $1.00 to get you this far, I'd love to know why you are not interested. I will throw away your email address unless you tell me not to. Negative responses are welcome, since being ignored is even worse.

       

Pass it on - the odds are that you are the only one in your group that has gotten through the maze on the net! While I can't pay you, I sure would appreciate the favor.

 In any case, here is a key word summary of what you would find inside, selectively quoted from the series, It is here for the spiders, but you are welcome, especially if you are interested in coal bed and other thin bed inversion topics, as well as non linear inversion solutions.
 
First, from the intro - "Seismic contractors shoot expensive, overlapping 3D prospects, processors fine tune the velocities, run automatic statics and migration, all to achieve maximum seismic resolution. But - given the coherent noise I'm seeing we're fooling ourselves if we think this detailing has really provided any seismic resolution when input data is bad -- Coherent noise distortion will garble structure & stratigraphy. Where present it makes such things as AVO analysis a virtual joke. Seismic wavelet work is hit very hard {and it is the essence of seismic inversion). Since the offending energy is often coherent, 3D stacking can mistakenly focus on it. Seismic migration is even more dangerous, since it seeks patterns spatially, assuming all to be true reflections -- The PowerPoint shows make it easy to instantly compare the "before" with the "after" showing you how real events are covered up by long seismic wavelet tails, and that removing these tails changes the seismic interpretation picture."
  
 From the AVO module "THE AVO MYTH - In case you haven't heard, a large percentage of exploration drilling is based on seismic software that looks for "amplitude vs. offset" anomalies, on the assumption that these variations point to the presence of hydrocarbons. My problem with this oversimplified logic is that it deflects attention from effective seismic inversion. Because such non linear inversion is my present claim to processing fame, I tend to get a little sarcastic. I do not say that such amplitude differences do not exist, I just say they will be dwarfed by other 
factors, and that the success of such discoveries is better explained by what I say below."    

And from the coherent noise section "Introducing the omni-present central noise cone (and the refraction phenomenon). The cone is at "A" of course. Some call it an "air wave" and others pass it off as "ground roll". Neither of these descriptions fits what I am seeing on thousands of examples. Where a ground roll (or air wave) would pass after a few cycles, this phenomenon continues as a series of apparently independent, low frequency events. This, and other observations are made on the basis of literally looking at thousands of cones, using the new seismic software I developed to ease this burden. The shear wave explanation makes more sense. We know shear waves are present (where particle moveout is lateral rather than vertical). Of course the source of the noise is not as important as the recognition that it is there, and that it is a problem. To the left of "B" and "C" you can see seismic refractions. Note the variation in the one way seismic velocities. You can see the one at "B" butting up against what is the target energy. We will discuss this interference problem later. The probable cause of both of these coherent noise forms is the presence of strong shallow reflectors. Of course seismic refractions (energy traveling horizontally along an interface) occur when the angle of incidence becomes greater than the critical angle. It may well be that deep water results are better because the angle of incidence is lower for any ray path. Since the shear wave  energy travels at very low velocities, the increased depth may keep them o ut of sight." 

From the filter discussion: "The presence of more than one seismic wavelet shape within a time zone makes it very hard to do advanced things with the data. Time series mathematics essentially assumes there will be just one. In the picture below it is interesting to look at the interference zones. Even though that data was filtered with a non-linear prediction technique, one can see phase shifts in the target data. I discuss this below under "Filters and pattern recognition". Filtering and pattern recognition - My problem is that I do not believe that "classic" time series approaches get the job done. Years ago I spent time looking at actual filter operators from time series deconvolution programs. I found that all their significant "multipliers" were clustered right at the front. This troubled me because I had an instinctive feeling that such filters should have a "Pattern recognition" capability. No matter the design technique, to do this the seismic filter action would have to be spread over the wavelet length. The same applies to band pass filters. They are what I call "frequency sensitive". That is to say they ignore phase. The subject of the frequency composition of a wavelet is very complex (at least it seems so to me). When we are trying to discriminate against particular wavelets (such as the shear waves I am suggesting), doing it just on the basis of dominant frequency seems simplistic. This is the industry approach however.