Odd geometrical shapes on seismic horizon

Oil and natural gas exploration -- geology and geophysics
Post Reply
Seismic101
Gold Member
Gold Member
Posts: 66
Joined: Sun Feb 22, 2015 5:17 am
Location: I am from the Internet

Odd geometrical shapes on seismic horizon

Post by Seismic101 »

Hello there- I have just finished interpreting the seismic horizon shown in the attachment, but I am seeing some uniform geometrical shapes that I cannot explain (the square in the red box and the straight line denoted by the yellow arrows). I have checked my tracking there, and it is accurate. The horizons is indeed very continuous and nice through the whole area, but, in the locations of those square, when I view a vertical section, the reflectors behave like this ----_____-------______--------_____, but that up-and-down is very subtle. This horizon has been tracked on a 3D seismic cube, and it's the shallowest horizon in the area (~ 1700ft MD).

What caused these shapes? I am not familiar with seismic data acquisition in the field. Is this what people call 'acquisition footprint'? Or, are they related to static correction?

Color in the picture represent elevation in TWT with warm colors being the smallest (ie. shallowest).

[Sorry, I had to delete the image]
Last edited by Seismic101 on Sun Jan 17, 2016 3:00 am, edited 3 times in total.

GuyM
VIP Member
VIP Member
Posts: 582
Joined: Sat Mar 24, 2012 11:35 pm

Re: Odd geometrical shapes on seismic horizon

Post by GuyM »

The stuff on the bottom left is definitely an acquisition foot print - its a land survey, right?

You can see the layout of the source/receiver lines and the variation in offset etc. There's a bunch of ways to fix that in processing - a good quality 5D interpolation routine will infill the gaps in the near offsets efficiently and then when you run a preSTM (or even stack) you won't get this kind of variation.

The linear event with the arrow is more interesting.

- you can get this kind of thing in marine data, because they tend to shoot in one direction. but I'm the *other* pattern looks like land.
- if this is in the US (and somewhere flat) then you do tend to get man-made things on the surface (like roads) that go in a straight line....
- hard to see how a processing artifact could do this

I once spent three days trying to track down a linear artifact on a marine survey, assuming it was a tidal variation in shooting or a change in acquisition for one swath, and it turned out to be a thin basalt intrusion aligned *exactly* with the shooting direction that "popped up" a fault block by ~6ms TWT.

I'd suggest looking on googlemaps/googlearth as a starting pint to see if there is a surface feature - like a road, or a fence, or a pipeline - that could have blocked access for the crew so they had to "undershoot" a little as a starting point. It could even be a drainage canal, or something like that?

If there's no surface feature could something have been "trenched" and then infilled (a pipleine say?) A different near-surface fill could do that?

If could also be a statics error, but that seems to pretty unlikely...

Seismic101
Gold Member
Gold Member
Posts: 66
Joined: Sun Feb 22, 2015 5:17 am
Location: I am from the Internet

Re: Odd geometrical shapes on seismic horizon

Post by Seismic101 »

Thanks for the comprehensive reply, GuyM.

Yes, it's a land survey, outside the US. The area where the survey was conducted is kind of an industrial city in the middle of the desert, so yeah maybe that linear feature is a road or a pipeline. I'm only doing research on this dataset, so I'd have to contact the company that provided the data and see if they are willing to provide more info. I will also have to explore what my software can do in terms of digital processing to fix these effects. I will update the post once I've got more info. Thanks again!

GuyM
VIP Member
VIP Member
Posts: 582
Joined: Sat Mar 24, 2012 11:35 pm

Re: Odd geometrical shapes on seismic horizon

Post by GuyM »

I will also have to explore what my software can do in terms of digital processing to fix these effects.
Well - a low pass spatial filter would fix the horizon, but not the data ;-)

These things are hard to fix post-stack, as a rule, because they gradually fade away with TWT.

Seismic101
Gold Member
Gold Member
Posts: 66
Joined: Sun Feb 22, 2015 5:17 am
Location: I am from the Internet

Re: Odd geometrical shapes on seismic horizon

Post by Seismic101 »

GuyM wrote: a low pass spatial filter would fix the horizon, but not the data
Yeah- that's what I meant, too. But I wonder: couldn't these effects be fixed during the actual seismic data processing?

GuyM
VIP Member
VIP Member
Posts: 582
Joined: Sat Mar 24, 2012 11:35 pm

Re: Odd geometrical shapes on seismic horizon

Post by GuyM »

Seismic101 wrote:
GuyM wrote: a low pass spatial filter would fix the horizon, but not the data
Yeah- that's what I meant, too. But I wonder: couldn't these effects be fixed during the actual seismic data processing?
Generally speaking, yes they can.

Of course - they have to be spotted on the quality control (QC) checks that have been used, and remediation of these things needs to have been a processing priority. While they show up clearly on time/amplitude maps along a horizon, that's not a standard processing QC check most of the time. They may show up on a time-slice, but more faintly.

You'd hope the processing team would check timeslices carefully as part of what they do, however if these are mainly "amplitude" anomalies as opposed to "time" ones then the scaling is a factor. Apply an AGC to the data and you would miss them. (Always check your data with and without AGC applied, by the way!)

There are "acquisition foot print" removal techniques which are largely based around 3D/4D spatial filtering on offset planes.

A good high quality "5D interpolation" using something like a POCS / Greedy Radon transform applied on "offset vector tiles" (where you split the data up by offset plane and azimuth) can address acquisition foot print issues, especially when combined with a good quality surface consistent amplitude correction and amplitude consistent processing.

This approach can also be used to infill obstacles that have been "undershot" to some extent - although its not magic, its really just regularizing and interpolating the 3D wavefield in a pre-stack sense while honoring azimuth.

This is what I'd described as "mature" technology, however its still on the boundary of "early adopter/early majority" on 3D land projects. Not all companies (or software) have this approach, and while people who process marine data routinely interpolate some land data processors don't view it as a valid technique.

Near surface velocity anomalies (roads, pipelines etc) should be resolved by a combination of shallow statics (refraction statics, perhaps with tomography) and surface consistent residual statics.

So - yes, it can be fixed in processing:

- if it was spotted in the QC stage
- if it was deemed to be a priority for whoever was paying for the processing
- if the processors have access to the appropriate algorithms in their software

None of which is a given.....

Post Reply
  • Similar Topics
    Replies
    Views
    Last post