460 Posts in 227 Topics by 67 members
If this is your first visit, you will need to register before you can post. However, you can browse all messages below.
|Page: 1||Go to End|
|Author||Topic: Shallow to moderate water demultiple||3268 Views|
19 June 2011 at 9:16pm Last edited: 19 June 2011 9:30pm
That attached powerpoint shows the impact of different demultiple approaches on a line with water depths running from shallow (100ms) to moderate (1000ms) in a framework that could be used as a test sequence.
When optimising a demultiple approach, one key factor to consider is the trace-to-trace spacing in the shot domain (ie the channel seperation or group interval) and in the CDP domain. If these are too sparse, steep dips may be spatially aliased, which in turn can lead to lower quality results and/or noise being introduced. Pre-stack interpolations (OFFREG/SHOTINT) can lead to a significantly improved result, especially on older, lower fold data.
15 March 2012 at 4:24am Last edited: 15 March 2012 4:07pm
I'm trying to demultiple multichannel data from a 240-channels 3 km streamer in about 4000 m water depth. The data were shot at 60s interval (about 150 m), giving a fold of ~10. The group spacing is 12.5 m. The information about the SRME processor is unclear. In the dictionary it says that the dataset should have "same number of shots and receivers". But in the demultiple section of the manual it says that "the SRME processor requires input data that has the same shot and receiver spacing". These two definitions are contradicting in my opinion. Which is right? What do I have to do to prepare my data?
15 March 2012 at 4:07pm Last edited: 17 March 2012 5:41am
I could not find where in the dictionary section on SRME it said "same number of shots and receivers" but I will double check to make sure our help is consistent - is it possible to give anymore details as to where you found this text? The SRME module requires the shot and receiver spacing to be the same. To achieve this you want to prepare your data using the OFFREG and SHOTINT modules.
16 March 2012 at 4:29am Last edited: 18 March 2012 3:54pm
It's in the demultiple_howto.help file in the help/ directory, section 2.11 SRME.
Does it mean that in my case I have to reduce my shot interval to match my receiver interval? To avoid increasing the amount of data too much I might first increase the receiver spacing from 12.5 to 25 m and then decrease the shot spacing from 150 m to 25 m. Unfortunately my shot spacing is not exactly constant at 150 m, since the data was shot at constant time interval of 60s, will this be a problem? Is there a way of regularizing the shot spacing, similar to OFFREG?
18 March 2012 at 3:54pm Last edited: 18 March 2012 4:33pm
From your description it sounds like your data does not really lend itself to the SRME process and you may be better off using a radon demultiple process (eg PRT_DEMULT). However, have a go at increasing the receiver spacing and decreasing the shot spacing to match at different values and see how the results turn out. A shot / receiver spacing of 25m should not cause too much aliasing, larger values of 37.5m and 50m probably will lead to aliasing but give them a go and compare the results. There is no module for regularizing the shot spacing. SHOTINT only interpolates by inserting new shots between existing shots.
18 September 2016 at 6:36pm
The contribution grid inside srme is related acquisition parameters. finer acquisition better srme modeling. contribution of 25x25 if acquisition not sparse, otherwise this need internal interpolation/extrapolation within srme to avoid aliased MCG (but time consuming). Its all related
18 September 2016 at 6:36pm Last edited: 18 September 2016 6:38pm
The contribution grid inside srme is to related acquisition parameters. finer acquisition better srme modeling. contribution of 25x25 if acquisition not sparse if fine, otherwise this will need internal interpolation/extrapolation in both XL/IL directions within srme to avoid aliased MCG (but time consuming). Its all related
|Go to Top|