I just started the end of my journey in university. I registered my Master’s thesis. This feels like quite the step and I’m very proud I’ve come so far.
Today I start the last chapter of my studies. In half a year's time I'll finish my Master's. Exciting times lie ahead.
— Jesper Dramsch (@JesperDramsch) July 23, 2014
I’m 25 now and I feel it’s time to move on. There’s so much to explore and new things to discover. However the coming 5 months will be a tough one. I tend to think “easy is for the lazy ones”, so I looked for the challenge and yes I found it.
For my bachelor thesis I have been working on trace interpolation using an algorithm that is known as the Partial Common Reflection Surface. Working with this has opened quite a few doors for me. The spin-off extended abstract got published at the EAGE and I got to present a poster in Vienna. I’m also fairly sure my two internships with Fugro FSI (now CGG) and Western Geco (part of Schlumberger) were only possible due to my prior knowledge of the Common Reflection Surface stack. So taking a challenge has paid off, why not take another one.
When you work in seismics you will assuredly come across a couple white whales, problems that will haunt you or even the seismic community throughout their entire career. One of these topics is subsalt imaging. Follow me into the depths of this little abyss.
What is Subsalt
Under very special circumstances it may happen that salty water will be evaporated to the point that only the salt is left. A forgetful cook might encounter this a bit more often, however this also happens on a bigger scale on earth.
One example is the Mediterranean Sea. When the straight of Gibraltar closed, it was basically becoming a big lake in an arid environment. The water leaving this place, left a lot of salt on the bottom (please dear geologists forgive my oversimplification). Salt is quite an interesting thing to work with. On geological scales it behaves more or less like a fluid. Through enough pressure it will also form a perfect seal to upper layers of sediment, therefore serving as an ideal top of a reservoir for carbohydrates. This is the main reason people want to be able to look under the salt. That’s where the term Subsalt comes from.
The challenge in Subsalt imaging
As you can imagine, salt is quite different from the surrounding rock. In seismics this difference means a strong reflection so a salt body will give us even two strong reflections. Usually this is nice, as the salt body can be imaged quite well. However, if you want to look below this salt body, you are in the tight spot of having only a fraction of the energy left to illuminate the subsalt region and then this energy has to travel through the salt body with the two strong reflectors again.
But this is not the only obstacle. Due to the crystaline structure of salt if may behave anisotropically. Additionally, the following happens when salt is deformed.
Salt deforms under pressure and will adjust to the rigid surround rock. Even the slightest deformation will leave a very non-uniform layer in the worst cases it will form a salt body that has lens shapes as reflecting layer. The non-uniform layer will defocus our seismic beam like a laser pointer with a badly crafted lens that will under no circumstances focus the laser beam. Thus, spreading out the left-over energy even more leaving us with a puzzle of reflection patterns to solve. The lens-like structures are just adding to this problem. Pointing a flashlight at a curved mirror will illuminate a lot of places of the room as the light is spread out along the reflection trajectories.
There are a couple of ways to tackle this problem. Since I have been working with the Common Reflection Surface stack in the prestack domain before, this will be one approach to do it. I will compare it to conventional CMP processing. Then there is this new method we developed in our workgroup. Sergius Dell worked on the Partial Time Migration process, which is a process based on Common Scatter Points that will give a prestack time migrated gather. So for comparability all of these gathers have to be time migrated and processed with great careMaybe this is the reason I am now working on Saturdays.
However, I’ve decided to let you be a part of this process as I’m taking up to much information to put into this thesis, I might as well have an outlet for it.
This will be an interesting journey. Not only are these three completely different processing approaches. It’s dealing with Subsalt imaging, which is one of the big problems in seismic imaging. Additionally, at least two salt layers show anisotropic behavior. Anyone else feel this tingling sensation?
Want to know more?
Let me know in the comments. I’m happy to share!
Latest posts by JesperDramsch (see all)
- Geoscience and Machine Learning – EAGE 2017 Workshop - 2017-06-13
- Lavafalls – Waterfalls are so last year! - 2017-02-13
- Getting started with the SEG Machine Learning contest - 2016-12-23