The seismic reflection method is the principal way in which we image the outer part of the Earth, and it can reveal surprisingly detailed structure. So how are these images created? Well that's something we're going to look at together over the next few minutes. The seismic method relies on putting acoustic energy into the Earth as a signal, and then recording it back at the Earth's surface.
Now this raises a problem. because there are many sources of acoustic energy in the earth, vibrations. So how do you discriminate between what is a reflection from a geologically interesting interface and what is just noise? Well, that's something we're going to address as we go. But we'll also look at the issue of what these images actually represent.
The geology that they appear to show is distorted. One of the principal reasons for the distortion in seismic images is that the vertical scale is not in kilometers generally it's in time and that time is the time it takes for the seismic signal to leave the source go down to a reflector and echo back to the earth's surface well let's see how this plays out let's consider a pile of rock layers shown here in color from the earth's surface going down deeper into the earth and clearly the depth scale in a conventional geological cross section like this would be in kilometers how would this be recorded geophysically by the seismic reflection method well we start off with a shot point sp which is our seismic source and this is going to be recorded at a receiver the record of the wavefront arriving at the receiver is recorded as the time taken for that signal to go from the shot point to the receiver. In other words, it's a delay time. Now on the graph on the right, we've got our shot point at the surface and time going off increasing downwards.
So the first energy to arrive from the source, the shot point, to the receiver is the so-called direct arrival. That energy goes directly to the receiver and it's represented by that big kick. early in the record at about one second delay. So if you look at the seismic record at the receiver we can see that after the direct arrival there's a series of other bursts of energy that are recorded and one of these here is the energy that has reflected off some interface in the subsurface. This is our first reflection.
So let's see how this might play out. coming back to our revealed earth structure. So in this illustration we've removed the direct arrival. We're not interested in that, we're interested in the reflections.
So here is the energy that's arrived from the shot point to the receiver after about one and a half towards two seconds and we've got our first reflection. As we keep going on in the record of the receiver, a bit later We receive another reflection on the diagram that's coming from the interface between the green and the blue packages of strata. And then another reflection somewhat later, what about four seconds into our record, which has come from the interface between the blue and the mauve units.
And then the record continues, presumably recording arrivals for seismic energy that's gone deeper into the Earth and then come back out again. So this is our seismic record. It's a wiggle trace. But it's pretty unclear isn't it where these reflections are.
If you just had the seismic record it would be quite difficult to pick the reflections from it. Their amplitude is not particularly high. This record includes these signals the reflections were interested in plus a lot of noise.
And the catch is that for a single record like this the signal and the noise are about the same size. So how do we improve the signal to noise ratio? This is really what we want to see. We want to have a nice prominent signal and a rather reduced noise.
So how do we get this type of record? How do we enhance the signal to noise ratio? Well here's our graph again with the time increasing downwards and here's our first record recorded from our single receiver and as we've seen this record is not particularly great.
But what would happen if we combined multiple records? Well, we'd hope that the noise would cancel out. The noise waveforms would destructively interfere, whereas the waveform representing the seismic reflections would constructively interfere and therefore be enhanced relative to the noise.
Something like this. This process of combining multiple records into a single trace is called stacking. So this is a stacked record combining multiple records and you can see that we've enhanced the kicks representing the reflections and the noise is still there but it's reduced certainly with respect to the signal.
Conventionally this will be displayed in a seismic profile by colouring in one side of the waveform. in this case the rightward kick and then we repeat the same thing from place to place building up a profile. So this is a real seismic profile and it shows if you look carefully a whole series of traces running up and down the image with their right hand kicks coloured in. Each one of these is a stack of multiple records so the seismic reflection that we identify on a profile consists of these multiple records.
So if we sit back, this is what the seismic profile looks like. And if you step right back, you can barely see the wiggles. You just see these continuous reflections.
So the question you need to ask is, how do we decide which of these seismic records to combine? How do we decide how to make our stack? There are various ways we can do this and the configuration we use is called a gather. We're going to compare a couple of examples.
So we'll use this cartoon which shows a hypothetical piece of geology, it's submarine, so we have the water surface represented by that wiggly waveform at the top, a seabed and then beneath the seabed two geological reflectors x and y. And the way this is going to be detected is by a marine geophysical experiment with a boat, and we have strung out the back a series of receivers. So a seismic vessel that has a source represented by that red triangle, which would be an air gun. Hanging out the back is a streamer of hydrophones.
In this case we've just got six, and this allows us to have a multi-channel record. So that's the configuration, a source and a streamer of hydrophones. We're only showing six.
In a real modern day geophysical survey this streamer could well be six or even 12 kilometers long with a spacing of hydrophones maybe every 12 and a half or 25 meters. That's a lot of hydrophones, a lot of channels. So that gives us the opportunity to stack a lot of records.
So here we have a situation where our seismic signal has been sent out from our vessel and it's been picked up by the first hydrophone in our streamer reflecting off the seabed. Let's see how this might look. So here we have a display showing the shot point on our vessel and the yellow blobs on the graph represent the positions of our various hydrophones out behind the boat and against this we'll plot the time in other words the arrival time of our reflection from the seabed as detected in these various hydrophones.
So here is the first hydrophone, the path shown in the diagram on the left, where the reflection arrives at some time after it leaves the source. So the vertical scale on this diagram is the time taken for the wavefront to arrive at the receiver from the source on the seismic vessel. The horizontal scale here is the distance commonly referred to the offset which is the distance between the source and the receiver.
Clearly that's going to be a conventional linear scale in in meters. So as we move out along our streamer the offsets will increase away from the boat. So here the ray paths for the energy that's left the boat and reflected off the seabed to those various receivers. Notice that as the offset increases, so too does the distance taken by the seismic energy along these ray paths.
So the travel time will increase with offset. Let's plot those. So there we go.
With increasing offset we have an increasing travel time for the energy to reach our receiver having reflected off the seabed. This pattern of arrivals picks out a shape and it's a half parabola that is concave downwards. So how does this parabola represent the seabed? and how do we combine our records to get a good stack? So what we do is we shorten or unify the length of the traces to the reflector.
So they line up here horizontally in the plot and of course we can colour in the rightward kick. So this shows that they enhance one another and we can use this enhancement to define the position of the seabed like this represented by a certain time in our record. This is called a common shot gather because the records all use the same individual shot or source from the seismic vessel.
The correction that we've applied to those traces to bring them up to horizontal is called the normal move out correction. We're adjusting the travel time with respect to offset. This adjustment of travel time is the stacking velocity. We're unifying the travel time to the reflectors for these various traces. For the configuration we've got here with a horizontal seabed, then the stacking velocity simply relates to the seismic velocity of the water through which the signal has passed.
If the seabed had a more complicated geometry, or was inclined, then the stacking velocity would not simply represent the seismic velocity of the sea. But we don't need to worry about that complexity for now. However, this approach uses a wide patch of the seabed to make the gather.
So clearly the seabed has to have a simple structure, otherwise those reflections would go off in different directions. What we really want to be able to do is to resolve small patches. the seabed or on the reflectors in the subsurface and join these up to define their geometry. So we want to identify small patches. So we can use the same survey arrangement of a source and a stream of hydrophones but we're going to change the way in which we gather the records.
So let's look at how this might work. Here we have our seismic vessel with its shot being picked up by receiver A. reflecting off the seabed.
Here is another part of that same signal and the ray path taken by that reflecting off reflector X and back up to receiver A and similarly we have another ray path and this is the one that goes down to reflector Y reflects off it and goes back to receiver A. So we've got three different ray paths going off three different reflectors. Now what happens next is we'll allow our seismic vessel to sail on like this and set off another signal another blast on its air gun to the same reflectors there but this time of course these signals will not be picked up by the hydrophone at the front of the streamer but by another hydrophone further back along the streamer and we'll call this receiver b so with respect to the seabed the shot point and the receivers have moved.
Let's do it again. The scientific vessel sails further on and lets off a blast on its air gun and these are the ray paths for the scientific energy again off those three patches, one on the seabed, one on reflector x and one on reflector y and these are picked up at a new receiver again which we'll call receiver c further out again along our streamer. So This is the configuration of shots and receivers that we've just looked at relative to those patches on the seabed, reflector X and reflector Y.
Let's remind ourselves on what these look like. Here is the ray path taken from shot point one to reflector A via the seabed. Here is the path taken by shot point two to receiver B again via the seabed. And here is the ray path taken from shot point 3 to receiver C, again reflecting off the seabed.
So we have three records of reflections from the seabed via these three different combinations. We can do this again for the ray paths that reflect off reflector x, and again for the ray paths that reflect off reflector y. All of these reflect off their same small patch.
So our gather we're going to do, is going to combine these various traces and they're going to tell us about the position of the reflectors at a common midpoint between our pairs of shot points and receivers. The gather therefore is called a common midpoint or CMP gather and you can see that the common midpoint lies halfway midway between shot point one and receiver A. So keep that in mind as we see how these plot up.
So here is our plot showing distance, that's offset, from the shot point to our three receivers relative to a single shot point. So we're going to plot that against travel time. Here's our record for the shot point to receiver A.
And it has that labeled offset. And we can see our three... arrivals on the record on the right representing the reflections from the seabed, reflector X and reflector Y. Here's the same thing for the shot that was picked up by receiver B, a similar looking trace but with a different offset.
Here again is the record for receiver C. There's our travel time plot against offset. And again, if we join these up, they define half parabola that are concave downwards. The upper parabola, the first arrival, is from the seabed. The next one is from reflector X.
And the final one on here is from reflector Y. So what we're going to do now is stack these three traces to enhance the signal. And these stack.
back to the midpoint like this. So we've enhanced our waveform, we've enhanced our reflection from the seabed and from x and y. Again we can colour in the rightward kick on our waveform.
So we've identified patches on reflectors, the seabed x and y. If we're building up a profile we'll do this again and again and again and again. And here's a real example.
So now we know how the image is created and we can see these nice continuous or reasonably continuous reflections in our profile. Could we simply color this in as geology? Well, think about what the vertical scale in this is.
It's in two-way time, the time taken for our seismic energy to go from the surface into the subsurface, find a reflector and rebound off it back to the surface again. So that's there and back, two-way travel time. To interpret the geology, it's useful to be able to think about what this means in terms of real depth.
And in order to do that, we need to think about the seismic velocity. In other words, how far the seismic energy travels in the time. So here's a real seismic profile.
It's... displayed with a vertical scale in two-way time, which is what we've been looking at, and we want to try and think about what this represents in terms of real depth, in terms of kilometers or meters, and how this might vary if the velocity at which the seismic energy is transmitted through the various materials also varies. So let's just consider for now just the positions of the seabed in that patch there.
and of a prominent reflector somewhere below the seabed at x. So first of all, how far down is the seabed? Well the seismic energy took two seconds to get from the surface to the seabed, reflect off it and get back again to the receivers on the sea surface, because the vertical scale says it says the two-way time is two seconds. So how far is that?
Well, The seismic velocity of seawater we're going to use is 1500 meters per second. So in two seconds, that seismic energy has moved. three kilometers, 3000 meters, but it's there and back.
So actually it only took one second to get from the sea surface to the seabed. In other words that's 1500 meters. So the bathymetry here is 1500 meters. Now let's turn our attention to the position of reflector X below the seabed and you can see that the separation between those two markers, the seabed and x, is one second of two-way time.
So in other words the seismic energy is taken one second to go from the seabed to x and then back to the seabed again. Well let's give this interval between the seabed and x a seismic velocity and let's say it's got a velocity of 4,000 meters per second. So in that one second the seismic energy in that interval has traveled 4,000 meters there and back.
So just to travel one way between the seabed and X, it's travelled 2000 meters. So the thickness of that unit there is 2000 meters, 2 kilometers. Compare that then with the bathymetry of 1.5 kilometers.
You can see there's a significant distortion in the vertical scale as represented in the two-way time plot. In order to move from two-way time to depth, we have to know or assume the seismic velocity. And to make the correction...
We halve the seismic velocity and multiply that with the two-way time. Now, the reason that seismic profiles are routinely displayed in two-way time, and they don't make these corrections, is it requires knowledge of the seismic velocity. And often that's not that simple to obtain.
So rather than add that additional uncertainty, the image is displayed simply in two-way time. So let's explore this a little bit further, the relationship between seismic two-way time and depth, and how that might impact on the way we perceive the geology. So let's imagine that in our geological case in the coloured diagram, that we do actually know the seismic velocities of the layers that are contained within it.
These seismic velocities are generally referred to as interval velocities. In other words, the yellow package at the top of our stratigraphy here has a seismic velocity of two kilometres a second that interval in contrast the blue interval has a seismic velocity of five kilometres a second now how would this be represented in a two-way time plot well it would look something like this proportionally geological units with high seismic velocities are revealed in a two-way time plot as a rather thin unit compared to those formations that have got a very slow seismic velocity. So in the case here our slow velocity layer is the yellow one on top and that plots with a very similar thickness to our blue package which has a seismic velocity of five kilometers a second.
In other words having rock layers with different interval velocities on top of each other generates a differential vertical exaggeration that distorts the geological image. Now that's relatively easy to consider in rock layers like we're showing in our cartoon, which are parallel and horizontal. What about features that are oblique? Well, consider a planar fault. That's planar in reality, and because the fault is planar, it plots a straight line when seen in a vertical section.
How would this plot in our two-way time version over on the right? It would plot like this. So the fault profile... is distorted on the seismic image. The dips we would measure in a two-way time image are not the dips of the fault plane in reality.
So to emphasize the point in order to understand the real geometry of that fault we need to convert our two-way time image into real depth and for that we need to know or make realistic assumptions as to the seismic velocities of the layers. Finally, let's look at a rather classic case of how a two-way time profile can be misleading geologically and this arises from lateral variations in seismic velocity. Well, this seismic profile is from the Southern North Sea.
You can see that it's displayed with the vertical scale in two-way time, horizontal scale in kilometers, and this image through sedimentary layers includes an important salt. formation. here.
Now salt has a higher seismic velocity than most sedimentary rocks, particularly only partially lithified sandstones. So in this diagram here we've assigned a seismic velocity for the salt of five kilometres a second, quite a high velocity, and it's overlaid by lower velocity material, three kilometres a second. So how does this play out? So the energy following a ray path here to the top of our salt body is propagating at a rate of 3 km a second through that shallower material.
It then hits the salt body so that seismic energy accelerates to 5 km a second. It's going much faster now. So consequently it intersects the base of the salt here at the arrowhead after a particular travel time.
In contrast, let's consider what happens just to the right, where the seismic energy remains in that slower material for longer. So it's going to take longer to get to the top of the salt. Then when it hits the salt, that salt patch is thinner, so it reaches the base of the salt somewhat later than the energy taken by the other ray path.
And we can see that in our plot. which is why the base of the salt there doesn't plot horizontally. Now of course the reason it doesn't plot horizontally could be because of geological structure and indeed if you're not thinking very clearly you may think that the fold structure that underlies the left-hand trace could be a real feature beneath the salt. But this part of the salt is plotting higher because of these velocity effects.
It's a feature known as pull up it's a velocity artifact the fold structure at the base of the salt here is not real it's not geological it's an artifact of the velocity structure interpreting this as a fold is a classic mistake and it can get you into big trouble because you may think that's a hydrocarbon trap whereas in fact there's no trapping structure down there if we were to use the seismic velocities now and the travel time through the salts we can reconstruct the geometry of the base of the salt and it would have a shape something like this. So the base of the salt is smoother in reality than it is portrayed on our two-way time section. Pull-up is a really serious issue in these types of profiles and it's important to be aware of it.
So we have to think quite carefully about how the velocity structure impacts on the size of image we're seeing in a two-way time section. So there we have it, a brief introduction to how the seismic image is enhanced through stacking, and the impact of seismic velocity on the image that is created. We can create wonderful images with spectacular detail, but always bear in mind that the geology we are perceiving could well be distorted in the image from its reality in the subsurface.