| 00:03 | OK. So what I've been doing the last uh oh half an hour |
|
| 00:08 | you and I fixed the stupid things have changed between 2022 and 23. |
|
| 00:19 | was basically the order of some of commands, the way they pop up |
|
| 00:23 | up differently. Also, the packages up in a different order. Last |
|
| 00:29 | , we had three packages and we pick number two and you guys very |
|
| 00:37 | followed my directions and picked package number , which really didn't have seismic interpretation |
|
| 00:43 | it. So I've captured that and in the notes which I'll resend to |
|
| 00:49 | . So you have it in the . OK? You may end up |
|
| 00:53 | patrol in, in the future for project. Uh Then we'll say, |
|
| 00:58 | OK. I gotta make sure it's seismic interpretation in the license. Otherwise |
|
| 01:02 | gonna be really frustrated. It's not do what I want. Um And |
|
| 01:10 | what else should I do? What had everything correct for generating the coordinate |
|
| 01:19 | system. But in 2023 it wants is immediately when you set a new |
|
| 01:26 | . And you're saying, well, it asking for uh and I had |
|
| 01:30 | the information in there, but in and before, after you imported the |
|
| 01:37 | , it asked you. So I noticed which I went over with one |
|
| 01:44 | two of you. Uh yesterday. a good Survivor skill that I don't |
|
| 01:50 | in the notes is how to scan trace headers. I don't think I |
|
| 01:57 | it in the notes. Do I ? No? Ok. I'll put |
|
| 02:01 | in. Uh so that you can sure that yeah, that looks |
|
| 02:08 | Because typically what you'll have you OK. Even back 1015 years |
|
| 02:16 | people were doing artificial intelligence. I , this isn't something new. And |
|
| 02:22 | one of the first projects we had Amaco back 30 years ago in artificial |
|
| 02:28 | was reading Segway data and it would , oh, if bite 17 is |
|
| 02:36 | integer between uh one and 10, probably a sample sample interval and it's |
|
| 02:48 | a high percentage that it was recorded Petty Rae Geophysical. And then, |
|
| 02:54 | , if fights number such and such very large numbers like greater than |
|
| 03:03 | 0, it's probably a coordinate. then if it's, if the next |
|
| 03:07 | is also probably a coordinate, then might have been recorded by CGG. |
|
| 03:12 | serious. This, we actually had big research project like this because, |
|
| 03:17 | know, companies would trade data or buy a company and then you got |
|
| 03:21 | these tapes. Well, what do do. So you have to have |
|
| 03:24 | way of looking at it. And patrol has a lot of buttons that |
|
| 03:30 | . I'd say 75% of the time you just say, ok, let's |
|
| 03:35 | to load it and it might do good. Uh But very often the |
|
| 03:42 | header and I don't think I captured too. When you loaded the |
|
| 03:45 | there was some information on the bottom said where everything is stored. And |
|
| 03:51 | always very rigorous to make sure that there. Sometimes people have it in |
|
| 03:59 | landmark and they'll just copy what the header was that said it was stored |
|
| 04:04 | certain bits. But now when you it from landmark, you put it |
|
| 04:09 | different bites and so it, it actually lie about where things are. |
|
| 04:14 | you gotta figure out where things So I'll try to capture that as |
|
| 04:19 | . OK. Lily. What you ? OK. What do you color |
|
| 04:27 | ? OK. So what color bar you like? So I I, |
|
| 04:35 | . So you're trying to save a bar or, or trying to read |
|
| 04:40 | in, read one in. So then here it's going to uh |
|
| 04:48 | called the colors. Yeah. And do do do do do do hang |
|
| 05:01 | . Yeah. Yeah. Yeah. on. So we're back in the |
|
| 05:05 | all the color. OK? OK. And you got a user |
|
| 05:12 | find OK? Go back to And see if I can right quick |
|
| 05:24 | you're looking at the template maybe I it that makes it discreet. |
|
| 05:38 | Yeah. Ok. Ok. Where our data set? Uh ok, |
|
| 05:45 | gotta go to input. Ok, it is. Ok. Ok. |
|
| 05:54 | what do you wanna do for the in line? Is that what you |
|
| 05:57 | do to kind of? Oh, . Oh, ok. So then |
|
| 06:06 | gonna go, there you go. ugly color. Probably just once I |
|
| 06:13 | um the Oh OK. So you a color bar now uh see up |
|
| 06:24 | see it says global color bar or . So you want to import |
|
| 06:29 | So now you're gonna go in here then these are ones you've used today |
|
| 06:35 | yesterday it green members and then there's of them. Now you're gonna have |
|
| 06:41 | of them and which one do you pick? How many there are? |
|
| 06:45 | . So you're gonna pick one if finalize history? Oh, so for |
|
| 06:54 | I would pick gray scale. Let's see what? Yeah. Oh |
|
| 06:58 | you made that OK. Good, . OK. And you say apply |
|
| 07:03 | but you're applying it to the seismic . Um So then you put if |
|
| 07:11 | , if you want just red, blue go back to click that |
|
| 07:16 | Now you got your red, white and then OK, that's OK. |
|
| 07:21 | just got the scale differently. So go to the Yeah. |
|
| 07:30 | So go to color make it a one and yeah, you can keep |
|
| 07:38 | . Oh yeah, you can do . Ok. Now you need to |
|
| 07:42 | let's put the limit so makes that 25 and make the bottom one minus |
|
| 07:59 | . Yeah. Ok. By that . Ok. Uh 01 other thing |
|
| 08:05 | minus 25. Yeah. Yeah, you. Ok. And uh hit |
|
| 08:12 | . Ok, that's good. And we go back to the wherever your |
|
| 08:17 | was. Yeah, so that Yeah, that's the Varian. And |
|
| 08:24 | I would do do do a uh quick on that guy. OK. |
|
| 08:31 | no um no right. Click on name and it should say info. |
|
| 08:37 | should be info someplace or maybe double quick on sorry, double |
|
| 08:44 | OK. And under info. she has this name, just erase |
|
| 08:51 | name and type in variants. The thing is forget the Corey because what |
|
| 08:59 | ? You can't read the stupid Iange. Yeah. Ge good. |
|
| 09:14 | , good. Now you know what is? Ok. Now you're |
|
| 09:16 | now you can uh again uh double on it and go to colors and |
|
| 09:27 | gonna, oh, you've got one . OK. You're doing pretty good |
|
| 09:31 | . We were playing with this um . Remember? So you got |
|
| 09:37 | You were in some funny, I know where you were. I wouldn't |
|
| 09:45 | about it. How are we You, Kathy? Ok. Uh |
|
| 09:57 | I'm going to the? Ok. you're ok. Good. Are we |
|
| 10:09 | ? Good. Good. How about ? How's your, how's your husband |
|
| 10:13 | with the sick? Two year Oh, I can't, you |
|
| 10:18 | I've got a two year old If he is sick, he didn't |
|
| 10:23 | there is no one. Yeah. , I know. Oh, |
|
| 10:32 | I'm gran, to my bigger grand . My four year old could never |
|
| 10:36 | Grms. So she called me grumps my two year old calls me |
|
| 10:43 | Jumps. OK. That's a classic . Yeah. And they scream when |
|
| 10:54 | want mom. How are we doing ? OK. All right. We |
|
| 10:59 | . Contento. OK. On a . OK. You're all right, |
|
| 11:10 | good, good. Um Did everybody the, the picture icon, the |
|
| 11:26 | icons? Let me show you interactively uh what is uh Jessica listening to |
|
| 11:38 | ? I am. Yeah. I'm gonna try to show something on |
|
| 11:42 | screen. Oh I can't show I'll show something on the screen in |
|
| 11:47 | bit. OK. I gotta load data we have here. We got |
|
| 13:51 | know it. Mm Pretty good. . It OK. OK. So |
|
| 20:53 | you can look at the screen a , can you see it? Let's |
|
| 20:58 | the I'm gonna turn the light off for a bit. OK. You |
|
| 21:06 | can see that. All right. I think uh Jessica and Felicia is |
|
| 21:12 | here. Um She can see Oh I just happened. I, |
|
| 21:19 | just happen to have a picture. ? Oh Where my, here it |
|
| 21:23 | . And I have a turbid coming here or if you can see my |
|
| 21:27 | thing. Oh, I wanna capture . You see up on this bar |
|
| 21:33 | . Bar has got a name. forgot what it's called. Got a |
|
| 21:37 | icon. So I'm gonna click the . It goes into the bit |
|
| 21:44 | I've got powerpoint over here. I'm hit control V. There you |
|
| 21:51 | OK. It's gonna be exactly that can click this. Uh I'll get |
|
| 21:58 | of this. I can click this while it's covered and it's, it |
|
| 22:06 | enough to do it. OK? um that's, that's quite handy. |
|
| 22:12 | , the only then you'll find annoying you go through your career as an |
|
| 22:22 | . There's lots of annoying things, one of them here, I'm looking |
|
| 22:25 | it. Oh, just this And then Utah says, hey, |
|
| 22:34 | , you see that and then he my hand, everything is rotate and |
|
| 22:37 | can never get back to that perspective . So I want to show before |
|
| 22:41 | after with interpretation without interpretation, one attribute three original seismic. I |
|
| 22:48 | get that same view that's like You gotta have to, you have |
|
| 22:53 | kind of save intermediate projects and then you can get that same view |
|
| 23:00 | time. But other than that, don't know, do you know of |
|
| 23:04 | Bob knows a different way of saving back to a particular view. |
|
| 23:10 | it's hard. I mean, uh just hard and it's kind of |
|
| 23:15 | But what can you do? And other annoying one is 3d visualization. |
|
| 23:22 | , and you're working with transparency, , not the exercises we're doing. |
|
| 23:27 | are pretty simple, but I'm talking like I've got five lines and I'm |
|
| 23:33 | put them together five vertical lines and gonna change the opacity curve on |
|
| 23:39 | the yellow, red, white blue color bar to look at the amplitude |
|
| 23:46 | then I'm gonna rotate them. So optically stack in a certain way, |
|
| 23:52 | the transparency and that fault shows up or that channel falls up beautifully. |
|
| 24:00 | then you come back on Monday, that channel? How do they do |
|
| 24:08 | ? Really annoying? So you'll find happens and then you'll come up with |
|
| 24:12 | own ways of remembering how you did , like writing, writing things |
|
| 24:16 | taking screen dumps, putting them in . That's about the only way of |
|
| 24:21 | it. So anyhow, OK, just want to mention that. So |
|
| 24:26 | you make for the, the lab that I'm asking open up powerpoints, |
|
| 24:33 | the figure in there, maybe write lines of notes or something like that |
|
| 24:36 | there's a color bar, I'd grab color bar and just put it to |
|
| 24:40 | side because if you publish a paper publish a, a thesis, that's |
|
| 24:45 | you wanna do. You have a bar that you don't want to have |
|
| 24:48 | little bitty color bar with all these bitty numbers that you can't read and |
|
| 24:52 | mean anything. Like if it's Amplitude have a color bar, maybe |
|
| 24:56 | white, blue, you're gonna say zero, negative. Keep it |
|
| 25:02 | Minus 25,000 plus 25,000. Simple. don't need all those little numbers in |
|
| 25:09 | . Nobody, nobody can read OK. Ok. Shout out if |
|
| 25:14 | got a question put on the part going to overlap his lane. |
|
| 27:23 | When laughing at this point depth. . It's not the right. |
|
| 27:36 | so you have one in front of other now. So but make this |
|
| 27:41 | four or is it gonna be Yeah, let me make it every |
|
| 27:46 | and then let's go go up a . Uh so it's 2020. |
|
| 27:53 | Ok, you can type it Um so and so they're both at |
|
| 28:00 | now or is one at one with one? Ok, so make the |
|
| 28:06 | one at 2000 uh 2820 16. I doing that right? My or |
|
| 28:20 | I guess make 14 milliseconds above the . That's what I want to |
|
| 28:25 | You can do that. I can't I'm looking at everything upside down. |
|
| 28:32 | . 2000, 2000. Well, now if you got them both at |
|
| 28:35 | and it's ok. Well, no this is the 2000? |
|
| 28:41 | ok. Ok. So that that works kind of. Ok. |
|
| 28:46 | . But when you put them on same level then it doesn't work, |
|
| 28:48 | ? Yes. Ok. So let's to the next thing you are going |
|
| 28:54 | . Which one is your base? guess it's gonna be the uh |
|
| 28:59 | So which one is the amplitude this ? 2016? Yeah. Ok. |
|
| 29:05 | cool. Ok. That's cool. you're gonna hit OK, Here, |
|
| 29:13 | this white blue thing. Click that display blue boxes. You're gonna put |
|
| 29:21 | check in the blue box. You're uncheck. What else do you have |
|
| 29:30 | here? Oh, none of OK. That's OK. So the |
|
| 29:35 | one you have displayed is this now click your VCR button and oops. |
|
| 29:48 | . Now you have one still displayed actually that where is the other one |
|
| 29:55 | was saying? OK. But Oh you had more stuff. |
|
| 30:04 | Where I only wanna have this guy the migrated data and OK. Is |
|
| 30:15 | what you have now? Yeah. . So take that. Yeah, |
|
| 30:18 | have that display. OK. So , so you can go uh |
|
| 30:23 | maybe, maybe animated. So they be locked together. No, it's |
|
| 30:30 | . OK. Wait. OK. that. Oh I don't. Oh |
|
| 30:35 | it's black on black. OK. on there. You another thing there |
|
| 30:39 | be a paint can here or some . Uh OK. Good. |
|
| 30:44 | Well, it's not there. Um me show you the thing because I'm |
|
| 30:50 | talking to here to Zach. In , I have a paint can of |
|
| 30:55 | on this particular picture. So, , let me go to my 3d |
|
| 31:09 | . Um, where was the paint at? It's the r uh, |
|
| 31:17 | to do. Oh, or after ? Ok. Ok. The paint |
|
| 31:26 | not empty and it's ok. It's upside down. You see the little |
|
| 31:29 | can icon, you gotta love icons patrol. You click that and you |
|
| 31:36 | make it go to a white Now, you can also change |
|
| 31:41 | So I might have it in the . I might not, you might |
|
| 31:45 | , click on and it'll see. right click and the de oh |
|
| 31:55 | Let's say you're writing um you're writing thesis for University of Houston and you |
|
| 32:04 | to print a hard copy. You're go through a whole lot of |
|
| 32:10 | a whole lot of ink with black . OK? If you wanna submit |
|
| 32:16 | for publication, the black background again just extra ink. The printer publisher |
|
| 32:23 | probably be angry at you. So know, you can change it |
|
| 32:26 | to uh I would call this a gray. This is white. |
|
| 32:34 | The touch screen. This is, anyone use the touchscreen here? Have |
|
| 32:39 | seen anyone use it? Who? . OK. What's he teach? |
|
| 32:58 | ? OK. OK. OK. modeling, geology, modeling, structural |
|
| 33:05 | . OK. Anyhow, it takes , takes a lot of choreography to |
|
| 33:08 | a touchscreen, especially in front of group that's really embarrassing. Anyhow, |
|
| 33:14 | is white, kind of a very gray. This is a great |
|
| 33:19 | OK? Because it's not stark. So uh just know that in |
|
| 33:25 | you know, you have to write report for work or something like that |
|
| 33:28 | you don't wanna like just have a done for black. You can put |
|
| 33:33 | and things a lot easier, you , little, little things. |
|
| 33:40 | So you should have them linked together , right? Oh, so |
|
| 33:45 | take this one off. No, quick unclick that I think that there's |
|
| 33:51 | issue. Nothing. Oh, you don't wanna be in a two |
|
| 33:56 | window. Let's do it in a window. Ok. Now you've |
|
| 34:02 | I see. Let's go. Let's look at the Z through the |
|
| 34:08 | Yeah. Take that off. Now you're gonna hit your little blue |
|
| 34:11 | button thing. Ok. And click variant blue button and, because they |
|
| 34:22 | us do all the, the majority the steps in one seat were in |
|
| 34:27 | . Oh, ok. Ok. fine then. Ok. I didn't |
|
| 34:30 | where you were and you hear you're at a True Cube. Ah. |
|
| 34:35 | right. All right. You're looking the, ok. That's cool. |
|
| 34:39 | think you're all right. Ok. . Mm. Yeah, once you |
|
| 38:27 | it, they can do it on computer. Now. I need to |
|
| 38:37 | the answer because the system will. . Right. And I, let's |
|
| 38:45 | at the answer. Ok. That's embarrassing. Ok, let's do |
|
| 38:57 | Like ch, let me make you . Mhm. Mhm. Ok. |
|
| 48:17 | what I did. Ok. ok, let me show is, |
|
| 49:23 | , Jessica on there yet. I'm still here, beer again. |
|
| 49:31 | still here. What you do help help the German economy. It's October |
|
| 49:36 | . Right? Um, actually Oktoberfest in September. Oh, your speaker |
|
| 49:46 | not working. OK. She's OK. So um I'll let you |
|
| 49:54 | her and then I'm still, you're recording me, right? OK. |
|
| 50:01 | . Tell me when to go and try to capture this image too. |
|
| 50:13 | There will be AAA PG explorer. any of you get the A PG |
|
| 50:18 | . It's like a, he writes Geophysical Foreigner, um which means every |
|
| 50:28 | he needs to have somebody to help write it. Uh And uh when |
|
| 50:33 | was with the big company, the two years worked fine because there were |
|
| 50:38 | 24 geoscientists there, he could draw then after two years, people started |
|
| 50:45 | their doors on them. They didn't to help him anymore. And uh |
|
| 50:51 | then he's drafted me to help and . So though, I think it's |
|
| 50:57 | month's issue, the little thing calling title. It's only a two page |
|
| 51:03 | in the A PG explorer. So , it's getting, getting bit by |
|
| 51:10 | many bites. Isn't that a cute ? OK. So if you did |
|
| 51:16 | I just did when I was playing loading the data and then realizing |
|
| 51:25 | I just use the default. Hey, everybody uses the defaults and |
|
| 51:30 | default is you take that big which is I think 200,000 and a |
|
| 51:36 | number, which is maybe 100 and if I recall what it is. |
|
| 51:42 | then I'm gonna choose to realize the . And that means I take the |
|
| 51:47 | bit data and I convert it to bit data so that it's easier to |
|
| 51:51 | at. Oh, it actually does things. Um And I could share |
|
| 51:56 | with you because that's what's in the . I'll try to put that |
|
| 52:00 | Um It saves the data as brick . So I think in the |
|
| 52:06 | the first exercise I have you look in lines and go through them and |
|
| 52:10 | cross lines and go through them and cross lines are gonna go slower than |
|
| 52:14 | in lines. These computers not too because it's got a ridiculous amount of |
|
| 52:19 | in it. OK? But my , it goes much slower and then |
|
| 52:24 | slice is slower still. So uh do a time slice, guess what |
|
| 52:29 | have to read in every single extract one sample and throw everything else |
|
| 52:36 | , not very efficient. So the format breaks the data into bricks of |
|
| 52:41 | size, typically 64 by 64 by boxes. OK. And so what |
|
| 52:50 | will do for uh maybe this data , if I can do my arithmetic |
|
| 52:56 | ? It will be 16 bricks deep about 20 bricks wide and about 24 |
|
| 53:03 | long. So if I wanna pick time slice, I just bring in |
|
| 53:08 | row of bricks. Not all of data. If I wanna look at |
|
| 53:12 | horizon slice, I just pick the that overlap that horizon. So you |
|
| 53:18 | see where it's much more efficient than in trace by trace in line, |
|
| 53:24 | by in line, followed by in . OK? So I just clicked |
|
| 53:28 | buttons. It said, realized it go eight bits. It's gonna take |
|
| 53:33 | number 200,000 minus 131,000. OK? divided by 255. And so the |
|
| 53:44 | of my bin is I can't do number in my head, but it's |
|
| 53:49 | 400,000 divided by 250. That's the of the bin. OK? So |
|
| 53:54 | is those data. Then I went same data volume and I said before |
|
| 54:02 | realized it, let me say I the maximum value to be 2500, |
|
| 54:07 | minimum value to be minus 2500. almost 10 times finer or 10 times |
|
| 54:15 | . OK? And so 200,000 to . So eight times less, then |
|
| 54:24 | gonna scale that value of 225,000 minus divided by 255. That will be |
|
| 54:31 | B touch. So here is the , same color bar, same range |
|
| 54:40 | the color bars that the color bar be 25 to minus 25. And |
|
| 54:47 | here is rescale. Default. We're , default, we're scaling, |
|
| 54:56 | we're scaling. So you'll notice I can't really use my mouth but |
|
| 55:04 | can see my hand here, this on that color bar. That's the |
|
| 55:11 | that's near zero. I can show color bar. He said you can |
|
| 55:19 | the color bar. See here, one is the default 200,000 to minus |
|
| 55:25 | 100 and 30. 0, it the number here minus 134 to 2200 |
|
| 55:29 | 12,000. Oh, here's my color . I'll put that over here. |
|
| 55:38 | says, well, I can't make a lot smaller, but I can |
|
| 55:42 | it a little smaller. Ok. how does, what does this |
|
| 55:49 | This is my a my data are binned incorrectly. Let me look at |
|
| 55:54 | color bar from here just to make I'm not making a mistake and colored |
|
| 56:03 | we go and, and, maybe that's the point no minus 25,000 |
|
| 56:13 | 25,000. Ok. So same color . So what's happened? Everything is |
|
| 56:19 | scared at those lower amplitude. It's , it's just not being mapped |
|
| 56:27 | So the, the peaks and the are doing the, the how would |
|
| 56:30 | call it? The high, high minimum are doing fine, but |
|
| 56:37 | losing the detail in between. So , you need to be careful. |
|
| 56:41 | can't just trust the defaults. In fact, what I'll do is |
|
| 56:46 | will, I think I have this here. I think Wednesday with |
|
| 57:00 | Um probably a raise it. OK, I'll put, put that |
|
| 57:20 | , I'll put that in the, the lab notes. That's the best |
|
| 57:23 | for it now. Um, everybody's doing. Ok. Everybody's happy |
|
| 57:38 | trying to figure out what, why one look like crap and the other |
|
| 57:42 | ? Ok. Right. Yeah. . Ok. Let me give |
|
| 57:47 | uh, I know I have this , uh, uh, I incorporated |
|
| 57:53 | in one, I did two weeks and Oman and I did lectures and |
|
| 58:04 | display. Same kind of thing I to you guys about but I may |
|
| 58:13 | added another pitfall. Where did I that pitfall? Mm hm hm hm |
|
| 58:21 | hm hm hm. I guess it's in there. Where would I put |
|
| 58:29 | ? Oh, well, that's because it's not the one on my |
|
| 58:37 | . There you go. No. ? Yeah, you're right. |
|
| 58:51 | I'll find it guys. Oh, am I gonna get that? I |
|
| 59:03 | why. Mhm. Because I didn't it. Yy. Yeah, I |
|
| 59:24 | to use the web, you Oh, yeah. Ok. |
|
| 62:54 | Hm. Why can I find Ok. I know I put it |
|
| 63:20 | . You want, I got it . Ok. I said that |
|
| 64:16 | So here, right? Ok. , you feel like uh everything is |
|
| 64:43 | pretty gone out and whatever? Ok. Hm. Thank you. |
|
| 66:21 | lost my, oh, maybe I was in the internet in Google |
|
| 66:28 | . Where did it go? This , no, this is, but |
|
| 66:35 | was just doing it. Um, that it? No, I got |
|
| 66:41 | gotta go back. No, you open it. Oh because I was |
|
| 66:47 | you need to uh got it. . Interesting. I know how we |
|
| 66:56 | . I know I was just I you know on the edge and Microsoft |
|
| 67:01 | ok, so we'll use edge. . That's fine. That's edge. |
|
| 67:06 | we need uh what its doctor? ok. I think no that's not |
|
| 67:16 | . So yes there um I don't leave it. Ok. Uh where |
|
| 67:32 | my edge go? Ok, here am. Ok. Mhm. |
|
| 67:47 | I was in office 365 logged in was just logged in. Oh |
|
| 68:03 | Ok. Many. Mhm. The what? There's no way that you |
|
| 68:47 | know. Yeah yeah it's perfect um thank you. Thank you. |
|
| 70:12 | All right. Hello. Well the . So just so when you start |
|
| 70:57 | um that yeah. Ok, so is ok. Yeah then I gotta |
|
| 71:17 | to all. I OK. Thank you so very much but that |
|
| 71:57 | so um you know what I know . Mhm. Mhm. That |
|
| 72:29 | Ok. You. Mhm. Yes, sir. I know. |
|
| 72:47 | funny. Ok. Yes, thank . So you um um so that |
|
| 74:59 | ok. Mm. Thank you. then we. Ok. Mhm. |
|
| 75:29 | . Mm. Ok. Ok. . Ok. I found the article |
|
| 79:06 | I can't remember my password from getting . Too many passwords. I'll figure |
|
| 79:13 | out tonight, but basically this is one getting bit by too few bites |
|
| 79:18 | September. And it showed that same of problem. I can show you |
|
| 79:24 | , a crude version of the same if you. All right with |
|
| 79:28 | I, I think I've got it . Uh Download tank. I gotta |
|
| 79:39 | to download, here's a crude This isn't it there it is. |
|
| 80:08 | And I know, damn it, me see if I can find the |
|
| 80:17 | image. Mhm. Oh OK. , no, no, no. |
|
| 82:50 | where it was. I remember What? Ok. Mhm. |
|
| 83:30 | Ok. Let's see if I can this. Uh Are we recording for |
|
| 83:35 | ? All right. So these are pictures from the book I'm writing and |
|
| 83:39 | is in the A PG thing, PC explorer for September. But what |
|
| 83:45 | trying to show on this picture is way we gener, we try to |
|
| 83:50 | data as bricks and all the commercial does this companies like blue wear, |
|
| 83:57 | have four and five dimensional bricks so , very efficient uh access to prestack |
|
| 84:06 | . And the point here is, , if I wanna slice an in |
|
| 84:10 | , it's pretty easy. If I slice a cross line, it's quite |
|
| 84:14 | bit longer. And if I want slice a time slice, it's the |
|
| 84:19 | of all. So if I have brick format, it's faster, then |
|
| 84:24 | paper goes back to uh 2001 from Earth from Texaco. People was developed |
|
| 84:31 | Texaco, was spun off as their company 3D Visualization. Then Landmark bought |
|
| 84:37 | . And that's when 3D 3D visualization commercial software uh really hit, hit |
|
| 84:43 | ground and uh they show uh pain still over there. Uh Here's |
|
| 84:53 | happens with it to 48 times. root means amplitude and the one on |
|
| 85:04 | right is the four times root mean of the data. OK. So |
|
| 85:10 | gonna, you're gonna pick some of numbers out of a hat as to |
|
| 85:13 | to scale the data. And in example, uh it's nice and |
|
| 85:20 | OK? Just like the data should . And in the previous and the |
|
| 85:25 | times it's quipped. Remember I well, it's OK to quip so |
|
| 85:30 | as you know, you're quipping. now this is what's stored in the |
|
| 85:36 | when it's converted da data. Now gonna compute frequencies. So in a |
|
| 85:45 | minutes, maybe 1015 minutes, I'll talking about instantaneous attributes. You're probably |
|
| 85:51 | with them right now and a sharp has higher frequencies. Than OK. |
|
| 86:02 | this is gonna have high frequencies to it. And so with this other |
|
| 86:07 | here and then what happens when you instantaneous frequencies on the left, they |
|
| 86:14 | a nice smooth frequencies and on the , they have a couple of artifacts |
|
| 86:19 | got created. OK. So his here is that when you got bright |
|
| 86:25 | and not, you can have artifacts the bright spot. So that's out |
|
| 86:32 | . Now, here's a data set to be Gulf of Mexico. Um |
|
| 86:40 | here's the way Petrell wanted to bring in ranging from minus 30 to plus |
|
| 86:48 | . And here is the histogram under around zero like we think it should |
|
| 86:54 | . And then I take the same here. It is in certain |
|
| 87:00 | they bought it but they changed the not to be 00 minus 25 to |
|
| 87:15 | . Thanks. So this looks fine bit data and this eight bit data |
|
| 87:24 | it looks OK, especially up but notice there's something different in this |
|
| 87:32 | box. I see a lot of areas. And then in this other |
|
| 87:37 | , I'm going to uns scale the minus five and five instead of minus |
|
| 87:44 | . And here's the default I'm gonna symmetric and I minus 25 plus 25 |
|
| 87:51 | minus 5 to 5. And then I plot them with the same, |
|
| 87:57 | same except it's, yeah, this is going to have, it's got |
|
| 88:06 | colors. So the beams here are be tighter, five times tighter than |
|
| 88:13 | . So let's look at those and go look in here and I actually |
|
| 88:17 | it in patrol because you put your on it, you click, it |
|
| 88:21 | you what the values are, what amplitude is. Well, there's a |
|
| 88:25 | of zeros in here. OK. I've got a zero, I've got |
|
| 88:30 | numbers in here and here because the are d are, the number of |
|
| 88:38 | per bucket is five times denser. don't have that dead space in |
|
| 88:47 | OK? Now I'm gonna run instantaneous , huh? Instantaneous frequency because that's |
|
| 88:54 | Myers did 24 years ago. Here's the original data, 32 bit |
|
| 89:02 | Here it is on the eight beat data carefully scaled. Not that almost |
|
| 89:08 | . And then here is the one improperly proper, too coarsely, too |
|
| 89:14 | scale using the deep ball. And can see all this black hair that's |
|
| 89:18 | trash. It does. It, thinks it's all zero. It thinks |
|
| 89:22 | data are zero and they're not. it gets worse than that because here |
|
| 89:32 | was careful and I, I say in the text, I probably have |
|
| 89:36 | picture when I scale it from minus to 25. I've intervened. I |
|
| 89:45 | it's symmetric about zero. OK? I let it use the default minus |
|
| 89:51 | to plus 26.4. It's not There is no bin that is at |
|
| 90:00 | . So my value of zero is , and this picture is like point |
|
| 90:11 | . All right. And you well, that's not so bad. |
|
| 90:14 | , it can be disastrous because in workstation software, they don't keep trace |
|
| 90:20 | . So they don't know where the traces are and they don't know where |
|
| 90:24 | mutes are. So something is constant and I export it into patrol and |
|
| 90:32 | want to do a bo analysis. here I've got him oh amplitudes increasing |
|
| 90:40 | offset for good bright spot. Yeah. Yeah. And then it |
|
| 90:44 | to 0003. Well, 0003 is a dead trace. It's got a |
|
| 90:50 | in it. So it's gonna try fit a curve that goes like that |
|
| 90:55 | you get garbage. OK? And if you have mutes or do |
|
| 91:01 | things that are sensitive to edges. , if I have a strict zero |
|
| 91:06 | there, I can say, all , that's a mute zone or no |
|
| 91:09 | zone, whatever it is, don't in there. And then when the |
|
| 91:15 | wants to go in there, be careful about it, you know, |
|
| 91:19 | it in some way. Well, it thinks it's alive, it's, |
|
| 91:23 | gonna die and give you problems. the paper, the picture I just |
|
| 91:28 | you was for the data you're working and um you know, kind of |
|
| 91:33 | careful that you gotta think about The conservative thing, saving his 32 |
|
| 91:40 | alternatively, save it as 16 So, but uh here's a problem |
|
| 91:45 | comes in, it's in that, that little September Geophysical corner thing |
|
| 91:51 | you know, I've had and Heather students coming up from Petrobras and they |
|
| 91:58 | , oh, we'll give you a and then you can do this |
|
| 92:01 | do the analysis on the deep stuff the basement. We were really interested |
|
| 92:06 | it, but it's not something we time. And here one of the |
|
| 92:10 | she came in and she had all data and up shallow, well |
|
| 92:15 | the top three seconds look beautiful. then where they wanted her to look |
|
| 92:19 | and five seconds. Well, a of the data was simply set to |
|
| 92:25 | because it was, they gave her eight bit data set from patrol or |
|
| 92:31 | wherever exported it as 32 bit. why? So she's thinking it's 32 |
|
| 92:38 | because that's what segu is. But with eight bit converted back to 32 |
|
| 92:43 | and she wasted like four or five not knowing what it was, what's |
|
| 92:49 | on. OK. So we will you know, kind of go to |
|
| 92:54 | place you can keep your projects open then I'll start lecturing maybe in five |
|
| 92:59 | . Is that ok? Ok. we'll go to close to six o'clock |
|
| 93:13 | I will send you that little article I get home and the church. |
|
| 93:44 | . It'll turn out that the eight using the default. It is bad |
|
| 93:50 | seismic amplitude data. Uh In contrast most, not all of the |
|
| 93:57 | for most attributes, eight bits going be fine. Our frequencies are gonna |
|
| 94:02 | from 0 to 100 and 25 It's fine. It'll be good to |
|
| 94:07 | half frequency coherence variants will go between and one. You'd be hard pressed |
|
| 94:14 | see details, you know, less 1 255th of that range. When |
|
| 94:20 | comes to R MS amplitude magnitude spectral , then we got to be more |
|
| 94:27 | . So I'll try to capture that we go through the course. Uh |
|
| 94:33 | I mentioned it just because recently a of our students said, oh, |
|
| 94:37 | have been burned by this kind of . OK. So most of you |
|
| 94:42 | started a little bit with grace right? So let's see what they |
|
| 94:47 | . What we want to be able do is visualize something called the Hilbert |
|
| 94:52 | of the data by looking at the seismic trace. Then second use complex |
|
| 94:59 | attributes as building blocks for interpretation and attributes use attributes computed from interpreted uh |
|
| 95:11 | structure maps to enhance subtle faults and . So we're gonna calculate like the |
|
| 95:17 | magnitude dip ASU of a P surface statistical measures of amplitude above or |
|
| 95:23 | picked horizons to map chaotic features that can't be picked. So we have |
|
| 95:29 | very complex uh channel system that's been three or four times. Uh So |
|
| 95:41 | no simple reflectors inside the channel. rather a bunch of heterogeneous packages. |
|
| 95:48 | , we can pick an area or may be parallel to the top of |
|
| 95:53 | channel system. Uh And then a simple reflector below it, not the |
|
| 96:00 | of the channel, but one below calculate RM amplitude. We're gonna be |
|
| 96:05 | to see that channel it well, . So here, let's start with |
|
| 96:10 | Hilbert transform. I've got uh this what the Hilbert transform looks like in |
|
| 96:17 | time domain. So I've got time from left to right. And here |
|
| 96:23 | my analysis point right here in So what I do is I take |
|
| 96:30 | sample that's one point to the And I multiply it by 11 point |
|
| 96:36 | the left, I multiply it by one, three points to the |
|
| 96:42 | I multiply it by three points to right. I multiplied by one |
|
| 96:49 | three points to the left minus one , five points to the right, |
|
| 96:56 | 5 points to the left minus 1/5 7th, 1/9 all the odd |
|
| 97:03 | So basically, we're taking a weighted of that grace where the weights are |
|
| 97:12 | over time with respect to where the of where that sample is. |
|
| 97:20 | Um Zach here, of course, me to use the word convolution. |
|
| 97:24 | I'll say yes, we can evolve little brown pattern with the seismic |
|
| 97:33 | That's all it is. It's just weighted average. OK. Now, |
|
| 97:38 | of all, I want to say not instantaneous, it's a weighted average |
|
| 97:42 | some kind. So the instantaneous attributes instantaneous. They're not measuring exactly at |
|
| 97:50 | spot. It's smeared by the Hilbert . And then because our seismic wavelet |
|
| 97:57 | you know, peak and sidewalks, also smeary. So nothing really instantaneous |
|
| 98:04 | . That's the word that has historically used. OK. Paper from Tory |
|
| 98:10 | in 1979 I might have included It's a classic and here's the seismic |
|
| 98:18 | , here's the data. This one here is a Hilbert transform. It's |
|
| 98:22 | three names. We call it the transform of the data. We call |
|
| 98:27 | the quadrature of the data. We it the imaginary part and, and |
|
| 98:33 | . Is it the quadrature or Hilbert quadrature? OK. So that comes |
|
| 98:41 | electrical engineering. And Hilbert transform is mathematics. People from Kant and |
|
| 98:49 | but also other uh sign analysis. they like to use complex numbers. |
|
| 98:56 | they'll talk about real and imaginary Three names are the same thing get |
|
| 99:01 | to it. OK? It makes head spin. You'll read the documentation |
|
| 99:06 | patrol and they'll talk about the imaginary . Hold on. What's the imaginary |
|
| 99:11 | to help her transform or I'm the quadrature. So they'll call it |
|
| 99:16 | different things in the documentation. Add confusion. OK. Let's look at |
|
| 99:23 | peak. A peak goes from a to plus zero crossing a trough goes |
|
| 99:32 | a plus to minus zero, crossing minus to plus zero, crossing goes |
|
| 99:39 | a trough and a plus to minus crossing goes to a peak. So |
|
| 99:46 | it does is it rotates the data 90 degrees. OK. Um Not |
|
| 99:52 | to do um If you have thin a thin reservoir, what will happen |
|
| 100:05 | that, that tuning effect is well, we'll go into that but |
|
| 100:09 | gonna go rotate those data by 90 . OK? Now I've got the |
|
| 100:14 | transform. I can plot the data a horizontal axis. I can plot |
|
| 100:21 | Hilbert transform on a vertical axis and can think of them as real and |
|
| 100:26 | parts. Um But first of let's take the data squared, the |
|
| 100:33 | transform squared, some of them take square root that's called the envelope. |
|
| 100:39 | got another name too. Reflection And what for I? And you |
|
| 100:48 | see the envelope is insensitive to So even though this one is 90 |
|
| 100:53 | phase, the envelope. Thank Here's my plot. So what we |
|
| 101:03 | then here I'm gonna calculate an angle the real axis and here is sample |
|
| 101:10 | milliseconds, 2002, 2004 2006, , 2010. Ah The phase is |
|
| 101:18 | and then here for whatever reason, , the phase is decreasing, it's |
|
| 101:24 | in the other direction, then the increases again. OK. So here's |
|
| 101:31 | phase and it's gonna be the art of the Hilbert transform of the part |
|
| 101:38 | the data and the original data. uh so here's zero degree phase. |
|
| 101:43 | increasing. Oops wraps around, comes , wraps around. Oops here goes |
|
| 101:48 | . What the heck is that? ? Now, if you think of |
|
| 101:53 | is, I mean perspective, just it. So that's what this bottom |
|
| 102:05 | is saying. We gonna take the of the phase and everything is kind |
|
| 102:10 | fine except what, what the heck negative frequency. And my vibrator only |
|
| 102:16 | to 60 Hertz. How do I 90 Hertz in the data? That |
|
| 102:21 | make any sense? What are computational ? And then here is the um |
|
| 102:30 | those artifacts occur. I've got a one negative frequency here. Oh That's |
|
| 102:36 | my phase went backwards. This is my phase went backward phase went |
|
| 102:41 | What about the positive ones? Oh phase went particularly fast. Well, |
|
| 102:46 | do they occur? Uh at the of the envelopes, see how they |
|
| 102:52 | the red wines all are at the of the envelopes. So what we |
|
| 102:56 | is the wave packets are interfering with other and that's what's causing those, |
|
| 103:03 | uh artifacts and phase. So if want to fix that and I recommend |
|
| 103:09 | do fix it, we're gonna take weighted average frequency. OK. This |
|
| 103:14 | what most people do. But there , there are instantaneous instantaneous frequencies out |
|
| 103:21 | . So we're gonna take a window 11 samples and it's stable at the |
|
| 103:28 | . So it's gonna have a nice answer. Uh It's gonna be biased |
|
| 103:33 | the peak of the envelopes by this of weighted average. OK. I |
|
| 103:39 | down to Terry Tanner's uh office while was still on this earth because |
|
| 103:44 | he died maybe 10 years ago. this is how we used to display |
|
| 103:48 | at seg meeting kind of like uh kids might do in uh junior high |
|
| 103:54 | . You get a big piece of board and then you carefully get some |
|
| 104:00 | , try not to gloop it too and then and wetter gas and plastic |
|
| 104:08 | with something called the Leroy pen real . You're very fortunate to have to |
|
| 104:12 | that anymore pain. His, his are perfectly well. OK. But |
|
| 104:23 | we are for envelope instantaneous frequency weighted . And the apparent polarity is that |
|
| 104:34 | envelope finds the kind of the OK. Let's look at the bed |
|
| 104:41 | and the wedge model high impedance, impedance, high impedance, negative reflection |
|
| 104:47 | the top positive reflection on the bottom peak trough for the zero phase part |
|
| 104:53 | the reflection peak trough peak for then degree phase part of the reflection and |
|
| 105:02 | they interfere, we have tuning and it gets weaker and weaker goes to |
|
| 105:08 | . Here's the Hilbert transform. It the data by 90 degrees. So |
|
| 105:13 | I have a trough here, I a zero fraud. And where I |
|
| 105:19 | a fee on the bottom, I there it is uh is always a |
|
| 105:28 | crossing at time. Uh 50 in picture. Well, you can see |
|
| 105:35 | was there. So for a thin , it's not uncommon if you're gonna |
|
| 105:40 | a horizon that's at the top of thin bed. Let's pick, let's |
|
| 105:46 | pick the zero crossing on the original . But then display the Hilbert transform |
|
| 105:52 | to actually show how, what's the change in that thin bed? Otherwise |
|
| 105:59 | get zero. That's not very OK? All right. So here's |
|
| 106:07 | answer for the not talking to There's too much information in there. |
|
| 106:28 | ? It's just not telling me much the wavelet. It's telling me about |
|
| 106:32 | particular part of the wavelet. What's face? So there's another thing we |
|
| 106:40 | . Mm W what attributes? So gonna as the zones by the |
|
| 106:52 | No stop the envelope and then the the same like we saw and then |
|
| 107:10 | and the drop times where, where the, the minimum in the envelope |
|
| 107:17 | ? OK. Here, here, , here, here and here. |
|
| 107:23 | . So we're gonna go then hello, extract the instantaneous frequency where |
|
| 107:33 | nice and stable and apply it to . We're gonna gonna go to the |
|
| 107:39 | low one constant value per energy We're gonna do the next one. |
|
| 107:45 | then the next one is to get idea how we're taking the value of |
|
| 107:50 | attribute at the envelope peak and say represents the range of the envelope and |
|
| 107:56 | kind of stabilizes the thing. You get any negative frequencies which don't mean |
|
| 108:01 | . We don't have any erratic positive , we get very reasonable ones. |
|
| 108:07 | here is the envelope. Um that's OK. Maybe I'm losing some |
|
| 108:13 | I still see the tuning frequency or the tuning is strong and here's the |
|
| 108:19 | phase. Now this one I like than that previous. It had all |
|
| 108:25 | to put out here. And what have is the top here is kind |
|
| 108:30 | like purple. Oh that's minus 180 plus 180 and here is yellow. |
|
| 108:38 | that's zero. So remember I had zero phase wavel at the bottom and |
|
| 108:43 | negative reflection or 100 and 80 degree at the top. Then as I |
|
| 108:50 | it read oh 96. OK. wavelength frequency OK. Uh kind of |
|
| 109:01 | moderate frequency of let's say 25 or Hertz here, then it gets a |
|
| 109:07 | lower because of the side of and it gets higher here at 30 after |
|
| 109:14 | I mean, there's the instantaneous. in my book, the uh wavelength |
|
| 109:20 | are better finding you use these Um They're in every commercial software package |
|
| 109:31 | was really easy to write. They their impact when he was at the |
|
| 109:37 | Service Road up, uh happen to Fort Worth Station and he's got the |
|
| 109:47 | frequency. So uh high frequencies and is in yellow and lower frequencies are |
|
| 109:57 | blue and then he's only got some the wells shown. He's got actually |
|
| 110:01 | 100 wells. So he's got a map. Ok? And I remember |
|
| 110:09 | or maybe. Yeah. Yes. I shut out blend colors and |
|
| 110:14 | That's what I did here. These the kind of things you do. |
|
| 110:21 | ? Now another attribute that's used particularly a Strat ofer. OK. The |
|
| 110:31 | don't tell Jay. I did OK? I just took a drink |
|
| 110:36 | water. Damn. You can put on your hands and say because it's |
|
| 110:41 | you're, you're worried about COVID. Oh, if you're a Strat |
|
| 110:52 | any of you guys, photographers or ofer wanna be Strat first. Absolutely |
|
| 111:00 | . OK. She's got good. you're a geophysicist man sequence Strat |
|
| 111:06 | Really painful. Where's my geology folks us about why? Why do we |
|
| 111:17 | about sea level rise and fall? , it was OK. Good. |
|
| 111:23 | you know something. So OK, share with us, please. Why |
|
| 111:29 | we care about sea level rise and in terms of the depositional environment, |
|
| 111:33 | say right here in the Gulf of is like creating the entire for |
|
| 111:44 | OK. OK. OK. So say I wanna look for turbines in |
|
| 111:58 | Gulf of Mexico. Where do I look? I mean, I got |
|
| 112:02 | seconds of data. There are a of places to look for, of |
|
| 112:22 | . OK. So right now 2023 at a high stand going higher. |
|
| 112:32 | anybody live in Queer Lake where I to live, I was 10 ft |
|
| 112:35 | sea level. Anyhow, as sea rises, we're gonna flood more of |
|
| 112:42 | earth. Oh, there the ocean transgressing against me like sinning against |
|
| 112:48 | That's how I remember transgressing. It's like doing something against me. |
|
| 112:53 | right. So sea level is rising it's gonna dump more clay and stuff |
|
| 112:58 | top of everything. If sea level , that regresses. If sea level |
|
| 113:06 | , we'll have a falling stage And then we have a low stand |
|
| 113:10 | sea level is an absolute low. during the ice age, then all |
|
| 113:15 | the sand and uh sand fans like Mississippi Delta all of a sudden that's |
|
| 113:22 | above ground and it's unstable and we'll a flood, heavy rain and it's |
|
| 113:29 | take all that stuff and send it into the deep Gulf of Mexico and |
|
| 113:35 | a turbine. OK. So where look for Turbos is that low |
|
| 113:42 | So you were, you were you, you, if you had |
|
| 113:45 | more positive attitude about Strat gray, could have answered that very cogently. |
|
| 113:53 | . So one of the pieces of , oh, this is a turbid |
|
| 113:58 | and it's gonna be sand rich is know where does it fit in the |
|
| 114:03 | boundary? In carbonates, high stand low stand. If I've gotten a |
|
| 114:09 | platform and if I can tell when a low stand, well, then |
|
| 114:14 | Bahama Islands are gonna be exposed to , they're gonna dissolve and have more |
|
| 114:24 | in them. And if I can the level high and preserve that |
|
| 114:32 | that understanding of when things happen when migrated when they were exposed is all |
|
| 114:39 | of the exploration system. And that's what your seismic and sequence ders |
|
| 114:44 | and mechanically what they do on this with the arrow. Now, that's |
|
| 114:50 | you don't wanna do it. I that Stephanie, you don't, you |
|
| 114:53 | wanna pick all those fingerprint. Ok. So my brother worked, |
|
| 115:00 | my brother, uh, Alan, worked for FBI for a bit and |
|
| 115:07 | they were doing finger based databases, it, they didn't start by matching |
|
| 115:12 | the fingerprints they matched at the on and a lapse, the intersection of |
|
| 115:18 | swirls. Yeah, you've got a on your hand. You're real |
|
| 115:24 | You know, you cut yourself as knife. Remember when you were |
|
| 115:27 | Yeah. Become a blood brother with , your best friend, you got |
|
| 115:31 | little scars makes you really easy to , but that's how they did |
|
| 115:36 | OK. So here, uh let's , instead of the, the amplitude |
|
| 115:45 | , let's take the phase and compute cosine of phase. And what that |
|
| 115:50 | is it equalizes everything and it be it in a gray scale. It |
|
| 115:54 | it really not really easy. It's tedious, but it's easier to see |
|
| 116:00 | the terminations. You see the weak and the strong terminations. So you'll |
|
| 116:05 | those folks who do that use this quite a bit, then we can |
|
| 116:11 | things. So you might call this meta attribute. I would prefer to |
|
| 116:16 | it uh with color instead of but it's OK. And this is |
|
| 116:22 | that made it into the into the uh was originally developed at Texaco and |
|
| 116:29 | called it sweetness. OK? And they called it bitterness, nobody want |
|
| 116:34 | use it, right. Yeah. sweetness. Oh These are the sweet |
|
| 116:39 | because they recognized in the Gulf of and offshore um uh West uh Northwest |
|
| 116:49 | that if I had a sand channel in a shale, well, the |
|
| 116:53 | channel gave a stronger reflection. It be a negative amplitude but a stronger |
|
| 117:00 | . So high envelope and it expressed it because it was more homogeneous. |
|
| 117:07 | didn't have the higher frequencies in So it was also a lower |
|
| 117:12 | So let's put the strong amplitude in though lower uh frequency in the denominator |
|
| 117:21 | call it sweetness. And then they a square root in there just for |
|
| 117:25 | you know, just what they felt good. OK. So you'll see |
|
| 117:31 | I think in picture. All OK. So instantaneous attributes provide a |
|
| 117:38 | but very fast estimate of envelope phase frequency of sic reflection. They degenerate |
|
| 117:44 | multiple reflections interfere with each other. , remember the little spikes and the |
|
| 117:54 | frequency. So some good friends of here in the city using machine |
|
| 118:06 | they'll put these instantaneous attributes in for survey like the Eagle Ford. And |
|
| 118:12 | they'll put an instantaneous frequency in uh and other things phase and then they'll |
|
| 118:19 | , oh look, I now have millisecond resolution because of uh artificial |
|
| 118:26 | machine learning. No, what they is they have some attribute that's got |
|
| 118:33 | in it where you're interfering. So not giving you better resolution. It's |
|
| 118:39 | you where the edge is. when you start working with var, |
|
| 118:44 | , you're working with variants. So is plotted the variant and you'll notice |
|
| 118:48 | you look at it carefully, those edges are one vel thin. |
|
| 118:55 | they're 25 m. OK? They're vel thin. It's not smeared, |
|
| 119:00 | pretty sharp. So well, hold . What about resolution? It may |
|
| 119:07 | one vel thin but guess what? could be 75 m to one side |
|
| 119:13 | 75 m to the other side, don't really know where those data |
|
| 119:18 | OK. So you think of if did a physics class and might have |
|
| 119:23 | about Fornell zones and limits of resolution so forth? So there's a uh |
|
| 119:30 | difference between what the, what the ? OK. Good example. Mm |
|
| 119:44 | test madam Gis, right? So something profound. So that Stephanie is |
|
| 119:55 | . OK? Don't, don't oh don't say I there's something profound |
|
| 120:01 | satellite uh mapping with Gis about what the resolution between me and Bob and |
|
| 120:14 | where Bob and I are. you're both in the lab. |
|
| 120:26 | Triangulation? Well, they, they triangulation. So there's relative distances which |
|
| 120:34 | really, really good at. So when they say like 1 m 2 |
|
| 120:40 | resolution, it can tell that Bob I are two or 3 m |
|
| 120:45 | But if it, if you asked where Bob and I are, it |
|
| 120:50 | know I'm in this room or it know I'm in this building. |
|
| 120:54 | That's absolute location. So you OK, we're always this far |
|
| 120:58 | But are we here or are we ? So we got that kind of |
|
| 121:02 | thing in there too. So when see the coherence is oh man, |
|
| 121:06 | fault is perfectly resolved. That looks . Yeah, but that fault may |
|
| 121:11 | be there. It's here. But those are the kind of trade |
|
| 121:17 | we have. So you look at giso talk, you, you homework |
|
| 121:21 | the weeks, you're gonna go look say, what's the, the relative |
|
| 121:26 | and the absolute. Now some of may be in a different class. |
|
| 121:31 | ever do surveying? OK. So surveying. We're really good with taking |
|
| 121:38 | tape measure and saying that, you , I'm 2015 ft from Haydn because |
|
| 121:48 | can measure it. But now if want to put us on a |
|
| 121:51 | well, then I've got to do these measurements until I find a benchmark |
|
| 121:57 | I know from many, many people tens and tens of to know where |
|
| 122:04 | benchmark is. What's my relationship or relationship to that benchmark? Those are |
|
| 122:09 | questions. OK. All right. response. Wait it out to be |
|
| 122:16 | general. Um, let's raise that gonna see everywhere. I'm, I |
|
| 122:31 | didn't say the effect but they, ones you're gonna see and then complex |
|
| 122:36 | can be used as auxiliary volumes and voxel interpretation to constrained auto pickers. |
|
| 122:43 | I would say, uh, not can it, but I would say |
|
| 122:48 | default, it's in all the auto today. So does, uh, |
|
| 122:54 | have any question? I don't think . Ok. Good, good. |
|
| 123:02 | see. I have to think of name too because I have a, |
|
| 123:05 | spouse's name is Stephanie and my granddaughter's is Stephanie and my daughter's name is |
|
| 123:14 | . So you can see how I confused and that, that's what am |
|
| 123:17 | thinking of them for? They're not this room? All right. Uh |
|
| 123:25 | I talked about this. That's Oh OK. So this is a |
|
| 123:32 | slide from earlier now, shaded I've got a horizon and here's my |
|
| 123:41 | coming in, here's the normal if angle of incidence equals the angle of |
|
| 123:48 | view, that is specular reflection. is what you're going to have in |
|
| 123:53 | mirror. So, cars, how you say mirror in a spec? |
|
| 124:01 | . So specular means mirror in, Latin. OK. So what you |
|
| 124:05 | is exactly what you get out of angle. And then from this |
|
| 124:10 | it's more of a diffuse reflection. I see the sun coming through, |
|
| 124:18 | a little. So I see that coming down. It's a little shinier |
|
| 124:22 | , but it, it's a little blurred. Maybe you guys can see |
|
| 124:25 | the light is coming down as So it's more diffused. I'm still |
|
| 124:29 | a reflection but it's more diffuse. fact, that's how I see the |
|
| 124:34 | brown. Like what's the color of mirror? Well, well, I |
|
| 124:37 | know. And you don't see the of the mirror, you just see |
|
| 124:41 | else. All right. So in uh the commercial software and uh I'd |
|
| 124:47 | they all do it very well. does a nice job. Um You're |
|
| 124:52 | look at non specular reflection. So gonna have to right the F secular |
|
| 125:10 | , very poorly documented as to what do in the code in patrol. |
|
| 125:17 | we get that far, you have have a horizon picked first a |
|
| 125:21 | Uh, the choices you get are . Ok. I understand that. |
|
| 125:35 | , plastic, I don't understand Woody, I guess that's like the |
|
| 125:42 | and they might have another one. they're using these real soft words or |
|
| 125:46 | equation are they using? I have idea. OK. But I think |
|
| 125:51 | know what a plasticky surface looks like what a woody surface like and a |
|
| 125:55 | surface looks so you can play with . OK. So here is uh |
|
| 126:02 | picked horizons. I picked the top this uh pro gradation surface. I |
|
| 126:08 | the face of it here at the of the surface and I am looking |
|
| 126:16 | it from north to south and then moved to school and pare and I |
|
| 126:30 | look at it a different night as, as a musician. I |
|
| 126:35 | also look at it at different angles second. Ah, that one, |
|
| 126:43 | what the tool looks like. You this guy. Uh And now I'm |
|
| 126:48 | play it and it's a system from Texas. So north central Texas, |
|
| 126:56 | got cars collapse in it, pick surface. So you can see the |
|
| 127:01 | cheese holes and you can see what did here or what I did. |
|
| 127:08 | just moved this mouse around first looking aim you then uh changing the incident |
|
| 127:16 | the white and what it really It's an interactive film. It's pretty |
|
| 127:22 | , you know, it's a whole of fun. Uh Definitely picking the |
|
| 127:26 | is not a lot of fun. part, you know, you, |
|
| 127:29 | spend two weeks and then you have five minutes of pleasure, but that's |
|
| 127:38 | . All right. So conventional interpretation . Um One of the first papers |
|
| 127:42 | 3D interpretation from uh Reich and Gilford uh Shell identify the horizon of |
|
| 127:50 | OK. Usually your boss or your of team will say because of seismic |
|
| 127:56 | photography, we wanna look at a stand and want you to pick that |
|
| 128:00 | or we are in uh the Johnno's . We want you to pick the |
|
| 128:05 | shale formation there or maybe the one filled with channels, et cetera. |
|
| 128:10 | So there's typically gonna be some focus what you're interpreting. Then you're gonna |
|
| 128:17 | a horizon on a grid of every 10 or 20 lines in line |
|
| 128:21 | cross line. There's a method to . You're gonna start with the bad |
|
| 128:28 | and then you're gonna quickly realize, I better use a more systematic method |
|
| 128:36 | ties the lines as I pick. I don't have to re pick them |
|
| 128:41 | do you don't wanna do all the lines first? Because then the cross |
|
| 128:48 | , they're not gonna tie. So have to kind of draw squares and |
|
| 128:54 | to make sure those squares tie and up confidence if you got a salt |
|
| 128:59 | or in, in our case, got a die up here or not |
|
| 129:02 | volcano in the middle. See if can kind of go around the darn |
|
| 129:06 | , don't try to go across it there's nothing to pick in the |
|
| 129:10 | OK. So this is the kind thing you learn by doing. And |
|
| 129:15 | and I notice even here, some you guys have quite a bit of |
|
| 129:19 | . You're helping the others you learn . You learn from your friends, |
|
| 129:23 | office mates, et cetera. I , that's how you learn. |
|
| 129:29 | Then you're gonna, you're gonna have grid and then you're gonna fill in |
|
| 129:33 | blanks between, let's say every 20th and you're going to use an auto |
|
| 129:39 | for that. Thanks extract horizon So time structure, amplitude extractions, |
|
| 129:48 | magnitude, dip Asmus combined, dip dip Asmus interactive sun shading like I |
|
| 129:54 | showed you horizon based curvature, those of uh horizon based. I would |
|
| 130:04 | one other term on here and maybe , I'll stick it in now. |
|
| 130:10 | the next time I do a class this, I have it. |
|
| 130:17 | I would apply the data conditioning first you'll notice that lab five, which |
|
| 130:25 | a couple of you got, did get to lab five? Maybe? |
|
| 130:31 | . Maybe during a week, whatever five is simple data conditioning. Not |
|
| 130:37 | to do. It's just pushing you're gonna find out that it, |
|
| 130:42 | helps the data. It includes, increases the vertical resolution and it definitely |
|
| 130:47 | the continuity. And by increasing the , it's going to make the horizons |
|
| 130:54 | to pick. And by increasing the of terminations at faults is gonna make |
|
| 131:02 | faults easier to pick. So I say data conditioning is, is |
|
| 131:11 | So maybe let me, let me put that in here. So the |
|
| 131:15 | time I remember and D and Ok. Ok. I'll make that |
|
| 132:01 | different color because that's not from, him. Ok. Fact, |
|
| 132:23 | Oh, somebody from Texas is calling 832. I moved from Houston 18 |
|
| 132:32 | ago. This is my daughter's old number. Have to use the |
|
| 132:38 | anything from Texas that I get. not a name that I know it's |
|
| 132:45 | trying to sell me something uh for car that I bought in Queer Lake |
|
| 132:52 | years ago. My, my, know that car that I sold 15 |
|
| 132:56 | ago, its warranty has run I get that over 65. Uh |
|
| 133:04 | of the calls are about Medicare and , stuff like that. And somehow |
|
| 133:09 | people know about my compromised romantic capabilities they're trying to sell me stuff for |
|
| 133:17 | too. Anyhow. Having the phone sounds like, you know, this |
|
| 133:26 | gonna be Partridge is good. All . So here's an example from central |
|
| 133:31 | platform of West Texas time structure Uh simple piece wise uh kind of |
|
| 133:39 | red down to blue. So red shallow and blue is deep. There's |
|
| 133:46 | dude plotted against the gray scale. got a reverse fault here. Reverse |
|
| 133:54 | there. This happens to be a slip fault that happened to be a |
|
| 133:57 | swp fault happened to be cars over picking it was hard to pick |
|
| 134:02 | So you can see Mard bad picks uh 1520 years ago when I did |
|
| 134:07 | . OK. In this room. right then. Oh, come on |
|
| 134:15 | . OK. That's my better But she doesn't know uh I got |
|
| 134:23 | , put it on. Oh Come mute. OK. That was |
|
| 134:40 | one of my Stephanie. OK. as mute. OK. Going around |
|
| 134:45 | color wheel. That's what, that's I think of it. This is |
|
| 134:50 | it looks like in the software. is done using Geof frame and then |
|
| 134:55 | OK. So now uh a paper Bruce got it. Uh things |
|
| 135:05 | let's say a carbonate. I'm gonna it. The fractures are gonna be |
|
| 135:11 | intense where or in this place where flexure is greatest. So here's his |
|
| 135:22 | form. Here's his line a a and uh more modern we would say |
|
| 135:27 | at the time he just used it . So here's a survey from uh |
|
| 135:34 | Colorado and he picked three horizons. ? And now you see this like |
|
| 135:44 | pink guy here. OK. I a fault here. Got a fault |
|
| 135:51 | and then up here it looks So the area is on that uh |
|
| 136:01 | hand picture and even here as well I have discrete discontinuities. Is that |
|
| 136:09 | seismic noise problem or is it geology ? I haven't picked on you |
|
| 136:20 | Do you think it's seismic noise or ? I mean, here I have |
|
| 136:23 | resolution. More glad, I don't how to undo that. I have |
|
| 136:31 | lateral resolution deeper in the data than shallow. What? That doesn't make |
|
| 136:38 | sense? Ok. That doesn't make sense. So, what, what's |
|
| 136:43 | on here party? You think it's ? Good. So, you're a |
|
| 136:53 | , right? Can't have one Give me three hypotheses as to why |
|
| 136:59 | is every geologist has to have three for anything. None of which they |
|
| 137:05 | all conflict. That's fine. So me one that you might think is |
|
| 137:17 | . Have you done any structural geology ? No? OK. Well, |
|
| 137:22 | of them I've got, this is , maybe not clear. But to |
|
| 137:27 | it's clear this is a reverse So, oh, maybe I got |
|
| 137:31 | fault then fold structure. So I'm down, I'm falling down deep and |
|
| 137:43 | it's folding and then it's just simply . Ok. That, that's real |
|
| 137:52 | for tectonics, like in the rocky . Certainly in Colombia you're gonna have |
|
| 137:56 | ben folds everywhere um down deeper than have more brittle rocks. Maybe very |
|
| 138:07 | rich brittle rocks and a shower. have more shale in the system. |
|
| 138:15 | the, the brittle rocks are breaking discreet faults and the shells are just |
|
| 138:24 | smearing. They're more ductile. More plastic or I'd have a single |
|
| 138:33 | down deep and then a sway fault above maybe five or 10 little falls |
|
| 138:41 | the shallower part of the data like the yellow and the red horizon |
|
| 138:45 | And seismically, I can't resolve that m offset on each fault, but |
|
| 138:51 | can certainly resolve the 25 m offset the one fault. OK. So |
|
| 138:58 | , so its geological, which one the opposite is true for seismic, |
|
| 139:03 | tend to have more resolution up shallow less resolution down deep. And the |
|
| 139:10 | is our resolution is coming from the that we can look at the |
|
| 139:16 | So if I've got just think of common spread, I've got more angles |
|
| 139:23 | uh up shallow and near parallel angles deep. Now, we have, |
|
| 139:31 | fix that a little bit with seismic , but the migration aperture up shallow |
|
| 139:37 | gonna be 100% and the migration aperture deep, we may not have data |
|
| 139:42 | enough out to get a good So anyhow, so then what he |
|
| 139:50 | is done 20 years ago now, it's still a good paper and he's |
|
| 139:55 | dip magnitude to production. So he's high production fier production. Oh It's |
|
| 140:05 | this flexure and then along this flexure here were the best one. |
|
| 140:12 | the flexor intersect, how, why that be? OK. And then |
|
| 140:21 | I just moderate production. Not nothing write home about. OK. Um |
|
| 140:31 | an outcrop from uh Lao de And we got a nice fold. |
|
| 140:42 | . OK. I curvature as the of the radius for a circle that |
|
| 140:52 | not only tangent to the surface but . OK. So I can have |
|
| 140:59 | ones that are tangent, but this fits it. And then I'm gonna |
|
| 141:04 | a dipping plane. A sin co decline a flat point and they purpose |
|
| 141:16 | I'm plotting. OK. So here going to say it for a planar |
|
| 141:21 | . Even if it's dipping, the is zero for a sin coin, |
|
| 141:25 | gonna negative for an anti cliental feature is positive and for a flat plane |
|
| 141:32 | is zero. Here's the equation. , that looks hard. Let's think |
|
| 141:39 | it this way, I defined it one old radius of the circle that |
|
| 141:44 | fits it. Now think of a I go around and I've got an |
|
| 141:52 | of two pip and I go around I have an art plank or circumference |
|
| 141:59 | two pi R two P over two R one over R. Let's take |
|
| 142:06 | little arc measured in radiance and an W measured in meters, take the |
|
| 142:13 | over the arc, W that's OK. So D the D uh |
|
| 142:22 | now I got DS. Well, , here's my surface and they're going |
|
| 142:27 | and down think topography. OK. , I don't wanna know the curvature |
|
| 142:33 | the surface. I wanna know the in the X direction or the Y |
|
| 142:37 | . I wanna put it on a time. So the, I'm |
|
| 142:42 | first of all have to change the axis or variable to A X to |
|
| 142:49 | yeah, S to X. And am I gonna do that? I'm |
|
| 142:53 | use the chain roll. So I to have a DS DX in |
|
| 142:57 | OK. So I'm gonna go from something DS D theta DS, I'm |
|
| 143:02 | put a DS DX spot. Not to do then. Oh I gotta |
|
| 143:09 | D theta DXD. How do I that? So I'm gonna go ask |
|
| 143:14 | who's much younger than me who did not too long ago, which is |
|
| 143:24 | . OK. You remember the OK. So the dip data, |
|
| 143:33 | the arc tangent of the height over distance, right, opposite over |
|
| 143:43 | right? All right. OK. that's simple. Now, I know |
|
| 143:47 | that is. What is the derivative of the arc tangent? OK. |
|
| 144:00 | remembers that? Nobody remembers that. why we got Google. That's why |
|
| 144:06 | have all textbooks. So you look up and guess what? You, |
|
| 144:10 | all studied this in school in the . We're gonna have a one plus |
|
| 144:16 | , that's the slope square to the half power. OK. That's, |
|
| 144:23 | what the derivative of the arc canon . And then, well, we've |
|
| 144:28 | another power of one of that term the denominator from the chain rule. |
|
| 144:33 | then we're gonna get D squared, squared and the numerator. So what |
|
| 144:38 | wanna get across here is curvature is the second derivative of elevation. It's |
|
| 144:46 | this correction term. OK. So I have a curve, sure, |
|
| 144:53 | curvature of this guy is the same of this as of this, of |
|
| 144:58 | , it's rotationally in bearing it. . That's the dip of the |
|
| 145:08 | Well, here's a, I got couple of little cartoons in here and |
|
| 145:13 | one picture is uh in red is topography and in blue is the |
|
| 145:21 | OK? So hopefully this works. . So I'm gonna put a, |
|
| 145:25 | circle tangent to it but not tangent it. OK. So positive. |
|
| 145:35 | . And here it's gonna be the circle and now the circle is gonna |
|
| 145:39 | a little bigger. All right. now it's gonna change orientation. So |
|
| 145:52 | gonna go from anti clal to syn . OK. So it becomes negative |
|
| 145:58 | then if we go along, it'll OK, it's gonna get bigger. |
|
| 146:01 | a big circle, big circle, small curvature. And then clearly the |
|
| 146:09 | who made this movie will, he an interactive thing maybe in MATLAB and |
|
| 146:13 | skin is not very, very steady he does. OK. And then |
|
| 146:18 | positive again. All right. And . So football season, I got |
|
| 146:25 | kinds of notes about not being able park on campus or whenever it |
|
| 146:31 | Uh So we're gonna have you got football, I'm gonna have curvature along |
|
| 146:38 | , thanks, the tightest part of football and I'm gonna have one perpendicular |
|
| 146:45 | , along the long side of the . So we were in, in |
|
| 146:48 | dimensions, we're gonna have two not like that one outcrop picture I |
|
| 146:53 | you and then we're gonna have an of the maximum and minimum curvature. |
|
| 147:00 | . So here's AAA cartoon from Hama he drew a, an image like |
|
| 147:08 | here is, and I don't want to spend much time on this |
|
| 147:13 | Here's my positive curvature and I here's something called, this is the |
|
| 147:17 | derivative and it's called K positive. , that's the um the crest of |
|
| 147:29 | topographic feature. But the hinge line is gonna bring into uh fractures is |
|
| 147:36 | be the most positive principle curvature. it's gonna be called K one and |
|
| 147:42 | hinge line of the most s inclined , the tightest negative curvature is going |
|
| 147:49 | be here. But the one that associated with the lowest part of the |
|
| 147:56 | actually has less curvature, it's less , it's less strained. OK. |
|
| 148:02 | we got a bunch of curvatures out . Um We have most positive principle |
|
| 148:09 | most negative. And this is what gonna use almost exclusively in this |
|
| 148:15 | Um There are others out there, mean minimum curvature, maximum curvature. |
|
| 148:22 | might show you curved ness which is sum of the squares of the two |
|
| 148:27 | . And yep, that's it. . So this is most bi most |
|
| 148:38 | one is biased towards and forms, other is biased towards scent forms. |
|
| 148:43 | is useful in machine learning classification. just wanna know what part of the |
|
| 148:57 | . Yeah, I mean mathematically good . But other than that these guys |
|
| 149:05 | , are less useful and cause too confusion because half of the literature is |
|
| 149:11 | with positive curvature. The maximum curvature you have a pencil. Good, |
|
| 149:32 | . OK. Just hold it, it. Here's my curved surface on |
|
| 149:39 | horizontal plane which direction shows the most of the structure you're showing up and |
|
| 149:50 | . No, but I mean in plane. So I want you to |
|
| 149:52 | like this way or that way. area W OK. She's saying this |
|
| 149:58 | that mathematically is the first eigenvector, eigenvectors and eigenvalues like when you're |
|
| 150:10 | Why? How am I ever gonna this? I never gonna use eigenvectors |
|
| 150:13 | eigenvalues. Why do I need to that? That was me. |
|
| 150:17 | That was me use eigenvectors and eigenvalues the time. But anyhow, that's |
|
| 150:23 | first eigenvector. What do you need know about an eigenvector is what it's |
|
| 150:26 | to measure, it's trying to measure direction best represents. In this |
|
| 150:35 | the most change in depth and it's way. So the maximum eigenvector is |
|
| 150:43 | have a negative side. OK? direction shows the least change in de |
|
| 150:49 | be parallel fact though, having a change, that's going to be the |
|
| 150:56 | curvature. Now, I got an , maximum curvature, minimum curvature. |
|
| 151:02 | ? So it's mathematically defined. Now say, oh, I'm a |
|
| 151:06 | Well, maximum's gotta be bigger than , right? So let's forget about |
|
| 151:13 | minimum curvature because it's hopelessly messed up half of the software. One |
|
| 151:19 | half of the software, the other , half of the publications, one |
|
| 151:22 | half of the publications, the other within the publication by Andy Roberts. |
|
| 151:26 | showed you a picture yesterday. I it was the first paper on horizon |
|
| 151:30 | curvature. Great paper. His figures curvature like eigenvectors and his text and |
|
| 151:39 | show curvature like a geophysicist would do . So in the same paper, |
|
| 151:45 | inconsistent. OK. Let's not do . OK? Just think, oh |
|
| 151:52 | a positive, you know, OK. Now, so here's the |
|
| 151:58 | of confusion, but I just captured and I'm gonna use a K one |
|
| 152:06 | the most positive principal curvature. K , the most negative principle curvature uh |
|
| 152:13 | . Why? Because they have eigenvalues eigenvectors or principal components if you |
|
| 152:18 | So K one is always gonna be than K two. Mhm. |
|
| 152:23 | But here he stop. Risk and ball stopped and you got the |
|
| 152:47 | We've got a school of differential Yeah. OK. I gotta update |
|
| 152:57 | do too. OK? Curvature picked outcrop in Wyoming in the US rocky |
|
| 153:08 | . So this fellow we, one the famous theologist of 19 nineties. |
|
| 153:13 | went out there, he liked his and then he got a yardstick or |
|
| 153:19 | meter stick and he measured how many per meter. OK. And he |
|
| 153:25 | fractures per meter to curvature. So Gaussian curvature, so Gaussian curvature, |
|
| 153:33 | lived uh in the 18 sixties, the other traits. OK. So |
|
| 153:40 | lived in the era in the 18 and he uh a lot of |
|
| 153:46 | one of his contributions to what's geology metallurgy today is the theory of thin |
|
| 153:53 | . So he said, yeah, I got a thin plate like |
|
| 153:57 | yeah, I'm gonna have a little on the bottom of the arch. |
|
| 154:01 | have a little extension on the But yeah, the average strange about |
|
| 154:09 | same on the top for a thin , for thick point is different for |
|
| 154:12 | point. So I'm not gonna have lot of open fractures instead, let's |
|
| 154:19 | that as a thin plate. Now I've got curvature in both |
|
| 154:25 | Ah So he's gonna multiply K one K two. That's what he |
|
| 154:30 | And uh you know, and that's there's cool, there's nothing wrong |
|
| 154:34 | that. And so here are the , OK. In terms of uh |
|
| 154:43 | practice for neuter. Uh So hi of fractures on this side, here's |
|
| 154:49 | curvature, high curvature on that So good visual correlation. OK. |
|
| 154:56 | one with the P. So this like doing this. But OK, |
|
| 155:04 | comfortable with the idea of a parent , right? So the true dip |
|
| 155:09 | this direction and the parent dip uh in that direction, we have the |
|
| 155:13 | thing with Kish, the true We got two principal curvatures this one |
|
| 155:20 | this, but we could have an curvature at that angle uh named after |
|
| 155:26 | . OK. So it's called Oiler . And in Petrella might be called |
|
| 155:32 | curvature. I don't know. But , and uh so the, it |
|
| 155:43 | thought, sure, here's my curved , here's my playing. I've got |
|
| 155:52 | strike and a dip. What is curvature in the direction of the striped |
|
| 155:57 | and in the dip direction? So idea is, and it's not that |
|
| 156:02 | of an idea early and you're gonna it, is that the rocky mountains |
|
| 156:11 | under compression. So any of the that are parallel to the rocky |
|
| 156:24 | those fractures, they're gonna be a of fractures, they're gonna be |
|
| 156:29 | they're, they're gonna be closed. ? Because the rocky mountains are under |
|
| 156:34 | in contrast parallel to that, they're be relatively tensile. So, ah |
|
| 156:40 | fractures I'm interested in are gonna be to the uh the transfer faults that |
|
| 156:50 | the folks in front of the rocky . But, and here he's working |
|
| 156:55 | the, in the rocky mountains and a structural geologist so he doesn't know |
|
| 156:58 | he's doing and he's got eight wells they were, they were locations and |
|
| 157:09 | his curvature anomaly from the surface and well, one and well, two |
|
| 157:15 | they are excellent producers at this horizon they communicate to each other. So |
|
| 157:22 | the simple explanation? Oh, I a lot of natural fractures in that |
|
| 157:27 | . OK. That's what he There's another one, this is down |
|
| 157:31 | the Maga Jess Maga Ji's basin there Southern Argentina and another horizon based |
|
| 157:44 | So here's a P horizon. I a little grob in here, a |
|
| 157:49 | grab in here. Structure map are , blue and purple are beat. |
|
| 157:59 | here on this time structure map, don't know what the real row |
|
| 158:05 | a good magnitudes. Oh, that's . And then here is his maximum |
|
| 158:16 | . Now he's using curvature the way just said it should be displayed. |
|
| 158:20 | like I said, half of the are one way, half of them |
|
| 158:23 | the other. So where it's that means it's Syn Kinal and where |
|
| 158:30 | positive, it's anti clal. So , we've got a feature that's kinda |
|
| 158:42 | kinda like that. OK. That of thing. So anti clal on |
|
| 158:48 | outer edges. Sin clal on the edges. And here's what it looks |
|
| 158:54 | in the vertical seismic. OK. you can map stratigraphy as well. |
|
| 159:01 | this is the Soto Canyon in the of Mexico. Here's an amplitude |
|
| 159:05 | I see this fine uh a time map. It's OK. I see |
|
| 159:12 | levy levy because Park Park O La Lede, Pasco Leve, right? |
|
| 159:23 | it's raised, raised, raised up Le Matin. I get up in |
|
| 159:28 | morning, I raise myself, my brush themselves, right? That's French |
|
| 159:35 | . Uh OK. Um We can dip as mute dip magnitude. |
|
| 159:44 | here I see the edges of the on the magnitude curvature. Wow. |
|
| 159:52 | see all these reworked edges of the due to the uh currents at the |
|
| 159:58 | and then has one. He calls . He doesn't apply. OK. |
|
| 160:04 | here, see the edges, they these notches notch. So here in |
|
| 160:14 | , you're not gonna see many notches , but uh go someplace where you |
|
| 160:20 | a road cut. So drive to or drive to New Mexico and you |
|
| 160:26 | through a road cut and here's the road cut and it's level on the |
|
| 160:31 | , but there was a heavy rainstorm there's this like triangular notch and all |
|
| 160:37 | dirt comes down and the walk. . So we have the backwards version |
|
| 160:42 | that submarine things are going in the , not the canyon, but down |
|
| 160:47 | channel in the Turbid flood stage. have a high flow, it cuts |
|
| 160:54 | and comes out. Where are the is gonna be at the cut bank |
|
| 161:01 | ? Ok. So uh Henry Postman , this guy, he's like the |
|
| 161:08 | of seismic geomorphology. So I did , not. Yeah. Yeah. |
|
| 161:27 | . OK. What the thing was you moved horizons along that fault during |
|
| 161:35 | time. So we've got the different . We know the top of the |
|
| 161:40 | division was a one time flat. know the top of the Cambrian was |
|
| 161:46 | time flat. The top of the was one time flat. So we're |
|
| 161:50 | have different thickness on different sides of fault of a normal fault if you |
|
| 161:58 | in a deep voice, Synon OK. That means you have deposition |
|
| 162:06 | the fault is moving. OK? then you want to put these |
|
| 162:10 | But now darn it and my fault like this no big deal, but |
|
| 162:16 | fault is like that it's corrugated and parts are quartz rich and more |
|
| 162:24 | sandy sandstone and other more ductal So when things are rubbing against |
|
| 162:32 | it's gonna cause fractures. So what did is she did a fine element |
|
| 162:37 | and measured strain and where are the that are going to be under most |
|
| 162:44 | during the geologic past? And then the well controlled, she correlated them |
|
| 162:49 | high fractures cool. Um so called and geom Mehan properties here, I've |
|
| 162:59 | three geom Mehan units. And this , he's got uh I guess these |
|
| 163:07 | shales and then he's got sand. he's gonna call it a competent |
|
| 163:11 | All sand. This one is almost all shale and this one is |
|
| 163:17 | sand and shale. He's gonna call a mixed mechanical unit. And you |
|
| 163:22 | see here, I have doctoral deformation . I have quite simple. |
|
| 163:32 | because a competent unit and in this it's something in between and noticed different |
|
| 163:39 | changes. So everybody here, who's geologist? You probably did, did |
|
| 163:47 | build sandcastles? Like you ever try build a sandcastle underwater? Doesn't work |
|
| 163:54 | well. OK. So you've got internal friction coefficients. You might have |
|
| 164:03 | studied about more circle. OK? uh coefficients of friction and so |
|
| 164:12 | So different rocks are gonna have, gonna more circle will say that depending |
|
| 164:17 | , you know, like cool coefficient that a rock is gonna break |
|
| 164:23 | 60 degrees or 80 degrees or 45 . You can figure out what that |
|
| 164:29 | of weakness is. All right. you can go back to your uh |
|
| 164:34 | textbooks, but that means the the shape of the fault or |
|
| 164:42 | or the, the dip of the will change depending on what I now |
|
| 164:48 | gonna move things along this stuff. , I can't, if I'm gonna |
|
| 164:53 | along there, I'm gonna have aosis, gap, aosis, you're |
|
| 164:57 | have holes in there. How am gonna accommodate those gaps fracture? |
|
| 165:03 | So this is what your geology your structural geology teammates bring to the |
|
| 165:08 | and put together and what, what do as interpreters. Well, we |
|
| 165:14 | that darn fall and sometimes darn I wanna make that fall straight because |
|
| 165:21 | my preconception of a nice clean fall sometimes it's not that way, |
|
| 165:30 | So here uh same guy and here got the different falls, high |
|
| 165:40 | the wave green, then just wait how you OK. Uh The magazine |
|
| 165:53 | still in bye that we talked I don't, here's his folk. |
|
| 166:00 | is the word he's using. We a couple. Her name clearly wasn't |
|
| 166:25 | . It was Yo Deli. Olo Yomi. Anyhow. Good, |
|
| 166:33 | Nigerian name. Yeah. And she studied here. Uh So she |
|
| 166:39 | the top and the base of this based of this formation. Here, |
|
| 166:45 | the top, here's the bottom and can see uh it's a little higher |
|
| 166:50 | . So it may be a little . There's an in size channel here |
|
| 166:54 | size channel there, inside channel that go give me uh a second snap |
|
| 167:02 | thick. The middle one, this little less thick and this one the |
|
| 167:09 | it was picked, we really don't this little incise there. That could |
|
| 167:12 | over here someplace. Now, uh my, my line OK? Everything |
|
| 167:20 | up. Now, what? So , my attribute is bottom line, |
|
| 167:27 | top, let's do R MS So the, but some of the |
|
| 167:34 | of the amplitude in that zone divided the thickness and here it what? |
|
| 167:41 | . I've got low R MS Ir MS on the sides, fairly |
|
| 167:48 | R MS here. And this one see in R MS. So notice |
|
| 167:55 | thick part of the channel has low MS amplitude and the sides have high |
|
| 167:59 | MS amplitude. So why is, is that Newton? Why is |
|
| 168:15 | Let's see if Jessica Jessica, you there? Yes. Yeah. So |
|
| 168:22 | , why do I have high arms on the flanks of the channel and |
|
| 168:27 | in the middle? Mm uh I know. OK. Anybody want to |
|
| 168:37 | a guess. Well, we're gonna a little bit of a levy on |
|
| 168:43 | side. So we'll have a little filled levy. It is filled, |
|
| 168:50 | charged with gas and the size channel was just filled with shale. |
|
| 168:57 | So one is mapping the levy and other is mapping the channel axis which |
|
| 169:02 | us to the next slide. So here she is mapping other sands |
|
| 169:08 | the Gulf of Mexico where we have salt tectonics. OK. So we've |
|
| 169:14 | salt tectonics. Offshore Brazil, North Sea. Lots of places, |
|
| 169:20 | ones and other places like in, Libya. And that, so what |
|
| 169:25 | is the salt doesn't really push its up. It's the basin is building |
|
| 169:33 | So the Mississippi River is dumping dumping sediments, dumping sediments. So |
|
| 169:37 | pushing down, pushing down, pushing , the salt layer is being squeezed |
|
| 169:44 | comes up where it can't. So the basin is building down and |
|
| 169:50 | as the salt comes up, the the the mini basins fill. |
|
| 169:55 | we're forever having the mini basins building and the salt coming up, changing |
|
| 170:02 | topography. So what happens? I these little bitty mini basins and then |
|
| 170:08 | spill into the next knee base and fill that and I spill. |
|
| 170:14 | That would be great if it stayed but nothing staying still. Everything is |
|
| 170:17 | around. So what did the geologist this bill and spill? Right? |
|
| 170:27 | cool, doesn't it? So, and Spill. So here are three |
|
| 170:32 | sections through the data kind of ugly but still. So here is just |
|
| 170:41 | , here's a thickness map. So thicker here and then thinner. And |
|
| 170:46 | here is her interpretation of um the where she's got distributer the main |
|
| 170:54 | et cetera. And then here is data on the right thickness and here |
|
| 171:06 | a uh research uh fish tank filled colored sand. OK. So Strat |
|
| 171:13 | like to do these things. the fish tanks could be oh 10 |
|
| 171:20 | by 10 m, might be a 5 m by 5 m. I |
|
| 171:23 | they're big. OK? Then maybe high and you put sand and different |
|
| 171:28 | sand and you add it and see does it go. How does it |
|
| 171:31 | , et cetera? I'll show you pictures uh next week on that. |
|
| 171:35 | notice the morphology very, very OK. Another one. And I |
|
| 171:47 | we're gonna stop there and see if can do a quick practice quiz, |
|
| 171:55 | ? You wanna show him how to it? Not any history? |
|
| 172:38 | Oh. Mhm. Yeah. So uh repeat the answer when my mic |
|
| 180:53 | off, this is the, this test counts for absolutely nothing other |
|
| 180:58 | for you to see the style of questions I ask. OK. |
|
| 181:05 | you know, typically you're gonna have or four pictures to choose from. |
|
| 181:12 | found that when I used to do tests by and I would pick like |
|
| 181:26 | of the questions with the synthetic, would pick bad answers that people had |
|
| 181:32 | drawn. The most, most common answers is drawn from here. And |
|
| 181:37 | would say the reliability is almost exactly same as drawing by hand, just |
|
| 181:43 | lot easier to grade. And for of you guys who like quick response |
|
| 181:47 | find out how well or poorly you immediately, pardon? You said? |
|
| 181:58 | , so I didn't have to. you don't, you don't have a |
|
| 182:03 | out. It's better that it's Oh, yeah. Yeah, you |
|
| 182:10 | . That's right. Yeah. because sometimes when you're making, |
|
| 182:13 | uh, the 80th copy the ink and some people get a bad |
|
| 182:20 | especially for color. If you got big class, you get good |
|
| 182:24 | good quality. Yeah. Pardon? , officially it's over. If we |
|
| 182:41 | , I'll go to 555 we'll I'll go quickly over the answer. |
|
| 182:47 | , you know why? And, , makes sense. Oh, except |
|
| 182:53 | might have, hopefully I didn't turn computer off. I have, we |
|
| 183:04 | quick. In fact, I couldn't out how to log off. How |
|
| 183:14 | you log off? Yes, for plan. Yeah. Yeah. How |
|
| 183:24 | you do that? There's no power or anything you need to, you |
|
| 183:33 | name? Oh, I gotta click name. Oh, and then it |
|
| 183:37 | , logs me off. Oh, see. Ok, cool. I'm |
|
| 183:40 | right with that. I don't know , OK. Ready to go over |
|
| 185:33 | . You don't have to finish But, so we'll go over |
|
| 185:35 | Let's do the first one. And got uh I've got four images that |
|
| 185:41 | instantaneous face and, well, and one is the best one? |
|
| 185:48 | So we can rule out, remember phase at 180 is the same as |
|
| 185:53 | phase at minus 180. So we to have a cyclical color bar and |
|
| 185:59 | is a rainbow color bar going from to red and this is the blue |
|
| 186:05 | , red, uh dual polarity color . So neither of these two are |
|
| 186:12 | in the upper right and the upper , the lower right. One is |
|
| 186:18 | cyclical color bar. So this is same color here is down there and |
|
| 186:22 | same color here is down there. one's OK, but here the interpolation |
|
| 186:28 | on the default. Hopefully you did in a lab exercise today and you |
|
| 186:32 | , it's got all these little artifacts in patrol. It shows up |
|
| 186:37 | the zero maps, the green in page color bar. It's got all |
|
| 186:43 | little green stripes here between the blue the red. I turn it |
|
| 186:49 | All that goes away. OK? you wanna use the one up here |
|
| 186:55 | in the upper left. The second you haven't started picking. So maybe |
|
| 187:01 | isn't quite that you need to know can figure this out. So if |
|
| 187:07 | ask it next week, you'll but we've got auto pickers. So |
|
| 187:12 | wanna like, oh I'm gonna pick , drop, drop, peak, |
|
| 187:15 | , peak, zero, crossing zero . If you are picking an angular |
|
| 187:20 | conformity, you're gonna have different reflection I'm sorry, you're gonna have different |
|
| 187:27 | coming up in an angle to let's a constant impeding of the shale |
|
| 187:33 | So sometimes you'll have a positive sometimes you'll have a negative reflection. |
|
| 187:39 | if I have a standing formation that to have the same impedance as a |
|
| 187:45 | formation, seismic doesn't care, you're have zero reflection. OK? So |
|
| 187:51 | can't pick peaks and troughs. And same thing is true. And this |
|
| 187:55 | why, um, Stephanie here hates photography because you can't use an auto |
|
| 188:01 | . You gotta pick everything by hand everything is on lamp and awful labs |
|
| 188:05 | you can't pick peaks and drops So, uh, you have to |
|
| 188:11 | turn auto picking and auto, turn picking off and then pick the terminations |
|
| 188:19 | hand. I mean, that's all can do. And it's pretty |
|
| 188:25 | Pardon? It would be eve it, it would be number E |
|
| 188:38 | in Edward, turn off the auto CND you can't do. So that |
|
| 188:45 | be a zero crossing. CND would a zero crossing. Uh Yeah, |
|
| 188:52 | doesn't, it doesn't matter. It's the reflectors aren't continuous. I'm |
|
| 188:58 | the reflection coefficient is not continuous. if you were to look carefully |
|
| 189:04 | you say, oh, here I've maybe I'll call that positive, then |
|
| 189:06 | changes, then the positive again, positive, negative, positive it |
|
| 189:13 | OK. You'll get comfortable with that you start picking. OK. Now |
|
| 189:20 | one takes a little bit of skill it, it doesn't require any understanding |
|
| 189:25 | geophysics and understands, requires more of . OK. So my geology guys |
|
| 189:33 | gonna be comfortable with that. All . So we're gonna pick, |
|
| 189:38 | this one kind of goes up. all right. And then it continues |
|
| 189:41 | go down. What about that What about that phone? What about |
|
| 189:45 | phone? Oh, I went up I went up. Ok. This |
|
| 189:49 | goes down, down the toilet. . This one goes down the |
|
| 189:56 | This is like my petroleum engineers would and this one says, oh, |
|
| 190:01 | got district faults and this block number explain everything on the computer. |
|
| 190:12 | Yeah. Yeah. Yeah. Sorry Jessica. You're quiet. |
|
| 190:18 | So, so this one kind of down the toilet. This one follows |
|
| 190:25 | blue troughs that goes down the This one tries to go up the |
|
| 190:31 | but doesn't do a perf a consistent the faults first. That would be |
|
| 190:40 | best way. Then you don't poke eye out. So you would pick |
|
| 190:44 | falls and you would say, aha are sliding them, the foot |
|
| 190:51 | I'm this is the foot wall and is the hanging wall. So the |
|
| 190:56 | wall is gonna slide down on top the, I'm sorry, the hanging |
|
| 191:02 | , the slide down on top of foot wall. So this layer is |
|
| 191:07 | to go down with respect to the on the other side and that's going |
|
| 191:14 | be consistent along. So this, block has slid down. This block |
|
| 191:21 | swayed down. This block has slid . This block, I slid down |
|
| 191:27 | both sides. So maybe this is the pick is and then you got |
|
| 191:32 | block here and this block so they directions a bit on you. But |
|
| 191:37 | I'm picking the same red, red pattern and trying to match the |
|
| 191:43 | across the fault as best as I by eye. OK. So what |
|
| 191:49 | need to do is when I, I was talking about reading the |
|
| 191:54 | you wanna honor the seismic. So , I'm always gonna pick a trough |
|
| 191:57 | this example consistently, but it needs be geologically reasonable. OK. So |
|
| 192:05 | got to integrate the two together. . This one comes from the book |
|
| 192:10 | Don Herron. Well, couple of asked, should I buy a book |
|
| 192:13 | you're gonna buy a book? This the book to buy. It's a |
|
| 192:16 | book to read, easy. Everything it is worth knowing. OK? |
|
| 192:20 | superfluous information. And he's got this of using an auto picker and what's |
|
| 192:27 | auto picker doing? It's connecting peaks troughs and it's cross correlating and he |
|
| 192:32 | right across there and maybe at this , it looks reasonable. But I |
|
| 192:37 | this guy up here that's clearly dropped here. It's dropped down. So |
|
| 192:45 | understanding of what's happening above where everything dropped down below everything is dropped |
|
| 192:52 | This guy should correlate to this negative in white. OK? And if |
|
| 193:01 | just use the picker and you don't it to stop at fault, how |
|
| 193:06 | you do that? Oh, you pick the fault in that line? |
|
| 193:10 | . But you can pick it and erase it and then start it |
|
| 193:13 | OK? Perfectly permissible if you uh it by itself is gonna pick right |
|
| 193:20 | there. And if and variant right at that level where the red pick |
|
| 193:25 | the variance would be zero, it not see a fault. OK. |
|
| 193:31 | . This is the more and you to think of reflection coefficients. The |
|
| 193:37 | thing, what's the polarity of the ? So I think on the first |
|
| 193:41 | I gave you the top reflection and have the uh impedances there. So |
|
| 193:47 | your head or on a piece of , you multiply 2.2 times 3.2 and |
|
| 193:53 | you multiply two by three. that's six, I can do that |
|
| 193:56 | my head. And uh so if going from high impinges to low |
|
| 194:01 | I have a negative reflection coefficient. means my source wavelength is zero |
|
| 194:08 | It's got a at that. Now, well, that's bigger than |
|
| 194:19 | times three. So I could have OK, that's good about to |
|
| 194:36 | I happened to the right one, can see correct one on the |
|
| 195:01 | No, that's not going from going high. It's one. I, |
|
| 195:18 | , I mean, because I there's some I should give zero reflection |
|
| 195:23 | thanks. And it's a kind of but it's not expensive. Definitely ask |
|
| 195:34 | a question just like this. Maybe was like, OK, this one |
|
| 195:40 | what we uh covered maybe this year uh direct hydrocarbon indicators. So here's |
|
| 195:48 | horizon. It's from the paper by , which I recommend you take a |
|
| 195:52 | at, I'll pick a picture from that for sure. On uh |
|
| 195:58 | Friday. So it's getting brighter as go structurally higher. When you are |
|
| 196:05 | structurally higher, they're getting much structurally or much higher. A much stronger |
|
| 196:12 | . So these are, they're just , they're simplest kind of bright |
|
| 196:17 | right. So that's kind of the I don't want anybody to be |
|
| 196:22 | like, oh, this guy's got or five pictures to look at your |
|
| 196:27 | , man, that really stressed them . They didn't like. And when |
|
| 196:31 | did it by hand, they I've never used colored pencils before. |
|
| 196:39 | , think engineering. Ok. um, you've got my email pretty |
|
| 196:45 | to find, um, and, , send me a note if you |
|
| 196:52 | a problem and uh, send, , you tie a note if you |
|
| 196:58 | a, a problem as well. , especially if you're in on the |
|
| 197:03 | and you get stuck at some You know, you come in, |
|
| 197:05 | commute, you get stuck on the . I know Stephanie is probably gonna |
|
| 197:10 | right after church tomorrow and tell her . Yeah, baby is still |
|
| 197:14 | You need to take care of take care of her. Um, |
|
| 197:20 | just, just send me an email , um, I'll see you next |
|
| 197:24 | . Ok. |
|
