0:05
Good morning, good afternoon, or good evening.
0:09
Wherever you are, welcome to the 2022 online: High Split Workshop. This is the first day.
0:18
Hello Aerial.
0:20
Ariel's Stein, the Director of the Air Resource Laboratory, is going to provide a welcome to us. Ariel, take it away.
0:31
Thank you Mark. And good morning, good afternoon, wherever you are in the planet.
0:37
Welcome to the High Speed workshop for 20 22.
0:43
This is really a pleasure for me too.
0:46
Welcome to this workshop that has been going on for, I think, more than 10 years.
0:54
We started with very we're a very humble beginning with only a few participants and in person.
1:04
And remember, back into 2007, before 2010, we only had, like 25 people coming into the building.
1:14
Back then, Ron ...
1:16
was teaching most of, of the workshop.
1:23
And then we evolved to a hybrid environment.
1:27
And today, we have this format that has allowed many, many people worldwide to access this information, and these educational platform that Noah provides, to all our users. So I'm very proud of how we have evolved throughout the years. We now have more than 300 participants.
1:53
And I'm really proud, also, of the high speed team that has taking over the, the work of teaching on this, this workshop.
2:08
I hope, that you enjoy these four days, that are coming up.
2:16
You have, you will have you in front of you the best of the best in terms of transport and dispersion modeling. These are word world experts. So I hope that you can take advantage of that, and also ask as many questions as you can.
2:37
So with that, I just wanted to say, thank you for participating in. this. Law is really proud of our products. On high speed is kind of like the crown jewel of moths.
3:00
So with that said, I hope that you have been very productive workshop.
3:06
Thank you.
3:08
Thank you very much area, and I wish you for helping us teach this workshop, but you do an excellent job, so thanks a lot for that. Welcome.
3:19
OK, well, what we're going to do is start out with a brief introduction here to get us going in the workshop, and I'm gonna go over these things.
3:30
I'll try to go fairly quickly, so we get into the meat of the workshop itself, but I wanted to go through some logistics and some of the ways that we'd like you to interact with us and some of the ways that you'll be able to get information.
3:43
Throughout that, the workshop, and I'll say all the slides I'm showing are available in a handout.
3:51
And in your goto Webinar control panel, you'll see a little tab called handouts and there's a PDF there with all the slides so you don't have to write anything down.
4:01
You can, you can download all these slides whenever you want.
4:06
Um, so, um, the first thing is that we're putting all of the information, um, that we want you to have on this workshop webpage. and It has the information that you probably saw before, the workshop.
4:25
In terms of what what to do in terms of installation of the of the, of the, of the model.
4:30
And also, as the workshop's proceeds will be putting in these, these various handouts, as well as links to the videos each day.
4:40
So we'll update this each day with, with that new information, and so this is a place that you can come, maybe keep open or go back to each day to see if there's something new that you need to get, or you might want to get.
4:53
This is the schedule for the week. This is actually the first real day we, yesterday we had a day of where we were doing the installation sessions. We had a couple of sort of lecture sessions.
5:05
Where are we going to the first two sections of the tutorial, and we had a we had several one-on-one installation sessions, and today is, like, I say, the first real day.
5:16
We'll be going through, um, four sections of the tutorial about meteorological data and beginning the trajectory calculations, and going into the trajectory, different options and statistics.
5:30
Um, the other next days, you'll see, we'll be going into the air concentrations and then some more in terms of special topics about source attribution and different kinds of emission sources like smoke, and volcanoes, and radioactive.
5:47
um, radionuclides.
5:51
Um.
5:54
So today, we'll, like I say, well, we having this brief introduction and we'll have some break start during the day as well.
6:06
Now, um, you are all using the GoToWebinar interface, goto webinar interface but only if you're live today and we recognize that some people will just be given the recordings. This next section doesn't really.
6:19
um, It's not really relevant if you're just figure recording, but for those of you who are present live, you can basically, you can be accessing this goto Webinar in two different ways. There's a desktop application I think it's called goto Launcher. And if you do that, then you have some kind of Interface that looks like this.
6:41
The, uh, this little arrow here, can either, or Hide, or pop openly, the, the interface or the control panel.
6:51
Or you can join through, ugh, oh, browser, and I think they may only support Google Chrome or Microsoft Edge.
6:59
And when you go into the browser, your interface might look something like this, and I think it depends on which browser you're using, and which version of the desktop you're using.
7:09
So there's your, uh, interface may look a little bit different than this.
7:16
Um, if you are using the browser interface, you'll see that you can toggle and hide the webcams.
7:28
If you want to not see any webcams, you can see more of the content full screen.
7:35
You can access the materials, like the handouts. From here, you can ask questions.
7:43
And then there's the settings thing, Settings, widget as well. You can take a screenshot, which is a nice thing if you want.
7:53
And, by the way, in terms of these hand, Racing's, you can do that, but we're not using those to ask questions. It'll just be like a reaction, or, you know, we may ask you to first show of hands for something at some point.
8:07
Um, when you use the browser interface, it seems like there may be some complications that when I tried to do what I had to do, a bit of a complicated thing, where I had to and get deep registration e-mail that I got from goto Webinar and add this thing to the, to the link, and then paste it into my browser, but it may not.
8:30
You may not get required to do that, and I guess if you're listening in, you were able to join the webinar one way or the other.
8:40
With the desktop application, I said you can click this arrow, to hide, or hide the control panel.
8:50
You can.
8:54
Um.
8:58
Look at the Control Panel in different ways.
9:01
If you toggle, you can toggle one of these little drop-down menus to open up a panel, like the handouts or the questions and the chat.
9:10
And, um, when you ask a question, we will get it, even though it will be trying to answer it.
9:19
But, I'm going to tell you in a second, there's another way we'd like to ask questions as well. I'll get into that in a minute.
9:28
You can undock any section.
9:31
So, if you want to make your question sort of docked from the, the main control panel. You can do that.
9:38
And if you click that button again, and it will bring it back and talk it back to the, for the control panel.
9:44
Um, Here's just an example, if you're asking questions, you know, you're typing your question and you're sending it, it'll show up, and then when we answer it will, we'll try to answer it.
9:55
When you ask, a question only comes to the organizers, it just does not go to the whole audience.
10:01
When we answer, you will generally answer you privately, unless the answer we give feels like it's a very general, and we'd like to give it to the whole audience. But you won't be identified as the questioner.
10:16
You, when you ask a question synonymous to the rest of the participants.
10:21
Like I say, well, sometimes ask for a show of hands, and when you raise your hand, we'll see that, but it's not a way for you to ask questions.
10:31
It's just a way for you to, to demonstrate something that you're responding.
10:36
Um, you can actually change the global that the, the control panel language for this little at this little globe up here, you can change it to here. I changed it to Portuguese.
10:48
Although the questions are still going to be in English, or at least our answers will be in English.
10:56
OK, so asking questions, if you're in this webinar, excuse me, if you're view your recording, or you're having trouble accessing something, and you can't even get into the webinar, You can send an e-mail to the REL Webmaster, um, or you can ask a question in this goto Webinar control panel and so what we'd really like, questions here would be questions about the webinar logistics, about the handouts, about something about the audio, about, you know, something sort of more general.
11:32
If you have a question about the High split model itself, or something that we're talking about it in the forum, excuse me, in the workshop. We'd really like you to ask it in the High split forum.
11:43
And I'll explain why in a minute.
11:47
Um, the main reason is that it's the the goto Webinar Question is very question box is very limited. You can't put screenshots in With a couple hundred people listening in.
12:00
We just get hundreds of questions coming in all the time here, and it's difficult to kind of keep keep up.
12:06
Um, if you want to just look at people's answers and questions, you can do that without registering in the bulletin board. But if you want to ask a question, you have to register and it's free, it just takes a second.
12:20
When you register, it does ask you to do this one thing to keep bots that you have to put a bunch of cities, either in the in the United States or other countries. And when you're doing that, just think of the most important version of the city. So, like, like, I think, Rome was one of the cities. And even though there is a Rome, in the United States, it's clearly not, You know, the most important Rome is in another country. And same with Paris.
12:44
I think a couple of it might seem ambiguous, but no, just think of the most important version of the country. 1 1, I'll warn you about a San Jose.
12:53
We're referring to the one in California and the Silicon Valley even though there's a larger San Jose and as the capital of Costa Rica, that's the only tricky one when you're answering these questions.
13:05
So when you, when you get into the forum and you've registered, actually you can see this without registering.
13:13
But you can go down to the Hospital Workshop area of the forearm, and go to the 2022 workshop area. And when you get there, you'll see a whole bunch of these different sections.
13:26
And when you ask a question, you can ask it in the appropriate section that we're dealing with up till now.
13:34
Um, people have been asking questions in the installation section.
13:38
Um, and yet, here's actually this last year. Some of the questions that came in.
13:43
and, you know, when you're in there, you can look at other these questions, and you can see the answers.
13:48
and so, it could be that the question you have already been asked and that's one of the advantages of, using, of using this forum. Also, when you ask a question, it's generally something that other people are dealing with, and so that'll help other people as well.
14:05
Couple of things.
14:06
You know, when you're, when you're asking a question, if you can try to avoid putting any links in your post, that's helpful, Some thinking about the system. Don't start aligning the two dashes in some kind of system.
14:20
Issue there.
14:22
And, But you can add attachments, and sometimes you want to put a screenshot or something in that can be really helpful.
14:29
When we're trying to diagnose what might be happening.
14:32
You can also add files as, well.
14:35
Control file or setup file, which you'll understand what those are in a little bit.
14:40
And I guess the reason why we're doing this is that, you know, you can add these screenshots, you can, we can go back and forth, and other people can see what the answers are, and then when people ask questions, we can refer it them to these answers in, in.
14:57
OK, a couple things about screens, it's a little tricky, because you know, you're looking at the webinar, which you're looking at right now and when we start doing things, you'll start seeing us doing, you know, things with high split and with the tutorial in the webinar.
15:13
But then it's a hands-on, uh, workshop, And so you're going to have your own version of tissue tutorial up on your screen, because it's going to be guiding you through the different things you're going to be doing.
15:28
And at the same time, you're going to be doing things with the model, and you'll start to see what I'm talking about shortly. But it's a little tricky to have, you know, all of these things up on the screen at the same time. I would say it's almost impossible if you're just doing using a laptop with one screen.
15:45
But, you know, if you have another screen, then it can be quite helpful maybe to have the webinar in one screen And then your own work on, on another screen decide how you want to do it, but, um, so the recordings each day will put them on the workshopped web page.
16:05
It usually takes about 4 to 8 hours after the session ends on a day for the web or the video to be processed.
16:13
But when it's processed, it will post the link on the workshop webpage.
16:17
So, actually, there's the link to the, um, the sessions from yesterday on the installation.
16:25
And the transcripts are produced. It takes awhile for those transcripts to be produced and they're not really perfect.
16:33
So, but we want post them for now, even though you can see, if you look at them, that's definitely not perfect in terms of the way the machine translate the names at the words, will try to go in later and fix that, but for now, we will put the transcripts up even if they're not not perfect.
16:54
Um, now the installation went through, we hope you've done this, because it's a hands-on workshop.
17:03
We did give, try to give instructions on the workshop webpage of what to do beforehand.
17:10
We hope most of you followed, you know, these various steps.
17:15
We think that it's been, hopefully, easier this year. Because of the problems that people have had in the past with installation.
17:21
We've tried to make it easier for people. We hope that that's the case, this, this year.
17:30
Um, the first thing that we hoped you did was download the tutorial, and when you do that, you'll get a ...
17:39
aye folder on your computer that looks something like this.
17:44
It's a bunch of bunch of subdirectories.
17:48
This is what it would look like on a Windows machine, like it's something similar on Mack or on Linux Machine.
17:55
And when you click on the index dot HTML, then up comes this, what we call the tutorial, and we're going to be going through this tutorial throughout the the workshop. And you might want to have a version of this open on your screen, because it has instructions and has the different screenshots and things that you'll be doing.
18:15
Um.
18:18
In the tutorial, there are the first sections about the installation, And so there's, you know, if you haven't installed things yet, you could go through these steps.
18:27
It kind of takes you through step by step, for either a Windows installation, a Mac installation, or a Linux installation. So those are these first section, is that we went through.
18:40
There's some of this, yesterday.
18:44
Very brief, brief overview of the installation. In order to look at the graphical user interface, which is this interface that we'll be using.
18:55
Like a point and click type of interface that we're using, you have to have this command interpreter, TCL TK language, and that has to be installed on your computer.
19:04
Or you won't be able to look at the graphical user interface and for Windows you either have to install it from the TCL TK website or we have a pre compiled executable.
19:15
You can download for Mac and Linux systems. It's generally already installed, so you didn't have to do anything in most cases, in that, in that case.
19:24
We're also, um, recommending that you install image magic and even more optional just Python.
19:31
But I'll say that if you weren't able to do the image magic for one reason or another, um, you can do almost the entire workshop without image magic it just you won't be able to do an animation and you won't be able to convert the graphics to other formats, but you can do almost everything in the workshop.
19:47
So if you weren't able to do the image magic, don't worry about about that. You do need to have TCL TK.
19:54
And, again, only the Windows people needed to figure out one way or another to get that on their machine. And then you one, of course, have high split.
20:04
There's two ways to install it. You can download that self extracting executable.
20:08
But if you didn't have privileges to run that, you can actually just download a zip file and unzip it and, and put those folders high split folders, like, in your users directory or anywhere you want, and run the model from from there.
20:25
For markets, fairly similar, and for Linux, we have pre compiled executables, or you could pilot locally.
20:33
Um, um, we hope that most of you have been able to install the model or will be able to install the model.
20:41
Although, if you haven't, you can actually, you know, still participate in the workshop and, and view this as more of a demonstration.
20:51
What does it look like if you've been able to install things well. You would have some kind of icon on your desktop.
20:58
When you click on it, You'll come this initial screen of the graphical user interface, and we call it Graphical user Interface GUI. That's the G UI, the GUI.
21:12
And when you click on the menu button up, comes, you know this basic thing, and you'll start seeing this throughout this workshop. So if you're able to get this, then it means that you've got that TCL TK.
21:25
Um.
21:28
Program installed, or it was already installed on your computer, and also if you've got haislip because for the most part, because this wouldn't even come up unless you had the high split model installed as well.
21:43
Um, know, a few other things. You know, you'll be clicking will be clicking on these different menu items and then other, other types of screens will show up.
21:52
And like I say, if you've seen these kinds of things and then things are probably installed properly, Um, um, one issue regarding the for Windows users, if you had to install the TCL TK and it perhaps it came.
22:12
Perhaps you downloaded the disk, the pre compiled executable and just put it somewhere.
22:19
Then you have to basically, when you make that shortcut on your desktop, you have to find where that TCL TK.
22:28
It's the actual executable is called wish 86 T dot E X C, and then you'd have to, you know, tell it where to find the high split, Um, TCL script, that runs the graphical user Interface.
22:43
So, the target in your shortcut, if you're a Windows user, if, if you did this download the executable, without installing it in the normal way, it would be a little bit more complicated but it could certainly be be done.
22:59
Here's an example, I just put that executable like in my in my user's Drive.
23:05
Um, and, uh, no, it didn't have to have administrative privileges to do that, and, but it will still work.
23:15
We're going to be working into tutorial this this entire time And a couple of things like I mentioned. You click on this index dot HTML and then this comes up.
23:23
Um.
23:25
You can navigate around in the, in the tutorial. We're going through the first section today will be about gridded meteorological data files.
23:33
In fact, these are the sections that we'll be doing today.
23:37
You couldn't go back and forward, You can, um, Aye, go back to the index of, up to tutorial.
23:49
On the tutorial pages, you'll see this plus sign, or a magnifying glass, and those are screenshots of things that you might be doing during the this page. And it can be useful to have this up, because it kind of shows you some of the data entry and kind of a abbreviated form of what you'll be doing on that page.
24:13
It's kind of in addition to all the words on the page, and, and in addition to the words on the page has also these links.
24:22
That will also show you images. So a lot of times, we'll say, do this, do this, And you should get this image on your screen, and then we'll show the image on this screen as well.
24:32
And so you can see, for getting the right right answer or not.
24:37
And this is why you can actually do the tutorial on your own in a self paced way.
24:41
You know, we're walking through this in the workshop, and we're hopefully adding some additional context and additional explanation, but the tutorial is designed to be able to you can go through it on your own. And, like I say, it tells you, that, you know, do these various steps and then it says you should get something like this and you can see if you did get something like that.
25:03
All right, and then, finally, um, we're almost done then, almost ready to start, But, I wanted to mention, the context of this workshop.
25:12
So, there's different ways that you can use high split, and, a lot of you have already used high split online with our ready system.
25:20
And, this is a system where you can go into the public, can access it, anybody can get into it.
25:26
And you can run um, trajectories and and dispersion calculations.
25:33
Um, there's different modes that this can be run into some sort of a public access, and there's also, you know, some specialized applications that you can get to as well.
25:50
The advantage of doing it on they on this. Ready, Website. The online is that the meteorological data that's used to drive high split.
26:00
The wind Direction, wind speed, All the net data that's used, you could basically use directly from our server, you don't have to download it.
26:07
What we're doing in this workshop is for downloading the model and installing it on our local computers.
26:14
And we're running it through this graphical user interface, this gooey.
26:18
and it's a lot more powerful.
26:21
You can do a lot more things, then you can, on this online version, There's a lot more menu items, and a lot more options that you that you can choose, but the disadvantage is, you have to download that data. Now, it's free.
26:36
We have all this metadata on our, on our website, and we're providing it to you in this tutorial, so you don't have to do anything other than download the tutorial for the workshop, but when you do it, when you're doing your own work, you do have to download that, that data.
26:52
And, again, we have it available to you, but you have to download it.
26:58
And, by the way, if you've downloaded the model, and you're getting to the point where you're running the Graphical user interface and realizing you want to do things faster, or you want to do the same thing over and over and over again, you can do scripting.
27:11
And, um, we're not going to be talking a lot about scripting in this workshop, but I can say that.
27:20
A lot of you will be doing scripting, and, but using the Graphical User Interface is a great way to learn how to do scripting, because it's the same commands.
27:30
And, actually, on the tutorial, you'll see these symbols on the on the top of each page. on most pages, a little penguin, and a little gearbox.
27:40
The Penguin actually shows you a script, um, of what's happening on that page in a Linux shell script.
27:47
And the little gear box shows you a script of what's happening on that page in a dos patch.
27:53
Ah, script. And there's other scripting languages that you can use, but these are examples.
27:57
So I would say, many of you will end up using scripting eventually, but that the Graphical User Interface is a great way to learn how to use high split.
28:05
And a lot of us, and with all of us in the in the ... group at the ...
28:10
Research Laboratory, do most of our work in scripting. But when we're doing something new, or something we haven't done before, we often go back to the graphical user interface and see how it's done.
28:20
Do it once with that, see that we're getting the right answer, in our script. And also, even looking at the script examples on the page to see how one would do the scripting.
28:32
This kind of shows you the same idea here that there's the model with a bunch of sort of auxiliary programs, graphics and and conversion programs and pre and post-processing. There's just Graphical User Interface that sort of sits on top of it.
28:49
There's the tutorial that kind of works with this graphical user interface and this workshop that we're that we're doing.
28:55
But again, many people will be running high split and the degree programs from command line and scripts.
29:02
And then the other ways that people, you know, use the model, is this ready system and there are other models as well, like a locus forecasting systems, that we have a little bit different than the ready system. And then actually, there's a lot of other uses internal to Noah for emergency response.
29:23
Where the Hyslop model is used as well.
29:26
Um, and so, finally, you know, we hope that you've installed the, the, The model, um, but if you haven't, it's totally fine to watch this workshop as a demonstration. Because you will see what the model can do, and you can go back then later.
29:42
And when you when you have installed the model and do all the things that we've done here, you can watch the video. if you want or do you can just watch the tutorial yourself. And this is also true.
29:53
If you're in a particular section of the workshop, and things aren't working perfectly, like you've made a mistake or if it's just it's, been, weren't getting the right results, it's totally fine to sort of sit back for a minute and just watch.
30:06
In some ways, it might be better just listen and watch, know, what's happening and what we're trying to get get across than to be getting super frustrated about what might be happening on your computer.
30:20
It happens to all of us, It'll happen probably to the instructors in the course, including me will make a mistake, and Won't easy to figure out exactly why we made that mistake.
30:32
So, from time to time, you may decide to not do the hands-on part. And just watch, and that's totally fine to do.
30:42
OK, so a quick recap: We're hoping that you'll ask most of your questions the technical and scientific questions, in the High Split Forum. And to ask a question, you have to register quickly to do that. Actually, we asked you to do that even before the workshop started. Some of you did that already. You can ask questions in the GoToWebinar question box, and we'll try to answer them.
31:02
But there's only going to be, you know, limited number of people that are gonna be answering questions, Only one person doing this, and I see people doing this. So, we'll do our best to.
31:11
To keep up the handouts like this presentation, you can you can download recordings, will come up about eight hours or so later. And again, if you get too far behind, you can view as a demonstration.
31:25
And now with thanks to the rest of the team and also our fabulous IT team at at the air Research Laboratory. I'd like to begin the workshop.
31:36
And to do that we're going to start out with Section.
31:40
Section three of the tutorial and Chris Lochner, doctor Sloppier, will be doing that.
31:47
And Chris, I am going to turn over the, uh, presentation to you.
31:55
Give me a second, make.
32:02
one second. presenter.
32:04
Yes.
32:06
OK, turn off my, Karen.
32:14
OK, Chris, and I think you might be, No, I'm OK, OK, great.
32:24
Say, do you see my screen? I don't see it on my screen where I'm, what I'm showing.
32:31
Yes, we see your screen, Chris, and there's a thing called an audience view that you can click on if you want, but yes, we see your screen. See it, OK, OK, OK, Sure.
32:41
OK, so, after installing, highest, split kinda, the next step is to get Meteorology files which are used to drive the High Split Model.
32:55
I think before I get into this, I just want to stress something that Mark just said, is that if, if you fall behind or throughout this tutorial, if you're not following along in and doing what, what, we're going over as well, I mean, I strongly recommend just taking notes, and then you can always look at your notes and go through this tutorial again.
33:22
And even if you can't follow along, it's, it's hard to kind of follow along, intake notes, all that, all at the same time.
33:31
So, as, as Mark just said, that for it might be just as good just to take notes and then you can always go back and do this Do this later or another time. It's available for you.
33:44
And if you've run into problems, there are, you can always ask questions on the forum, and it doesn't have to be, you don't have to ask them this week, and ask them next week, next month, next year, whenever we're, we're here, to help you, answer questions, help you get started.
34:02
OK, another thing I want to go over, is, when I go through the tutorial, I like to have to kind of windows directories open. The bottom one, I'm showing is the, the the main high split directory, and when you have this installed, which you should buy now. You'll have a bunch of these directories, and.
34:28
When you are Creating files in the GUI, everything gets created in this working directory.
34:38
And if you ever get in trouble, you're not getting the right results. It could be something is not set properly in the GUI or there's an input file that should not be there. They're always there these reset buttons throughout a bunch of these menus in the in the high split menu. And if that's still not working the surest way to just kinda start fresh.
35:05
Is to do this advanced and then clean up working directory and hit continue.
35:10
Then, this working directory changes to working dot 407.
35:16
And then, you have a new working directory that has nothing in it.
35:21
And then, if you open hi split again.
35:24
It automatically populates these, these, kind of default input files.
35:29
OK, the second window I like to have open is the tutorial directory, and.
35:36
In here, we have basically all the input files that you need to, to perform this tutorial.
35:45
Oh, one more thing I want to point out in the high split directory, is that in the documents directory, there is a user guide.
35:53
That's a PDF of everything of how to run high splits, if in the high split model. Through.
36:00
All these help buttons throughout the code, if you hit help, it goes through a specific, ah, help screen, and links to different places of what you need help with.
36:13
And the, in the in the documents, the PDF of the user guide, is just one document of all of these help pages, all altogether in one place.
36:26
Another thing is this obscurely named Arielle, 2 to 4 document are the technical descriptions of the different high split schemes used in the model.
36:38
OK, so, what I mentioned before after installing the next thing to do is to get meteorology files. And the question you might have is, how do I get them, where do I get them, how do I create them?
36:54
Which files are better to use, OK.
36:59
So, throughout this tutorial of a bunch of us at the bulk of the tutorial uses, meteorology meteorology files covering the second release of the cap tax Experiment and here is a brief list of some of the files. We're going to use.
37:19
We have the wharf weather research forecasting model the UW version that's just a name of a PBL scheme that is used in warfare many different ways to run wharf. And this one uses the University of Washington PBL scheme.
37:34
And then an older versions of Wharf 27, 9 and 3 kilometer simulations.
37:45
We also have A wharf, 27 M C. this stands for. This was run with it.
37:54
Time average mass coupled winds were saved, whereas the other warf runs instantaneous winds were saved. So, it's it's it's fairly similar to the 127 UW. Just the winds are time averaged versus instantaneous wins.
38:11
The North American Regional Re-analysis, this goes back from, Do About the late 19 seventies.
38:20
We have archived, and when the ECM W, or 40 raya analysis product, which is, uh, that's, that's also available for you in the end car and sap, sap, and car, Global re analysis. And we have a couple of extracts for covering that the cap text to experiment.
38:41
And if you click on any of these folks and I click on this one, this just shows a picture of the of the domain for this particular meteorology file in.
38:51
This is actually an extract of a larger meteorology file. That covers just the cap Text to Tracer Experiment Region.
39:05
OK.
39:06
And some other general information on some of these files are located down here, telling you at the time period, the number of grid points in the X and Y direction, number of vertical levels, the horizontal and the temporal resolution for this waterfront is 27 kilometers in one hour and some notes on each of these meteorology files.
39:37
And we can also look at, if you have a meteorology file, you can also look at these kind of details or summaries using the high split model. If you have a file, not sure what's in it.
39:49
We can, we can look at that using the high split code, split jewelry. So that's, that's what I'll do right now.
39:58
So if I, if I go, if you click on the Meteorology tab and then display data and check file, then hit the set file name of Arielle format Data.
40:13
And now, these meteorology files for the cap Texts Experiment are located in the tutorial cap text.
40:20
So I'm going to navigate to Tutorial, cap Taks.
40:25
And then let's look at the cap tax 127 UW file.
40:33
Then it has loaded up the directory and the name of the file.
40:37
And if I hit run file.
40:40
And this is kinda what what shows up here shows the the starting and ending time of the meteorology file tells you what model it's the advanced research wharf model, a wharf and the grid size.
40:55
There's 80 grid cells in the East West Direction 78 in the north south and 34 vertical levels.
41:06
Then down here, with a poll point latitude longitude, down to Sync Point Latitude longitude, this describes the projection of the meteorological grid.
41:20
And then Farther down 34, this shows the the meteorological variables on each level.
41:31
So here this is the top, the highest meteorological level, and this is N sigma co-ordinates zero point zero one two five.
41:43
And if you're confused about what a sigma co-ordinate is, you can read about that in the air L two to four document the technical document or the Wharf User Guide or Wharf technical document.
42:00
So it tells you what meteorological variables are in here. P R E S is pressure as temperature wind, velocity, and the U V and W, When when components and also.
42:13
Specific humidity, TK and then also to diff variables and what we call ... and ....
42:20
And in warfare, as we, we may have these diff variables to add extra precision to certain meteorological variables.
42:30
We have diff variables for the vertical wind velocity, which is ... TK, which is diff T, and also rain, which is diff, are, that would be a surface variable.
42:43
And we have these diff variables do increase precision for meteorological variables, that can change, very abruptly horizontally.
42:53
So you can have a sharp, sharp change and for example, precipitation from one grid cell to another vertical velocity and one grid cell to another as well well as TK.
43:03
So this shows, again, as you scroll down, all of the variables on each model level, and then at the bottom on the level. one, this is showing the surface, the met variables on the surface, server side, surface pressure, and so on.
43:22
OK, and so this that that menu just popped up when I hit Run File Program.
43:29
But also, if you go to, If you look at in your high split working directory.
43:38
We also see a file called Check File.
43:40
If I click on that, it's basically the same file that just popped up that I went over with you. It is the same file. So this.
43:47
So when you run check file, if that screen pops up with the information, but it also writes that same file into your working directory. So when you're running doing this tutorial, all the files that are created are, go into this, in this working Directory.
44:06
OK, and I mentioned, if you click on any of these, here, it shows you the domain.
44:13
Now, if you also have a file, want to see the domain yourself.
44:18
If you don't have a cheat sheet right here, you can can't just click on this, and see what you can do is go to display data in Grid domain.
44:29
And then let's navigate back to one of the one of these meteorological input files.
44:39
So that's, it's located in a tutorial cap tax.
44:44
Let's look at well, we could look at the same thing. Worth 27 UW file.
44:50
A grid point plodding interval.
44:52
So you can choose to plot, you know, every grid point are not every grid point are skipped one or every other grid point, skip to grid points and so on.
45:06
Let's see what happens when I show every grid point.
45:10
And the lat long interval lat-lon every one degree.
45:17
Hit Create map.
45:18
We'll see what happens.
45:21
So it created this every.
45:25
Oh, every lat-lon, it actually didn't show, OK, I was wrong.
45:29
The grid point plotting interval this is zero, means that it's not going to show you any. So let me, let me go back.
45:36
So there's, so if I hit one, this would mean I'm grading, plotting every single grid point. This is the two would be every other, So let's see what happens if I hit.
45:47
So there, every every plus sign is, is is the center of a grid cell, and it's really messy.
45:55
So if you want to skip a few, so let's only show every 10th, uh.
46:04
And here's a similar plot, OK.
46:10
OK, and, again, at the Working Directory, it created this same file, here, called the show grid, show grid.
46:22
This is, This is the file that, that was, that was saved, and notice it's in the HTML and dot HTML, It's a, it's, uh, Yeah.
46:33
If, and if we want to convert this file into a gif or a JPEG file, too, and in some other document, we have a tool to do that too. So let let, let us do that.
46:49
So if you hit meteorology utilities, and then SVG to image, it automatically populates the show grid because it was the last figure that was just just created.
47:03
That's keep it. That's make a gift file.
47:07
We just hit execute conversion.
47:11
So it read in that HTML file, the SVG graphic, and then created show grid dot gif.
47:18
So now we have, if I click on that, this, we have the same figure as as, uh, image as a gif.
47:30
OK, the next thing I would like to show you is, you can also look at a profile of of sort of some of the data at a certain location.
47:44
So, if I go to Display data, tax profile and hit.
47:53
So, let's go back to that.
47:55
That's same file.
47:57
So let's go to the tutorial cap tax.
48:04
And the Wharf 27 UW.
48:08
And the Wind Display. You can either show it in vector format or polar format. Let's just keep it polar.
48:15
Time offset.
48:16
So if D zero, We're going to show A starting at the the first time of the meteorological file.
48:26
Weekend.
48:27
Show, every other. Or we can skip to the, the second time period.
48:32
The third, or the sixth.
48:34
So let's just, let's look at the, the first time of, of this file, time increment.
48:41
We can, zero, would be, we're only going to show a profile at the, only one profile.
48:53
Or this would be show a proof of, if you click on one, it's showing profile every one hours, or two hours, or three hours, or every 24 hours show it, show a profile here. Let's just look at that one profile, just right at the beginning of the met file type period.
49:13
And then you can choose a location lat-lon.
49:16
If you leave it at zero, this will just use the by default, Then it will show the the center value, this the grid cell at the very center of the met domain. So, let's just do that. I'll just leave it at zero to look at a profile at the center of this, this domain.
49:40
And I hit run profile, and this new menu came up.
49:46
OK, and what this shows you, it shows the file start time, and ending time at the top.
49:54
We told the profile program too, create just one profile.
50:00
At the very start time, profile time is at the same as the start time.
50:05
The center of this domain happens to be at latitude 42.9 SEC 6 and longitude . Here it's showing that surface values, the surface sites, surface pressure and so on, followed by upper level fields, the three-d. fields. From near the surface to the top of the model.
50:32
For pressure 1013, millibar to 111 Hello bars, so.
50:44
And again, since that created, that file that popped up, we can also have that file already in the working directory, that same file called profile dot TXT, So it's saved for, for you there.
51:03
OK, so, uh, went over a little bit about what that variable map files are, available, how to look into inside these met files, to check to see, the time, the projection, what variables are in there, one particular, a particular profile. Now, question is, how do you get get net files?
51:33
Well, in the GUI there, there are, um, several, a couple of different, uh, types of Matt thousand get archived net files, appended files, or forecast files.
51:49
So if you go to Arielle data FTP, know, there's this forecast option to Download Forecast, arel, high split input files, appended files, as well as archived files, OK?
52:09
So?
52:14
Then I also did the, there is also the Sett Server that kind of just defines where the files are coming from, and there's a default alternate backup for these buttons for forecast data and archived data.
52:36
So for the default, we have the forecast data coming from the RL FTP site.
52:42
And here's the directory within the subdirectory within the FTP site.
52:47
And the alternate and default are the same, the backup is coming from an sap FTP site.
52:56
So you can get forecast data from either the Arielle server or the NCP server. It's really up to you.
53:03
The archived files are only available at the RL server So if you can change these buttons but there, they're all set for to get data from the Arielle server.
53:19
OK so to get data, start with forecast data, click on that yeah what I just said meteorology meteorology, Arielle, data, FTP, then Forecast and then this menu pops up.
53:35
And from here you can click on, what forecast data you want your GFS data, it's Global Data Product a one Degree and zero point twenty five degrees.
53:51
North America that there's the nam forecast the nam: uh, on the sigma levels The her.
54:03
And so I'm not going to actually download them, but it'll take some time but for example, if you want I want GFS data or the name on the Sigma co-ordinates. Just click here.
54:17
You click that, make sure you have the year, month, day, and cycle that you want, listed here.
54:26
And then you have a choice of doing FTP from the data from the Arielle server or there's also, Nomad Server, which is HTTP site, and W get Command is issued to download download the forecast data, so you have a choice whether you want to get it from an Arielle server or the NCP server.
54:52
It's it's, it's your choice, OK.
54:56
So highest split whether you have, if you have the registered version of High Split, you can download forecast data, and then also run high split with forecast data. If you have the trial version of high split data, high split, you can still download forecast data, but you won't be able to run high split with today's forecast data. You will have to have to wait a day. So if you want to run with actual forecast data, you will, you will need to get the registered version of High split.
55:32
OK, and then there's also the Appended data.
55:38
So this is essentially multiple, uh, forecast data, uh.
55:49
Sequentially, in time.
55:52
So, for example, let's say, there was a chemical release, 12 hours ago, and you want to run a high split dispersion model to see when, where, where the pollutants may have gone.
56:14
It's not necessarily forecast data time anymore.
56:18
But numerical weather prediction models utilize the data simulation system that, for the first guests they use.
56:29
For example, may use a six hour forecast from the previous forecast cycle to help initialize, give that first guess of the met files at that particular time. So, for example, GFS has run four times a day the GFS bottle, the Global Forecasting system, at 0 6, 12, and 18 UTC.
56:53
And for, at, to run the GFS model, initialize that zero UTC.
57:01
it would use forecast from our GFS model simulation from the previous day, initialized 18 UTC.
57:10
And then it uses that met forecast for that day at zero UTC to provide a first guess of met variables and then also along with observations at UT at zero UTC we'll perform a data assimilation technique to help minimize the model error to create this, these initial conditions for for the met model.
57:39
And so what we can do if we want to run a simulation from previous, say, 12 hours, uh, we can use the initialize initialization time from the previous forecast time periods.
57:55
So, what this, which is basically, what this appended?
58:01
Appended.
58:03
Data is.
58:04
It is the appended initialization data for these particular time periods before the actual forecast time starts.
58:14
So, if there was a, there was a chemical release, 12 hours ago, you just found out about it.
58:21
You have the forecast data for now, but, in the past, you don't have anything. So let's get this appended data for the previous 12 hours, And then we can run high split starting from 12 hours ago with the Appended data.
58:34
And then, once you get to the current time, then you have the forecast to move on farther in Time.
58:42
OK so that's a long description of what the Appended data is. Basically initialization time periods for previous model runs and you can choose between these particular models the GFS Nam different nests of nam nam over Alaska or Hawaii Choose. Your year, month day, and forecast cycle.
59:07
Then also choose where you want to get the data from the Arielle server or the nomads Server.
59:17
OK, and then the next would be the archived data.
59:25
OK, so here, this is coming from the High Split, I mean that the Arielle Server you have a choice from.
59:35
You can look at all these choices of where this data is.
59:38
what you have the nam G das GFS EDS and you choose your month, year, day, hour, and then hit the FTP site and then FTP data file and then and then you'll get, get your data.
59:53
OK, so I'm not going to do that, but it will take some time, so I'm not going to do that.
59:59
Now 2 one other thing, if you want to learn more about Different meteorological datasets, if you just Google High, Split Ready.
1:00:14
The first thing that comes up, is the ready site, OK, And if I scroll weight, close to the bottom, on the left side, there's link.
1:00:27
It says Archive Data Information, OK?
1:00:31
Then this is, shows a description of all of the archived data available for the nam es.
1:00:38
G dash S GFS, ends up and carry analysis, her name and so on. You can look at this to see what data you want. What time periods is it available?
1:00:48
What is the resolution, what domain, you click on it for each of these kinda sections. There's a readme file that has specific information about what's in the file, the projection, what met variables, the time period.
1:01:07
And also, a map of the actual domain.
1:01:13
And then, I went back to far, sorry, oh, that's happening.
1:01:21
OK, so that's for archived data, and then also on the left, here's the archive, there's also forecast data information, we have similar, a similar kind of menu for this, the forecast.
1:01:32
There's the her three kilometer on pressure levels signal levels, which are the Native Model levels, and then Nam and so on. So you can look through these on your own.
1:01:45
And then also, it shows you what variables are in the files.
1:01:52
The surface fields, upper level fields. And you might look at this, like, what does S HTTP? And I don't know if you click on certain either surface fields or upper fields.
1:02:04
This well kind of describes all of the possible variables that are in RL files in these high school input files, not all of these variables are actually used in high split.
1:02:16
But these are all the variables that her, where once, in, a high split in input file.
1:02:21
And if I search for this S HGT, you see the S H G T is surface height, with a units of meters.
1:02:34
OK, I think I'm getting a little behind.
1:02:36
So, let me try to catch up a little bit.
1:02:41
OK, the next section is to convert Meteorology Data. I'm not going to get into this at all.
1:02:47
This is, I think, in my opinion, this is too advanced, and also, in my opinion, that we have this this Menu for Convert to Arielle, Options for Converting Wharf. Global that lawn data that which are on crib one format and the ECM W F era.
1:03:08
in the GUI.
1:03:09
But I think we probably shouldn't, it's a little more advanced.
1:03:14
And the Wharf era W Converter has has gotten updated more more recently then the actual converter within the GUI. So, input files telling us to how to convert it.
1:03:32
It doesn't have all of the options that a user might want.
1:03:36
So put what I recommend for people that, ooh, there are reasons for people that may want need, or have reasons to convert their own met model or climate model into high split input file.
1:03:57
But for most users, I think we probably have, for most users, we already have the files that you need exist on the RL server. So, you can just download them, FTP the data. But there are a few users that will have a need. If they ran wharf at a very high resolution for various persons particular domain, they'll probably want to convert their Warf meteorological output files into high split input files. Or a thought.
1:04:26
Climate Modeler ran a climate simulation and wants to run high split over along period from, from, from their model output, they can convert that.
1:04:38
So what I recommend for that, the people that ran these mette models, they're familiar with, uh, running, and on Linux, and using their own scripts, they should feel comfortable being able to convert these programs using directions that we already have available.
1:05:06
Yeah.
1:05:07
Um, in it, it's within the High Split System.
1:05:14
Let's see, don't have it year, so in, in the high split, there is this data to Arielle.
1:05:25
Folder.
1:05:26
And this shows you all of the different meteorological Converter converter, as we have.
1:05:32
So if you look at the air W two RL file, so on all of these files there's there's a Readme file.
1:05:39
So if, if someone really wants to convert their own file, I highly recommend just looking at the readme file, following the directions very closely, and converting, converting their data to convert their net output into a high split input file. So I'm not going to go over this now.
1:05:54
But, again, if people have questions and have problems doing this, there's always the forum that you can ask those questions from.
1:06:03
OK.
1:06:04
So there are I went over the different servers that we have the RL server, the NCP server here.
1:06:11
I'm showing the actual paths of the FTP sites for the RL forecast and sap servers NSF also has this HTTPS sites where you can get data.
1:06:27
So let's just go to that, We can look at it, Sorry.
1:06:33
If we go there, we see that, too, directories, high split, four, today and yesterday, click on that.
1:06:40
And here, we have all the high split files that are available to download now her an app, GFS, G dash, and so on.
1:06:56
So we used to be able to just go to FTP Sites on the browser. You no longer can do that so.
1:07:04
Well I showed you that you can FTP sites with the GUI using don't ...
1:07:10
data FTP but actually don't recommend doing that.
1:07:15
It's much easier to download data using FTP client.
1:07:21
So here, I'm showing, if you go to here. I use files that you can use whatever FTP client you want.
1:07:29
So if I go to the FTP client, FTP to RL dot NOAA dot gov, it's an anonymous site.
1:07:38
Username would be anonymous.
1:07:39
Password is your e-mail address, and this is OK. So that's this is for the RL server.
1:07:51
I'm not sure if I'm still connected. Let's say then the Arielle server for.
1:08:00
Yeah, I'm still, I'm still connected.
1:08:01
So to get the Forecast data and Arielle server slash pubs slash forecast, and then you have all of these different folders with different dates. So you see there are more dates for the Arielle Server. The forecast go back five days. The NCP server, it's the last two days.
1:08:21
So if I click on June eighth, we can we have the same. I split file's. Name slightly different. But they're essentially the same files as what's available NCP get at Get here.
1:08:37
And also on an Aerial Server, you go to pub slash pop slash archives, we have all of the data that's available that's already converted.
1:08:51
Uh, again, for descriptions of these, you can look at the ready site and look at the archived data information for that description of all of the different different net data available on the Pub Archive site.
1:09:14
OK, so, one more section to try to finish up on time here, OK?
1:09:24
Next thing, the other thing you can do to create data is create user defined data.
1:09:32
So, if we go to, let's see, convert to Arielle User Enter, OK.
1:09:42
So this creates a high split input meteorological file based on just one observational location at the surface.
1:09:53
OK, so one reason you might want to do this, is if there was, or accidental release set of power plants of, or somewhere else, I don't know anywhere you want.
1:10:08
And you want to see where it went, where this pollutant is going?
1:10:14
And you have a meteorological station, very close to where the accidental release occurred.
1:10:21
Then you can use met data, right at the release location, two, create an input file to calculate the dispersion of the pollutants.
1:10:33
Very close in near the site.
1:10:36
Now, this would be huge.
1:10:38
Of course, we know meteorology changes over over space and time.
1:10:42
So we only want this file to be a very small area. So you can run high split in a nested domain where at a very high resolution will use in very small location around the power plant.
1:10:57
You can use this station data, user defined data, that we have observations right at the right, at the release location.
1:11:07
And then outer most demain, that would be from a meteorological model, which would obviously be and it could be at a course coarse resolution covering a much larger area.
1:11:19
You can have nested many, many, many different multiple meteorological invested.
1:11:26
Nest within, within High Split.
1:11:29
OK, so, I'm just gonna go over this example here of how to create a user entered data. For a simple, simple test case. So, I'm just gonna copy kinda what, what is, shown here on the, on the Windows screen.
1:11:43
So, let's use a latitude of 40.7 longitude of . And then, we don't, this file doesn't exist.
1:11:56
So, we're going to hit create file OK, so this, this window pops up and let's put in this first line here.
1:12:06
So, we're going to create this file for 1983 September 25th Yet Zeros E.
1:12:18
But nope, too, obviously.
1:12:22
zero minutes.
1:12:24
Direction: 225, Wind speed of six meters per second, mixed layer.
1:12:34
5500 meters, and a stability.
1:12:39
Classification of three.
1:12:41
Now, you might be question of, what? What does this stability? What does that mean?
1:12:46
So, for this this particular case, if I click on stability category, it uses the Pascal Stability classes, which are there lettered, but in high split, we number them. So, A is one, B is two, C is three. And so on 4, 5, 6, 7, from extremely unstable conditions to extremely stable conditions. And the high split stability classes are based, generally, on the classical stability schemes, and are, based on the range of the KZ.
1:13:26
The estimate of the KZ, within the model.
1:13:30
So, for this case, stability class, we put three, which is C The KZ is in between 40 and 85.
1:13:42
And there, there are also, on this page, are listed, how other people define the Pascal stability classes based on daytime insolation and nighttime conditions, cloud conditions, and fluctuations in wind direction in the bird vertical temperature, temperature, gradient.
1:14:08
OK, so back here, So here, for this particular case, I chose Stability Class three, and then I hit the repeat button, just copies all the data down downward.
1:14:20
And now all I have to do is assume that these meteorological conditions are the same throughout the entire time period.
1:14:28
But this doesn't have to be the case, if you ever met station, and you can very, very, you have the data, you can bury the the Matt parameters.
1:14:39
OK, so, 25th 18 UTC and now, next slide, 26th at 0 0, 26th at 0 6, 26, 12 and at 26 at 18.
1:15:01
Let's save data to file and then I hit the run convert button.
1:15:08
That window disappeared.
1:15:10
Then if I look in my working directory, let's go look in there.
1:15:20
I have a couple of new files, one is station data dot TXT as basically the file I just same thing I just created.
1:15:30
And then I have station data dot bin, This is the input file that we used for high split, and data config, uh, this describes the file that we use. So, so, the projection, that was, that was used to have met variables that are now in the file. So, is the surface pressure sir at heights temperature, the wind velocity, and the UV.
1:15:58
Component. And the vertical velocity variance in, in the different directions.
1:16:04
The east, west, north, south, and the vertical for four different heights.
1:16:10
And one other thing I didn't mention before, is that the, I'm going to open this binary file, this Arielle file in a notepad So we can, we can look at it a little bit.
1:16:24
So at the very beginning of the file, it has some information on the date.
1:16:32
How has created site? That means I created it myself. Other ones, it would say, like, a war for GFS ditch to show you what model, Some of the projection information I'm not gonna go over really.
1:16:43
And then the the met variables that are in it for each level, and then a whole bunch of gibberish, because it's about the actual. data isn't binary, but the header is an ASCII.
1:16:58
So since it's a combined ASCII binary file, when you are transferring data via FTP, make sure that you transfer it in a binary format.
1:17:09
Because it is a mixture of ASCII, FTP, ASCII, and binary.
1:17:13
Don't have the don't transfer it, and ask. You don't have the FTP client automatically detected. Because if it sees this ASCII at the beginning of the file, it might transfer the file in ASCII format and then the file will be corrupt.
1:17:30
OK, Let's say.
1:17:35
So, then, that, I'm not going to go over this, this.
1:17:39
This basically does the same same thing. I just did. it did earlier. Yeah, it ran well.
1:17:47
They ran the profile Program on the file that I just created, the station data dot ben. and we can just look at the results here. And.
1:17:59
Then, you can see the profile at this particular location, 40.7, -83.
1:18:09
So, is that yeah, that that's just the center of the demand. And actual profile is the same for every grid cell in this particular station data program.
1:18:21
That has the temperature, the heights and the the values of the verdict velocity variance. And you can see that where the ... Velocity variance changes to very small. That is the boundary that that's the PBL height is, is in between one thousand or 1500 meters.
1:18:38
OK, So that's the station to air L file. We also have a couple other user defined files. one is a sounding to RL file. It's not in the GUI.
1:18:48
It is you have to run it and using the command line, and then also a program called Drone to RL, which uses drone data for the Bottommost layer, where, however, high the drone flies, and then appends, a forecast model on top of that.
1:19:09
But again, that's, that's not in the GUI either, but it can be run from the command line.
1:19:16
So in the High split, there's the general high split dispersion in transport and, and trajectory codes. But there are also many, many, many, many other other utility programs such as this sounding to air our drone to Arielle station to Arielle.
1:19:34
And let me show you quickly.
1:19:40
If I go to, if you pull off the user's guide, you'll remember sitting in your, in your In your documents, but also in the Document subdirectory, But also if you, if you just Google, I split user guide, there's the first thing that comes up, OK.
1:20:06
I split user guide first thing that comes up.
1:20:08
So if you don't feel like, going to your document directory, just go here. You'll, you'll get it and.
1:20:16
We got the table of contents down here.
1:20:19
And then at the very end of the table of contents, there is the Utilities.
1:20:24
And this shows all of the many different utilities there are.
1:20:30
And if you scroll down the ones that have to do with meteorology these are the meteorological data to RL formats. These are the converters the meteorological. Data editing.
1:20:41
So a couple ones that would be interested that, I think, are most used, in my opinion, are, the extract grid.
1:20:49
This is if you have a met file that is, you know, get a global map file, but you only need you only want to run high split four, a small chunk of the north-east of North America.
1:21:02
You can can take a subset of, of that that demand to have as much smaller demand.
1:21:09
So it will run one hospital run faster, because as soon as a particle is off the meteorological dread, that particle has no no longer are no longer transported if there are no other meteorological grids for them.
1:21:25
Particle to go.
1:21:27
And another is we, we use some of these, the check file, for example, the profile, but another one is V mixing.
1:21:38
This will give you different estimates for PBL height, based on different turbulent, different PBR schemes that you might choose within the high split scheme as well as extract station. This will extract meteorological data at the surface for a particular location, which is helpful for comparing with met data for a meteorological station.
1:22:03
OK, then, to run any of these programs, let me just let me just look at extract station.
1:22:13
There's, there's just a quick tutorial of, like, how to how to run it.
1:22:17
and also if you run any of these programs through either run interactively or it would be a prompt.
1:22:25
That tells you exactly what to do or if you, or help, uh, statement will come up.
1:22:33
So if I just run it for example extract citation there'll be a prompt that says enter the meteorological directory at the met files. And then you type in the meteorology directory and then what's the name of the file? So so and you just follow the prompts.
1:22:48
Other ones, other utility programs, may not be a prompt, but if you just click, type it, um, Likely, Help statement will come up that which shows, tells you how to run it.
1:23:02
Uh.
1:23:04
OK, I'm already past my time.
1:23:07
Sorry, sir.
1:23:09
I'll, I'll kind of end there, but also in this chapter, you guys can do this kind of on your own. This isn't meant to really be done during this. So you can have these exercises that kinda go over after each chapter. These exercises that you can, you can go along with. If you have trouble, the hints and all of these exercises are very good to help you get there. And then also, again that hit this button to help you. The graphical, kind of graphical hint of how to how to do these problems.
1:23:40
And then if you have questions or problems began, high split form is there for you.
1:23:47
So, Mark, I guess I'll turn it back over to you. Thank you very, very much, Chris. Really excellent. And don't worry about going a little bit long.
1:23:56
You had to cover a lot of the basic stuff about the tutorial and about ty split that normally would have been covered in the installation section. So, thanks very much.
1:24:05
I'm going to change the presenter back to myself and see. Second.
1:24:15
Sure.
1:24:24
There we go, OK, so what we're going to do is go into a break right now, and we'll go for about 15 minutes.
1:24:32
We'll come back at 10 40 Eastern Daylight time, or 20 minutes to the hour, about 15 minutes from now, so we'll see you back in about 15 minutes. A bye.
1:24:49
Baser experiment is that, we have a real answer. two.
1:24:53
What the air masses for doing, what pollutant releases would be doing? So, it's nice to be able to have some some objective data to compare the modeling against, and in this particular experiment, there were six releases.
1:25:11
Uh, we're gonna be looking at one of the releases kept X. Experiment number two. It was a three hour release of a particular fluorocarbon and released, happened in this case, from Dayton, Ohio. And, you will see that when we start doing some of these trajectories will be releasing the trajectory, starting the trajectories from that location.
1:25:34
And then when we start doing dispersion runs, will also be doing the same kind of thing, and eventually, we'll be sort of simulating this entire experiment.
1:25:44
And you can see how well the model does, just to give you an example of what this looks like.
1:25:50
Um, this is an animation of the model, results that will eventually be getting to do. And then also the numbers on the screen, or the concentration measurements.
1:26:04
Actually, and it's the app, it's the total, 3, three day average of those measurements.
1:26:10
So, you see that they pollutant, you know, started going up to the north east, and then started to spread or out, and then come down a little bit to the, to the south.
1:26:22
So anyways, that's the context of a lot of these calculations that we will be doing.
1:26:29
And now, without further ado, let me go now to the next section where we're going to be starting to do trajectories.
1:26:38
And the first thing we're going to do, and we're going to be doing this a lot today, is we're going to go to the the High split menu.
1:26:48
And in the trajectory section, we're going to click the setup.
1:26:57
And when we do that, we get A many that looks like this, And the first thing we're going to do is put in the starting time.
1:27:09
And actually I.
1:27:12
We'll put in the correct time here, so it's 1983.
1:27:19
2 digit ear, um, September the 25th.
1:27:26
And our 17. And just note that all the times in high split are UTC or Greenwich Mean Time.
1:27:35
Um, and um.
1:27:39
This is these time when the tracer was released for that kept text and for to experiment, and then we're gonna put it in the starting location.
1:27:48
Andy, you click on that starting location, button, ear for an application button.
1:27:55
And we're going to type in 39.9.
1:28:01
Oh, that's the latitude and longitude , and the height of 10 point O meters.
1:28:13
And Note that the west longitudes nicking, the United States, are going to be negative.
1:28:26
Just note that, and, um, Let's see, one second. I realized that I need to reset something here.
1:28:35
I did something a little bit earlier.
1:28:37
Let me just save this for one second.
1:28:41
I'll show you all of this later. I made a mistake here. Let me reset.
1:28:48
All right, in fact, save situations run, oops, OK.
1:29:01
Locations, OK, there we go, Yeah, the reason why I knew I had to reset is because it had the mean sea level and we want this to be above above ground level, and I'll explain a little bit about that later. OK, the next thing we're going to do, is we're going to put in the run duration.
1:29:20
And the Tracer experiment took 68 hours and, um, we're going to go forward with the trajectory and, um, leave the top of the model at 10000 meters.
1:29:35
I'll explain a little bit more about that, about that later.
1:29:39
Um, we're going to call the output just a simple um output file name called Teton, like trajectory dump.
1:29:49
You can specify whatever name you want and this little dot dash.
1:29:58
That little dot slash means esthete working directory.
1:30:03
And as Chris had said, all of the files that you're creating in the GUI, the input and output files are all ending up in your Working Directory.
1:30:13
Um then the other thing we're going to have to do is add the meteorological data. And I actually had already added this, But let me clear this.
1:30:23
So, if you have something else in there, like the October 16 to 18 dot band or something, you can clear.
1:30:31
And then we're going to add ecological files for this simulation.
1:30:37
And, um, you browse to a tutorial wherever you have it on your computer, into the cap Text sub directory.
1:30:49
And we're going to pick the cap texts were 47 UW dot B, and that was the one that Chris for showing you, as well.
1:31:00
And so this is kind of what your setup screen should look like.
1:31:04
You have the right, I'm starting time.
1:31:10
You have the right starting location and height.
1:31:16
You have the right runtime going forward. We have the right output, and then the right metadata. And then what you do is, you hit Save.
1:31:25
And whenever you do these setups, you always have to hit save, because that then saves this information, too, a file in the working directory.
1:31:35
It's actually called a control file, and that's, sort of the basic file that haislip needs to, to do.
1:31:41
It's run, So you save it.
1:31:44
And, um, the next thing you're going to do, after you've set up a run, as you go trajectory, run model, we're going to be doing this a lot. We set up the run. Then we run the model.
1:31:55
So, you click this trajectory run model, and, I had done some extra stuff here, and it found this advanced thing. I'm going to delete the file, explain this later. Sorry. I had done.
1:32:10
Some things earlier, all right.
1:32:12
And when you do that, run the file, run, run, the model goes pretty fast, and when once you get to that end of that window, you can hit exit, and. Um.
1:32:26
Like I say, all of these files are ending up in that, In the working directory here, I'm showing you the working directory here.
1:32:33
Here's that key dump file that was created here versus control file that we, that we wrote when we saved the GUI, and here's a message file that Chris showed you.
1:32:44
some a little bit about, er, I'll show you a little bit about that more actually message file a little bit later in the workshop.
1:32:53
Now, we could, we're going to show the results of the trajectory, we go trajectory, display, trajectory, So it's trajectory, display, trajectory, and up comes this menu.
1:33:08
And it already had populated.
1:33:10
Because it knows, the last thing we did was this T dump, trajectory, dump, file, and you leave everything.
1:33:19
Uh, actually, let me, with respect to pressure, um, I think you can leave everything the same.
1:33:28
And you hit execute display, And when you do that, up comes a, up, up comes a graphic should open up in a web, web browser And, um, it, you should see something like this. I, I may have had a slightly different labeling interval than you, I put six hours showing the endpoints, every six hours.
1:33:52
It might have been 12 hours in your, in the default version when when you first started, but.
1:34:01
You could see that the release point was in Dayton, Ohio. And the trajectory kinda went up along the border between the United States and Canada.
1:34:10
And after about three days, Direct Directory stopped after 68 hours. You can see that.
1:34:17
See, starting pressure was 989, hec two Pascal and stayed pretty close to that pressure the whole hallway.
1:34:31
We're gonna quit out of that, and, um.
1:34:36
As long as you've got that result, and I hope that you, that you did, then, what we're going to do is, we're going to save that control file that setup information.
1:34:48
So, you've got trajectory, setup, run, and we're going to Save As.
1:34:55
We're going to save this setup as trash face, control dot text.
1:35:04
That means that we're going to save this, this setup information. Then we can retrieve it later if we want to.
1:35:09
And we're going to hit safe and it'll automatically save in the networking directory.
1:35:16
And, here, you can, let's see here, if.
1:35:23
There, you can see that I had just saved that trash based, control that text and working in the working directory.
1:35:31
And, by the way, in the tutorial directory, ah, When you downloaded it, there's there is a folder called files.
1:35:43
And in that folder are actually all of these files that we'd be saying, we'd be saving, So, trash.
1:35:50
That's a trash base.
1:35:57
It's in here, fast based control dot text.
1:36:00
And depending on where you put your tutorial and where you put your high split program, it's possible that you might make a few little changes too to this.
1:36:11
But this won't have most of the information that you would need if you wanted to try to take a shortcut and do a simulation without, by just retrieving something like this, OK, that's the working directory, OK, the next, um, staff is, we're gonna look at the, the, actually the equation of how we do these trajectories and in order to do that, I have A a little slide here that might be helpful, and let me bring that in here.
1:36:48
Um, I hope you can see, see that.
1:36:53
The trajectory starts at a certain position in time, and at at P, then, we're trying to figure out what the position is, that, at T plus, delta T P.
1:37:05
Prime position at that after the one time step happens, in N equals that initial position, plus some velocity that the trajectory is moving at times, the time step.
1:37:19
And um, the way that that we do this is it's called a predictor corrector approach. You start with this position.
1:37:31
At P, this is like a three-dimensional position, and we have a velocity at that location that we got out of our metadata.
1:37:40
And, um, you move the particle, or the, the trajectory, essentially, this so-called first guess position.
1:37:49
Which is essentially just that, the velocity at that initial position times the time step.
1:37:56
Then you look at the velocity at that position, which is gonna be different, because you're in a new position now, you get this from, from the metadata, and you then move that.
1:38:08
that position, that initial position with that velocity over to this so-called second guess position.
1:38:17
And then the final answer is just the average of those two positions.
1:38:21
And it's called a predictor corrector method and it actually works really quite quite well.
1:38:27
You can use more advanced approaches, but they don't really give you a better, better answer, and this is basically what the trajectory, um, model is doing.
1:38:38
It's essentially just integrating the wind forward in time, um, from the metadata that you're, that you're providing it.
1:38:46
And the time step that is used, um, varies in the model. The minimum time step is a minute, and the maximum time step is is an hour.
1:38:58
But what the model does is it says if we have mythological data that has a certain grid spacing, we never want to go more than three quarters of a grid.
1:39:09
Um, space in any one time step.
1:39:14
So, if you're using met data with with very fine, very fine resolution, you're going to end up having a smaller time step and might go down even to a minute, or more or two. Whereas if you're using that data with a very coarse resolution, you can often get away with time steps of 30 minutes or even even an hour since sometimes.
1:39:35
But the model will set that time-stamp based on the size of the met data.
1:39:41
Um, Chris had mentioned this the co-ordinate system in horizontal, actually, high splits using these grid units us depending on on the size of the mythological data grid. But in the vertical is using this sigma unit.
1:39:59
And sigma goes from if you think of the model, think at the top of the model where receive would be the top of the model, then see top minus C, 1 equals 0 and so sigma.
1:40:14
Sorry, Yeah, that would be equal to zero at the top of the model.
1:40:19
And when Z equals zero, um, then sigma equals one.
1:40:24
So, it goes from zero to, to why, Um, sigma is it's sort of a terrain following, for falling co-ordinate.
1:40:33
And let us see the top of the model. Scaling is Inst internal, the model is 25, 25 kilometers and in fact, when we go, let me go back to this.
1:40:50
Net data here is top of the model.
1:40:54
This actually is not the internal scaling of the model. The internal scaling of the model is always 25 kilometers unless you do something special to change that.
1:41:06
This top of the model that you said here tells the simulation how much met data to process. And if you're doing a trajectory in the boundary layer, and it's never going higher than maybe 1 or 2000 meters at most, and you have a big met dataset that goes all the way to the stratosphere.
1:41:25
You can basically tell the simulation to only process like the first, in this case, 10, 10 kilometers. When you do something like that, the model will run much faster because it's not trying to process all this data that that might be there. If you want to be safe, you can always make this bigger.
1:41:43
Let's say 25 kilometers or more.
1:41:47
Um, let's see, let me go on now to the next section.
1:41:56
Um, so this section, we're going to be.
1:42:02
Estimating the next layer depths and, um, you could see that, that trajectory that we found.
1:42:14
See the director that we found didn't really look like, let's see. Where is that?
1:42:20
Didn't really look like the simulation that I showed you at the beginning. Remember that?
1:42:26
Tracer kinda went all the way up here and then kinda came down. So we're not, we didn't get a lot of this area here in this one trajectory. And that's sort of typical with what the trajectory.
1:42:37
No, we started it at one heights and we just went forward. And this would be you could think of this maybe as the central line of one part of the plume. I guess would be one way to think about it.
1:42:49
And so, um, what we want to do is, the first step, to determine the mixing, the boundary layer height at the at the time of the trace release, is going to open up the meteorology, and Chris had done this before, display data, text profile, and actually, Chris showed you a little bit of this, as well.
1:43:16
And when we do that, we're going to open up this menu, and we're going to then find that tap texts, um, 20 70 kilometer UW data again.
1:43:33
And, um, notice that these buttons that Chris talked about, these offset buttons and increment buttons.
1:43:44
This particular file, take, takes data every hour, but we want to get the starting time of the Tracer Experiment.
1:43:55
And we knew, the file starts at zero UTC, but the trace somebody started at 17 UTC.
1:44:02
So, what we're gonna do is, we're going to set the offset to 12, then set the time element to every six, then, therefore, we'll get the 12 hour and then we'll get the 18, our.
1:44:13
And the other thing we have to do then is to set the time the location of the Tracer Release, 39.9 O minus 84, zero point twenty two.
1:44:29
Then we can run this profile.
1:44:34
And when we do, up comes this output screen and, um, couple of things.
1:44:42
I want to point out from this screen, you can see that.
1:44:47
Let's see here.
1:44:53
Oh, OK. This was the first time step now is this.
1:44:57
Uh 12 hour, OK and you can see the PBL height is fairly low and then we get to the second time step, but this is it's 18 hour, which is close to the trace release.
1:45:08
The PB L Heights is about 1200 meters, which is typical for in the day.
1:45:17
And um, you can see that.
1:45:25
A couple of things I'll point out about, about this.
1:45:30
See, Boundary Layer is such that the, the potential temperature, it is constant throughout most boundary Layer, and then it increases above the boundary layer.
1:45:41
And so you can see, the boundary layer is kind of ending it around, you know, 8, 30, or 8, 40, act of pascals, pressure.
1:45:52
Anything with it.
1:45:53
Like kinetic energy, you have a fair amount, air, and then the player eat very little, um, even the wind, relatively constant, in the boundary layer of we start to start to increase.
1:46:13
So, this tells us that the boundary layer heights at the, when this, when this tracer released was about 1200, 500 meters, and we're going to use that a little bit later, when we run trajectories for different.
1:46:30
And also, what we do dispersion as well from, from that location.
1:46:39
Um.
1:46:42
Let's see here.
1:46:46
Let me exit out of that and let me quit out of this window.
1:46:52
And, um.
1:46:58
Next section here.
1:47:01
So, now, we're going to start deploying.
1:47:04
So, more realistic trajectories are used for different heights.
1:47:09
So, basic trajectory calculation.
1:47:13
That we did last time, so we're going to open up trajectory, setup, run.
1:47:20
It'll remember what you had from before.
1:47:24
But, we're now going to make a case.
1:47:28
So, the first thing we're going to do is, we're going to change, Um, the number starting location is a three.
1:47:37
And when we do that, and then, when we click this Starting Location button, now, the menu that pops up has space for three different, um, three different locations.
1:47:50
And, um, it's actually in here. This May or may not have come up for you. I think maybe I had put this in before.
1:48:01
But, what you want to do is have the same latitude and longitude then three different heights 10 meters above ground level, 600 meters above ground level, it should be about half of the boundary layer height.
1:48:17
About 1200 meters, above ground level would be near the top of the of the boundary layer hit, OK, once you've entered those in, and, um, Well, now we're going to save, this, you need to save it, and then we're going to run the model trajectory run model.
1:48:45
It's a little bit slower, but still pretty fast, an exit, and like we're going to be doing all day long, once we, once we save this setup, and once we run the model.
1:48:56
Then we're gonna display the results, display, trajectory, and we can just go ahead and execute displayed.
1:49:06
Right now, we can just use the defaults, and we get These three trajectories, and let's see if we got the right right answer here.
1:49:19
Yeah, that's what we're supposed to get.
1:49:21
And let me Um, I'll pull this out.
1:49:25
And It's just more or less what we got.
1:49:28
OK, um, Let me move this over a little bit.
1:49:36
This over a little bit, OK.
1:49:39
And so if you did get this answer, then You can save this, uh, setup file to trash level three control dot tech select menu.
1:49:54
Let me.
1:49:55
But that will go back to, set up right.
1:49:59
And now I'm save.
1:50:01
This set up as, oops, sorry, directory.
1:50:08
Set upfront, save as, we're going to call it trash, aye.
1:50:15
Level three control dot text.
1:50:21
Level three.
1:50:23
Control dot text, because we had three different, Um, heights that we're releasing, all right.
1:50:33
Now, um.
1:50:39
Before we go any further, I'm going to show you something else about, about the model, which is for it. And we're going to go into this so-called.
1:50:52
Advanced menu.
1:50:54
And, when we don't go into that, actually, Chris might, well, let's see, no, heat, showed you that too much of that, go into the Configurator, and then, the trick.
1:51:07
And, why don't we do that?
1:51:11
Cecily of these different menus and we're going to pick for this this time. We're gonna pick sub menu number two, this MSL.
1:51:19
Um, AGL units, we're going to click that and this lets you set the height units for the Model SV either being the height above ground level.
1:51:33
Which, the default was, or Heitz, above mean sea level, which is one choice you could make, or as a fraction of the boundary layer.
1:51:44
And, um, let's select this fraction of the Mixed Boundary layer, because this is something that we're going to do next, could be very helpful.
1:51:54
Suppose you want to start your trajectory at the middle of the Boundary Layer, but you're not sure what the boundary layer height is.
1:52:03
You can set this, make this setting in the model. And then the ...
1:52:10
will calculate the Height of the Boundary Layer and pick this fraction at the Boundary Layer for that, for the starting height.
1:52:17
So let's go ahead and pick that Fraction of the Boundary Layer, and Save.
1:52:23
And, um, then we save, uh, this OK menu, and by the way, when you use this advanced menu, it creates another file called setup dot C F G, which is, we call it a name list file, but it's essentially an advanced settings file.
1:52:46
The setup CFD file is optional.
1:52:50
You can see it was just created when I, when I saved, saved this.
1:52:54
Um, if you don't have a setup dot C F G file, the model run, and it will just use all the defaults, that, that would be, that would be set.
1:53:04
If you do have a set C, F, G filed, and the model will read that, anything that you changed in the advanced menu will then be utilized.
1:53:14
Sometimes it's really important to do that.
1:53:16
And sometimes you don't want to, to do that but in this case we change the one thing we changed was that the heights will be interpreted now instead of meters above ground level as fraction of the boundary layer.
1:53:33
So now when we go in to do the setup, set up run, and we set up the starting locations, Let's just go ahead and put one back again.
1:53:43
And instead of 10, now, we would have to put, if we want to have started at the middle of the boundary layer, we'd have to go zero point five and you could see, it actually says, fraction of the boundary layer here. That was, when I first started, you might have noticed I, I realized that I had had the MSL up here.
1:54:02
That's why I knew I had to go back in and change that, But it does reflect that we had made that change.
1:54:08
And in fact, here, we want to put if you want to start at the half of the Boundary Layer height with the zero point five.
1:54:14
If we had left 10 in there, it would have started at 10 times.
1:54:19
The Boundary Layer height, which would have been way too high, in fact, would have been probably higher, than the meteorological data that we, that we would have.
1:54:27
Um, so now we say, OK, and we're going to Save, Ah.
1:54:37
We're going to save this.
1:54:40
Um, setup file, and we're going to run the model now.
1:54:47
And now, when we run the model, up comes this menu, and I had come up before, because I had left something in there.
1:54:54
It's saying that it's found this setup dot CSV file in the working directory, and it's asking, do you want to run with that, those advanced settings that you did, or, where they just left over from a previous calculation that you had done, and you don't really want to do that? So if it's leftover for something, then you would hit delete file, and just run with the defaults. But in this case, we know that we did set it. And so we're going to go ahead and run with a setup.
1:55:20
CFE file, and the model runs, and if we go to trajectory, display, trajectory, and go ahead and just, go ahead and hit Execute Display.
1:55:32
Here, we get this trajectory, and it's a little bit different.
1:55:36
The one that it started out at ground level, kinda stayed up, actually kinda went up, on the border between the US.
1:55:44
And, and Canada.
1:55:46
And this one that started at about the middle of the boundary layer, showed a little bit more of the behavior that we saw in that initial animation of the of the tracer.
1:55:58
And, you also could see that it did, It varies in its height, throughout the trajectories, which started in the middle end, um, and moved around vertically.
1:56:12
one thing we want to do, is we want to save so of these files.
1:56:18
So, we go back to the Advanced Menu Configuration setup trajectory, and we're going to save that setup dot C G file as um trash, mixed D setup dot T X.
1:56:32
I had already done this, but, you could type in those words just as sort of as we're showing here and trash, mixed layer depth Set up dot text, save and, um.
1:56:51
Let's see here.
1:56:58
We also want to.
1:57:02
OK, so yeah, so that that's right, and that's the result that we saw All right.
1:57:08
Close out of this.
1:57:10
Um.
1:57:13
Throughout the workshop, we're going to be doing a lot of this, you know, setting up, and the trajectory, and also in the concentration, and also in the advanced menu.
1:57:23
And you can see that when you do, um, these runs's things are showing up in the, in your working directory.
1:57:34
And Let me see if I want to show you any of these files. Let's look at this Teton file for a second here.
1:57:47
Currently open with Notepad.
1:57:53
OK, the ... file is this trajectory dump file and it's created whenever you do a trajectory run.
1:58:01
And what you see basically are, this is this initial time step, and you could see the, uh, initial latitude and longitude, and the last one we did, started from a height of around.
1:58:19
half of the boundary layer, which was 639, in this case, meters above, above ground level.
1:58:28
And so you could see that, um, as the trajectory proceeds, every hour, it goes out to 68 hours.
1:58:37
Um, and we see the latitude longitude throughout the the run. And we see the height, I was telling you about how the heights, Because kinda changing you see it, it's going up, and then down, and then back up.
1:58:49
Um, then down, and then, and then, and then back up. These kinds of text files are created. Every time you'd run a trajectory, is going to be really helpful. if you want, to. You'll see, in a second, actually, when we're going to use this, but also if you want to export this data to another program, or to a GIS program, or something like that.
1:59:10
OK, let's go on to the next section, and here we're going to do.
1:59:17
We're gonna look at the kind of errors that you can, uh, get.
1:59:22
And one way we're going to check the air, is we're going to run the trajectory forward to the end, and then start at the very end, and run it backward.
1:59:34
And see if we get the same answer.
1:59:37
And if the, the wind field is being integrated, um.
1:59:45
Accurately, with good fidelity, then we should basically get the same answer going forward and, and backward.
1:59:53
And so, the first thing we're going to do is you're going to retrieve that, we're going to go to trajectory, setup run, and we're going to retrieve that trash based control that we had, um, saved before, so if we're going to retrieve.
2:00:13
I'll browse here into my working to Trash base control dot text.
2:00:19
I had saved that before, and I'm going to say, OK.
2:00:25
And when I do that, it had brought in those, those Settings, and, um, we will change starting Hight from 10, 2 600, So we'll start about in the middle of the boundary layer, and since we want to compare this with a backward trajectory, we're going to actually rename this now forward.
2:00:54
Keep them forward that way, we can find this file later, and it'll be identified by this name.
2:01:02
And, um.
2:01:07
I think that's all we need to do, And so I'm going to save this, and as we typically do, I'm going to trajectory, um, Run model.
2:01:18
Now, in this case, we do want to delete, because that had that setup, that advanced setting up of the Boundary Layer.
2:01:29
Interpreting heights as fraction of the Boundary Layer, and we don't want that. So, I'm going to delete the file, and then run. Just want to run a plain old run there.
2:01:36
And we say, Exit, let's see if we get the right answer: Display, trajectory, Execute display.
2:01:45
And, um, see if you get the right answer here. And they should get something like that.
2:01:54
Victory, Schoen.
2:01:56
Yeah, so we got the right answer, If that is the case, if you did, then we do want to save this.
2:02:07
Setup domain, and the exit out of this, quit out of this trajectory display.
2:02:12
We want to save this setup now, save as, um, trash forward.
2:02:19
Control, dot text, or Word.
2:02:24
Control dot text, Um.
2:02:29
And I will save that.
2:02:34
So, now what we're going to do is we're going to try to start a trajectory at the end of, of this and go backwards.
2:02:43
All right, And so what we're going to do is?
2:02:49
we can open up this T dump file, actually, this is when I was actually looking at before it opened with Notepad.
2:02:59
OK, and it was this last line we're looking at here that the trajectory ended up at this latitude and longitude, and you can see there's a small rounding error, I guess, 369 ...
2:03:14
and a height of 216 meters above above ground ladder around level. So we're going to start a trajectory at this latitude and longitude and this height and go backwards and see what happens.
2:03:29
So, now we're going to go and we'll go into the trajectory setup run, and the first thing we have to do is we have to change the time to the end.
2:03:40
And it's so, it's 19 83, September 28th.
2:03:47
UTC 13, ours, we have to set up the, starting location now, to be 38, 665, and point 3 6, 9.
2:04:13
And a height of 216.4 meters above of ground level.
2:04:21
Let's see if I don't have any typos in here, OK, um.
2:04:30
See, give me one second. OK, now, the next thing we want to do is we're going to go backwards.
2:04:37
So, we're going to tell the model to go back.
2:04:41
And when do you do that, it enters a minus sign in front of the total runtime.
2:04:50
And, the other thing we're going to do is, we want to save this now as a different teatime Phi, also call it T dump.
2:04:58
Um, back, we can find it later, and I think that, that is all we need to do.
2:05:08
Um, we changed the starting time, we change the location, we're going backwards and we change the yacht the file name so, it will save that.
2:05:19
Now, we'll do, we'll run the model, they'll exit.
2:05:24
And we'll display the result, trajectory, display, victory, execute display.
2:05:32
And you can see, it started at this ending point and went backward. And it does kinda look like what we had before, but we're going to check that in a second to see how, how well it actually matched.
2:05:43
And you can see down here, the, the starting point, and now the times are going backwards Before it started, and the times are going forwards. So, this is a so-called back trajectory, that people do back trajectories all the time. Suppose you are making a measurement, maybe you were on a boat or something making a measurement from this point, and in the Atlantic Ocean and you and you make something at this time, and you wonder, where did the air come from?
2:06:12
That arrived at my measurement device at this time.
2:06:17
This showed you a back trajectory of where the air came from that arrived at that at that time.
2:06:24
And so, um, if R result matches it, it looks like it did, then, let's save this, make me quit out of here.
2:06:37
And, we'll save this as, we go back to this setup run, save as, Trash, back, control dot text.
2:06:50
So, we can use this again.
2:06:55
Um, so now, let's see if we can compare these.
2:07:00
And, in order to do that. High school has a really nice feature.
2:07:06
that lets you plot more than one trajectory at the same time.
2:07:10
So, we'll go to trajectory, display, trajectory, and it pocket.
2:07:18
The last thing we did was that back, one. So, it thinks that's all we want to do.
2:07:23
But you can do something here.
2:07:25
You can go plus a plus sign, and no, it no space, um, T dump.
2:07:35
Forward.
2:07:37
And so, that was that.
2:07:41
Other trajectory we did going going forward, and let's go ahead and change this. Vertical co-ordinate two meters above ground level.
2:07:49
We've been showing pressure all this time, but let's, let's show the vertical co-ordinate meters above ground level, would be a little bit easier too to interpret, and, um, let me execute display.
2:08:04
And the result you can see, there is so close that you couldn't really hardly even see, um, the red line underneath the blue line if you kinda see a little bit at the beginning here. And a little bit at the end.
2:08:19
But they're really pretty quite close, both in terms of the latitudes and longitudes and also the heights that, um, that the trajectory ran, ran at.
2:08:36
Um.
2:08:40
OK, so this is showing us that, at least in terms of the integration of the wind field, you can go forward and you get an answer, and you can go backward and you can get an answer, and we're not introducing any errors in the way we sort of calculated the wind and how we've integrated the winfield in this, in this trajectory.
2:09:04
All right.
2:09:07
Um.
2:09:11
If we, If we actually, we can look, I won't do this, but you can look at the output file, and you can look at that last point in the output file. You can see it's actually quite close to the initial point.
2:09:26
It ends up being maybe a few kilometers maybe as much as four kilometers away.
2:09:32
That's pretty small air after going three days and going, you know, a thousand kilometers or more in, in distance.
2:09:45
Um, OK, let's go to the next section and I think this is, uh, this is a very important section because, um, the computational err, I think with high split or really any other model, light like this.
2:10:03
It's actually very, very small.
2:10:05
The main source of uncertainty is the meteorological data that you use to drive the model.
2:10:12
And you know, its crest talked about in the last section. There's lots of different meteorological datasets that you can use.
2:10:19
Each one is basically the output of the methodological, whether a weather forecast model, or a weather analysis model, um, but it's a model output, and, you know, just, like any model, it's not going to be perfect.
2:10:32
And, and even if the model was perfect and had the exact perfect answer at the grid points that it was being output at, there would be the sub grid phenomenon. There be, you know, the wind speed, wind direction in-between the grid points that you wouldn't necessarily know perfectly.
2:10:50
Um, what haislip does is it reads in that mythological data set on the grid points.
2:10:57
And in between the grid points it interpolates and just assumes basically some kind of linear interpolation between the grid points both horizontally and vertically.
2:11:08
And so you can imagine, if you have relatively coarse data, you're doing a lot of interpolation.
2:11:14
And you know you don't know in between the grid points again even if they were perfect at the grid points, how accurate with the in between the finer the data you have, you know, the less AAA's you're doing but, you know, in reality, the models aren't even perfect at at the grid points.
2:11:31
And so, what you find is that high split, will faithfully move a trajectory or a plume in the direction at the speed and in the characteristics that the meteorology that you've given it dictate.
2:11:50
And if that meteorology is a little bit off, and again, no, no model is perfect, then the trajectory will be a little bit, a little bit off.
2:11:59
Um, this is not to criticize, you know, weather models, quite difficult.
2:12:05
And they actually doing, you know, quite well, really. We need all things considered, but this is just an important thing. And we're going to demonstrate this now, and in this section, Basically, if you use different metadata. You're going, to do, you're going to get different answers.
2:12:18
And that's to be expected, because the met data really are a little little bit different.
2:12:24
Um.
2:12:25
All right, so what we're going to do, to show this, is we're going to open up the trajectory, setup menu, and, um, we're going to retrieve the trash forward control dot text, and I'll browse to go to my working directory to get trash forward, control dot text.
2:12:50
I'll say OK.
2:12:52
And when we did that, we see we go back to that initial starting time.
2:12:57
We go back to that initial starting Location and heights.
2:13:00
We go back to now, the 68 hours, and it's going forward, um, and we have a, um, name of tea them forward.
2:13:09
In this case, we're going to use a different set of metadata.
2:13:14
And so we're going to use this NARA, the North American Regional Analysis.
2:13:19
So I'm going to call this teed up NAR and in order to use that data, we're going to have to specify this mythological data.
2:13:27
So we're going to clear we're gonna hit the clear button.
2:13:32
And then we're going to add a different mythological file.
2:13:36
And so instead of that, cap decks are 27 UW, we're going to pick excuse me.
2:13:47
Kept text naar dot bin OK, and then our data is similar. It's 32 kilometer as opposed to 27 kilometers. And actually, the ... data was used to initialize the waterfront that that we were using.
2:14:06
So there's actually, you know, a fair amount of similarity.
2:14:09
But one difference is, the star data has data only every three hours, whereas the worst data we were using had data every one hour.
2:14:19
So, right there, you're gonna see, a pretty big difference, because, again, even if the data were perfect, at those three hour grid point values, you're now interpreting, not only in space, but you're also interpreting in time more than you would be doing, without that one hour.
2:14:39
So, let's go ahead and save this, and let's run the model, trajectory, run model, It goes pretty fast.
2:14:49
That's excellent, and let's look at the trajectory, display, trajectory, execute display, and we're gonna, we're gonna look in a second to see how this compares with the other one, Um, I'll quit this and that's actually what we're going to now do is superimpose.
2:15:14
Those two, uh, trajectories using that. same technique.
2:15:21
That we that we use last time.
2:15:23
So, we'll go trajectory, display, trajectory, and we're gonna go, in addition to the ..., we're gonna go plus cheatham forward.
2:15:38
And, um.
2:15:42
Let's see, and let me execute display, OK, and, um, you can see that they are a little bit different, and, you know what, you would, You would expect this.
2:15:55
And, by the way, if you're, uh, graphics, looks a little bit different, I'll say, I mentioned to you, that the, when you make one of these maps, see map, domain, and boundaries are set by, like the first trajectory that it's plodding.
2:16:14
And so, if we had put the t-dap forward first, and then our second, then it would have used, um, the forward one.
2:16:27
In fact, that's I'll show you that here, if I put teed up forward, first, ketones are second.
2:16:41
And execute display.
2:16:43
You see it Now it shows that the fault Teton forward trajectory.
2:16:51
And then also there are there's a trajectory and also the NAR trajectory.
2:16:56
Um, and, you know, one of the things that you can do when you're doing a trick trajectory analysis, you know, one of the sort of a sanity check you can do is you can run your trajectories with different metallurgical datasets and see how similar or different they are, and if they're really different.
2:17:13
It tells you that, you know, at least one of the met datasets may not be representative, and maybe both aren't necessarily very representative.
2:17:21
If you do, you know, two completely different ideological datasets, and you get for very similar answers, chances are better that your result is pretty robust.
2:17:36
OK, let me quit out of this, and let me go back to our, Um, um, No, like I was saying, a lot of it really has to do with, sort of a sub grid phenomena. Like, so, even if the met data were perfect at the grid points, There could be gradients that are happening from hour to hour, minute to minute in between the grid points.
2:18:07
And depending on how, uh, find your grid is, you know, you will either be, um, having less of those sub grid phenomenon, because your grids are closer together, or, or, or not.
2:18:22
And as I said, this is probably the most important source of the trajectory on uncertainty.
2:18:29
All right, let me go to the next section.
2:18:34
So, here.
2:18:37
We're going to actually use a little bit of the Tracer Experiment data.
2:18:42
So, now, as I mentioned, we did this Tracer experiment and we released the tracer from Dayton, Ohio, and we followed it downwind.
2:18:50
And so, we know where it went.
2:18:53
And we're going to use that information to see how well the trajectory computation works, We're going to actually use the fact that, you know that the plume went in a certain direction and we measured it in a certain direction.
2:19:06
And we're going to see how close that that got.
2:19:11
Um, and in order to do that, we're going to actually use one of the flight datasets.
2:19:17
There were ground level measurements, but also flight level level measurements, measurements in flight, and one of the flight's flew at 914 meters above above sea level.
2:19:32
And um, you click on this link here.
2:19:40
You'll see um, the time, latitude and longitude and concentration of Tracer.
2:19:51
And you can see really high concentration.
2:19:56
Or the peak concentration here at this.
2:19:59
At dislocation was the highest concentration of Tracer, and what we're gonna do is we're going to use this location and go back and see if if we can get near the source, maybe if we go forward. And, see if we get to that location, I'm not sure how we do this. All right.
2:20:18
So, um, yeah, this was that, uh, highest, um, concentration, And, we're now going to include this.
2:20:28
Yeah, we are going to go backwards, so don't worry that we're going to go ahead now and open up that trajectory set up run, and, we're going to now start at that location and end time.
2:20:39
Um, so, that time, it turned out was about 10 hours, roughly after that really started.
2:20:48
And so, um, let's see. Chris, really 17.
2:20:55
Sample location list.
2:20:58
0, three On the.
2:21:05
Um, 26, so nice.
2:21:10
September 26, 0 3, we're not going to use the minutes here, doesn't really, um, add too much to the, to the calculation.
2:21:22
And, um, we also have to just change the starting location now, so instead of the the starting location of the Tracer Release, we're going to start at that high concentration.
2:21:35
So it's 41 oh, 9 -82 point 5, two, and the height, it's 914.
2:21:50
Point O. Now, remember.
2:21:54
We had actually set this, this, as Fractured Boundary Layer, we'd actually deleted that, but we're going to have to go go in and change that now. So, we'll say, OK.
2:22:05
Um, before we do that, we wanna go, we're gonna go backwards 10 hours, and you see, when I put the negative sign here, it flipped that to the back.
2:22:16
If I had just 10 here, and then click the Back button, that would add the negative sign in.
2:22:21
So, you can either do either way, either put a negative sign in or or toggle it.
2:22:26
Um, this way, and, um, we're going to call this T dump, O 9, 14 to kind of remind us that we're going backward from this 914 meter flight.
2:22:41
Let's also change that metadata back to the that or University of Washington Boundary Layer Scheme Run.
2:22:49
So I'll clear this and add me to logical files.
2:22:54
I'll go back to that or F 27, U W.
2:23:02
Um, hmm, and let's see if there's anything else we want to do here. I think that is it on in terms of the setup file, but as noted here, we need to do one last change and we have so we will save this.
2:23:18
We'll go to the Advanced menu, configuration setup, trajectory, advanced, configuration setup, trajectory, and we're going to go again to that menu to, that we used before this MSL AGL units.
2:23:35
And this time, we're going to pick, relative to the mean sea level because that 914, um, height was thought to mean sea level, A lot of times when you have aircraft data.
2:23:51
The aircraft usually give elevations and above mean sea level as opposed to above above ground level. And so, we'll say, OK, we'll say save.
2:24:01
And we'll save, And now we're gonna go, go ahead, and we're gonna go ahead and run the trajectory victory, run model.
2:24:14
We are going to use, that's setup file because we did make that change and we run the model.
2:24:22
By the way, when it, when you ran it, you notice at the very top here it says it's using that setup that CFD file. So that means it found it in there and it's using it so that can give you a clue.
2:24:32
Sometimes we do a run, maybe you didn't want to do that, but it tells you that you did, so you might wanna go back.
2:24:39
You might need to delete that file, and do that run again, and, um, let us see what we get, trajectory, display, trajectory, execute, display.
2:24:55
And what we see is a short trajectory, um, starting at the location, where that high concentration was observed and going backwards for 10 hours.
2:25:09
And now we're going to see if that came close to that emission, Are they released site or not?
2:25:16
And, in order to do that, um, we're going to use another technique in the, I split model, instead of, when we got the display trajectory.
2:25:32
Instead of using the plus sign to display two different trajectories, we can actually use a file, and I'll show you how to do that in one second.
2:25:42
First, we're going to create the file, let me open up a notepad instance, and that makes those smaller.
2:25:53
We're going to create a file, and the file simply didn't have to names in it T dump.
2:26:04
Or Word, which was that, um, the trajectory going forward from the release site.
2:26:14
And the next line is going to be this trajectory that we just did, T dump oh 9, 14, and we're going to now save this file.
2:26:25
Um, I'll save as Trash, um, files, dot text. I actually had already done that.
2:26:34
But if you haven't, you're going to be saving it as trash files dot text. You can name it whatever you want, because all we're going to do, is, we're going to use that file name in this display program.
2:26:46
And I already have created it, so Yes, All right.
2:26:50
And now, so, instead of then, these two trajectories, we put the plus sign at the beginning, and we go now trash files, um, Trash, ooops, Sorry, let me get the right spelling.
2:27:09
Isles, dot text.
2:27:12
And, um, let's execute the display.
2:27:18
And, you can kind of see what you see here that, when we went forward, we started at the release location and we went for 68 hours.
2:27:29
And then we went back, where we started at about 10 hours downwind.
2:27:35
And we went backwards and became pretty close, actually back to the measurement, the release site.
2:27:41
So, this is one.
2:27:43
Wait for you to see how well that trajectory might might have done.
2:27:52
Now, of course, the, the release itself was then ended up as a plume of material, that was going at different heights, and it was disbursing horizontally, and vertically down downwind.
2:28:07
And so, no one trajectory, either forward or one trajectory going backward, is not going to be able to capture the entire phenomena of this plume dispersion.
2:28:18
And yet, you know, we started sort of like at the peak concentration, you know, one of the highest concentrations measured. it may go backwards.
2:28:26
We do get approximately to the, to the measurement sight.
2:28:32
Um, you could, Ah, you could use, uh, we have other data that are in the ...
2:28:46
directory At different heights, 12 to 19 leaders Above sea level 50, 2429, 2134. You could do the same kind of thing.
2:28:59
Um, you could find the, uh, no, the highest concentration, and go backwards from that. Looks like nice.
2:29:09
ear, Aye.
2:29:13
Go backwards from, from that, and see how close you get to the beginning. And let's see.
2:29:20
I think that we, I'll show you the result of that.
2:29:28
Here. are we starting at different these different times?
2:29:30
And you can see that.
2:29:35
Not some of the trajectories do kind of go back too, near the source, But this one, the heights actually started at a very much higher height.
2:29:46
That doesn't go back all the way to the source. So what?
2:29:50
What that's telling you is that as the plume is going forward, it was mixing horizontally and also vertically.
2:29:58
And then as if it's mixing it vertically, It started going in different directions, and spreading out, But, um, when you start from somewhere where it had spread out too, you won't necessarily go back to the same place.
2:30:17
Ah, And let's see here.
2:30:22
I think that is all we want to talk about there, and, um, we're not going to really do these exercises, and this workshop.
2:30:36
You're absolutely or completely welcome to To do them.
2:30:42
and here, I guess we are.
2:30:45
The first one is, you know, you can look at, um, different, using different met data, so we showed how, when we use the ...
2:30:54
data, we got a different result than when we use this particular ..., 27 UW Data. And here, we're going to add two other ones to the graphics.
2:31:04
So you basically run it would clear.
2:31:10
Trajectory, setup run, you would clear and then add, no, one.
2:31:16
Do the run and then add another one to the run.
2:31:18
You rename the TDM files so you could get that result later and then at the end you might what you might do is?
2:31:29
Go into this ... files dot text it maybe rename it something else.
2:31:34
It would now put you know these other TDM files that you've done and then you would display the result.
2:31:42
Um another uh, thing that we do here.
2:31:51
Is looking at the Time time-stamp and, again, this is a little bit like we talked about what the difference between the ... data and the worf data that in our data had data. Every three hours, and the Wharf data had that, we were using a data every one hour.
2:32:06
And actually, one of the, uh, datasets has actually dated at 15 minutes.
2:32:15
And so that's a three kilometer dataset.
2:32:17
So it's not only, um, a finer, horizontal, um, resolution of three kilometers rather than 27 kilometers but it also has data every 15 minutes, rather than every, um, every hour.
2:32:38
So again, when we do this, we're going to get a different answer as well, and you can imagine that, you know, when you're starting to use three kilometer data with 15 minute.
2:32:47
You know, in principle, that should be more accurate because you're doing less interpellation, in-between the in-between, the arc, that spatial and the temporal data that you're using.
2:33:02
And by the way, I had mentioned earlier that that high split, well, always.
2:33:10
If you don't do anything special, high split well, pick the time step at any given point in the simulation based on how fast the wind is going.
2:33:21
That's how fast the trajectory is going, and relative to the grid spacing of your metadata, so that you don't leapfrog over grid points.
2:33:29
and you only get, in this case, the default is zero point seventy five of a grid.
2:33:35
That means that if the wind is going pretty high, fast, that the time step can be quite short.
2:33:42
And by the way, that might be a reason. I meet me.
2:33:46
Ah, save and get out of this.
2:33:50
That might be a reason when you go to the trajectory setup menu.
2:33:55
Um.
2:33:59
Oh, nevermind. Vermont.
2:34:01
Well, it when you run several trajectories at once, you can get different answers than when you run trajectories individually.
2:34:11
And that's because if one of the trajectories is going to up like a much higher height, and the wind speed is much greater at a higher height, then the model will choose a time step, which is much lower, and it will do all the trajectories at that lower time step, and you'll get a slightly different answer for even those lower trajectories.
2:34:38
Then you would have had you run that trajectory all by all by itself.
2:34:43
The other thing I want to show you here is, in the Advanced configuration trajectory, you can set the time step to be fixed, and that's that first menu here, and that's actually what you would you'd be doing in this exercise.
2:35:01
You can set the value, um, to be, I think, in this case, to actually telling you to set to 15 minutes.
2:35:11
Um, the maximum time step that you really can use is an hour, and the minimum time step that you put that you can use is a minute.
2:35:22
And by the way, this was that zero point seventy five of a grid square that you actually could change that, if you wanted to make it more restrictive or less restrictive.
2:35:33
If you wanted to. I'm gonna quit.
2:35:36
I'm not going to do that, and let me quit out of out of that.
2:35:44
Let's see, this in this last one.
2:35:52
We're looking at, again, adding different metadata to the.
2:36:00
To the analysis and doing trajectories with different with different metadata. And I don't think it doesn't add too much, I think, to this.
2:36:10
Alright, um, we are getting to the end of this section.
2:36:14
and, um, we'll be going to, to lunch break soon. See if I'd like to add anything else to this.
2:36:29
To this.
2:36:32
Section here, let me look a little bit in the, um, working directory, and let's look at this message file.
2:36:39
It's sort of from the last one that we ran.
2:36:42
I'm not sure exactly what we're going to see.
2:36:45
Forgot which one we ran last.
2:36:48
Um, open a note pad.
2:36:51
Um, See here.
2:36:55
So the bigger.
2:37:00
And we'll go down.
2:37:05
The beginning is showing you, actually, these are showing you those variables that would be set in the advanced settings.
2:37:14
And if you didn't have any advanced settings, no, it'll just give you all the defaults.
2:37:20
But if you did sent something, then it will, It will show you that, I think we did one thing, might have been the K M S, L.
2:37:28
Yeah, I think we did that.
2:37:30
one thing we changed, zero would be above ground level.
2:37:37
one is above mean sea level, and two, is the fraction of the boundary layer.
2:37:43
But I think everything else is the would be the defaults, and show you a little bit more here when you're doing these runs.
2:37:58
In this case, you can see it was actually picking about a 15 minute, um, time step.
2:38:04
In this case, I think this was that short run that we did of 10, 10 hours going backwards.
2:38:10
Um, and actually then it became to be an hour time step, and yeah, 10 hours went back, 600 minutes.
2:38:22
Um.
2:38:24
And you can see these the message files are a little bit cryptic.
2:38:31
As you get to use high split more and more are gonna start to recognize things. You know, it's, you know, this line is saying that it was reading in metadata of this ...
2:38:40
data at different times. And in fact, I split always reads in data at two different times.
2:38:49
Um, or it has data from two different times, because it needs to sort of interpolate between the times, like, our 17 and 18. So as it was doing this hour, run, it needed data for those two hours.
2:39:02
Um, this number is a weird number.
2:39:05
This is the number of minutes since, I think, in 1900.
2:39:10
This is actually the internal time unit, in High Split. But here is the actual.
2:39:15
You know, the, um, the actual date here, 19 to 3 September, July 25th 18, UTC.
2:39:26
Um.
2:39:30
And by the way, the message file is something, when you ever have a problem with high split.
2:39:36
And if you, for example, if you even direct during the this workshop, or later on, and if you do ask us a question, in the forum, often what we'll do is we'll ask you to send us, see the control file, the setup dot C F G file, Andy, message file, because those with those three files, we can kind of get an idea of what happened during your run. And we can try to try to reproduce the mistake.
2:40:04
And while we're doing that, let's look at second one of the control files, um, open with Notepad.
2:40:15
This is what a control file looks like, and you'll actually recognize a lot of these things. Well, pretty much everything from the setup file.
2:40:25
Here, let me trajectory set up, run.
2:40:28
What you see is that, we set this starting time, and that's that first line, the number of locations, Um, then the, uh, latitude, longitude, and height of the location, the number of hours that it's running, um, forward and back.
2:40:51
I think that's that I think that that's that, that's the line there. I think it might be.
2:40:58
The height of the of the model.
2:41:01
Oh, no, this is sorry. This is the vertical of input model model Data, which I didn't talk about very much.
2:41:09
We'll talk about that.
2:41:10
In subsequent sections of the default is that if the model, if the meteorological data that you're using has vertical wins in it. That you use those wins.
2:41:23
Some beta logical datasets do not have those vertical wins, but there are other ways to estimate the vertical whens. And so actually, there are these other choices that you can use.
2:41:35
And also other, um, vertical motions schemes that you maybe you want to always go on, a constant pressure, which might be the case if you were flying a balloon, Um, know, you might not have meteorological vertical winds. But you can estimate them from the divergence in the horizontal.
2:41:56
Um, wind.
2:41:58
Anyways, we'll look at that a little bit later, But generally, you use the model input data.
2:42:05
And, by the way, if you do select this, and if your model doesn't have it, it will figure out, you know, a way to get the vertical motion, I won't just leave you hanging.
2:42:18
Um, um, and then here we're seeing the number of met data files.
2:42:27
Here received the directory of the where the met data file is, the name of the metadata file, the directory of the output file which is just a working directory, and then the name of the output file. And again, this is a little bit cryptic.
2:42:44
Right now, you wouldn't, necessarily, if you looked at one of these files, know everything about it, but, as you start to use the high split more and more, you start to recognize, you know, you can see, no, if there's some weird typo here, you know, you'll see, Oh, gosh, that can't be right, You know, the name of the file is not, it's not that. So, you can often see.
2:43:07
Little errors that, that might, that might crop up.
2:43:11
All right?
2:43:13
I think, we will, um, stop now and go to A, um, lunch break, which, I'm sorry for, for us here on the East Coast. It's, it's lunchtime and, I guess, from where you are.
2:43:28
It may not be lunchtime, probably isn't lunched. We're going to take a 60 minute break, and we'll come back in one hour.
2:43:37
And so, until then, enjoy your break, and I hope to see you in in an hour.
2:43:57
Little button on my microphone that got pushed by mistake, pardon me? Well, welcome back.
2:44:02
Um, and I wanted to mentioned one thing that we're basically going into the next section of the, of the workshop with trajectory options.
2:44:15
And before I, we'll go into that, I want to mention one thing.
2:44:19
At the end of the section that I was giving, we were switching back and forth between fraction of the Boundary layer.
2:44:27
or above mean sea level or above ground level in terms of the Heights units. And at one point when we were doing that, I think the 914 above mean sea level.
2:44:39
See, we had the graphical user interface at indicated in the heights field that we were entering that.
2:44:46
It was still fractured boundary layer, essentially the, the graphical user interface had gotten behind or confused or didn't get updated. And so the simulation worked because we had set it properly in the setup. In the advanced menu. But, so, basically, what the lesson there are. Sometimes Sometimes you can reset the graphical user interface.
2:45:08
And that's probably a good thing to do, because every once in awhile it can get a little behind on the settings at your, that you're doing.
2:45:16
All right, without further ado, um, I'd like to turn the D presenting duties over to .... Doctor ...
2:45:28
and ..., I'm going to make you the presenter, you will get prompt in a second.
2:45:39
Yeah, and when you're ready, ..., take it away.
2:45:45
Yes, I'm trying to Can you see my screen?
2:45:53
Yes, we can.
2:45:54
We see Earth screen OK.
2:45:59
Pop, yeah, **** There we go. Now we see the screen that you want I think.
2:46:04
Yes, oh yes, OK, great, thank you rock.
2:46:07
Sure Um See, OK, Hi everyone. I'm going to take you through Section five of the tutorial.
2:46:23
The first topic is about a vertical motion in a pure resection.
2:46:30
Marc has already explained to us they how the model how to model calculus trajectories.
2:46:41
The three-dimensional winfield, I would cry and to do that horizontal when you and me and then the vertical wind taboo.
2:46:49
Usually the input may touch meteorological data can provide a vertical wind field. If not available, high speed calculates it internally from the rusty divide, but last he divergence. That is not as good as using the vertical velocity directly from the Metallurgical Data File, but it can give you pretty much similar ourselves.
2:47:20
In some cases, we want to we compute a vertical velocity based on the assumption that the apostle is transported on, some odd surfaces for sample, choosing the ISO barrick option.
2:47:38
So the equation here, add up becomes P, pressure.
2:47:45
The motto computes the vertical velocity required to keep a trajectory traveling on a constant pressure level.
2:47:54
Then you can simple, simply replace pressure P with like potential temperature for an ISO ISO tropic trajectory or, um, the density replaced at hub with density in this equation, that is an ad, so picnic trajectory, husband uses touring, following, so-called sick, I'm a co-ordinate system. So, for choosing, I, so, I so sigma, the trajectories stay a long been mapped, sigma suffers.
2:48:31
So the question is like, Um, Well, why? Why would we Why would we want to use one of these options? For instance, following up but learn you may want the trajectory using the ISO picnic option because the balloon might stay on a constant density surface and that example you slay following an aircraft.
2:48:58
The trajectory may stay on a constant pressure surface, Mmm hmm.
2:49:04
And then another situation is, in some of the ...
2:49:10
matos, as the space show resolution goes down, the matatu off, the vertical velocities may go up.
2:49:22
So when you have high spatial resolution ..., you may not, at the same time, have high temporal resolution met mechanical field. So this kind of like inconsistency may cause the output and the sampling depth to the field.
2:49:41
So let's go to this tutorial. I will use this graphic Chae to try to illustrate a situation.
2:49:52
On the left, we have a relatively coarse grid spatial resolution that on average in the grid it is showing downward motion.
2:50:04
On the right for a higher resolution grid, let's say a quarter of that best solution, you may have one of these grids.
2:50:16
one of those slight high resolution quit sow's, showing clouds and strong upward motions, But, later, a drawing grid sow's have somewhat, right.
2:50:27
We cut down what motion, So, the overall average being like this, so, if the output interval is not coming across in a sufficient frequency, that will be a problem.
2:50:41
So, this situation may cause you to want to use a different approach to pivot a vertical but ..., but of course, the better approach would be actually have consistent spatial and temporal resolution in the ... data.
2:51:01
So, let us go back here.
2:51:05
So it will be doing a bunch of backward trajectory, cut calculations, to show the effect of changing the vertical motion methought. We will do that, That will calculate, I had mentioned that. That is the exercise we we had earlier in Section 4.5.
2:51:25
So let's go to the hospital GUI. Click the reset first and then bring up the trajectory setup run window.
2:51:41
And then let's see.
2:51:46
We can set up all the information lie here, but less to less less use.
2:51:55
A file so call, try back on the ... dot T X T, you can find it.
2:52:07
You can find it in the tutorial Directory and in the files sub directory.
2:52:16
And then click OK, so, we load it.
2:52:20
Um, hmm.
2:52:21
The backward trajectory calculation information, um, and then let's save the output.
2:52:30
Name two, TI dum, underly Theta.
2:52:37
And then, Save Change.
2:52:42
Dang I save it, save, just make sure, and then Random Model.
2:52:50
And then less display trajectory, so the input ...
2:52:58
teat dumb, underlie data.
2:53:01
And then let's do six hour interval.
2:53:06
Then, let's meet her meters AGL, then click Execute, but, so this is the backward trajectory. So now we will repeat the steps.
2:53:22
But Conficker simulation using a different vertical motion max dot so the option is here. Click select. You'll see a bunch of options available for you to choose.
2:53:40
So we just date.
2:53:42
OK, let's follow here this we just did the simulation using the Vertical Velocity field in the Metallurgical Data File.
2:53:51
Now we are, we're going to do ISO barrick trajectory, which means it will, the trajectory will, will stay on a constant pressure surface, click OK, and then remember to change the file name too, ISO B, and then Kick safe, The Random Model, That is the second simulation we do, bring up the setup manual again. Now let's do.
2:54:26
Ayesha, Tropic, trajectory, OK, and then change the file name output file name, Save Random model.
2:54:44
The fourth one is, but recall, my foot number, three constant density.
2:54:59
Yeah, and hence.
2:55:03
Model.
2:55:08
And then the fifth one is, is I so sick them up, Option number four.
2:55:24
And then the, um, last one we will do is from diversion. The vertical motion is compared up by, but a co integration of the velocity divergence.
2:55:42
Gee, save and run the model.
2:55:49
OK, if you Oh, Your working directory, Dan, you will see less Hurricane Directory.
2:56:03
And Yo will see odd is TI dum underscore.
2:56:08
I'll put trajectory output files, So for this, hey, we want to use the Super Impose feature.
2:56:20
So for doing that, we would need a file listing, order, trajectory, output, file names. So, let's create a file here.
2:56:31
In the working directory, caught it, try underly.
2:56:41
Files dot TXT.
2:56:45
So, it's just a text file, so you can edit it with the notepad. So, and then, you put all this, trajectory, output, files, names, on this text file, so I would just copy it.
2:57:00
Fun, tutorial.
2:57:06
So, um, so we have 1006 trajectories simulations.
2:57:14
And we will be putting one more, that is the forward trajectory.
2:57:21
Oh, we will probably do it later. Tests, just to odd, is six trajectory first.
2:57:28
Save this file, and then bring up the display window, and correct, OK, put duh.
2:57:43
That's, try, check.
2:57:46
Try the lie files dot TXT.
2:57:52
And then click Execute it.
2:57:55
And then, so you would see all these trajectories, six tracked trajectories display in this figure.
2:58:07
Um.
2:58:10
So, it's hard to see which you switch, and then, so let us look at the, um, yeah, so in the tutorial, we can see that all of that backward treacheries, having quite different paths, that two trajectories actually They are using the data from WTH, a call file, and also to divergence calculation.
2:58:48
So these two options, the chorus, so add us as somehow somehow somewhere else.
2:58:58
So slightly, a slight changes in the vertical motion may quite may make quite a difference in the endpoint.
2:59:07
So, now less, As I mentioned before, last spring, the open that try and the light files detests TXT file again, and then put one more file kita underlie.
2:59:23
I think, yes, default, by default, one re, generate it in the previous section, save it, and then execute.
2:59:36
So, you can see that I trade the forward.
2:59:41
Trajectory, actually match very well with da, the backward using the vertical velocity data from the ... data file.
2:59:56
So, less, go to the next section, close this.
3:00:07
Um, the next section, we're going to look at them metrical Data to exam.
3:00:14
What kind of flow field may have larger uncertainties for the trajectory calculation, because in the periods section, you see all this, like, Oh, only to network trajectory, actually close to each other, and then ought to forge trajectory, even using different but co team methods that they end up somewhere else.
3:00:42
Still, we want to exam the mythological data to see what kind of like uncertainty causing the tragic tree calculations so different.
3:00:56
The wolf meteoric Hall file use in our trajectory calculations does not really have enough spatial coverage out to the arm, Atlantic.
3:01:11
So the cut off of the data is really just like to reduce file size but that's not do a good job to to draw contours. So instead of using the the wolf data file, we will use the global data to examine the flow field for this case. So the cobol data will give you a sense, like, what WCAG cook audition out there, where the starting location?
3:01:44
Yeah, OK. Over here where the starting location of the pebble trajectory, we want to look at this area.
3:01:52
So let's close the window here, and then.
3:02:01
And then that's open, Open them, then ...
3:02:07
called Meta Archie Tab, and then Display Data Contour Map.
3:02:18
And then, go to select the file, um, in the tutorials directory, kept exited sub directory, is car kept text to, Gee, B, L, R?
3:02:37
Has that been, and then select, um, Heights fiel.
3:02:45
This is the height of the pressure surfaces, and then we want to see, not at the surface, yeah, I think we will do, um, the time offset.
3:03:07
Thomas: 48, O R two they downstream.
3:03:15
And then, we'll just look at the one time step and then set the map.
3:03:21
Send her to 40 Pawns 0, and then minus 80.
3:03:31
And then set the radius too, 25 degree.
3:03:42
OK, also the Level we want to look at Level four, which will be about 815 minute bar.
3:03:50
And then Run con Con tool.
3:03:54
So here, this is the card on me.
3:03:59
Yeah, not a window.
3:04:04
And then let us go back to our trajectory, trajectory, plats.
3:04:11
And you can see that, um, those bet what trajectories that we just computed from dislocation here south of the Long Island.
3:04:27
You can see that in this figure the flow layer here, the flow.
3:04:34
The fall through is liddy ill defined, so you can make a slight change in height.
3:04:41
And you can end up with a total, totally different trajectory direction. That is the difference.
3:04:48
That is the difference between running a forward trajectory, from this area, in the mid-west, which took us around the high pressure system.
3:05:02
Versus the pepper treachery, we're now in this ill defined area.
3:05:08
So this, another, this is another consideration, in terms of, like, looking at the potential flow errors and trajectory errors.
3:05:19
Um, hmm.
3:05:23
Yeah, the vertical and horizontal resolution of the metadata and the township data matter, so as the temporal resolution.
3:05:31
So this graphic shows gridded mythological dataset with.
3:05:40
Oh, let me go back. Sorry.
3:05:43
I'm trying to talk about this graphic.
3:05:49
So this graphic show out created ... dataset with this little green pestis they, so they are great points. So, at time T A low pressure system is sitting over here.
3:06:06
Then, the next time step at T plus, delta T, the low pressure system moved to dislocations. So when the ...
3:06:14
Random model interprets between these two time steps, adding sing error may occur. So so you need to have a consistent between the temporal and safe Spatial Resolution.
3:06:29
That's **** of this sexual we are trying to highlight, OK?
3:06:40
And then we are going to look at the trajectories and terrain.
3:06:47
So ... is another factor contributing to the uncertainties of the trajectory calculation. So, in this exercise, we will configure high speed to run using ISO Barack.
3:07:02
Vertical motion methought, at the location of just, some, some lach touring area that will foster trajectory to intersect the graph. That's what we want, what we want to see in this exercise. And then, we want to see like, how that will contribute the uncertainty in the tragic to a calculation.
3:07:27
So, we will do a trajectory.
3:07:34
To come by the westerly fall, going through the Rocky Mountains, let us also hit the reset and then go to the trajectory setup manual.
3:07:48
Let us look at the Yeah.
3:07:52
So, we will due September first Siraj ...
3:07:58
UTC and then, OK said the location Who won and then?
3:08:11
Let's copy this.
3:08:19
So, this is our starting location.
3:08:23
Then, let's do 72 hours.
3:08:31
Forward trajectory, Select the vertical motion methought, number one.
3:08:42
I saw barrick, and then save the output us, the filename to have them.
3:08:52
Then we will be using a global dataset.
3:09:02
That is, in the, in your working directory.
3:09:05
It's call.
3:09:09
O P.
3:09:13
19, 83, 0, 9.
3:09:24
See that here.
3:09:30
Then, we can go to the tutorial, try to find a file.
3:09:50
..., I think it's an ...
3:09:52
directory, OK, thank you, Mark, Sure, this is very helpful, thank you.
3:09:59
So, we'll use this global, um, data file, And he's safe.
3:10:13
And then, let's also go to the Advanced Menu, Configuration Setup, trajectory, and then we'll open Menu Number six.
3:10:29
So, um, This manual actually can give us like HR variables in the trajectory output.
3:10:45
So, we want for now, at the couple of options here potential temperature, Precipitation, mixing height. Relative humidity are these options are available for you to choose for your application. But not for now. We want to have deterring height in the alpha file.
3:11:06
Silk Cook Safe.
3:11:10
And save. Less run, the model.
3:11:15
And then you will have this, uh, map window pop up, and then run using Excel File.
3:11:24
OK, and then less display.
3:11:33
OK, one more thing we need to do is, um, display, So let's close it and then Display here.
3:11:52
And click Display.
3:11:55
OK, here we go.
3:11:57
This figure shows the, the West to East Treachery and, uh, Rocky Mountain is shown in the black lie. So you can see that This is Touring Hide.
3:12:14
Showing black lie. Once the elevation has been raised to this point, the trajectory continues on the ISO baddeck surface.
3:12:28
So, next, we're going, so, this is forward trajectory.
3:12:31
So, next, we will do the backward trajectory, um, similar thing you have all written half done, like, in the previous section. So, we need to open the M on file.
3:12:46
The F dum.
3:12:52
Yes.
3:13:00
So you can use N Notepad to open it.
3:13:06
So in this endpoints, trajectory output file, you can see we have now the last two columns, the pressure and the Turing height.
3:13:18
So let's go to the end of duh, and on file, and then you will see, this is the position.
3:13:30
So, Chad Pepper Church, who was that, ah, at September fourth, C was zero UTC, and then the System, Let Location, and then the height.
3:13:45
So you can copy this, and then, let's bring up.
3:13:52
Trajectories App Manual.
3:14:01
See, we'll see all your D C, And then the Location, Copy this.
3:14:20
OK?
3:14:23
And then the duration na'vi come Um, Back work 42 hours.
3:14:31
And then save the output file too.
3:14:37
Be dumb.
3:14:40
Thence, use hyperbaric.
3:14:45
And ... is still the same.
3:14:48
So safe, and then us go to Advanced trajectories App menu.
3:14:59
Make sure we have the terrain, height option turned on, because we want the touring hide in the in the in the output file.
3:15:11
Save, and then run the model.
3:15:18
So let us do the display.
3:15:23
Then we will use here. Put two files to get at the F thumb with the earlier, and then the plus beta.
3:15:36
Execute.
3:15:38
So we don't we don't get to we don't get the complete overlap at the End Pawns.
3:15:47
After bet my trajectory, in terms of capturing the position on the follow a trajectory, you can see that the calculation match much quite Wow, right matches up like perfectly up to this point of the terrain.
3:16:03
And, but that will, but once we continue past the terrain, has, we cannot, cannot know, like, cannot know that this, like forward trajectory, was forced upward. So so the backward trajectory just continue on the ISO bouts of its surface.
3:16:22
So, now I can see that laid the intersection with ... actually is a point where the trajectory, computation loosest information and no longer become reversible. So this is how, we understand the terrain and trajectory.
3:16:47
O K, plus closed days, peiris, simulations, and then the next.
3:16:59
So, in the, in the, in the following sections today, we will look at different methods to do multiple treachery fall that as his purposes.
3:17:10
So you can, yeah, you can run the model multiple times with different starting times and put them together, put them, like, put either individually trajectory files together. In some cases, we want that. But we can also do multiple trajectories in time in the same computation.
3:17:34
So, that's, I'm going to show like this section. We're going to see how to do that.
3:17:41
So, let's click the reset button for us.
3:17:46
The reason we had doing, like, correct, the reset reset button so often is because like we are doing the tutorial so that we have working on different examples, but in a reality, like usually, you work on one particular problem, You would not, like jumping back and forth between different things, so your puppy will not need to click the reset button this often. But for now, we need to make sure I'll be reset order thing, let auto things in the manual.
3:18:23
And then let's do the trajectory.
3:18:28
Again, the setup menu.
3:18:34
And then we will look at we'll bring will load a chai forward control dot T X T file.
3:18:45
Try for Work.
3:18:49
So, this is control file puppy, save it earlier, or you can use the one in the Tutorial Files Directory, OK.
3:19:10
And then, just double check the starting time and the starting height and then click Save, and then go to the Events Manual trajectory.
3:19:27
And then click ah, Menu number three, So, here, so, we will be start, we will set the restart interfere interval 2 to 1, so that means we will start a new trajectory for our, from the starting point.
3:19:48
And then set the restart duration to four.
3:19:56
So, this, yes, this means that we are going to restart a trajectory as the 7 17 UTC days, the first first time step and then 18, 19, and 20 UTC. So, four hours duration.
3:20:17
So, actually, there will be four, um, trajectories. So we're going to do Font for Level, we'll adjust, to one level, shall live at 2, 0.
3:20:32
Hit Save, Save.
3:20:36
And run the model.
3:20:45
And display.
3:20:55
So, here is the result.
3:20:56
So, you can see that there are four trajectories one hour apart.
3:21:04
Chaser, released late. This is kept tax example. So, we, the ... period from 17 to 20 UTC here.
3:21:14
So each successive trajectory is further south, and west, indicating that changes to the flow field, doing this for our worship period for hours, could account for at least some of the chat chase up surf over Pennsylvania.
3:21:34
Yeah. So a single trajectory level and Tom is sufficient to capture is is insufficient. I can see that it's insufficient to capture the plume sculptures so we can come by multiple starting highs with starting trajectory in tap. So how to do that.
3:21:56
Close this and then quit and then start with the same trajectory setup, plus, bring it up.
3:22:06
again.
3:22:09
The same trajectory set up.
3:22:12
But we want to have three starting location.
3:22:23
So, let us look at the tutorial.
3:22:30
So, that's three starting occasions.
3:22:34
The same location for Ohio Day Daytime Ohio.
3:22:39
with but with three different heights.
3:22:46
So, 30 and I upon my allotted 2 -84 22 lambda to the first one ad, 10 meters height, known as carpet, to the second occasion and felt location, but then changed the height.
3:23:05
6, 600.
3:23:10
and 2000 meters, 1200 meter. Sorry.
3:23:20
And then click OK.
3:23:24
Save.
3:23:26
And then open the offense menu for ... Directory.
3:23:33
Number three, doing them multiple trajectories in time.
3:23:39
For now, for this runway, we set the interval 2 6.
3:23:47
So, start a new trajectory, F is six hours, and then the duration, too, 24, start new trajectory trajectories for the full 24 hours. And then level we have 3, 3 levels.
3:24:05
So this means F is six hours.
3:24:08
We will start three trajectories, We're going to do this for 24 hours and have three laptops.
3:24:16
three heights, save, safe.
3:24:26
So, and then, let's run the model.
3:24:38
And then, it's play.
3:24:46
OK, so instead of starting three news trajectories at six hours at the initial starting location, new trajectory as data at, at the end of each of our previous trajectories.
3:25:04
So you can see that, like, the trajectories starts every six hours, but only going out for 24 hours, then six hours later, we start a new sets off trajectory Ed.
3:25:21
This location um, in space like this downwind, not just at 1, 1 height of the meat bounder layer, but but also we have, the trajectory starting at the top of the boundary layer. Green lies on the screen lies at the surface or this red lies.
3:25:45
So, so we can understand this as up, as a trajectory, represent representation of mixing that occurs during the daytime transport. With with this trajectory analysis, we can have a sense of like of the speed and irrational variation of the flow at different levels and us as going downstream.
3:26:14
This accounts for Tracer Distribution and the single tragic to Teri was not like it was not sufficient to show.
3:26:25
This.
3:26:25
We saw ourselves shows that the surface tracer concentration distribution could be exploring in, in basically quite a lot light by the change, by the changing wind direction, with height and time within the boundary layer.
3:26:49
So, so, if we look at the multiple trajectory in time, and then we next section, we will be looking at the multiple trajectories in space. That is another application for the trajectory calculation.
3:27:07
This may take us like to help us to, to not waste a good location, to start the trajectory for your analysis.
3:27:17
If best thing about, like, if we can have, imagine if we can have tragic to remap all over your domain, and then auto trajectory, a lot of trajectory over your domain, that we may be able to get a better representation of what's going on in the domain.
3:27:36
And you can two days late by creating a control file in a certain way fall for this purpose.
3:27:44
So, let's see how we can do that.
3:27:52
OK, again, click the reset button and then bring up the Trajectory Setup menu.
3:28:02
We are going to do forward trajectory.
3:28:06
So last loader, try underlie forward and ally control, that tasty file that is basically the information we want.
3:28:18
Starting trajectory, starting at September 25th, 17, UTC.
3:28:25
And then for the location, we're going to do three locations. I mentioned that this control file will be set up in a certain way to do multiple trajectories in space. So the location, the three location.
3:28:44
The first one, actually distributed occasions will be representing the grid.
3:28:51
We want you to trajectory multiple trajectories in space. The first location.
3:28:59
It's ops.
3:29:06
The first location is the lower left corner of the map of the of the domain you'll want to do your trajectory, so.
3:29:20
So, here, 38 on Sira, pounds. zero.
3:29:28
Starting high.
3:29:30
six hundreds, 600 meters.
3:29:36
Then, the second location is the upper right corner, so, I will just copy it in sap typing.
3:29:46
The same height, 600 meters, then the third location.
3:29:52
Just like you can be any coupons, like inside your box, inside your domain, but now we will be using a, uh, agree ****, right?
3:30:07
one degree of the what is the offset of the lower left corner?
3:30:24
So, click OK. And then we'll set the duration. Now for six hours. Just shot trajectory, six up.
3:30:34
Save and Run.
3:30:36
Motto.
3:30:39
Let's display it.
3:30:46
So you can see three locations.
3:30:49
Um, oh, it was set three locations in the control file.
3:30:58
And then, and then you you, you're seeing light, this thicker that we had three trajectories.
3:31:06
So the red one is the lower left corner, and other blue one is the upper right corner.
3:31:12
And then the green one is like is the quick **** one degree offset of the lower left corner.
3:31:20
So that is the basic setup sister domain. And then we, oh, great. We want to to do our multiple trajectories in space.
3:31:31
So now go to, To create metrics off grid of a magic grids of trajectory.
3:31:41
We go to trajectory, special space, a special runs, match checks.
3:31:51
Then you will have this popup window asking you, If we want to do it, 10. Just hit Continue.
3:32:00
And then the run is done.
3:32:02
Hit display trajectory.
3:32:07
Execute, OK, So, we had three locations in the control file, but the special runs, launches, preprocessing, step.
3:32:19
That actually converts straight, OK, Starting location into a control file with 77 starting locations.
3:32:29
So less, as tried to open the control file with the note hats.
3:32:38
Paul Graham.
3:32:40
So, odd is 77 location, so this is generated by uh, pre processor.
3:32:52
So this is the results show.
3:32:54
You can see dad, um, yeah, you basically, with this, this run, you can have visual picture of what to fill with.
3:33:07
Be like sample that the day is strong fall to that Knopf Wes up, north-west of this region, north-west of the region, and dies sort of like weak flow, in this area. So we started our trajectory as 600 meters above ground level.
3:33:37
You'll notice this, this the height we show here that's pressure.
3:33:43
So obvious that terrain is different in all three locations the starting location.
3:33:49
So, we had we are going to be starting at slightly different Heights, if, if we turn on the meters HCL for the display.
3:34:05
Now, you can see that in this figure, Arda trajectory start at 600 meters above the ground level.
3:34:20
OK, so this is how we do multiple trajectory eastes in space.
3:34:27
Um, the next section we're going to see how to do trajectories with me.
3:34:40
Touch a mythological great zambo, let me close this windows.
3:34:51
Um, this is another approach that may help us to understand the source of trajectory error, um, due to the ... RG.
3:35:08
And, then, we, so, you may, you may do the trajectory computation for a point that is maybe a little bit like north or south of the location of your interests. Or, maybe you can create a box of four points about that location, and then if like after 48 hours, 72 hours of the transport, like your trajectory, audi's trajectory, if it's still stay like a box. That means like you have confidence in your trajectory, The uncertainties puppies, Not a lot.
3:35:47
But, if light, you have strong radiance, all strong winds schiele or the ill defined region like the pyrus examples. So, then your trajectory will be very much like, put he divergent.
3:36:06
So, in high speed we, this methought has been implemented like directly in the intuitive code so, that trajectories are automatically computed about three-dimensional cube.
3:36:22
Let me bring up that picture, It's something like this.
3:36:26
So, that this three-dimensional cube center about the starting location, you set up in the control file.
3:36:36
So, when you select a starting location and hospital create a cube.
3:36:43
And then, so, there's that would be 27 location, because that night, her face. And then we have a face upward and downward.
3:36:56
So when normally you do the calculation that as starting location, using the metallurgical battle at at, at that point, that's normal practice.
3:37:08
But with this, we have tiny with ... them.
3:37:14
The you do the calculation using the meteorological data at one of these 27 locations.
3:37:22
So essentially, you will have 27 different possible possibilities for a solution.
3:37:31
So, if if the gradient and the ...
3:37:35
coalfields wet fairly weak, then you will you know that we would get very similar solutions that would not be a hole lot of divergence in the computation.
3:37:51
But if the gradient will show or isa ill defined region, then probably see it very diverse, diverse light aversion trajectories.
3:38:04
So let's see how we can do this, click the reset button, and then let's retreat, uh, default trajectory, Pack's file, cheat.
3:38:24
Less load a tri underlie forward underlie, control dot T X T file, OK, safe.
3:38:35
And then bring up the events menu on the main menu, the trajectory, spatial, special ones, ensemble.
3:38:54
Click OK. Continue.
3:39:09
Can, let's do the display.
3:39:24
Yes. This is the fire trajectory using the ... Grit and assemble methought, so you can see that.
3:39:35
Um, even even a good **** offsets in the ... is very similar to what we have.
3:39:45
When we were doing at the trajectory from multiple ... code data sets that was done, like in the previous sections.
3:39:57
So, this is the forward trajectory. So, let us do the same thing, but with backward trajectory.
3:40:08
Plus, bring up the, so, the manual.
3:40:12
And then load, try back control dot TXT file.
3:40:24
So now we are doing that backward trajectory.
3:40:28
Then the same thing goes to the trajectory, special rands ensemble and continue.
3:40:47
And then display the result.
3:40:58
OK, let me bring this window two here.
3:41:05
So what you are seeing on the left a step backward trajectories using the ... ensemble methought.
3:41:14
And then another on the right is the forward trajectories.
3:41:19
So, um, so as I mentioned earlier, doing that back works You get quiet divergence.
3:41:32
Will you we look at this case earlier, that the field in in this region in the off coast, up the in the in in the Atlantic. So this is ill defined region.
3:41:48
So you get a lot more uncertainties in the backward trajectories considering just those trajectory not caught in the anti-psychotic saturation day. So you're an artist.
3:42:07
Trajectories there is considerable variation in the upwind region.
3:42:14
So, I think that's the end of this section. Let's see. Do I miss anything?
3:42:26
Yeah, so, this concludes the Section five on the exercise, at the end of this section, that you can do, but, for now, I'm not going through this exercise, but these are good exercise.
3:42:43
That can help you to get more in the intersection, in this section, and learn more about our top brand. Has to be trajectory for different applications.
3:42:58
Yes.
3:42:59
Now I would turn it back to Mark.
3:43:04
OK, thank you very much, ..., really excellent and we're going to now go into A 15 minute break and we will come back at 15 minutes past the hour for the next section.
3:43:20
Let me stop sharing.
3:43:23
Give me one second here, OK?
3:43:37
OK, we are even we don't know the the ..., you can have 90 though. The summed up your kitchen for delta one.
3:43:50
Though, for those that we have three of the kitchen and that that is on the FTEs that the kids are not fully cover these guys about the. They can discuss the rhythm of both, though, but how to utilize those information.
3:44:09
Oh.
3:44:13
Again, oh, how weak?
3:44:16
What kind of kicked Katie, if we needed this kind of application, The user actually start the animation?
3:44:23
But this morning, Margaret showed the example of the cafeteria campaign and so we have one source location and go about the fifth location for the netbook measurements. And, if we need to, the giant is kind of pretty much where you have a kind of a third location. And we wanted to design the how to.
3:44:52
distributed generation for the most efficient way of capturing the information you need, is why there will be some general information of a flow path. Because of this kind of experimental work, the ...
3:45:12
take a lot over the surface.
3:45:13
So, you can You cannot simply do that.
3:45:19
And then, go to ...
3:45:22
kind of impediment to you need to design it in a very the rest the way and they need it.
3:45:30
It takes more than four years, you know, the math Oh oh. We need to know the why they would be the general information about the flow, between the information about the gender upload from desktop climatological data, in, this uptick in, what we really need to do is, we will use this one month simulation.
3:45:57
For the everyday every six hours, and then we have a multi ... handle. We've made the senators, say some general information out of that one.
3:46:09
So, that's our basic goal of all this location.
3:46:15
And, let the therapist some sure, simulation something. Oh, well. We've started from the thread of three, for the simulation we did this morning. And if you have said that the controller fat as Mark mentioned, that this morning, though you can find that the tragic the Trinity overcome the controller that TFP.
3:46:44
Well, you didn't say that, that the Control Fire, you can be creative.
3:46:51
From the tutorial five from here.
3:46:56
You can go through that.
3:46:59
And some fires.
3:47:03
Then, then, there's even the twitch board couture.
3:47:10
And we can look at that information. And, actually, it's not really complicated. So, you can type in the information from.
3:47:23
This one, we need to change the system information and but at least the one, we will use a six hour simulation for four months Over 83, September Start the from the timber? Both zero.
3:47:42
And so, we already know the roof kasia the date on. Hey, yeah. I'm really sorry to interrupt but I wanted to mention to people that, if you retrieved a file from the files directory, the directory structure of where you put your meteorology may not be the same.
3:48:02
So in this case, it was assuming that you had deer tutorial in the same directory tree, as your high split file. And that may not be true.
3:48:14
So the one thing that might not work if you, if you bring this file in, is the directory of the metadata. So you may have to clear that and add that.
3:48:25
Yes, taking the ferry.
3:48:28
Yeah, it started from this one, MW to update that book. But definitely we will use a different than the theoretical gift by Soviet and to updating this one.
3:48:40
So, for the time we have a 48, 48 hours.
3:48:45
And then we're also changing the alpha five for the app to five and then we really do the difference the met data and you can see that that's net data file from the tutorial, kept texts and then we will use the IP 19 83 timber data.
3:49:16
So, the phasor?
3:49:19
That thing we need and though we will say myth Actually where you can one Can then why simulation with you? When you run the matter, but what we will do is we will do a market or simulation.
3:49:39
And cold dead one, We're gonna go to the pressure loss, and then we choose the daily.
3:49:47
Then you can see the start of the simulation at the 83, or nine from the first day, to 2 0, time. And for every six hour. And the accurate, the default number is 30 days.
3:50:06
But if we run for 30 days because we are just using just the one terabyte good day for one month, so we don't have October data gets the alleviate the problem for the simulation. So, we will use the 28th day.
3:50:30
And zero hours after start, though, when we get the script, we start to learn the multi person relation 1 by 1, Let's do that.
3:50:45
You can see there are large of them.
3:50:48
five, being generative.
3:50:57
The veterinary care co-ordination is a time. And you can see now, though, though, that's tough. The tough the really happened.
3:51:08
Oh, hi, for the working directory and we can see that they don't pass generative, so basically defeat after fail-over the Python files.
3:51:22
Oh, we have they don't die.
3:51:29
OK, so basically, the Kindle Fire for the one trajectory in time and for the 48 hour.
3:51:45
We have a lot of a trajectory by extended in this K.
3:51:51
Oh!
3:51:59
Actually, we can say that this.
3:52:02
Control five, so.
3:52:14
Just any case for the nephews.
3:52:27
Uh, prompted a vertical pipe, then we can actually play a long and productive prey offset, we can go to the frequency. And it is really common practice. For the kind of a hyper vigilance we, if we need to put a fire at the input for the certain staff, we owe a lot of cases, we use the generated, ask if that has the reef though, all the five though.
3:53:07
In our case, yeah, you can, From this option, we can generated, the, reached over 3, 3, 5, be just the generic teeth.
3:53:19
So, we can, when we click the create file over the file name and we can see the, the using, the based name for the F dump is actually regenerate the gunfire. And when you click it, what is the only data that it lists out? The older, the file name starting with something that I've done.
3:53:46
For the Using for the Why the Cut Stuff, ...
3:53:52
first it says it in five is a creative, and if we check the Working theta three, boom, we can see in the file itself, that's OK too to see what's inside. So, whenever you generate, the ... inside, and used a lot of K 15, period, so, whenever you use demanded the Earth, it inspires you about a 10th of what to eat, that. Either way, it's a very different testing entity from unnecessary finite included in the list, then the output can be different from this here. We can see it a lot of bias and that's what we really Just the general gift from the timber for the ... and defend.
3:54:47
But we have a September 28th 18, though we have a full credit for it the 28th day.
3:55:01
Oh, and then, ha, bit can skip display.
3:55:10
So, basically, this is a picture of distribution of frequency.
3:55:17
Because we have the 4 times 28 days in total 125.
3:55:24
Oh, 5, 3, 3 Oh, it's a counting number of trajectories for each grid tariff, how manual that 33 is the path to buy, and then divided by the total number of threads, if the highest number could be the 100, ah.
3:55:46
So, when you did the plot that, when you check, the maximum number is really 100%, though, at the location, on August fifth, though, we're in all the cases, If they really think through it, though, the, where it's starting from, the location. So, the counting all the trajectories are 100%. And the other locations, you can see the thumb.
3:56:15
Oh, something like the over the Tempo tend to work around ..., Folks can see the fifth dimension of both of those pre-conceived path.
3:56:28
With the phone, these both shown with, we can't change it, though.
3:56:35
The resolution of a great resolution, we are using the high Priest that we are using the treasure through a thing that the trajectory has both cheaper end points and the, in the alternate. Then, we are using the 48 hour simulation, and for every hour, it has defined the, but one trajectory here, the 48th endpoint.
3:57:03
So basically what the data is counting the number of N deposed point, though, if the ... think that, that it's increased death. And the two options are to co create. But when you calculate those end points for each, If there, if we have a, uh.
3:57:28
And you end the point in their career, that we just counted as a one.
3:57:32
And then if the flow is very slow, we can have a multiple endpoints inside that, that, each increase there.
3:57:42
But in that case, we can calculate the depth path, and the voltage for that one hour option needed. So, We can change is the residence.
3:57:55
Time has a yes, then we have a, oh, slightly different shape over the kinds of plots.
3:58:05
And, when you compare them with our previous, Well, this scenario, but it may be different, but the biggest gift differently when you see the maximum number of the pizza, uh, over 400%, then these.
3:58:23
For each agreed that we really counted if the total number over the end, the point. Then means is the kind of checking though how slow the actual flow is. testing the decrease death and the stoke code, the residence time weight to the division of the frequency.
3:58:47
So, we can use both of the information.
3:58:51
Oh, just the for the case, and, for this, I finished inaccurate the, just the, counting, the number of end, the point to end.
3:59:06
In this case, we're handling those, and put it just, just the quantitative, then, that's not really right.
3:59:18
Depending on the of Peter, about flow and the stage of a grid resolution, there can be some mismatch If we try the reader with higher resolution of the, Great.
3:59:33
Oh, we can see a slightly different pattern, and they, If we get the really higher resolution.
3:59:45
The Internet thing is that there are some, totally the pet sounds. So, that means that the flow is a testing through that, that location, they didn't kick clearly, and those are the really high at the end that curve it, in a one hour. You can pass through the thumb ..., though. It is not the only connect, here the end.
4:00:10
The next hour, it can pass to the next degree, if they're so. Cool.
4:00:15
To me, my, that problem, though, you can really, we have the option to change it.
4:00:21
Those are great resolution so that we can get them the nearly optimized for the or the dirt.
4:00:33
Oh.
4:00:38
This one is, for example, to get the frequency of Omega three led to eighth.
4:00:49
And then we then move to go Crossroads Narrative.
4:00:55
Oh oh, four, the cluster. Another thing, if it's some kind of a very good way of summarizing information for that, to get the general path to the end.
4:01:11
What do you really do in-depth analysis of the ...? We make us them grow well with the trajectory that can represent, where, each cluster. So, oh.
4:01:28
If someone asked how many patterns of gender appropriate sending the location for a certain season of the year, we can, we can expect some kind of video or prototype TensorFlow, Python, Though, some kind of, when he dove, flip everything, does some of the campaign, we can ... catching the thing. I know the wildly the dominant pattern in that location and that path be done.
4:02:05
Let's check what the really doing, that kind of thing, in case Oh, oh, what do you do, either?
4:02:13
But we have a lot the trajectories and the full.
4:02:17
For each treasure three, they're slightly different, though from we can calculate how those different threads.
4:02:28
If there is a kind of a spread it out, and they'll be the sum for the patient through the variation, we can calculate the spatial variation over.
4:02:43
It's a trajectory, but what we really do is we, our goal is that we really want to minimize the patient videos, but also, we want to get some good, the present number of the cluster of, oh, oh.
4:03:06
If you have 1, 2, or three clusters, showing the dominant type of the flow patterns, that the cool thing to youth. But if we have something, we have something like a more than 10 cluster of that.
4:03:22
Yearly, we can update that we have some good come in packs.
4:03:27
So our goal is that we really want to minimize a special area, but also we want to kind of minimize number of cross. But basically what we want to do for ... and uh.
4:03:43
To do that though, we, we need to calculate .... Fisher variance from the hope that it's clusters from the. ... from the cluster. And then we have the ... at the, at the end. We have tipped me, the total space a very tough but that is also the ... or the ... area. We will check, the, total, ...
4:04:21
entity is, and what they really wanted to do is, we went to see the, how, with them, some minimum values, and the, with, the core of, the, getting the small number of a cloth.
4:04:37
Oh, oh. No.
4:04:42
Let's go back and.
4:04:47
Something we need to know, when you calculate either for a specific area, we need to know how those, those, a number of the changing the shear is a very temporary ....
4:05:01
Trajectories.
4:05:02
But actually, these are the you can see the line saying, hey, if Dave, the pre-k. system trajectory, where, luckily, we can estimate that there are two types of alternative.
4:05:13
It's the one either going to the north-east north-west and then another one is a boon to the thumb some work today if the direct shot. So when you start co create special events between them, and we begin to minimize the group down and the mochi into the bars.
4:05:36
But when you find a very predatory, there is a very close. We can go to that, and then we kept the Calculating. The fish are very at both.
4:05:49
Oh, prove it.
4:05:53
If we have with some of the main threat, The threat of trade corridor sharing, what we can see, though.
4:06:02
The actual Tisha variant of the really small, when you calculate the insight that this trajectory and the ... of the clusters. .... within this group, the ... really small. And then, we minimize those number over the last spring. And at the end, for me.
4:06:26
The first breath, before, the, just the 1.
4:06:31
1 cluster, But it did some kind of, if we average the gift of shim: one and the shame to, at the end, we have the ... one tool that these are actually not really providing the good information.
4:06:48
Uh, folder, the actual, the flow path.
4:06:52
Though, in this case, when we calculate, especially that anything from the one and the same two.
4:06:59
Oh, if the really changing to the changing, to the different locations, And then it's a .... There are getting bigger book processing, We really see that while we calculate this F N. So? when you target from the, though.
4:07:22
Uh.
4:07:26
When you started, especially for the OK, if the ... really high. And the way when you begin to minimize it merges the ... close to each other, then the ... reality creating. And then, at the end, we have just the pure cluster threat.
4:07:51
For a finisher variant is the ... begin to increase. But why did you really need to do is to catch the point when the total variance is begin to increase very quickly.
4:08:04
So, when you get the point O leak, we can be, though, that, either from the path to optimize a case for the thing number, oh, we're still really can do this case, but for the kind of thing or narrative that journey with you.
4:08:29
We will review the creative three simulations we used for the frequency analysis, and although they're gone, we view, both, took a patient alone, and clustering, understand that to some options that we can use. And the trajectory, that we, as the 48 hours simulation, and the Louisville, for the youth, that information. And, they can specify the time into parent at 23 to skip.
4:09:07
But, we will not change that option, though, oh, the first option, we can change, our clustering is that we have multiple trajectories up to 48 hour.
4:09:20
And we can tie that up to which our how many and the quality we used ah oh from the toe of the trajectory.
4:09:33
And then be able to put forward. Therefore, the end point for further, actually, that means that the directory that fails, if they find out OK.
4:09:49
And the, oh, we either change the working directory, and also.
4:10:02
That's the point that we need to specify is, actually, the wrong We have a working directory, and the folder of the clustering, we can specify. the different location for the cluster. will keep only. But also, we can now use a thin A working directory, though vehicle theft here, and then again, we need to generate that, the so-called in five file that has it the risk.
4:10:33
Trajectory fire, So we will catch up all the OFDM ...
4:10:43
Then we can turn it off.
4:10:46
in fact Oh, Let's see We have you in the restive inside?
4:11:02
So, we can check the where it has the right the rest over.
4:11:10
... fires, we used the previous case.
4:11:18
And then we click the low cluster analysis.
4:11:24
The K, in this case, the depth, the cluster calculation was done because we took the using, the one hand the trajectory, but some cases we can use the more tragic stories, but sometimes it can be more than a thousand files.
4:11:49
2000 were phased out and then it nearly takes time.
4:11:55
It goes up when you click If the ...
4:11:58
Why The Genetic David, The old It, co-created all The case is oh.
4:12:08
The beginning from the individual trajectories into mulch, and it has all the history of what though the creation of this process. It really takes time and.
4:12:23
people, we have a fifth of the direct either the any kind of processing can be done pretty easily.
4:12:31
And, so, once we have a depth calculation that are, We can see the aha.
4:12:40
The fourth patient variance, How the total area changes. So, one is the speed therefore, the number of a cursed. Oh, we can see the total pressure very app and our goal is.
4:12:59
We need to catch the sudden increase over picture variance for this K, so when you see that this graph that you can see, the number of the, cross, the number five, from the 5 to 4 vote, there is a fee increase. So each one of our candidate to be used for the cloud thing number and the another candidate detail, when you see the change between the correct number, 3, 2, 3, 2, 2, Also, big change is still.
4:13:33
Currently, number three can be another candidate did, but they didn't know clear solution for it.
4:13:43
I will miss the reality.
4:13:46
Recommend, oh, don't we?
4:13:49
The what kind of a thing, number we can try. And then another inflammation we can utilize that there is neither. The comfort of the criterion, and you can see the ....
4:14:03
When you click this one with a 30, it generated some numbers.
4:14:08
So actually, what it, that's it.
4:14:12
It's just showing the thumb.
4:14:14
That's the number of a cluster number we can utilize. And you can see the there is a thumping number of standing for, the crafting five and the clerk things today. Just 73.
4:14:26
So it can be another candidate, and that, actually, those numbers don't change from the previous the changes for the number to the left one.
4:14:38
Actually, do, When, you see the, I'll put We can.
4:14:51
Thank you.
4:14:52
From the working directory, we can see the delta P Delta P tissue five deadly demanded that stayed for a special very honest. And the actually for this graphic, we are just a huge, showing the number for the low number of clothing number, but actually it has.
4:15:17
The ... rated the number for the number of a cluster of the next to it. For the very high across the number one.
4:15:27
Hit some high number if and when the theta with open the door to change its peak, it really decreasing and at the end, they begin to increase again, and from, whereas, for example, for this plus 5 into 4, or we can see the tourist fisher variant, calculate the number, is changing from the 16 to 16.4 to 33, all nine, and.
4:16:06
This number, we can see just the percentage of a change. If and when we have with the high number, 10 needed, some southern change. So, that means that we can use.
4:16:20
Did this thing number as a candidate for the cluster number?
4:16:26
Oh, well, We really don't need to youth.
4:16:30
It does this thing number and we can easily check that that number from the diff graph you go. It's much easier to use and get the information and that this, the TIF, the change.
4:16:45
We have a discount for the group proof the five.
4:16:50
Then we output the number of clusters, five here, and click the Lung.
4:17:00
both time.
4:17:01
And then, again, quickly, it had released over, cluster, is actually what the daily show user, oh.
4:17:13
It is just showing the full list of a threat actor is regenerative.
4:17:16
And, at the first column, you can see, there is this cluster number, If the, from the 1 2, but, he, well, five, but then, myth. It's time to close the number for its trajectory.
4:17:34
Then, on the next table, we can check the output when we clean, didn't really show the cluster traffic, to me all the five a cluster.
4:17:51
So, when you say this one, we have Fannie, 29% of their three points.
4:18:04
So, north-east direction and another, Next, time in the flow pattern, I need to going if the world and around the 27% of the profit in supporting their book. Though, when he had with this kind of information.
4:18:25
So, this means that, we use the information from the timber and the teeth that show the ...
4:18:36
in this location for September, the PICO that this kind of ... can change by Suzanne.
4:18:45
So, when you compare this one, the spatial distribution over, that the cap tech heavy month, the call, depending on if, therefore, they're very well designed to people. We have the flow pattern to give the gift, or the north-east and we have our, most of the.
4:19:11
observation, though, that's the work located in that location.
4:19:17
So that way, this came onto the ... and there was still this kind of spinning entity Giant, we can utilize inflammation.
4:19:29
Now, another thing, what? We can try, that actually, we can change the number of clusters may be, because we have another can do the for the B, and the run again.
4:19:42
Then it can also, oh, what are the things the case?
4:19:50
Oh, if we want to the cluster group, the thread is 3 into 3 groups.
4:19:58
We have a kind of a paradox in this case, what, when you use the five cloth director, we have two types and subtypes that. That. That was done over. them was emerging in this here. So in this case, where we have three groups of permanent types, if the cemetery is inflammation.
4:20:24
But the bit different, depending on the choice of the cross number, though, when you have with this kind of a notice and when ready to prepare, the kind of communication from both the come on kitchen is, or how you define, what is the reason why we say that, if the number of clusters, and for those cases, you really can show the.
4:20:59
If the change of the ...
4:21:00
change, then, uh, this is why we choose the cloth number, and that that can be a very good nation for the work, that choice.
4:21:17
So, that is helpful to customize this.
4:21:27
And further that crossing the occasion I already covered.
4:21:37
The next application of those chiro cation, and the in this case, we have using different ..., a real use, the word of the tragic this simulation and alleviate.
4:21:54
But, what if this is the case? When we have a thumb, not sure of division?
4:21:59
And we, if we don't know the location of the source, we can look at the estimated whether it neither best suggestion for that. That's the first case. And, Oh.
4:22:20
To do that.
4:22:26
Let's take a picture of data. We really use this case. What we know is, we know the time over the list, but we don't know the location of the store.
4:22:40
Then, we have our observation, then, means we have a location and the time or what the debtor, though, we know the location and time of the fifth Thursday, and we know the kind of over the third, but that don't know the rotation of the rotation of the Earth, and therefore, this application, we will use peptic data.
4:23:12
OK, we will use the data, that they're, actually, it is the Formatting, the Data format.
4:23:20
And the, basically, it has, well, they implement idea, that, the giant to provide the power of the measurement information. And actually, it has our own information, for, the time, the starting time, and the duration, and the location over the texture. And the amount of, imagine a month, and in the rest of Colombia, the station number, actually.
4:23:49
It took the for the in-depth thing, and where, so 40 from the station, we have different ... information.
4:23:59
Oh, let's try that.
4:24:08
With you?
4:24:15
Yay!
4:24:16
Better than the review.
4:24:21
Q: The same.
4:24:29
Very forward.
4:24:42
They took the information.
4:24:45
We have was 13 time and a 13 kesha.
4:24:49
Actually, this certain location, we don't know.
4:24:54
And duration, Change it.
4:25:00
Oh, put.
4:25:01
Good. Good to go.
4:25:07
Yeah.
4:25:08
And then we need to state why we set this one is we need to keep an inflammation of the cutting time for the key rotation or flip case on.
4:25:25
And then we go to the based on them, zero cation.
4:25:34
As I mentioned that we utilize that, that kept test measure months, we need to find the date of birth.
4:25:48
Of K And also, we can start Chiro where you actually from your data set, ...
4:25:59
occasional indication If the sub TO use the location, Imagine a month, from the non zero number, and from here, you can see the general Imagine a month, and then some actually met a month.
4:26:17
But we if we have a thumb, actually met a month it, it knows the pebble trajectory.
4:26:25
And, uh, this, zero value means we're going to zero, well, robot. Actually, we can increase this number if the data to some kind of, uncertainties, dementia, month end date, that is, just the noise is something that the fourth one is meaningless, tell you, We can remove them.
4:26:44
So you can change this number, But for the application, we will just use a 0, 0 number for the first thoughts.
4:27:00
Then we will use the order number that is not zero, and here we can set thing the, how many people say they have to redo the youth, and we can cut a user beginning medium, and the end of the day.
4:27:21
Imagine what that means. These are all these observations.
4:27:26
For this case, it had this information about the starting time, and the duration is a three hour.
4:27:35
So for this one, we can rarely do the three F tragically, uh, hmm.
4:27:44
The beginning from the 18th and then it had, though.
4:27:50
three hours later, and then the immediate over, that duration.
4:27:55
Another, some other cases, we have another baby, The observation deck has a duration of the hour, then the, the timing for this trade, if they were the only two hours.
4:28:11
So, from this observation data, that we have actually the world from the 19 ... measurement and you don't see the effect trajectories from the eternal death, the beginning, middle, and end.
4:28:28
So, at the end, we even have a 592, Devlin trajectory.
4:28:40
So what we do is based on the observation data and the basic information from the control file, I just mentioned about the, we will do the 1497 simulation. And the in this case, it generate the older, controlled fire in decades.
4:29:08
And when I click the data, it can be checked.
4:29:14
Working directory picture you can see data really a rug over control. five generative from the 2 0 1 little bit to see what's inside.
4:29:32
And then, uh.
4:29:35
So, it's a refined, the trajectory of the simulation. And for this one, it had the time of the measurement and the location of the measurement.
4:29:48
Then it has a negative word today today, because A, is it doing?
4:29:54
It will do the, the threads if the creation, and that's the way we have negative number here and why wait for our, because at the beginning, we gave with the wasn't compatible with that chemical delays.
4:30:10
And we needed, that time, the difference between that time and the, the.
4:30:21
And at the time of the Imagine a month that the different shouldn't be used to refer to the trajectory. So that's why we have the number here, and those.
4:30:34
For the case, Though, we have a different control five.
4:30:40
And then we go to this tip speed.
4:30:46
Then, it begin to generate.
4:30:50
Sure.
4:30:54
Terrific, the output, Paretic, check.
4:30:58
Just generating and edit, check, though.
4:31:02
OK, so, where are we, We are using the ... for the tragedy of Biased and you can see, we have a lot of new to them. five Generative Thatcher. did they still another system fire and always want to see?
4:31:22
Boom.
4:31:26
Because it is to being generative.
4:31:31
So, now you send it to the almost 100 trajectories.
4:31:39
Keep generating, though, what they really do here is, we generate the data, whatever, the trajectory.
4:31:48
And then, at the end, though, even though the display utility, if we use the for the frequency narrative, and the show, though, did the revision of the Oh, oh.
4:32:03
Of the directories, and When you have, From the each measurement point, we are using that trajectory. And, basically, it is a point that to the source location?
4:32:18
Oh, oh, although, the, if the 597, 3, 3?
4:32:27
We can now have with the reactor, OK, seven, both, the, the, towards the ...
4:32:36
crowd that you can hear, the potential case. That's what they want to see.
4:32:46
Linear ..., with.
4:33:00
Oh, when I took the row, the video that, that, that role and the last year, he wanted to say about the co creation time and ....
4:33:18
For this case, I use the word three point, the beginning, and then the middle point, but if we use that as the middle point, is the number over.
4:33:30
This is a trajectory.
4:33:32
Oh, to the ones, you really can pull, begged the calculation time, and also if you increase the value of their thing, and they move without them, none. Or very small number, it also can deduce the total number of the simulation needed, then you can also save time.
4:33:58
But in this case, how to create tutorials done.
4:34:03
So all the calculations of the only company. Conflicted.
4:34:08
Hmm.
4:34:12
And, we can display this output. What this means.
4:34:18
We have almost 600 ... generated in this folder.
4:34:27
And then we can display that output.
4:34:31
But, again, we do frequently, and here we need to generate that insight, OK.
4:34:39
And here we are using the Git thing for the right path, and then click here.
4:34:49
To be well worth 250.
4:34:55
So, we have, we just have this narrative, then, let's check.
4:35:02
So.
4:35:09
It's a full list of our current trajectory than the rest.
4:35:16
I'm actually just moved and tested, but hopefully, towards a different trajectory, It has the chicken and the trick that we used, and also those at the number of, either. It is the name of a patient from the observation, and so it had that, the timing a little above to Asia, but you can generate some kind of a unique ID path.
4:35:45
Though, we can get the Trump month, the Whew!
4:36:07
OK, so, this is what we can have a ... application. So, what we did was we used that, that, the death of division and the ... trajectory from the location of the detection, and then, we'll go forward trajectory, is going back to the location and the most populated locations around the here, the height, the value comes here. And thank you to share a TIF Paramount.
4:36:43
The reason was the prompt him intake, Tom.
4:36:47
Uh, so, it doesn't provide a very ... to lift a very good suggestion of the potential ..., so, this is what we can use.
4:37:04
Yeah.
4:37:09
To get the, uh.
4:37:13
This is just a kitchen, how? We utilize fifth of February three, with the observation that took care of the case.
4:37:31
And here, this is another example, or the, uh, 10 meter village height, Let's check the simulation, because we target Rondo original simulation and the village height, the width of that 600 meter.
4:37:53
And, uh, but they're not sure height of measurement put the the near the therapist.
4:38:01
And maybe tomorrow could be the more number of the when we confer the output.
4:38:13
It really didn't change that much so.
4:38:18
So basically, that means the core of the reserve could it be that it would be robust and we have a kind of a suggestion for the source to go cation.
4:38:38
Well, OK.
4:38:44
And we have ...
4:38:47
and the ..., What these guys want to do is, well, we haven't that.
4:39:03
But basically, what we did, the words of buffalo kept at the pediment.
4:39:10
If we did this kind of, the, uh, the perfect candidate fit in this region and the location, they've done a lot of the total for the location of the data that is based on that, though, If it's personal and met with network, and it will be: D, the delay is the, the, the, the chemical leads from the different location. In this case, if we used the very in Canada.
4:39:49
Why did it'd be the difference?
4:39:53
Well, they can feed in our, from the gift of the cluster analysis. We use the data on the phone vocation and if we use the different location for the clustering.
4:40:11
Although with this, the thing is not the case.
4:40:14
If it did really been at this trading augment the highest frequency is really going to save and it's the Confederate dead one.
4:40:26
Sure, you need to look at.
4:40:40
Kept to experiment, OK, This one, OK.
4:40:48
So, the point is, it related to chemicals in different location. There is a less frequent, though. That's the.
4:41:01
It left.
4:41:02
The, uh, likely perfect.
4:41:06
passing through the Thatcher location, although the detecting, uh, the third fifth network.
4:41:16
This.
4:41:18
or this timber, it can just nearly, though, just the part of the testing through the north of the detention location. So that means it can be designed in. This way there is a high risk.
4:41:35
Very highly scaleable misting the actual big enough.
4:41:40
So that the way of the the campaign.
4:41:43
for the third, the third location of, uh, think that they've done for the rotation and that that's that's how the paper capture that if they're going to predict where and if we choose a different location, it could mean there is a system.
4:42:01
It could have been a friend that kept the stig enough and it also same folder.
4:42:18
The theme of the third thing I notice is, different than the previous one was just the ..., and if we wanted to capture the weird taking, her father, had a month. We have a lot of working side. These are just the in this location, but there is high probability, the probability.
4:42:47
There is a high chance of coming in the flow pattern, the quantitative way, though, it can. It could have a myth that extra big enough.
4:42:59
That does some lessons we can learn from this kind of publication. And basically, this shows why we need a different kind of application to provide the downright inflammation of which ended up flow peptides.
4:43:23
And this one shows that example.
4:43:33
Oh, another thing I want to say here is a booth.
4:43:40
Let's just say the, our working directory, and we chatted a lot of work.
4:43:48
There's actually, you can see that the simplified entity than five, laugh and laugh.
4:44:03
At the end, then, we need to creating our working directory.
4:44:06
So the first one of the things that we want to save that, the output, uh, well, actually, we can just the Earth.
4:44:16
They move, delete the whole directory about the collective. you want to say, that this output, and the another property that really the cleanup.
4:44:26
And the start of ..., The indicates with youth death, clean, both so on.
4:44:35
And you can see the power a working directory name was changed, if we did some.
4:44:44
Some random number, and when you go back to here.
4:44:50
So we have another working directory, very clearly in-law, and our previous one, that was a move to other working directory, though.
4:45:01
High speed is catching the information from the work, from the point of a final word, from the data file. And, Keith, if you can mix, you can use a thumb information, you really don't want to use and to avoid the kind of K thief.
4:45:22
one could practice theater.
4:45:27
I think this Working directory to the different name, and this test, again, and when they use this option, you can see that that is, it is really clean directory.
4:45:41
And when these tags, again, you can see to immediately generate the more five believe, that the first five has been done, and then these types of games. So it's a more powerful way of the 13 aptitude mutation in a very clean way, actually to Pope, would vote for content, the reset button.
4:46:07
And?
4:46:12
Oh, today, I display the introduces the way over for using the market portfolio today. I hope for, And they actually go, the cluster analysis is a wonderful program.
4:46:27
The popular way of presenting the general pattern with the general flow patterns, and, uh, that's just the, just equipment for a couple minutes, to see locally, what.
4:46:47
But the actually, get from the close down another thing.
4:46:53
Can I, I mentioned about the when the, when you do the analysis, the thumb, something, we can see it, from the working directory.
4:47:05
we can see you size the quota.
4:47:11
She control, it's not not contribute to sick control, and basically it has the information over.
4:47:22
Told the crowd to consider a firewall the killer thing.
4:47:25
So we use the 48 hour today, three for the clustering, and it has that information inside. And also, some information we'll want to check later is that we can see the cluster underqualified.
4:47:42
But it's another ASCII file.
4:47:46
And what you can see from here, either from, from this part, we have a full list of a thread of three files to be used for the clustering. And they had the number at the first column. You can see the numbers for the 1 2, 3 oh, five.
4:48:05
That means it's close the number.
4:48:09
So, whatever days when you are reading ... number, and when you're reading the file name, you can see the which finally. really. Oh, thank you. To the which cluster number, when you combine the information we've got out there.
4:48:28
K for these crosscutting we're just using the information for the wind, the fear and the photo, the trajectory.
4:48:36
But if we want to see some other variables for each cross there, how they change is the one you need to do is check that this information and, uh, get the information for the different file, and they usually hazard the kind of information, though with this kind of information and the credit information, you can use today that the information.
4:49:05
Then you can go the calculate to the sum different number.
4:49:11
I mean, you can use this information to get the ..., the theoretical condition, for each cluster.
4:49:20
What is this temperature? So, for those different clusters, it had a totally different level of interaction. And then, these synaptic, conditioned, for those are different closely. It's a totally different.
4:49:36
Then, usually you'd have a very, uh, clearly a distinct difference.
4:49:43
The difference is in the kinetic condition, depending on that, though, the Synaptic Pattern, though, when you want to summarize that kind of information, you need to check if the file and that was kind of trajectory to the crest number.
4:49:59
And this, you can see the similar five.
4:50:02
Oh, before I did the 3 to 5 cross number, the fifth one is alpha for the ... number, and this stage of another output.
4:50:18
Because I also did the cloth number using that study cluster, this is the output for a decade.
4:50:25
So, we have a lot of Alpha Phi Alpha here, and the checking out that this information is a really good book for the better narrative.
4:50:40
So, I think that's it for today, and.
4:50:46
Right, back to you.
4:50:49
OK, Thank you very much, ..., really interesting, and I'm going to change the presenter back to myself here.
4:50:57
And, let's see, sharing.
4:51:02
Oops.
4:51:05
one second.
4:51:14
OK, there we go, OK.
4:51:19
Um, so, um, I'm going to give a short wrap up of, of what we've done today, and it's been a long day, depending on what time zone you're in, it might be early in the day or very late in the day.
4:51:37
Let's see.
4:51:37
The first thing I wanted to talk about briefly was the ... data, but we started out with and the key point here I think, is that high split is driven by just Gritted meteorological data.
4:51:50
It's required to run high split, and it has to be this R L X format, this high split format.
4:51:58
um, and as we as we learned, we have conversion programs to convert most met model output into that format, and we. We do that conversion for you, in many cases, and so we. you know, we have a whole website full of this met data.
4:52:11
that you can that you can download when you want, but if you do create your own that data, you can convert it yourself as well.
4:52:20
I'll note.
4:52:21
We didn't talk about this, but there's also an inline version of high split that's within the wharf model that can also be done.
4:52:30
So like I mentioned, we have a lot of datasets that that that we have available to you.
4:52:36
We have some datasets for the CONUS, The continental United States, I'm usually goes a little further, but also some like three kilometers, 12 kilometers 27 kilometers.
4:52:47
We have some global grids, like, for example, the quarter degree, global forecast system data.
4:52:53
And the important thing I think is that it's high split interpolates both spatially and temporally to estimate the metallurgical variables at any given point.
4:53:03
So, you know, you can imagine for a trajectory, you need the wind speed, wind direction, you know, vertically and horizontally.
4:53:13
And when you have a gridded data set, you need to be able to estimate those values in-between the grid points.
4:53:21
one thing we didn't talk about very much is that you actually can define multiple grid grids in a high split simulation. And this is often done. We may show some examples later on in the workshop.
4:53:35
But lot of times, you might have a high resolution mythological dataset near your source or that you get your brand or maybe even just a high resolution data set over the continental United States.
4:53:48
But you want armada along trajectory or a long concentration run.
4:53:52
Um, that might go past that what'll happen is if you if you don't have multiple ecological grids and if the what's the trajectory or the concentration plume goes past, the grid will literally stop.
4:54:09
So, you don't get any information past that, but what you can do is you can define, like, uh, a local grid, or a regional grid, or a national grid, and then add into your simulation, a global grid.
4:54:22
What will happen then, is it will use the finer grid while it's it has that data.
4:54:28
But once it goes off that grid, it will then pick up on this, start, the larger grid, the course or grid, the global grid. And in that case and high school, it's actually pretty smart. It.
4:54:38
It figures out where it is and says what's the best meteorological grid that I can use at that location at that time? What's the finest grid?
4:54:48
Um, and you try to tries to use that, that grid.
4:54:53
So, then, finally, this question about complex terrain.
4:54:58
You know, if you're in the mountains or if you're, you know, a lot of, uh, Where the, where there's a lot of complexity. Clearly you.
4:55:08
No, it's gonna be difficult for any met data to really capture those phenomena, and you know, clearly if you're using, you know, like a 27 kilometer dataset, or even a larger dataset, And if the mountains, you know, are a couple of kilometers and part, and there's a big variation in elevation, It's going to be difficult, too, To model that.
4:55:32
And, again, high speed, it can only do what the met data tells it to do.
4:55:37
So, it's not, no, again, it's not a criticism of the media logical models, But it's also not a failure of high split if, if you don't give it the right wind direction, you know, to, to use.
4:55:51
Um, let's see. So, you know, the question then becomes, what data should you use? And here, I've listed a whole bunch of the ... datasets that we have.
4:56:00
You can download, and you can see, we have a lot of data sets in the continental United States, at different resolutions in different different time periods, and then these global datasets as well with different resolution and different time periods.
4:56:15
And one of the things I mentioned at the very beginning is that when you're running the model on your own computer, at your local computer, you need to download these data.
4:56:25
So one of the issues is going to be how big the datasets are.
4:56:30
And so here I have the total size of the datasets for one month.
4:56:34
You know, you might say to yourself I want to use this three kilometer data.
4:56:38
But you'd need 390 gigabytes, you know, for one month, which may be fine for you.
4:56:44
Um, um, maybe for that same month, you just can use the Global Analysis data of 2.5 degrees.
4:56:53
It's only 100, 110 megabytes you know for one month, so it depends on the analysis you're doing and it depends on the kind of answers you're trying to to achieve.
4:57:05
And as we mentioned earlier also that it's sometimes good to run at least for a couple of different datasets to see the kind of variation before we did the example of the 27 kilometer and the 32 kilometer nor.
4:57:18
And showed that there was even differences, you know, between, between that.
4:57:22
Um, here's the domain of this that we're 27 kilometer dataset. It's sort of bigger than the continental United States.
4:57:30
Um, and, no.
4:57:33
Like I said, there's not one right answer.
4:57:36
one nice thing about this 127 columnar data set over over the United States is that it's a It's true risk analysis dataset that we've gone in and recalculated the met model based on observations. It's not quite the same as some of the other datasets that we have.
4:57:54
Some of the other datasets are these so-called appended datasets where we take just the first couple of hours of the forecast Assuming that those first couple hours are pretty good.
4:58:04
And then just keep appending the first couple of hours to get sort of a quasi analysis dataset.
4:58:12
And again, if it's, you know, even if you're in complex terrain, 27 kilometers is probably going to be too coarse.
4:58:17
And and even three kilometers might might be, might be too coarse, but depending on on the on the domain.
4:58:25
Again, you can use different datasets.
4:58:27
And then finally, one thing that you might do is if you have meteorological data in the area that you're interested in like from a remedial article data station, you can actually compare what the measurements are.
4:58:43
two, what the meteorological data you're using for your analysis.
4:58:47
And What you want is, for those two match up, pretty good.
4:58:52
If they don't, then it means that likely, there's some complex terrain or coastal phenomena or other phenomena that are changing the, um, making them that data that you're using, not representative. So you can check this, and you can see if what you're using.
4:59:12
Makes sense.
4:59:14
Um, another question people ask, I'll try to go through quickly of this is that, you know, when you're doing a back trajectory, what height should you start from?
4:59:23
So suppose you're making measurements near the ground.
4:59:26
You might think that what you should do is start that back trajectory at the height of the sampler, like two meters above the ground.
4:59:34
But what we saw that ...
4:59:36
showed earlier, is that, when the trajectory hits the ground, you kind of start losing information.
4:59:43
Or it kinda just goes along the ground, and you don't get the full, um, description of the flow field, and so I'm gonna give you a couple of quick cases here, in case one, sort of simple terrain and you're fairly far from sources, and then Case two, on the top of a mountain.
5:00:02
In the case, case, one, suppose you're making measurements, and you're pretty far away from this, from a source maybe 20 kilometers or more.
5:00:11
Then what happens is, um, during the day, the height of the boundary layer is pretty high.
5:00:16
Maybe 2000 meters, something like that, and the plume know, maybe go this. maybe the central line of the trajectory of the plume.
5:00:24
Um, and in fact, you know, this might be the plume.
5:00:28
It might, um, after, you know, maybe 10, 15, 20 kilometers, it's pretty much spread.
5:00:35
It's about the whole Boundary Layer.
5:00:37
What that means is that what you're measuring on the ground, it's kinda like the average concentration throughout the boundary layer.
5:00:44
And what that means is that if you were going to start if you're wondering where the air came from that hit this sampler, you're probably better off starting near the middle of the Boundary layer.
5:00:55
As long as the Source's aren't too close, As long as they're far enough away, you're better off starting at the middle of the boundary layer.
5:01:02
Then you would, starting at the, at the height of the sampler.
5:01:05
If you start here, the directory might hit the ground, it just kinda kinda scurry along the ground, and you kinda see the spreading of the of the wind and the dispersion that you might see.
5:01:16
Um, in a real case.
5:01:19
Um, this is what I'm, this is what I'm saying there. Sorry. You know, at night, it's important to recognize that.
5:01:27
The boundary layer is very shallow, could be 100 meters, could be last maybe 200 meters, and it's interesting to note that at night, the stack of A of a power plant, or source, can actually be above the boundary layer.
5:01:42
And so, if you're making the measurements at night, and if the stack height, what's above the boundary layer, you might not actually see anything.
5:01:50
And so, again, in this case, if you started at the middle of the Boundary layer, that would still be OK, because the middle of boundary layer would be actually be, you know, maybe 50 meters or something like that.
5:02:01
And, uh, then indeed, even, you know, you wouldn't necessarily go back to the, to the source, but you do have to be careful.
5:02:11
Thinking about interpreting, you know, your, your data with respect to the boundary layer and at what times you're seeing your sampling.
5:02:21
Um, by the way, at night because things are very close to the ground and even if you have right a position talking about tomorrow, at night you can get a lot more dry deposition because everything is really close to that to the ground.
5:02:36
Um, oh, let me go back for a second here.
5:02:39
The bakke trajectories and trajectories in general do not include deposition.
5:02:46
So if you have a case where deposition or for that case chemical reactions are important, you need to remember that back trajectories or farber trajectories do not include phenomenon.
5:03:00
Um, so what I'm saying is, started at maybe half the Boundary Layer height, then that way.
5:03:06
And you could make that setting in the, in the model where you could set the starting height to be, relative to the boundary layer that fraction the boundary layer, and use the zero point five.
5:03:19
Then it would always start the trajectories at the middle of the boundary layer, whether it be at night or the daytime.
5:03:26
And that might be a better way to go. If you're only going to start to do one set of back trajectories, that might be the best one.
5:03:34
Then the other case is let's suppose you're taking samples at the top of a mountain We have a sampler for example the top of Mauna Loa.
5:03:42
Volcano in Hawaii, and what would you do in that case, and it's a little different.
5:03:50
The boundary layer is probably, you know, much lower than the, than the the top of the mountain, and what we're really seeing up here is the freeze spheric errr coming across, and we sometimes see upslope flow or downslope flow.
5:04:06
And so what she would do I mean, just let me just shows the sampler there.
5:04:10
Let me just fast forward here.
5:04:13
What you do in this case is you probably kick the relative demeans sea level and um up sorry, and then just take the height of the summit.
5:04:25
Um because basically what you're seeing there is the air coming in at about 3000 meters or you know, approximate the height of the of the sampler.
5:04:35
Um it clearly wouldn't want to know that the height above the ground level. Sampler might only be 10 meters above the ground.
5:04:42
But you put the air you're seeing that's hitting that.
5:04:46
That sampler is really coming from, you know, pretty much the height of the of the of the sampler above sea level.
5:04:56
Um, then, finally, I'm going to show a couple of quick examples of sort of, simple approaches to source attribution in the last section.
5:05:07
I'm sure all sorts of really interesting examples of how we use high split to get source attribution information, and we'll show some more examples of that as well.
5:05:17
But I want to show some simple examples, Know, before you even have to do a lot of work, you can get some, some simple examples. And here's an example where we were making mercury measurements in sight in Beltsville, Maryland.
5:05:32
And one day, we noticed a big peak in a certain kind of mercury.
5:05:37
So we did back trajectories, you know, during this day to see where the error was coming from.
5:05:43
And what we found is that here's the sampler here, and these are the back trajectories that are coming backward arriving at the site.
5:05:54
Um, at a certain time that I'm that I'm showing here, and the back trajectories were done at different heights of the boundary layer. So I did not, only zero point five, but also point 1.3, zero point seven zero point nine.
5:06:09
And what you can see is that, as the day went on, and by the time we started to get that big peak, we were basically, the air was coming from this area, north, east of us in Baltimore that had a lot of big mercury sources.
5:06:26
Then, as we continued on to the day, it went that the air is coming from these other areas without, without many sources. Now, this is not quantitative.
5:06:35
It doesn't prove that the mercury came from those sources but it's very suggestive.
5:06:40
And if you did this enough times with enough data, but enough samples, you could, you know, either convince yourself, or not convince yourself that a certain area, it might be a source.
5:06:50
So, you know, again, the air was coming from areas with big sources, and we got a peak.
5:06:56
So, it's suggesting that perhaps those sources, the to the mercury at that site. And then one other example I wanted to show you.
5:07:06
Similar, Suppose you have a whole bunch of samples, and at a site you could take, you could run trajectories, say for every sample.
5:07:20
But then, you could remember those in files that ... was creating, that had lists of trajectories.
5:07:28
Well, what you could do is you could separate those trajectories out and take all the sample, all the trajectories that were associated with high values, and put them in one in file, and take all the trajectories ever associated with low values, and put those in another in file.
5:07:47
Then, you could do that frequency analysis that he showed you, um, for those two cases, and, you could see, is there a different pattern between the the air Mass Source regions. The air came from when it was high, and low.
5:08:03
And I'll show you an example of this, in the next slide here, here, we had a, um, a sampling site off the coast of Florida, um, where they were sampling dust. And, they were wondering if any was coming from the Saharan Desert in Africa.
5:08:18
And here were the, here were the, is a frequency analysis of all the trajectories?
5:08:24
Um, this is, you know, the percentage of end points that showed up in each of these squares are the samples where if we had low dust concentration at this at the sampler, and here's the trajectory mapped.
5:08:40
Let me add hi, dust concentrations. Go back and forth below.
5:08:45
And, the high, you could see that, the, in, that, when we had high desk concentrations, were definitely getting more of the flow from the Saharan Desert, whereas, when it was low concentration, it was more coming from the North Atlantic and wasn't getting as nearly as many of trajectories from there.
5:09:01
And, again, this is not proof, um, anything, but again, it's suggestive that that, perhaps, uh, when we were seeing the high dust samples that could have been coming from, from that, that area.
5:09:17
All right. That's just a couple of quick examples.
5:09:20
And I want to now go to just finishing up for today, just reminding people that we put stuff on this workshop webpage, including the recordings, oh, sorry.
5:09:33
And any kind of handouts, or things, that we're, that, we want to give you, um, tomorrow, we will be going into the air Concentration calculation. So, that the real dispersion.
5:09:45
And then we may be ending up the day with Pluton Transformations Deposition, or if not, we'll start the next day.
5:09:53
With, with that, um, let's see, this is what we did today.
5:09:59
And I've reminding people that we generally are asking questions some questions so you can put into the goto Webinar and then other questions, it's better to ask, and in the form.
5:10:09
Recording is the recording from today.
5:10:12
We'll probably be ready in about, oh, 4 to 8 hours, something like that, and when it is ready, they'll try to put a link.
5:10:19
I'm on the workshop webpage.
5:10:23
And, Oops, sorry.
5:10:26
We have talked about a little about the different ways to use high split.
5:10:29
You know, the workshop is dealing with this GUI.
5:10:32
Um, you can run it on online and if you have C model downloaded. You can also run it with scripts, but this GUI is Graphical User Interface that we're learning in the workshop is a great way to learn how to do scripting.
5:10:47
And I think that is it. I think I've said this enough.
5:10:52
And with that, I will leave everybody to go back to their regular work and the regular day, and I hope to see you tomorrow.
5:11:03
Start at 9 0 AM, Eastern Time, or wherever that that time is in your location, OK? Bye, Bye.