Polar Radar for Ice Sheet Measurements
|Home>Prism Update>Student Presentations 2005 - Joel Plummer|
My name is Joel Plummer. I am a graduate student in the Department of Geography, as well as CReSIS here, and I've been here in Kansas about a year.
I'm going to be talking about GIS (Geographic Information Systems) today. I just want to lay out a little disclaimer that if you were to go over to the Geography department, you could probably take 2 or 3 semesters work of GIS classes. So when I say an Introduction to GIS in 30 minutes, this is going to be a real introduction. It's not going to be anything in depth.
Hopefully by the end of the session here, you'll be able to answer a couple of questions. Whether or not they are simple is up to you.
First, what is a GIS? Also, what can GIS do for you? I'll show you examples of what it can do out in the real world as well as how I've used it.
And then finally, where can you learn more, if you decide that GIS is something that might apply to your work. We can get you on the path to using it.
For good or for bad, this is what most people think of GIS. It's just a software package to make pretty maps. I don't want to understate the ability of GIS to do so. It's a very, very important function, but cartographic visualization is not the only function within a GIS
. It's very powerful and there's many more things you can do as I'm going to show you.
What Is A GIS?
This is an ugly textbook definition of GIS. It's basically a system (hardware, software, data, people and infrastructure) that you use to manage geographic information. And what I mean by manage is a whole slew of components, from collecting, storing, visualizing (which is the mapping or cartographic component), to analyzing and then redistributing that data.
So we've got two questions really.
What does this mean? As I say this is kind of a textbook definition. Basically, in a nutshell, GIS is a database system that is very highly specialized. It has specialized functionality, enabling it to deal with geographic information.
And the second question off the bat is, "What is geographic information?" and "What is the big deal with that?"
Geographic information is data that can be located in 2-dimensional or 3-dimensional space. So on the left here, we have a database. Everyone has seen something like this. We have it in one form or another, a simple phone or address book. This is not a geographic database.
However, once we add the next component in, an address, it becomes a geographic database. That attribute ties each of those records to a physical location on the surface of the earth. Now it becomes a geographic database.
Why Is GIS Important?
Why is graphic information important? Why do we spend so much time and energy to analyze and look at it? That's because locational attributes provide inherent spatial relationships which, in turn, provide an opportunity for analysis that's not normally found in regular database systems.
These spatial relationships imply a common attribute, a common function, some connectivity or flow. Also inside geographic information we have the ability to map it. And we all know the power of charts and graphs to convey tabular data. Cartographic visualization, as I've mentioned in the beginning, is a very, very powerful tool.
As researchers, you know, what good is our information that we develop if we don't provide it to somebody? Now I'm going to come back to cartographic visualization in a second. But I want to hammer home the importance of spatial relationship with some examples here.
In a normal database, if you have some data, we have the ability to extend that data using what's known as a join. So if we take our data on the left, we've got a column with phone numbers. That's a unique attribute. We can then go and find other data that has that same sort of attribute information and then we can use it to extend our information.
So now Bob, whereas previously we had his date of birth and phone number, now we have information regarding his phone account. You know, hypothetically, we could go out and we could do many of these things.
Spatial Joins and Queries
In spatial analysis, we can perform what is known as a 'spatial join." We have the information about where these people live and then we can go out and add more information based on that locational attribute. So for example, in this table on the right, you see where the county is, that's the locational information, but in the table on the left, you're not going to find county names. You're not going to find Douglas, Johnson or Shawnee any place in there. But if you plot that information on the left, it will co-exist with some of that information on the right. And because of that we can perform what's called a spatial join.(All these address attributes are located within Lawrence, so we know that they're in Douglas county). And we can extend our database that way without having to share any sort of attribute information.
Similarly, a very basic database function - you query some attributes. So if we wanted to find all people less than 30 years old, we could enter in a pseudo-SQL query and we could return, based on the attributes, some records.
You can do this in a GIS. You can perform a regular join in a GIS, but in addition to an attribute query we can also perform spatial queries. For example, how do we find all people located within a mile or two of campus in a conventional database? You can't do that without some extreme processing going on. But in a GIS it's very natural. We just take the addresses (and these are hypothetical addresses, I don't know where they're located) ... just take their addresses and we can plot them in 2- or 3-dimensional space and we can find where they are.
Other Things We Can Do With GIS
We can build on these things and we can come up with a bunch of other actual uses for GIS that you don't find in conventional databases. It goes without saying we can measure distance and area much more accurately than we can on a paper map. We can overlay two different data sets as I've shown. We can perform joins, unions, intersections. We can buffer points and we can find proximity as I've shown in the query. We can create networks of roads, streams, and on those networks, we can find the shortest path, and different other functions. We can create 3-dimensional surfaces and then from any particular point on that surface, we can create a viewshed.
What objects on the surface can I see from where I am? And, in turn, what objects can see me? And then another big one is spatial interpolation. We can estimate, you know, unknown values based on the distance and direction of other known values. And this just touches the very surface of GIS. And I don't have time to really explore everything, but I just wanted to give you an idea of what the basic purpose of a GIS is.
And kind of a recap, you know, to answer that first question, "What is GIS?" Again, it's a database system that has some very powerful functionality built into it to take advantage of the spatial relationships of objects; whether it be adjacent objects, one object may contain another, or various other relationships.
Some Real-World Examples
I now want to show you just a few basic applications for GIS and this is real-world stuff. Very simple stuff that after a class or two you could probably perform on your own if you felt the need.
First, telecommunication companies - Sprint uses GIS to locate their cell phone towers. They use the viewshed operation so that they can have maximum coverage.
Wal-Mart and other retail establishments - they track their customer base. If you ever go into a store and when you're checking out, they ask you what your zip code is, that's what they're doing. They want to find out where their customers are so that they can plan a future store or they can plan marketing operations in your zip code.
Hydrology or utilities managers - they use flow networks to find problems upstream and police dispatcher finds the fastest route to an emergency on a road network. Those are very basic examples of how GIS works in the real world.
This next one is a little more advanced, but it's also really cool. Land use planners use 3-dimensional models of cities all the time and with this we can kind of plan how a city will look. A realtor can plot a house on that landscape and say, "What will the view be?" We can plan where buildings are going to go around road networks, utility networks.
Using GIS To Plot Ice Thickness
Next, I really just want to tie this in to what we do here at ITTC. And what I do specifically, so I can bring it home to a more real-world example.
I've been working for the last year with data in Greenland - ice thickness data -and this has been my study area. It's West Greenland. This is the Jakobshaven area.
And this kind of demonstrates some of the functionality of GIS as well. You have many different kinds of data sets. On the left we've got a vector, just a shape, and on the right we've got a raster satellite image. And they're all very compatible. And you can bring them in and work right on top of each other.
I inherited two data sets. On the left we've got ice thickness from the radar depth sounder. That was taken between 1997 and 2003. On the right is LIDAR, which is ice surface elevation data. And I believe that was taken throughout the same time frame.
And my goal with this dataset was to build a model of the terrain underneath the ice. So first I took that point data and I created a surface. This is ice thickness. The white areas represent areas of great thickness and the darker areas represent areas of minimal thickness.
Now we have ice elevation. Again, light areas represent areas high up, dark areas represent areas of low elevation.
Now theoretically, with two surfaces - one surface being on top and a thickness underneath - we can then calculate what the bed elevation is going to be. And this is the result. Dark areas represent areas of low elevation, light areas represent areas of high elevation.
Visualizing Under the Ice Sheet
I can render that in 3 dimensions and we get a nice profile. This is exaggerated 2 times or 3 times, I can't remember which. So it's not a real world picture but it gives you a good idea of what the surface looks like underneath the ice sheet. I can turn it. Give you a little better view.
Then on a purely visual system. No analytics going on here. This is part of a project for a class I'm taking on visualization. What I've done here is taken that DEM and built a model of what it might look like. I know it's probably not that craggy in the real world. And we can do a little animation. This is at no vertical exaggeration. And what's going to happen right in the beginning is it's going to bump up to about 2.5 times vertical exaggeration. So if you're not prepared for it, it might seem a little weird at first, so just warning you.
Another point to bring that visualization will help you in your analysis is that we have tabular data - latitude, longitude, thickness. This is something we've all seen.And we help it come alive a little bit more by using echograms. This is again something we've all seen.
The red line in this file is a 1997 flight line and the blue line is going to be a 1998 flight line. And we use this (animation) to look at the intersection between the two points.
Resources for Learning More about GIS
So finally I guess I've shown you a little bit about what GIS is and how it might be able to work for you. And the next question is, "If you've got questions on how you might apply it to your own research, where can you go and learn it?"
The first stop, I guess, might be the geography department. And I would be, or I could be, your point of contact there. If you have any questions about your research and how GIS might be applicable, you can just ask me and I'll help you out
The Geography Department offers a basic, intermediate and advanced course in GIS, and you're certainly welcome to sign up and enroll. November 13-19 is National Geographic Awareness Week and specifically on the 16th is GIS day and the department is going to be holding presentations all day long. Students, faculty and people from outside the department will be giving GIS presentations. It's open to the public and it's certainly open to other members of the University. You're welcome to come over and check it out.
Outside of the Geography Department there is the Anschutz library. Down on the very first floor next to the map library we have the GIS & Numeric Data Lab. That's their web address. There you can get training, advice, data, all sorts of help free of cost.
In addition to those sources, there are many organizations. Perhaps some of you are members of a professional organization. They might be able to help you out as well and get all sorts of information on how to apply GIS to your research.
PRISM © 2002, 2003 is brought to you by