Draft of 9/5/06

Introduction to Geographic Information Systems

ERE 450/550

Lab Exercise 2

Lab due Friday September 29 at 5pm in Bray 410

In this lab exercise you will examine, review, assess, and prepare the data that you’ll need to address a hypothetical problem in the city of Syracuse. The first step of the GIS process (Purpose, Problem, Objectives) is provided for you; this lab focuses on the second and most labor-intensive step, data management. This one can be tough, but then you’ll be ready for the rest of the labs. You will also notice that there is a little less direction than the first lab. If you need reference on how to do the steps in this lab, you will have to refer to the first lab and/or your notebook….you are keeping a notebook, right? To answer the questions embedded in this exercise, refer to your textbook and the power point lecture notes. We may not have covered some of the concepts in class yet, but that’s why I’ve given you the power points you’ll need.

NOTE: In order to be successful with these labs, read the exercise over IN ADVANCE of doing it; read and follow directions carefully; be deliberate; keep track of GIS procedures in your notebook; also in your notebook, keep track of filenames and where data is stored; and finally anticipate and prepare for problems with the software….and sometimes my directions. (By the way, if you see problems with directions please let me know ASAP so that I can alert the class)

If you haven’t noticed already, these exercises can get wordy with discussion of GIS concepts and questions (that’s part of the class goal…you’re not here just to learn software), but embedded within the exercise are simple GIS steps and processes that you’ll have to learn so that you can do them again more efficiently in the future.

If you want to have clear and concise software instruction, don’t use these exercises; use your notebook.

Step1: The Big Issue-Problem, Questions, Purpose, Objectives

Hypothetical Scenario: Syracuse Planning for Global Climate Change

Elizabeth Nifkin (yes, she’s a relative of Eustis B. Nifkin, and also a member of the “green mafia” – another fun title given to ESF alumni) was recently hired as an environmental planner for the city of Syracuse. She in turn has asked you to help her do climate-change planning and assessment using GIS. Her advisory committee (made up of forward- thinking Syracuse citizens) has requested the following information and has asked the following questions.

How many people are near or in flood zones within the City of Syracuse?

They would like information on home ownership relative to renters; they would like to know population densities of these areas; they would like the Meadowbrook area examined in more detail; and they would also like to know the zip codes of these areas for future mailings.

Where are the Armories, Railroads, and Roads relative to these zones? They want to know, so that they can evaluate evacuation and supply issues.

They would like to know about the dams above residential areas.

They would like to know the watersheds that are affected (because they are also concerned about the potential for pollution loading during a flooding event)

They would like you to map the wetlands, so that they can assess those areas beneficial to buffering the negative effects of flooding.

They would like you to assess potential soil loss using the Revised Universal Soil Loss Equation.

Finally, they would like an idea of relative urban temperature issues that will need to be addressed.

Since the purpose and objectives (Step 1 of the GIS process) are already done for you, you now need to start the project with the next step of the GIS process: Data Management. What this means is gathering information about the data you will need to do the project.

The specific objective of this exercise is to examine, review, and evaluate the data and to prepare for the next lab. To enhance your understanding of GIS concepts, you’ll also be directed to answer some questions along the way.

IMPORTANT INFO BEFORE YOU START

The deliverables (Step 4: Output) for this lab are attached and include a table (fill-it-in), a spatial model (fill-it-in), and the answer sheet. Those deliverables capture your Step3: Analysis (asking and answering questions about the data). Note that many of the blanks have been filled-in for you. In addition, the data is on the Greenfield/PROJDATA folder. That way, if you stop in the middle of the exercise, you don’t have to take all that stuff home. You should however have reference in your notebook so that you can pick up where you left off. You can also check your work against what’s on the data folder to see if you’re doing okay.

One more thing... check the website often for notices, announcements, and helpful hints that may be posted regarding this and future labs.

Step 2: Data Management

Before you begin, think about this a bit. Using the issues above, what are some of the datasets that you’re pretty sure you’ll need? The first question explicitly asks about the flood zones and population within the City of Syracuse, so you’ll need to find something close to those three data sources when you go searching. They’ll need some zip code information for mailings. We also know that they want an evaluation of soil erosion, so I’ll search for soils data plus other variables of the Revised Universal Soil Loss Equation (RUSLE) which requires slope, so you’ll need Digital Elevation Models or DEM’s. It also requires an evaluation of land cover (what’s on the ground vegetation, impervious surfaces, bare soils, etc…), and it requires precipitation and the relative intensity of that precipitation. What else? Well, transportation for evacuation and critical supply movement, armories for evacuation and National Guard troops for help, dams to assess upstream threats, hydrography and related data to locate wetlands (great flood and pollution buffers), streams, rivers, ponds, and etc… Finally, we may want some fairly recent imagery (like satellite or aerial) for verification of features, accuracy, and for updates to existing data sets. If this stuff isn’t available, then you’ll have to think of doing something else either by using an existing dataset and modifying it or by creating a new one. That’s basically how you do GIS, and how I created this exercise. Ultimately, I had these questions, but I wasn’t sure what data was available. This exercise specifically addresses that issue.

Folder Management

A big part of data management, is setting up the data storage structure using folder management. You are about to download a bunch of GIS data, so now is the time to organize. You will set up a folder and subfolders to hold all this downloaded data. This also makes it easier to figure out what you need to save before leaving the lab sessions. You don’t have to save your data from this lab, but it is so much easier to just save your whole project folder directly from the C drive to your portable media. This is pretty important, because an ArcMap project does not save your data.

Use Windows Explorer to create a folder using your last name on the hard drive (C:) of the computer. In this exercise and in the future, I will refer to this folder as LASTNAME.
Within this folder create PROJDATA.
Within PROJDATA create the following folders to help you keep the data organized: HYDRO (for all the water data sets), DEMS (for the digital elevation models), TRANS (for the transportation data), CENSUS (for the blocks and boundaries), SOILS, and COVER (for classified land cover).

If you need more help on how to use Windows Explorer, use Exercise 1 and your notes from Exercise 1.

Data Gathering: a.k.a Data Acquisition

You’re going to download a bunch of data at one time using the “basket” feature on the CUGIR webpage. This will produce one big zipped file of all the datasets that you will then have to unzip individually. Use the checkbox to the left of the datasets, but do not download the basket until you have them all. You can easily go backward and forward on the webpages that you’ll need to navigate to get all of the required datasets, and then “update” your basket as you go.

You should also begin filling-in your table using the metadata and info links associated with each dataset. You may not be able to finish the “Vector or Raster” field in the table from the first part of this exercise, but after looking and manipulating the data sets throughout this exercise, you’ll be able to complete it. You should also use the steps of the exercise to “fill-in-the-blanks” of the spatial model.

Some of the table information will be pretty-hard to assess, and Date is going to be one of those tough ones. We are ultimately concerned about the currentness of the data or in other words, how “fresh” is it? For example, the first dataset you’ll add to the basket is the freshwater wetlands. If you look at the metadata it has a publication date of 1999, but is that really how current it is? If you scroll down through the metadata, you’ll find a link under “Currentness Reference” that refers to specific county work. When you click on that link and look for Onondaga County, you’ll find that despite the publication date, the most recent update to Onondaga county freshwater wetlands is 1994, so that’s the date I put in the table. Yes, this is a pain in the butt, however metadata includes all those darn readme files and any other documentation that may accompany a dataset. For some dates (like the DEM’s), you may have unknown. As you do these labs, you’ll find the information you need is in some of these other places. This is maybe where you start cursing me (it certainly won’t be the last time)….and that’s okay; I’ve been there, and I understand and appreciate what you’re going through….and I hope that you are beginning to appreciate the role of a GIS person.

Why am I worried about dates? Well duh, “the one constant in life is change.” There are actual physical changes, like a stream system constantly changing its path, and there are changes in detection, interpretation, and science. I’m really appreciating the service of CUGIR over the New York State GIS clearinghouse lately. Later in the exercise, when you download the big CUGIR zip file, you’ll see a readme html document that has links to check on updates to the data sets you downloaded…pretty cool…they really know how to serve up the GIS!

By the way, the other item that may make you a little crazy is looking for the scale, but again, we’re looking for limits, so do the best you can to find the most appropriate scale denominator. The freshwater wetland data is most appropriately represented at the 1:24,000 scale, as is most data created with the 7.5 minute USGS quad sheets as the source document. Look for the source scale denominator within Lineage first, and then if it’s not there, you may have to look somewhere else in the metadata documentation. You may also witness one of the big concepts (in the lineage) that I’ll repeat over and over in class, when doing analysis with multiple scale or resolution datasets, you’ll always be “stuck” with the smallest scale and/or the most coarse resolution.

Let me explain that a little, and you should think it through. If I combine two data sets, one at 1:100,000 and 1:24,000, the scale of the analysis is the smaller scale, 1:100,000. That’s because the 1:100,000 is the limit on accuracy. The 1:100,000 has been generalized more than the 1:24,000 scale to a point where features are more abstractly represented beyond their actual location. A good example is the railroad, road, and river next to each other separated by 1m. If we actually used the scale to define where those features are in real space, they would literally be on top of each other. So, they are artificially separated to provide relative location, not absolute location. The same problem may also exist in the 1:24000 scale data, but it would be less abstract and more accurate than the 1:100,000. The same concept works with raster data as well. Resolution refers to the smallest cell that can be representative of an area. If we take the example of a picnic table in your backyard that is 2m long and 1m wide, I may be able to pick that up in a raster image of 1m resolution (it would be 1 cell by 2 cells), however if I had a 30m resolution image, the color and shape of the picnic table would definitely get overwhelmed by the generalized colors and shapes of the surrounding area.

So, here’s the deal. When you combine 30m and 1m resolution data in an analysis, you are stuck with the 30m resolution as your resolution of analysis. Again, I will repeat this over and over and over in class. To maintain accuracy, you can go from big to small scale, but you can’t go from small to large scale. And, you can go from high or fine resolution to low or coarse resolution, but you cannot go from coarse resolution to fine resolution.

Despite my pronouncement, we still may do analysis that goes the “wrong way,” so to speak, in scale or resolution. But, it’s like anything else in science. We sometimes over generalize in order to simplify. The best thing to do is to acknowledge and disclose these limitations, so that we don’t do something stupid in policy when considering the real complexity and unpredictability of life. Okay, I guess it’s time to get some work done.

For the resolution and scale fields in the table, fill out resolution for raster data and scale for vector. If there is a scale for raster data or a resolution for vector data, I will not hold you responsible for that information on the table. In most cases, we are concerned about scale for vector data and resolution for raster data.

NOTE: As you do the following, check the info and metadata links to fill in your table.

Go to the CUGIR website and use the MapBrowse tab to get to the datasets for Onondaga County.
Download the following datasets from the Current dataset(s) for Onondaga County page that first opens after you map browse (click on) Onondaga County: Freshwater Wetlands, Census Blocks (2000), Census Places 2000, County Boundaries (Census 2000), Hydrography (Census 2000), Railroads (Census 2000), and Roads (Census 2000). After you check them, click on “add checked map datafiles to basket.” You should have 7 data sets.
Now go “back” (click on the back button at the top left of Internet Explorer) and click on the link for “Mapsheets in other spatial series related to Onondaga County” near the bottom of the page.
Find and add to your basket, the DEMs “Elevation Data” from the “Data published by USGS 1:24,000-scale (7.5 minute) quads” by clicking on the following quads, Jamesville, Syracuse East, Syracuse West, and South Onondaga. You should have 11 files in your basket.
Find and add to your basket the NYS Large Scale Hydrography data (both the Large Scale Hydrography Network and Surface data sets) from the Oswego and Seneca Watersheds. Now you should have 15 data sets in your basket.
Go “back” or click on the Map Browse tab, and then click on the Statewide Data tab. Find and add to your basket the Major Dams of New York State, New York State Average Annual Precipitation, the 8 Digit Watershed Boundary, New York State, and Soils (STATSGO Statewide). You should have 19 data sets in your basket.
Download the selected data to your PROJDATA folder within your LASTNAME folder, and unzip the CUGIR zipped file. By the way, notice that the name of the CUGIR zipped file is dated and timed. Don’t close Internet Explorer…you’re going to get more data.
Use Windows Explorer to move the contents (the zipped files within) to the folders you set up earlier.

Move 067blk00s.zip, 067cty00s.zip, and 067plc00s.zip to the CENSUS folder.
Move p29elu.dem.gz, p30elu.dem.gz, q29elu.dem.gz, q30elu.dem.gz to the DEMS folder.
Move 067rds00s.zip and 067rrs00s.zip to the TRANS folder.
Move 067fwa.tar.gz, 067hyd00s.zip, n4140201a.e00.gz, n4140202a.e00.gz, s4140201a.e00.gz, s4140202a.e00.gz, 36huc8s.zip, 36_ave_annual_precip_s.zip, and 36_dams_s.zip to the HYDRO folder.
Move 36_so_a.zip to the SOILS folder.

Navigate back to the Current dataset(s) for Onondaga County page and find and click on the link to NYS GIS Clearinghouse data and imagery for this county.
Click on the link for Data Sets in Onondaga County. Wow, eh? Lot’s more data! But, a lot of it we can’t get to from here because of security protocols and privacy issues. Look over the data available for future reference.
Now find and download to your PROJDATA folder the following data sets: Armories (under State Coverage Datasets), Digital Q3 Flood Zone Data, both the tab and the shape for Onondaga County (under County Coverage Data Sets). Also be sure to fill in your table using the metadata links.
Move the armories.zip to the TRANS folder, and move the map_c36067.zip and shape_c36067.zip files to the HYDRO folder.
At this point, you now have a total of 21 data sets for this project. Eventually, you’ll have more, but I’ll let you know about that later.
Unzip all the datasets and make sure that they are extracted to the folder that they were in; e.g. n4140201.e00.zip will be extracted to the HYDRO folder.

Adding Metadata in ArcCatalog

This is an important step so that the metadata is kept with your dataset. These days, we primarily use .xml documents to keep track of the metadata. For this exercise, you do not have to add the metadata to each dataset, but try a few using the steps below. For future work, you can always add it when you need to, or you can look at the reference on the website. Your table serves as a nice summary of the metadata.

Start ArcCatalog and add the connections to your folder PROJDATA and the Greenfield/PROJDATA folder (refer to Lab1 or you notes for more direction).
First take a look at the data in the Greenfield/PROJDATA folder and subfolders. This is your back up in case you need it, and the prepared data you will use for future exercises. Note that my file names may not be the same as your file names after manipulation of the data sets.
Navigate to the CENSUS folder and in the catalog tree highlight the tgr067blk00s.shp file. On the right in the data view, click on the metadata tab.
Find and click on the Import Metadata button (above the windows and they look like pages). The import metadata window will open. Click on the option for automatic updates.
Browse to the metadata folder, which is in the “CUGIR_date/time” folder within the PROJDATA folder and find blk00s.xml (you may have to change the dropdown menu “Files of type” at the bottom of the browse window to allow you to see“all files.”
Open it, and then change the format (drop down window) of the Import Metadata to FGDC CSDGM (XML), and click ok. The metadata appears in the data view of ArcCatalog.
Click on the Spatial tab of the metadata in the data view.

Question 1: What is the “Geographic coordinate system name” under “Horizontal coordinate system?” (Put your answer on the answer sheet)

Click on the link for Details.

Question 2: What are the Geographic Coordinate Units? Are these units of measurement planar (Cartesian, rectangular, or projected) or spherical (angular, geographic, or geodetic)?

Question 3: Look under Geodetic Model. “What is the Horizontal Datum Name?” What is the “Ellipsoid Name?” Is an ellipsoid a spheroid? Is a spheroid always an ellipsoid?

Question 4: What is a datum? (You may use the very simple one- to two-word definition.)

Adding metadata can be tricky at times. You may have to change the style sheet (upper left), and sometimes you have to import the metadata as a different format. This example at least gives you an idea on how to use this feature. If I cannot get the import metadata feature to work, I at least download metadata (save as the HTML) and any other accompanying documentation to the folder where the data is kept.

Thankfully, the CUGIR data came with metadata files in the XML format, however, the stuff that you downloaded from the NYS GIS clearinghouse didn’t always have the metadata accompanying the data set. That’s when I have to “save as” the HTML document and then keep it within the folder with the data of interest. Don’t worry, you don’t have to do that for this exercise, because you’re using a “handy-dandy” table to keep track of some of the more important parts of the metadata.

NOTE: There is one small, but very inconvenient problem with using the ArcCatalog imported metadata that you’ll see later in the exercise. So, use the webpage metadata and info links to fill in your table.

IMPORTANT: To stay safe, limit file names of the converted interchange files and DEMS to 8 characters or less with NO SPACES!!!! Also, keep track of these new file names in your notebook. Finally, make the names intuitive. For example I named the s414201 file SEN_SHYD so that I would remember it as Seneca watershed, surface water hydrogaphy. I named the stream network file SEN_NHYD. You may use longer filenames for the shapefiles, but don’t go over 12 characters (including spaces/underscores). In the past we were restricted to 8 character filenames and field names in GIS, so I tend to stick to that rule because I’ve seen some trouble with longer file names….even when they “should work.”

Importing and Converting Interchange (a.k.a Export) Files

Interchange or export files are ASCII text files created for quick download and easy management of ArcINFO Coverages (just another data format of GIS). They will not “show-up” in ArcCatalog or ArcMap until they are converted or imported.

If you do not see a button/drop down menu for “Conversion Tools,” on the ArcCatalog interface, add the tool by going to the View menu, select Toolbars, and put a check next to ArcView 8x tools to make it “visible.”
From Conversion Tools (click on it) select Import from Interchange File. A window will open titled “ArcView Import from Interchange File” with a spot for input and output.
Input an interchange file by clicking on the browse button (looks like a folder) and selecting an interchange file, e.g. the 067fwa.e00 file within the HYDRO folder.
Name the output dataset appropriately, e.g. for the example used above, I named mine wet1a because there are two wetland export or interchange files. I named the other one, converted from 067fwaan.e00, wet2aan – by the way note that I used less than 8 characters to name the ArcINFO coverages…and NO SPACES.
Click OK. The processing window will open and show you the process in text within the window. When it’s done, click close after it’s completed.
You can convert your interchange files one by one, or you can figure out how to use the Batch button to do a bunch of them. (No hints here, trial and error is sometimes how you learn these shortcuts).
Do the rest of your .e00 files (total of 8 including the wetland sets). The soils one will be a little hard its found in the following path SOILS\jc55\Census90andStatsgo\STATSGO_Statewide\NY\SPATIAL and it’s titled ny_soils.e00
After you’re done, click on the network hydrography data for the Oneida watershed (within the HYDRO folder) and clicking on the Contents tab in the data view (you can see what I’m about to ask by also clicking on the + sign next to the dataset).

Question 5: What are the four data sets within this ArcINFO coverage?

Click on the Oneida watershed surface hydrography data set the same way you did it above.

Question 6: What are the five data sets within this ArcINFO coverage?

Hopefully, you also took note of the icons used to symbolize these types of data, especially note the difference between tic, arc, node, route, polygon, and labels. There are other types of data that can be in these coverages as well, such as annotation. (To take a look at that one, check out the flood coverage.)

ArcINFO coverages are topologically (ability to know what’s next to what) referenced data sets that hold feature and attribute data within a three-folder system. A root folder, or workspace, holds both a feature folder and info folder, both have files that are associated with each other. If these files are separated the whole thing falls apart; that’s why we use ArcCatalog and NOT Windows Explorer to move GIS data sets around. Tics hold known ground coordinates of the data; nodes are basically points that make up lines; arcs are lines; polygons are made up of connecting lines; labels are points within polygons created to reference those polygons (a similar data structure to labels is used to make just points); annotation is additional descriptive notes of a data set; routes are used for network and connectivity analysis (like Mapquest’s directions). Next, you’ll take a look at the structure for shapefiles. I’ll discuss more issues about GIS data formats throughout this exercise and in class.

ArcCatalog vs. Windows Explorer: More on GIS data structure

In this section, I will demonstrate why it’s easier to use ArcCatalog over Windows Explorer to manage GIS data. As you saw above with ArcINFO coverages, the biggest issue here is the GIS data structure.

Open Windows Explorer and adjust its window and the ArcCatalog window so you can see them together (side-by-side) on the computer screen.
Navigate in both Windows Explorer and ArcCatalog so that you can see the contents of your (not the one on Greenfield) CENSUS folder.

Question 7: How many shapefiles are shown in ArcCatalog?

Question 8: How many files (other than the zip files) are there in the CENSUS folder shown in Windows Explorer?

Shapefiles are non-topological vector data sets will have at least 3 and up to 11 files that make up one complete data set. They are VERY common, because they are cheap on memory and easy to create and manage. The three minimum files are the .shp file, which stores the shapes as lists of vertices (points that make up a line) as binary code, the .shx file, which stores the index of the shapes for locating the values in space, and finally, the .dbf, which stores the table or attribute values for each one of the spatial features.

So the reason to use ArcCatalog vs. Windows Explorer is to manage GIS data is so that you don’t screw-up the data management structure. By copying and moving one shapefile in ArcCatlog, you automatically grab all of those separate files. This is also the advantage when dealing with other GIS folder/data structures.

Question 9: Look for the two files that have the .prj extension (you may have to change your Windows Explorer View to see Details). What are they (names)? This will be important to understand the next part of the exercise.

Defining Spatial Reference

In this section, you’ll learn how to apply the GIS concept of spatial or geographic references in the software.

IMPORTANT NOTE: Use the webpage links for metadata and info to check the datums, ellipsoids or spheroids, projections, and coordinate systems of the data sets you have. Make sure that you rely on the metadata on the web to fill in the table, because you’re about to see a problem with the metadata spatial reference tab in ArcCatalog.

In ArcCatalog, click on the metadata tab in the data view, and navigate to your CENSUS folder in the catalog tree and look at the spatial reference (spatial tab in metadata) for all four of the data sets.

Question 10: According to ArcCatalog, what is the “Geographic coordinate system name” for each of the shapefiles? (Provide both the file name and it’s associated coordinate system)

Question 11: What are the Horizontal Datum Name and Ellipsoid Name for GCS_North_American_1983? (Use the Details link)

Question 12: What are the Horizontal Datum Name and Ellipsoid Name for GCS_Assumed_Geographic_1?

Question 13: According to the web link to the metadata of these four datasets, what is the horizontal datum name? (and the one you used to fill in the table)

Question 14: Is the North American Datum of 1983 (NAD83) a local or global (a.k.a. geocentric) datum?

Question 15: Is the North American Datum of 1927 (NAD27) a local or global datum?

So, what’s the big deal, and why do I care that these are different? First, have you heard that line about assuming? “Assume makes an xxx out of “u” and “me.” In this case, it really could, because, if I had to do analysis between two datasets with two different datums, the GIS would locate x and y coordinates according to their respective datums, which will take the same place in space and map it in two different places…..NOT GOOD! Around here, the difference between the NAD27 and NAD83 datums is around 30 meters. Don’t worry, there are ways to convert one datum to another (although not perfectly), and I’ll show you that later.

The other issue deals specifically with the spatial reference metadata. Even if you import the xml document of metadata, the spatial reference information will not overwrite what is on the “spatial tab” (it is actually kept in a separate file). Sometimes this file does not exist, or sometimes it’s incorrect. You will fix this problem. Most of the datasets you’ll use will have spatial reference information stored in a .prj file, or embedded with the data (like ArcINFO coverages or ArcINFO GRIDS), however most census data available on the web do not have the files that make spatial reference easy. Yes, county boundaries and census blocks do, but the other two do not. The way to deal with this is to “define the projection” and create those .prj files.

If ArcToolbox is not already present (the red toolbox in the middle between the catalog tree and the data view), activate it by clicking on the ArcToolbox button.
Find the Define Projection tool within Data Management Tools\Projections and Transformations. DO NOT USE PROJECT.
Double click on the Project Define tool (hammer). The tool window will open.

By the way, if you want to see help on this tool use the Show Help button in the lower right hand corner of a tool window. If it’s already open it will say Hide Help. Note that you can get even more information from help by clicking on the Help icon above the help window. To learn software, and especially GIS, help documentation is usually VERY useful. On the job, I use Help all the time to learn more about GIS stuff.

Browse to select the roads dataset (tgr36067lkA.shp). Notice how the coordinate system window says that it’s Assumed Geographic. Click on the button to the right of Coordinate System to open the Spatial Reference window.
Click on Select, find and select North American Datum 1983.prj which is in the folder Geographic Coordinate Systems\North America. Add it, then click on Apply and Ok on the Spatial Reference window. Note how the Coordinate System window changes to GCS_North_American_1983. Click Ok.
The process window will open and show you what’s happening (take a look by scrolling and reading the code). Close the window when it completes processing.

I would not suggest checking the box that says to close that window automatically upon completion, because when problem solving, we use all the information we can get, and the window reveals what the software is doing.

Check out the spatial tab in the metadata to see if it changed.
Using the same process, define the projection for the places file (tgr36067plc00.shp).
In the HYDRO folder you will have to do the same for the census hydrography (NOT THE LARGE SCALE).
If you were to continue to prepare this data, you would do others, but this is enough for this exercise. If you want to practice, try the files in the TRANS folder. By the way, the armories data set came without ANY spatial reference. Yikes, put unknown in your table.
Check Windows Explorer to see the new .prj files you just created (if you’re not sure if they are new, you can always look at the date of the files)
Close Windows Explorer.

Converting DEMs

The 4 digital elevation models need to be converted to something we can use, similar to the way you dealt with the interchange files…and actually these DEM’s are also only ASCII text files. First, we have to make sure we have the correct Extension “on.” The Extensions, are supplemental software to ArcGIS…and yes, they usually cost extra money. In this case, you will use Spatial Analyst, the GIS software used to deal with Raster GIS. You will convert the text file (“.dem”) to an ArcINFO GRID data format. The GRID format is very similar to the ArcINFO coverage format with a three-folder system.

In ArcCatalog, check to see if the Spatial Analyst Extension is on by clicking on the Tool menu and, and selecting Extensions. A window will open listing the extensions. If there is a check next to Spatial Analyst, then go to the next step. If not, click in the box to the left of Spatial Analyst to “turn it on.” Notice the check mark format. The check mark in this software generally designates something that is active or on….like the layers in the Table of Contents (TOC) in ArcMap too. Click close.
Find and double click on the DEM to Raster tool within Conversion Tools\To Raster in ArcToolbox. The tool window will open.
Use the browse button to select one of the four DEM files (within the PROJDATA\DEM folder) and make sure the output raster is being saved back into the same folder. Keep the Output data type as FLOAT (a real number as opposed to INTEGER), and keep the Z factor at 1(a multiplication factor for z values…z values are elevation values…x’s are eastings, y’s are northings, and z’s are the 3^rd dimension). Again, if you want to better understand what the other options do, check help in the window to the right by using Show Help. Click OK. The processing window will open. Close it when it’s completed.
Do the same for the remaining three DEMS.
Now click on one of the completed DEMs (look for the waffle), look at it’s spatial reference in the metadata in ArcCatalog.

Question 16: What is the Grid Coordinate System Name? What are the planar distance units? What is the resolution (with the correct distance units)?

Look at the metadata for this data, either by importing the associated .xml document (36dea.xml) into ArcCatalog, or by using the weblink for this data from CUGIR.

Question 17: What is the Altitude Datum Name? What is the Altitude Resolution? What are the Altitude Distance Units?

Question 18: Is the vertical datum used here for z-values (elevation) based mostly on an ellipsoid or a geoid?

Before we go further, there is another small problem. What were the units of vertical distance? Let’s check that using ArcCatalog.

Highlight one of the DEM’s in the catalog tree of ArcCatalog and click on the Preview tab so that you can see the geography. Use the identify tool to click around on the image.

I don’t know about you, but I seriously doubt that these are decimeters (1/10 of a meter). I know for example that Pompey Hill is around 1700 ft of elevation. Converted to decimeters that would be well over 5100 decimeters. The number I looked at near that area was just over 500 units. Aha! The units are not decimeters anymore; they are actually meters. The software converted them. Good to know, eh? That could really screw this stuff up….by a factor of 10. If I wanted to keep decimeters as the unit of elevation, I would simply change the z-factor to 10.

One more issue before we leave the DEMS. According to the metadata, USGS quad sheet contours were used to create this data. Generally, the accuracy of USGS contours are +/- half of the contour interval, so in most cases between 2.5 to 10 feet depending upon the contour interval for the individual quad map used.

Question 19: Given the known error discussed above, would it be wise to claim that an analysis using this data was accurate within a decimeter?

Evaluation of Competing Data

You downloaded a few hydrography (water) datasets that you better take a look at before we decide which one to use. You downloaded two different data sets of streams and rivers. One was the Hydrography (Census 2000) tgr36067lkH.shp, and the others were the Large Scale Hydrography Network and Surface i.e. the imported/coverted s4140201 and n4140201 .e00 files (whatever you named them after importing them). First, take a look at the table that you’ve been filling-in based on the metadata.

Question 20: What is the scale of the Census Hydrography data? What is the scale of the Large Scale Hydrography data?

Question 21: Which one of those has the larger scale?

Question 22: Based on scale alone, which data set would you keep if you wanted greater accuracy?

Now, take a look at the two data sets simultaneously in ArcMap to see how they compare.

Open ArcMap (see Exercise 1 if you don’t remember how).
When the first window opens, click on the option for a new and empty map and click ok.
Use the Add Data button to add the Oneida and Seneca hydrography network data and the census hydrography data.
Symbolize them with contrasting colors (so that you can tell them apart) by clicking once on the symbol-color below the name of the layer in the TOC. Notice all the options for symbolization that you can access from this window. The color option is on the right. You may also want to change the size of the symbol as well…whatever you need to do to differentiate the three layers.

By the way, because of ArcMap’s projection on the fly feature, you were not prompted about a problem with spatial reference because if you did the define projection step as directed, both datasets have the same datum, NAD83, even though they have different units of distance (large scale is UTM meters and census is Decimal Degrees). The units of distance and coordinates in the Map or Data View of ArcMap (look at the lower left) will default to the first dataset loaded.

Question 23: Are UTM meters a projected or geographic coordinate system?

Question 24: Are Decimal Degrees a projected or geographic coordinate system?

Right click on the Census hydrography layer in the Table of Contents (TOC) and select Zoom to Layer.
Compare the large-scale hydrography and the census hyrdrography by turning off and on (by clicking the check mark in the TOC) the census hydrography.

With some minor exceptions, the two different data set have comparable amounts and locations of the water features within Onondaga County. So, they look about equal here.

The U.S. Department of Commerce (Census) is accountable for one dataset, and the New York State Department of Environmental Conservation put together the large-scale hydrography. Based on the scale issue from above, and now the agency, the theme of this data (water features), and the most local governmental jurisdiction, the Large-Scale Hydrography would be a better fit for our purposes in this GIS project.

Close ArcMap, you don’t have to save the project.
In ArcCatalog, delete the Census 2000 hydrography data. If it doesn’t work, you may have to refresh ArcCatalog, by right clicking on a folder one level above the data set and selecting refresh. The easiest way is to refresh everything is to highlight “Catalog” at the very top of the catalog tree and do the refresh.

Fitting the Data to the Geographic Scope

Now it’s time to manipulate these datasets so that they fit our geographic area of focus and reduce our memory needs…and you thought this lab should be over by now, eh? Again, I’m really not doing this to torture you, but to give you a thorough understanding of how the concepts that we’ll talk about in lecture apply to actually doing GIS work. After you survive this semester, you will have no excuse for not anticipating the amount of time and money that you may have spend to get GIS work done in the future, especially the second step of the GIS process and generally the most expensive step, Data Management and Acquisition.

While a lot of this work will only apply within the boundary of the city of Syracuse, we may want to be sure to keep track of stuff upstream of the boundary, so perhaps using the Onondaga County boundary file may work as our “cookie cutter” so that you can limit your memory storage needs.

First some terminology: when you put together smaller vector data sets to make a bigger one, you’ll merge them; when you put together smaller raster data sets to make a bigger one you’ll mosaic them; finally when you want to make smaller vector or raster datasets from larger ones you’ll clip them, or you may export them based on selection of specific features of interest.

Mosaic’ing the DEMs.

In ArcCatalog, find and activate the Mosaic To New Raster tool within Data Management Tools\Raster of ArcToolbox.
With the browse button, add all 4 of your DEMs, make the output location the DEMS folder, and name the Raster data set SYRDEM. Do not change any of the other options, and click OK.
Check out your nifty new DEM for the city in the data view.

Merging and Clipping (VECTOR) the Hydrography

In ArcCatalog, find and activate the Merge tool from Data Management Tools\General of ArcToolbox.
Use the browse button to add the Hydrography Network data set arcs for the Oneida and Seneca watersheds (NOT SURFACE). You’ll notice that when you add the coverage, it will ask you to pick which part, pick the arc. Make sure the new data set (named something like arc_Merge.shp) is getting saved within the HYDRO folder. Click Ok, and then close the processing window when it’s done.
Find and activate the Clip tool from Analysis Tools\Extract of ArcToolbox.
Use the browse button to select the merged shapefile you just made and then select the Census county boundary file to use as the clip feature. Again make sure that the output feature class is getting saved to your HYDRO folder. Don’t change the cluster tolerance, and click ok.
Take a look at your clipped hydrography in ArcCatalog.
Check out the metadata for this data by clicking on the Metadata tab, and then the spatial tab. Look near the bottom for lineage and notice that your processing step was logged by the metadata.
If you want some practice, go ahead and clip some of the other shapefiles (like statewide data sets of dams or precipitation) using the county boundary.

Clipping (RASTER) the Classified Land Cover

Connect the Greenfield/PROJDATA/COVER folder in ArcCatalog.
Find and activate the Clip tool from Data Management Tools\Raster of ArcToolbox. NOT THE SAME AS ABOVE! This one is specifically for raster data.
Input the New York land cover image from my COVER folder (the one with the .tif extension and NOT the CLIPNLCD) and then click on Environments. When the Environment Settings window opens click on General Settings and scroll down to Output Extent. Use the browse button to select the county polygon shapefile in the CENSUS folder. Note that the Top, left, right, and bottom windows are filled in with decimal degree values from the county boundary shapefile. Click ok, which returns you to the Clip tool window.
Now press the clear button, notice how the numbers changed…and they are meters now. Make sure the output raster is being saved into your COVER folder, and name it with 8 characters or less. I named mine CLIPNLCD
Take a look at the clipped land cover data.

Exporting Selected Features

Start ArcMap and click okay on the first window that opens to select a new and empty map.
Add the Census place shapefile (tgr36067plc00.shp)
Right click on the layer in the TOC and select Open Attribute table.
Click on the Options button and choose Select by Attribute.
While you have this window open, read the information and the wording on the window to better understand this selection function. Keep the method as Create a new selection. Double click on “NAME” so that it appears in the bottom window. Click once on the = button, and then click on Get Unique Values. From the names that appear in the window above that button, double click on Syracuse. The expression in the bottom window should look like: “NAME” = ‘Syracuse’ and then click Apply.
You’ll see the city boundary of Syracuse highlighted in blue. Right click on the layer in the TOC and select Data and then Export Data. A window will open. Read this over too.
The top should say Export: Selected Features, select the option (radio button) for this layer’s source data, and then use the browse button for the output file to save this layer as Syracuse within the CENSUS folder. Click okay. You’ll be prompted to add the layer to ArcMap, click yes if you want to see it.
Now go to the Selection menu and select Clear Selected Features (I’ve been in the practice of doing this because a lot of the ArcMap tools function on selected features, and I’ve often made the mistake of processing only the selected records when I instead wanted the process to work on the entire data set.)

Clipping (RASTER) the DEM with adjustment for the correct Spatial Reference

This one will be a little different because we want this to be fairly accurate, and I want to show you how to change projection and convert a datum.

The Syracuse DEM has a projected coordinate system of UTM meters in Zone 18 based on the NAD27 datum (shorthand for North American Datum of 1927). You’re going to change the spatial reference of the Syracuse city boundary shapefile you just created so that it more appropriately fits the DEM spatial reference.

In ArcCatalog find and activate the Project tool within Data Management Tools\Projections and Transformations\Feature. For the input data set select SYRACUSE.shp, for the Output Dataset name it SYRUTM27 and make sure it will be saved in the CENSUS folder.
Click on the button next to the window for Output Coordinate System, and after the Spatial Reference Properties window opens, click on Select and find and select (add) the NAD 1927 UTM Zone 18N.prj file within Projected Coordinate Systems\UTM\Nad 27 folder. Click Apply and OK on the Spatial Reference Properties Window.
You’ll see the new Output Coordinate system loaded. Now use the drop down menu next to Geographic Transformation (optional) to select NAD_1983_To_NAD_1927_NADCON option. You will see it load below. I use this conversion method most of the time to convert between NAD27 to NAD83 or NAD83 to NAD27 datums in New York State. It’s basically an equation that “makes it fit.”
Click Ok, and click close on the processing window when it’s done.
Clip the SYRDEM using the SYRUTM27 shapefile, like you did for the land cover data set above. Make sure you are using the extent of SYRUTM27 and saving back into the DEM folder.
Check it out to see how it worked by opening ArcMap and adding the clipped SYRDEM (what did you name it?) and the SYRUTM27 data. Generally, ArcMap will always place vector data on top of raster data in the TOC.

While you have ArcMap open, play with some the ArcMap features for symbology at this point. There are a couple different ways to see through the SYRUTM27. Double click on the color below the name of the Layer SYRUTM27. The symbol selector window will open. On the right under options, you can use the drop down arrow and select no color. Now make the outline color thicker by changing the outline width to 1, and change the outline color to black. Change the color of the raster data (the clipped DEM) by clicking once on the color and changing the color ramp. The other way to change the ability to “see through” a layer is to double click on the name of the layer to get the properties window, selecting the display tab, and changing the transparency percentage.

Use the icons in ArcCatalog to help you fill-in the field for raster or vector data in the table.

Don’t worry about saving your project, and you can close ArcCatalog.

If you have to continue filling-in the table, spatial model, or answer sheet don’t do the next step.

I will give you data sets to start with on the next lab, so you don’t have to worry about saving these. Please delete your work off the hard drive.

Whew! There are some minor things to do now to prep for the next exercise, but you’ll do them then.

Name _______________________________________ Lab Section ________

ERE 450/550 Lab Exercise 2 Answer Sheet – Lab due 9/29 at 5pm

Hours spent doing the lab_______________ Score_____/44

Question 1: What is the “Geographic coordinate system name:” under “Horizontal coordinate system?” (Put your answer on the answer sheet) (1)

Question 2: What are the Geographic Coordinate Units? Are these units of measurement planar (Cartesian, rectangular, or projected) or spherical (angular, geographic, or geodetic)? (2)

Question 3: Look under Geodetic Model. “What is the Horizontal Datum Name?” What is the “Ellipsoid Name?” Is an ellipsoid a spheroid? Is a spheroid an ellipsoid? (4)

Question 4: What is a datum? (You may use the very simple one to two word definition.) (1)

Question 5: What are the four data sets within this ArcINFO coverage? (4)

Question 6: What are the five data sets within this ArcINFO coverage? (5)

Question 7: How many shapefiles are shown in ArcCatalog? (1)

Question 8: How many files (other than the zip files) are there in the CENSUS folder shown in Windows Explorer? (1)

Question 9: Look for the two files that have the .prj extension (you may have to change your Windows Explorer View to see Details)? What are they (names)? This will be important to understand the next part of the exercise. (2)

Question 10: What is the geographic coordinate system name for each of the shapefiles? (Provide file name and the coordinate system) (4)

Question 11: What are the Horizontal Datum Name and Ellipsoid Name for GCS_North_American_1983? (Use the Details link) (2)

Question 12: What are the Horizontal Datum Name and Ellipsoid Name for GCS_Assumed_Geographic_1? (2)

Question 13: According to the web link to the metadata of these four datasets, what is the horizontal datum name? (and the one you used to fill in the table, right?) (1)

Question 14: Is the North American Datum of 1983 (NAD83) a local or global (a.k.a. geocentric) datum? (1)

Question 15: Is the North American Datum of 1927 (NAD27) a local or global datum? (1)

Question 16: What is the Grid Coordinate System Name and zone? What are the planar distance units? What is the resolution (with the correct distance units)? (3)

Question 17: What is the Altitude Datum Name? What is the Altitude Resolution (with the correct distance units)? What are the Altitude Distance Units? (3)

Question 18: Is the vertical datum used here based on an ellipsoid, or a geoid? (1)

Question 19: Given the known error discussed above, would it be wise to claim that an analysis using this data was accurate within a decimeter? (1)

Question 20: What is the scale of the Census Hydrography data? What is the scale of the Large Scale Hydrography data? (2)

Question 21: Which one of those has the greater scale?(1)

Question 22: Based on scale alone, which data set would you keep if you wanted greater accuracy? (1)

Question 23: Are UTM meters a projected or geographic coordinate system? (1)

Question 24: Are Decimal Degrees a projected or geographic coordinate system?(1)