Look at the GPS Track in the Context of a Variety of GIS Data
The subject of this topic is using GPS for input to GIS. Now that you have GPS data in ArcInfo let’s look at it with other GIS data. We have avail able several digital maps and images of a portion of the Kentucky River plus a couple of vector coverages of a few arcs from two USGS quadrangle maps. The county to the north is Fayette; that to the south is Madison. Of the two 7.5 minute quadrangles, the western one is COLETOWN; the other is FORD. In the folder_:\GPS2GIS\RIVER exist the following:
• A line coverage digitized from the Ford, Kentucky quadrangle: FORD_VCTR
• A line coverage digitized from the Coletown, Kentucky quadrangle: COLE_VCTR
• A TIGER line file (shapefile) of the streams of Fayette County, Kentucky: FAY_TIGER
• A TIGER line file (shapefile) of the streams of Madison County, Kentucky: MAD_TIGER
Figure 6-22. Portion of ArcInfo GPS track showing point IDs.
• A digital raster graphics file scanned from the Coletown topographic quadrangle: COLE_DRG.TIF
• A small orthopohoto GeoTIFF image12 (it is a digital orthophoto quadrangle that is 1/64th of a regular 7.5 minute USGS quad): COLE_DOQ64.TIF
• A Digital Elevation Model (DEM) in the form of an Arc/Info GRID consisting of "pillars of elevation" that are approximately 30 meters square: COLE_DEM
• A vector line coverage of the elevation contour lines generated from the DEM: COLE_CONTOURS
• A Triangulated Irregular Network (TIN) showing elevations derived from the DEM, in the form of an ArcInfo TIN: COLE_TIN
• A vector (polygon) coverage of most of the Coletown quadrangle showing soils: COLE_SOIL
• A vector polygon coverage of the Coletown geologic (surface rock) quadrangle: COLE_ROCK
• An Arc Macro Language (AML) file named DKR. AML (Display Kentucky River) to display FORD_VCTR, COLE_VCTR, and the
![tmpF3-42_thumb[2]_thumb](http://what-when-how.com/wp-content/uploads/2011/10/tmpF342_thumb2_thumb.png "tmpF3-42_thumb[2]_thumb")
GPS track. You may run this file after displaying other GIS data. It will cut down on the typing you do. (There is also a duplicate of this file named DISPLAYKYRIVER.AML.)
In the steps below you will add these feature-based, grid-based, tin-based, and image-based themes.
The AML file is there to prevent you from having to do a lot of typing. You may use it or modify it as you wish. It says: {__ } Study the AML carefully; understand what it will do when executed by ARCPLOT; we will use it shortly.
{__ } CLEAR the ARCPLOT window. Set the mapextent to the GPS track, and change the workspace to __:\GPS2GIS\RIVER by typing the following
MAPEXTENT POSNPNT ARC W_:\GPS2GIS\RIVER
{_} Change the viewing area: Rearrange the ArcInfo and ARCPLOT windows so that the graphic display area has the lion’s share of the left side of the screen. Leave a piece of the Arc window over to the right so you can click on it when you want to type commands.
In the next steps you will use the AML to add data that have been converted to Kentucky State Plane coordinates in the NAD 1983 datum:
• the vector coverages based on two USGS 7.5 minute topographic quadrangles FORD_VCTR and COLE_VCTR, which contain a few arcs digitized (without great accuracy) from the topo quads Ford and Coletown (In particular, an attempt was made to trace the banks of the Kentucky River and some local highways from the quad sheets);
• the TIGER-based coverages depicting the streams in Fayette County and Madison County;
• the GPS track
{__ } To see some of the vectors from the topo maps, the streams in the vicinity of the Kentucky River, and the GPS track type:
&RUN _:\GPS2GIS\RIVER\DISPLAYKYRIVER.AML or, because you have set as the default the _:\GPS2GIS\RIVER workspace, because there is a duplicate file named DKR.AML, and because &R is a substitute for &RUN, you may type: &R DKR13
You should see the GPS track in red plus a few arcs from the Coletown and Ford topo sheets (some of the arcs outline the river) in green and some blue arcs from the TIGER line stream files from Fayette County and Madison County. Each county treats the bank of the Kentucky River that is in its territory as a stream—so green and blue arcs are sometimes coincident and sometimes not, depending on the accuracy with which each was depicted. The starting point of the trip is in the southeast and you can see a green topographic line outlining the island at the other end of the trip.
{_} Use the ARCPLOT measure command: You type MEASURE LENGTH and use the mouse cursor, clicking at the points along the path you want to measure. Follow the instructions in the Arc window. To complete the measurement and get the results, press the 9 key with the cursor in the ARCPLOT window; see the results in map units (survey feet) in the Arc window.
Run the AML again to put the vector data back on the screen. Click on the title bar of the ARCPLOT window to hide the Arc window. Then get the Arc window back.
{_} Further zoom up on the bend in the river: Type MAPE *. Then make the box so it includes the GPS track that shows in the DRG. CLEAR the window. Then redisplay the DRG and rerun the AML. What is the NORMAL POOL ELEVATION according to the DRG?_
{__} Display the orthophoto file: COLE_DOQ64.TIF: Type: IMAGE COLE_DOQ64.TIF and redisplay the GPS track.
You can get an interesting perspective on the images by alternating between COLE_DRG and COLE_DOQ64.TIF. (Use the up and down arrow keys to alternately bring up the commands to generate the two IMAGES.)
{__ } Display the digital elevation model that is in the form of anArcInfo GRID: COLE_DEM. A little wand waving is involved. If you don’t understand the ARCPLOT commands don’t worry about it. Type SHADECOLORRAMP 1 5 BLUE GREEN GRIDSHADES COLE_DEM VALUE EQUALAREA Redisplay the GPS track and vector coverages. Use the command ARCPLOT command CELLVALUE COLE_DEM * together with your mouse to find the elevations of various points on the GPS track. (The results are displayed as "VALUE" in the Arc window.) Do the answers appear to be at least roughly consistent with the Normal Pool Elevation you read from the DRG earlier? Check some of the surrounding elevations. Press "9" to end the CELLVALUE command.
{__ } Redisplay the DRG. Using the CELLVALUE command again, check out the topo map elevations. Look at the topo map where the text shows elevations of 600, 700, and 900 feet near the bend in the river. You should see some agreement between what the topo map says and the values of the underlying grid.
{__ } Display digitized contour lines of the Coletown Quadrangle:COLE_CONTOURS. The digital elevation model COLE_DEM was converted to an ArcInfo line coverage—another way to represent altitude. You can display the contour lines around the river by selecting those whose "CONTOUR" value is between 550 and 700 feet: RESELECT COLE_CONTOURS LINE CONTOUR >550 AND CONTOUR <700 ARCS COLE_CONTOURS &R DKR
{_} Display the Triangulated Irregular Network ArcInfo TIN:COLE_TIN. The digital elevation model COLE_DEM was converted to an ArcInfo TIN—a third way to represent altitude. You can display the triangles of the TIN in numerous ways. Since we want to look primarily at the river we will show the triangles by slope, as the slope of the river should be close to zero and the area around it is quite steep. Type: CLEAR
SHADESET COLOR.SHD TINSHADES COLE_TIN SLOPE &R DKR
The least slope category is shown by light beige and you can see the GPS track in the lightest areas. But the result is not too satisfying . To make real use of this sort of map you would need the legend.
Also look at the average elevations of the vertices of each triangle. CLEAR
TINMARKERS COLE_TIN SPOT &R DKR
The lowest elevations are marked by white "+" symbols. Again, not too satisfying. But our point here is a quick demonstration of how GPS and GIS data can be matched up.
{_} Display an ArcInfo polygon soils coverage: COLE_SOIL, using the category "Minor1." Type: CLEAR
POLYGONSHADES COLE_SOIL MINOR1. Run the AML. The soil types are considerably different around the river.
{__ } Look at the polygon attribute table: LIST COLE_SOIL.PAT.
{__ } The Coletown geologic quadrangle has also been digitized. Its name is COLE_ROCK. After clearing the window, bring it up as well. Type:
POLYGONSHADES COLE_ROCK NAME GQ_LUTBL followed by &R DKR. Because the river has cut down through the layers of rock, the exposed surface shows rather dramatic changes over short distances.
{__ } Spend a few minutes looking at the various coverages and images at different scales. This may not be too easy—and even if you can do it well, you can still appreciate why ESRI developed Arc View.
Of course, there is a long story behind each of the types of data that we have looked at briefly here. If you are interested in a particular type or source of data I suggest that you look first at the web for sources and then study web pages or texts for more detail.
{__ } Then quit ARCPLOT
{__ } Change the workspace back to _:\GPS2GIS\RIVER_AI\EXPORT.
{__ } Use the ArcInfo command RENAME to change the name of the coverage from POSNPNT to BOAT_SP83_yis. RENAME POSNPNT BOAT_SP83_yis
{__ } Minimize the ArcInfo Window.
PROJECT 6b-B
Blunders Caused by Using the Wrong Datum
One of the central difficulties in using GPS for input to GIS is mismatched parameters, as I have stated so often that you are sick of reading it. Indeed, the issue of getting the parameters of geo data sets to match is a general problem with GIS. Let’s look at what happens if we aren’t careful.
{__ } Bring up Pathfinder Office again. If you click on the icon for the Export Utility, the program will still be set to export the GPS file C111315a.COR to _:\GPS2GIS\RIVER_AI\EXPORT.
However we want to make some changes, so click on New Setup. In the Export Setup Name blank replace "Copy of AI_yis#1" with "AI_yis#2." Click OK and up will come our six-tab window.
{__ } Make no changes in Data, Output, Attributes, or Position Filter. However we want to make this export setup so it makes a coverage in the UTM coordinate system in the NAD27 datum. Also we want to set the Export Parameters specifically, rather than relying on the current display parameters. So, under the Coordinate System tab select Use Export Coordinate System and then click on Change. Make the System Universal Transverse Mercator. The Zone should be 16 North, the Datum NAD 1927 (Conus), Coordinate Units Meters, Altitude Units Survey Feet, and Altitude Reference MSL.
{__ } Under the Units tab also use Export Parameters. Use meters for all three options. Click OK.
{__ } Double-check the Export window, then click OK. You may be told that files will be overwritten. That’s okay, since you renamed the previously made coverage POSNPNT to another name, thereby preventing it from being involved in these new operations. A new set of AML files (which will generate a new coverage named POSNPNT) will be generated.
{_} With a text editor add the "..\" you needed before to make the AML work.
{__ } Go back to ArcInfo (you can switch back and forth between ArcInfo and Pathfinder Office simply by clicking the appropriate buttons on the Start bar, or by using the Alt-Tab combination). Run the POSNPNT AML to make the new coverage. Then RENAME the new POSNPNT coverage that is in _:\GPS2GIS\RIVER_AI\EXPORT, to BOAT_UTM27_yis.
{_} In Pathfinder Office again, export C111315a.COR for the thirdtime: This time make an export setup named AI_yis#3 that makes a shapefile in the same UTM zone but uses WGS 1984 as the datum.
{__} Fix the stupid AML.
{__ } Back in ArcInfo, make the new coverage and rename it BOAT_UTM84_yis.
{_} In ArcInfo, display the following coverages:
Points in _: \ GPS2 GIS\RIVER_AI\ EXPORT
\BOAT_UTM84_yis
Arcs in_:\GPS2GIS\RIVER\COLE_UTM
Arcs in_:\GPS2GIS\RIVER\FORD_UTM
{__ } Note that the boat is no longer in the river. Now add the points from _:\GPS2GIS\RIVER_AI\EXPORT\BOAT_UTM27_yis which is based on NAD27—the same coordinate system as the map from which the river was digitized. Observe the distance between the tracks. Note that the difference is much greater in the north-south direction than in the east-west direction. In this section of the country the UTM coordinates were moved about 200 meters along the meridians and almost not at all on the parallels when the new datum was calculated. Other sections of the United States will, of course, have different values.
{__ } Zoom up on a portion of the GPS tracks. Measure how much displaced (in meters) equivalent points are from each other in a north-south direction. _ How much in an east-west direction?_. Recall the results you got from the latitude.
{_} The banks of the river were digitized from topo quad sheets.
In which datum would you say those sheets were developed?
This is another illustration that while the difference between NAD27 and WGS84 may be only a few meters based on latitude and longitude coordinates, it may be hundreds of meters different when based on UTM or state plane coordinates.
PROJECT 6b-C
{_} Back in Pathfinder Office, generate the files for a new coverage that is an arc rather than a sequence of points.
Basically, you will repeat PROJECT 6b-A, using the Export Setup AI_yis#1 except this time choose the "One line per group of Not in Feature Positions" option under the Data tab. Under the Attributes tab select Length (2D) and Length (3D) as Line Feature attributes. The GPS fixes will generate files to make an ArcInfo line coverage rather than a point coverage. Those files will be POSNLINE.AML, POSNLINE.AA, and POSNLINE.GEN. The coverage name will be POSNLINE rather than POSNPNT.
{__ } Run the POSNLINE AML (after fixing it with "..\") to produce the coverage. Display POSNLINE with ARCPLOT. Look at POSNLINE .AAT. First observe that arc attribute table is very dull (only one record) compared with the (hardly exciting) point attribute table that showed each point. The time given is the starting time of the file. Also look at the two fields in the file labeled Gps_length and Gps_3dlength. Gps_length is the sum of the distances, in feet, of the GPS arc segments, considering only the two-dimensional coordinates of each point. Gps_3dlength is the sum of the lengths of the sequence of lines that connect the three-dimensional coordinate points. You would expect it to be longer, but not that much longer, unless the Kentucky was a major whitewater river. The large difference is due, of course, to the GPS errors in recording the heights of the points that make up the arc.
When you execute an AML (or SML) file using a GEN file as input, a single arc will be created. This will be true, even if the arc crosses over itself. However, if you had such an arc which intersected itself, due to errors in the GPS recording process, and used BUILD with the LINE option, it would fail. If you CLEAN the coverage, you might get several arcs (and polygons). Without going into detail, let’s simply say that you have to be careful when you translate GPS files into arcs. Examine the results carefully and be prepared to edit.
PROJECT 6b-D
{__ } In the __:\GPS2GIS\RIVER folder find a GPS file named KYCLAUTO.COR. This is a track from the marina, north on an access road, north onto I—75 to Lexington’s New Circle Road, and west on New Circle.
{__ } Convert this file into an ArcInfo LINE coverage compatible with FORD_UTM and COLE_UTM.
{__ } Graphically overlay this coverage with FORD_UTM and COLE UTM.
PROJECT 6b-E
{_} Now comes the real test! Can you integrate GPS data that you collected with GIS data that you digitize? In this project you might use the SSF file from PROJECT 2B, in which you moved the antenna through space. Whatever data you decide to use, you might want to plan to digitize the general area of the path on a USGS quadrangle.14
