Surveying with the Total Station: Creating a Micro/Meso Topography of Campus

Introduction

Continuing on with surveying methods, this week's lab focused on collecting data with x, y, and z coordinates. Working in small groups of three, we were given the task of surveying out the micro/meso topography of a one hectare plot area. This was done using a Topcon total station blue toothed to a GMS-2 GPS.

Study Area - For this project our study area was a hectare (100 meters x 100 meters) plot located in the campus mall area of UW-Eau Claire. Most of the study area was uniform in flatness, with a general sloping trend towards the stream which runs almost parallel to the path between flags 2 and 3, shown in figure 1. The area around the stream had a much greater slope which will be evident by the DEM images created in ArcScene, shown later in the report.

Figure 1. This panorama was taken from the corner of an approximately square hectare area. The study area,
highlighted in yellow, is located on the University of Wisconsin - Eau Claire campus mall area. The blue arrow
shows the occupied point where all the points were surveyed from with the total station. The orange arrow,
just to the left of the bridge, shows the location of the back sight used to orient the total station.

Methods

Figure 2 Shows how to collect a
point with the Total Station. One
operator will look through the lens
of the Total Station aiming the
crosshairs in the center of the prism
on the top of the pole. The other person
will try to remain as motionless as possible.
Once the point is "shot" the pole person
moves to the next point.

I have a more detailed step by step walk through of the total station on my "how to" blog. However, I have listed a VERY generalized process below.
1) Set up the Total Station....secure tripod into the ground, use the proper levels the balance the system.
2) Set up the Blue Tooth from GPS to Total Station.
3) Set up a TopSurv Job.
4) Collect GPS points
5) Set your occupied point and your backsight
6) Collect Data
7) Export points to a shapefile from TopSurv using Active Sync.
The exported file will be in a text format. Before a shapefile can be created the text file had to be opened up and edited.

Header>>Delimiter(,)FileFormat(Name,Lat(North), Lon(East),Ht(G),Codes)<<

The highlighted portions need to be deleted. After edited, the file can be exported as a text format in comma delineated format. From here the file can be converted into a shapefile. Figure 3 shows the results of the total station after the shapefile was created.

Figure 3 shows the one hectre plot of land we surveyed with the Total Station. As you can see the UW-Eau Claire campus has changed recently, we surveyed in the middle of the green space shown in figure 1. 

Once the elevation points were added to ArcMap, Kriging (3D Analyist) was used to create a elevation surface. The decision to use Kriging to create the surface was based off my analysis during a previous lab which analyzed the various types of interpolation methods: IDW, Natural Neighbor, Spline, and Kriging. My main decision to use Kriging was because it created a new surface and created new values even where data was not collected. Kriging creates the new surface based off the first rule of geography, assuming that closer points are more closely related than those farther away. Figure 4 shows the results of my group's kriging surface from the surveyed data.

Figure 4 shows the resulting kriging surface from the Total Station survey. Over 130 points were collected over this one hectre area. More points were collected over area's with high slope, all of which were along the river. The elevation change from this model was just under five meters.

Once the surface model was created for my group, the other groups survey models were added to ArcScene. Although each group had to survey a one hectare area on the campus mall, not all surveyed the exact same location. Even with the slight differences in the area surveyed you can still see the differences in elevation that the survey's recorded, figures 5 and 6 show just how well the survey's match up.

Figure 5 shows five of the groups in Field Methods all overlayed on top of one another. For a better
analysis of the area I have created some landmark features giving clues to the elevation changes on the
kriging models. The bridge over the Little Niagra creek was difficult to shoot and may look different if
more points were collected closer to the base of the bridge. Also take note in the gaps between the surfaces.
There was over eight meters between the lowest values in group 2 and those of group 6. The color scale on
the top of figure 5 is only applicable to the top layer of group 2. 

Figure 6 shows a view point almost looking almost parallel to the ground. This viewpoint better illustrates the differences in elevation between the five group's layers. Groups 1, 3, and 4 had similar elevation values; meanwhile, groups 2 and 6 were outliers, sitting either significantly lower or higher, respectively. This discrepancy likely is the result from the inaccuracies of the GPS unit used to collect the occupied point for the Total Station. This goes to show ho the accuracy of the survey, with the Total Station, is only as accurate as the GPS used to collect the occupied point.

Discussion

Most of the difficulties with this lab came during the initial setup phase, specifically when trying to set the back sight. We had briefly been instructed on the basics of the total station; which included, set up of the total station for both the physical assembly and leveling of the station as well as the technical steps necessary before surveying can begin. However, we lacked the hands on experience of operating the total station, which resulted in amateur mistakes. When we tried to set the back sight for the total station to orient north with, we made the mistake of having the lens flipped in the wrong direction. We thought since the knobs could be used to precisely move the lens that they should be facing towards you. It was not until 2 hours of trouble shooting that we finally called it a day and had to try again over the weekend after we had found the error in our methods.

Additional difficulties came with trying to accurately capture the elevation changes along the river. One of the problems arose with a broken knob on the pole which raised and lowered the height of the prism. It was set at 2 meters, so this wasn't a major concern. However, if the knob was intact we would have raised the prism higher when we were surveying behind the bridge. There were a few spots along the stream just behind the bridge that we decided to skip because we didn't find it necessary to move the total station to collect two or three more points. Also, just having a varied surface was a difficult trade off. Since we had never used the total station to collect elevation we were not sure of how many points would be necessary to accurately portray the elevation changes. We did take more readings along the river, where the most severe elevation changes were found.

In addition to individual group struggles, there is also the issue of accuracy that comes with the GPS. The Total Station itself is extremely accurate. However, the Total Station has to be given an occupy point. This occupy point is the reference point for all other points in the survey; unless the total station is moved, then the new occupy point will act as the reference point. So although each group may have followed similar procedures and surveyed a relatively similar area, the GPS gave different reference heights for each survey and as a result the kriging surfaces don't match up, in regards to elevation.

Conclusion

Surveying with the Total Station was an effective experience into another important method of data collection. Prior to this I had only used a GPS to collect data. Once the Total Station was set up, the actual surveying went much faster than using a GPS; also, the Total Station was much more consistent. With a GPS you can get high PDOP values behind buildings and your data will be skewed. With the Total Station, there is likely only a handful of GPS points you will ever need to collect, lessening the chance of multipath effects on the GPS signal. In the future I hope I have the opportunity to use a Total Station again, but with a more accurate GPS at the survey grade level.