Creating a Topographic Survey

 

Introduction

   In this lab we are tasked with creating a topographic survey of the UWEC campus mall using two different surveying techniques. The first method is by plotting points using the Tesla RTK with HIPER SR (Figure 1) attached. The second method we will be using the Tesla RTK and Topcon Total Station (Figure 2) to collect elevation points throughout the campus mall. These methods will then be compared to determine which method is more time efficient and produces more accurate data. 

 

Figure 1  Topcon HIPER SR GPS unit used in concurrence with the Topcon Tesla RTK to collect elevation points throughout the UWEC campus mall. 

Figure 2  Topcon Total Station used in many surveying applications. This unit using distance, vertical angle, and azimuth to create topographic points.

 

Study Area

   The study are for this lab, as previously stated, is the UWEC campus mall (Figure 3). It is located between between  the new W.R. Davies Center and Schofield Hall. This area has a relatively flat terrain that gradually slopes toward the West/Southwest. The mall itself ranges from 241.043 meters above sea level to 237.923 meters above sea level over an 84 meter distance. There is also a creek that runs East/West through the campus mall. This area has the greatest elevation change ranging from 235.884 meters above sea level to 238.627 meters above sea level over a 22 meter distance. 

Figure 3  The campus mall on the University of Wisconsin- Eau Claire campus. 

Methods

  The first portion of this lab was designated for survey collection using the Topcon Tesla RTK and HIPER SR attachment. For this process, a project was set up on the Tesla RTK device. Then some basic parameters were set, including coordinate system and height of the HIPER SR. The Testla RTK was then connected to the HIPER SR using a mifi connection. After the connection was set we were ready to start collecting points. For this process all that was required was the level the HIPER SR and hit save on the RTK to save the point into the job folder. After 100 points were collected we saved the job as a .txt file on the RTK, connected it to a computer, and imported the file into a geodatabase (Figure 1).

   The next step was to go into the field and collect elevation points using the Tesla RTK and Total Station. This process was significantly more time consuming than collection with the combination above. The first step in the process is to set up a job similar to above. Then we need to set up a the occupancy point and backsight. This is done to provide an azimuth reference for the Total Station to use while collecting points. The occupancy point and backsight are set up similarly to the process above where we level the Tesla RTK/HIPER SR and save the point into the job file. From there, we need to set up the Total Station directly over the occupancy point, measure its height from the ground, and level the Total Station. We then connected to Total Station to the RTK, input the occupuancy point and backsight into the RTK, and lazed to the prism which was set up on the backsight point. After the Total Station was set up with the occupancy point and backsight we collected 64 points around the campus mall. After the points were collected we saved the job file as a .txt file, similarly to the process above, and imported it onto the elevation geodatabase (Figure 1). 

Figure 4  Example of the .txt file that was imported from the Tesla RTK device. This file was then created into a point field with x,y,z coordinates. 

   We then used the Create Feature Class From XY Table (Figure 5) in ArcMap to create a point feature class with the x,y, and z coordinates for each of the .txt files and added them to the Arcmap display (Figure 6).

Figure 5  Create Feature Class from XY Table tool that was used to create each of the individual .txt files into the survey points.

Figure 6  Locations of each of the surveyed points collected  in the lab. The green points are those that were collected using the Tesla RTK and HIPER SR, while the blue points were collected using the Tesla RTK and Total Station combination. Also shown are the occupancy point and backsight that were required by the total station for accurate placement of the x,y points.

   This next step was optional, but I felt it provided a more accurate representation of the topography of the UWEC campus. To do this I created a creek feature class in the elevation geodatabase and digized the banks of Little Niagara Creek (FIgure 7). 

Figure 7  Little Niagara Creek was digitized and used a hard breakline while creating the TIN and digital elevation model. Breaklines are commonly used in areas where steep elevation changes occur, such as ridgetops or shorelines. 

   I then created TINs for each of the elevation datasets collected. In order to create the most accurate topography representation I set the elevation points as the input for the TIN and set the digitized creek polygon as a hard breakline (FIgure 8). The breakline is used to define interuptions in the surface, such as the banks of the creek. In this case, I erased to creek from the TIN because we were unable to collect points from within the creek and it would not give us the best representation of the topography. 

Figure 8  Creating a TIN is the first step in creating a 3D elevation model. In this example the elevation points surrounding Little Niagara Creek were input as x,y,z coordinates and the digitized polygon of Little Niagara Creek was input as a hardclip to remove that area from the TIN.

   After the TIN was created I used to the TIN to Raster tool as pictured below to create a continuous elevation surface (Figure 9). The TINs are represented in Figures 11 and 12 for the area surrounding Little Niagara Creek and 16 and 17 for the campus mall area, while the continuous raster surfaces are represented in Figures 12/13 and 18/19 repectively. 

Figure 9  After the TIN was created we use the TIN to Raster tool to create a continuous raster. 

   The process was also run for the points collected with the Total Station. However, there was no breaklines set because there was no significant interuptions in the elevation that needed to be taken into account.

Results

   The first 5 figures below are the results of the elevation survey using the Tesla RTK and HIPER SR combination. This method provided easier data collection and was overall more accurate than the Total Station. 

Figure 10  Elevation points collected using the Tesla RTK and HIPER SR. These points, along with the creek creaklines, will be used to create the TIN pictured below.

Figure 11  2D TIN created using the elevation points and creek breakline.

Figure 12  3D TIN that is diplayed in ArcScene. The vertical exageration in this photo is set to 3.9 to make the elevation change more noticeable.  

Figure 13  Continuous elevation raster created by using the TIN to Raster tool.

Figure 14  3D representation of the continuous raster elevation model.

   The next five figures represent the data collection that was done using the Total Station. This data collection method was much more time intensive and did not create an accurate output of data. All of the data points appear to be shifted about 40 meters to the East. The points in the Northwest corner of the map should align properly with the mulched planter ledges that are located in the very Northwest corner of the map.

Figure 15  Data points that were collected using the Total Station. This map also includes the location of the occupancy point and backsight that were used to reference the location of the total station.

Figure 16 2D TIN constructed using the elevation points from the Total Station collection. 

Figure 17 3D TIN that was created using the Create TIN tool in ArcMap.

Figure 18 2D continuous raster surface showing the elevation values. This raster was created using the TIN to Raster tool, as shown in Figure 9. 

Figure 19  3D representation of the TIN to Raster tool. 

 Data collection comparison

   This next section will include a comparison of the HIPER SR collection vs data collection using the Total Station. The two main factors that I am going to consider are accuracy and time consumption.

Accuracy

   The Tesla RTK and HIPER SR combination collected very accurate x,y,z locations. The majority of points were located within a .008 meter accuracy. In addition, the points were placed in the right location on the ArcMap display. While there was no direct way of determining accuracy of the points collected by the Total Station, by looking at the overall shape of the TIN and DEM I can tell that the elevation is fairly accurate. However, there is no way of comparing the data since it is not located in the correct location. If it were located in the correct location it would be possible to compare the elevation data collected in the field with elevations that were collected using Lidar data, which is extremely accurate.

Time Consumption

   The Tesla RTK and HIPER SR combination was also a lot less time consuming. All that was required was the start a new project, set a few parameters (such as the coordinate system that data is being collected in), and begin collecting. It was easy to move from one location to the next. A huge plus to this method is that it can be done with only one person. All in all, collecting 100 points only took about 90 minutes.

   The Tesla RTK and Total Station combination was very time consuming to use. The majority of this time was spent setting up the occupancy point and backsight that was required by the total station. It also took a significant amount of time to get the Total Station set up directly over the occuapancy point and level. In order to get the occupancy point and backsight we had to use the HIPER SR to get an accurate GPS location that could be used. 

Conclusion

   Overall, it seems unnecessary to use the total station to collect elevation data when you have a method that takes a lot less time and it just as if not more accurate. Therefore, if I had to set up a surey project I would not use the Total Station. Instead, I would stick to the Tesla and HIPER SR combination that seemed to work very well.  

Conducting a Distance Azimuth Survey

Introduction

    The purpose of this lab is to become familiar with creating a survey using direction and distance. Sometimes while in the field technology is going to fail. Sometimes a GPS will run out of batteries or will not be able to acquire a signal due to heavy forest overgrowth, for example. In this case it is important to be able to collect data points using semi-archaic methods. Therefore, we will be using distance and azimuth measurements to plot the location of vehicles, and some associated attributes, in the Phillips parking lot (Figure 1). The Phillips parking lot is the most commonly used parking lot on campus and is constantly full. It contains about 90 faculty parking spots, 40 guaranteed faculty spots, 144 student parking spots, and 19 spots reserved for visiting future Blugolds.

Figure 1 Aerial image of Phillips Lot located on the University of Wisconsin- Eau Claire campus.

Methods

    To begin this exercise we had to decide on some features that were to be collected. Since we had to collect 100 points, we decided to survey the one thing that was readily available, vehicles. We set up the TruPulse 360 laser on the corner of the parking lot and started collecting the horizontal distance in meters and azimuth to individual vehicles parked in the Phillips lot. We also collected two attributes for each vehicle, the type of vehicle and the color. After collecting data for 68 different vehicles in the first location we were unable to collect any more. This caused us to move to a second location outside of Phillips Hall near the "art" by the entrance to the courtyard (Figure 2). After the data was collected for all 100 vehicles we came into the lab and entered it into an Excel table (Table 1).

Figure 2  TruPulse 300 setup used to complete the vehicle survey.

Table 1  Sample of attributes collected using the TruPulse 300 distance/azimuth finder.

    The next step in the process is to use the Bearing Distance tool to create lines to each of the vehicles based on distance and direction. At first, I had some difficulty getting the tool to work correctly. After some frustration I decided to take another approach. I altered the Excel table to only include the x,y coordinates of the TruPulse laser points, the horizontal distance, and the azimuth (Table 2). After this new table was created I imported it into the distance azimuth geodatabase and added it to the ArcMap display. We then opened the Bearing Distance to Line tool and inputted the variables into the tool (Figure 3).

Table 2  Altered table that was input into the Bearing Distance to Line tool.

Figure 3  Bearing Distance to Line tool interface.

    After the tool was finished running it left us with lines based on direction and distance to each vehicle we surveyed (Figure 4).

Figure 4  Lines created showing the location of cars based on distance and azimuth.

    After the line feature class we created I joined to original table containing all the vehicle attributes to the altered table that was used in the tool (Table 3). This will allow us to create a map showing where the different vehicle types and colors are located.

Table 3  Original table joined to the altered table in order to map the attributes of each survey point collected.

    After the distance azimuth lines were created and the tables were joined together I ran the Feature Vertices to Points tool (Figure ) to create a point feature class of all the vehicle locations (Figure 5). 

Figure 5 Feature Vertices to Points tool interface used to create point locations of each of the vehicles surveyed.

Figure 5 Point locations of each vehicle surveyed.

    Combining the lines and points created into one map we can show the location of vehicle colors and what type of vehicle is present at each location. As you can see from the graphs below, 25% of the total vehicles surveyed are silver in color (Graph 1) and 54% of the vehicles are cars (Graph 2).

Figure 6  Map showing the color and type of each vehicle surveyed.

    Then, I ran the summarize tool in ArcMap to produce a table that gives a count field for each of the vehicle types and colors. I imported those summary tables into Excel and created a graph showing the number of vehicles by color (Graph 3) and the number of vehicles by type (Graph 2). 

Graph 1 Number of vehicles surveyed based on color.

Graph2  Number of vehicles surveyed based on vehicle type.

 Discussion

    As previously stated, the method of surveying using distance and azimuth is a fairly good method when extremely accurate results aren't necessary. After the data was collected I noticed two major errors that affected the location of the survey points. The first error was the inexact x,y coordinates being used. Not having the exact x,y coordinates causes that location of the TruPulse laser to be moved from its actual position. Figure 6 shows that location of the x,y coordinates that were input into the Bearing Distance to Line tool compared to the actual position of the laser.

   

Figure 6 Actual location of TruPulse 360 compared to the GPS location.

 Another source of error associated with azimuth is the magnetic declination at a particular location. Magnetic declination is defined as the angular difference between magnetic north (the direction a compass needle points) and geographic north (the direction perpendicular to the equator). Some areas of the world have up to 30 degrees of magnetic declination. However, in Eau Claire, WI the magnetic declination is only 1.05 degrees to the west. Therefore, in this distance azimuth survey all of the points plotted must be moved 1.05 degrees to the west. This error can be seen in Figure 7, where the location of the points are located east of where they were actually located in real life. 

Figure 7  Errors associated with the TruPulse 360 that are caused by magnetic declination

    While the location of the x,y coordinates or the magnetic declination may be causing errors in the data, there may also be other factors causing the points to not show up in the correct locations. I recently noticed, while collecting GPS data for another class, that this particular aerial image may not be orthorectified properly. During this GPS collection the points collected on the east side of Phillips Hall were constantly being placed about two feet west of the sidewalk even though we were collecting the points from the middle of the sidewalk. If the aerial is not orthorectified accurately, the survey data collected in this lab may actually be correct.

Conclusion

    In conclusion, there are many highly scientific ways of plotting survey points, such as survey grade GPS units. However, there are certain situations that high tech equipment will not work or may not be needed. In this case it is very important to understand how to use simpler methods to achieve a fairly accurate survey. A basic survey, as done above, can also be conducted using only a compass for direction and paces to estimate the distance to each individual point. Obviously, this method would not be nearly as accurate as using a distance finder. Therefore, one needs to know the overall project accuracy in order to determine which method is necessary.