Lab Ten: Total Station Topographic Survey

Introduction

The purpose of this lab is to set up a micro topographic profile of the area with the use of a Topcon Total Station. Measurements taken by this survey method were then used to create a digital elevation model (DEM) of the study area.

Study area

Map one. Study area for total station demo.

As this exercise was designed to introduce students to the concepts of total station data recording it was unnecessary to go far from campus. The total station was set up just a few meters west of Phillips Science Hall on the University of Wisconsin Eau Claire campus (Map one). 

Methods

The first step of this project was the initial set up of the Topcon total station. Students did not participate in this task as course instructor Dr. Hupy completed it before their arrival. When setting up a total station there are several expectations that need to be met in order for data to be recorded properly. The operators need to define an occupied point, this is place which the total station occupies; essentially the central hub of the study area. The occupied point is recorded by a high accuracy GPS reading in the same fashion as those done in the previous lab, the total station is then placed directly above it. The height of the Topcon unit above ground level is then measured and defined in the software. This is a very important step in setup as the height measurement is necessary to be able to accurately measure elevation of the stadia rod via relative comparison. The final step for getting the total station ready is to define where due north is in relation to the occupied point. This direction is entered into the machine allowing it to record accurate azimuth measurements. Once the total station is set up it is very important not to touch the sensor, or the tripod on which it stands. The slightest disturbance will throw survey points off by a very large amount.

Figure one. Students collecting survey points with total station. Note the stadia rod being held by the student in the foreground and total station in the background. 

After the station was set up students used the equipment in teams of two to collect several points. It is important to note that cooperation and solid communication was needed amongst team members to efficiently and accurately record points. The survey style was essentially random but attempts were made to fill in some of gaps toward the end of the survey.

Once all points were collected they were exported from the total station as a comma delaminated text file. This file was then converted to an Excel spreadsheet and uploaded as a new feature class to ESRI ArcMap with the create feature from XY coordinates tool. This new feature class was then ready to go through the processes of 3d interpolation to create a DEM of the area. A Kriging interpolation method was used to display the results in three dimensions through ESRI’s ArcScene program.

Results/discussion

 

Map two. Total station survey results shown in two dimensions with Kriging interpolation raster. Contour intervals generated from this raster at half meter intervals. 

Figure two. Total station survey results shown in three dimensions with Kriging interpolation, elevation data exaggerated five times. 

Despite the setup being longer than with a survey GPS unit there are several advantages of a total station in comparison. A total station doesn’t have any interference problems relating to picking up satellite signals. So, once an occupied point is defined other points can be recorded in areas where satellite signals would be impossible to detect with a GPS device. The operation of the stadia rod is also a lot faster than using a survey GPS for individual points. The elevation measurements are also far more accurate than those taken by a survey grade GPS device.

Total stations have several disadvantages as well, chief amongst them being the small area which can be recorded without moving the station to a different occupied point. The stadia rod will not receive any laser signal from the total station if there is an object obstructing the two devices. The stadia rod may have to be extended if the area being measured is at a significantly low elevation. This leaves a lot of room for errors in measurements as it introduces a human component to the survey process. As with other point based surveys, a total statin survey is still subject to the limitations of a survey which doesn’t equally represent the area of interest well. In this survey the lack of representation is most apparent on the edges of the Kriging raster, as one would expect the river banks to be of similar elevation for the entire length of the stream. In the survey it appears as if it is a depression and not a stream. In order to correct this issue, which is partially due to the interpolation technique, one would either have to accept a smaller focus area from the existing data or collect more points to correct fill in the gaps.

Conclusion

Students were introduced to a total station survey device built by Topcon by collecting points in pairs. As a class a survey of the area West of UWEC’s Phillip’s Science hall was completed. This survey data was then exported from the total station to a comma delimited text file. This file was then imported to ArcMap and used to visualize the survey area in both two and three dimensions by converting the point feature class to an interpolated raster file. From this analysis it was easy to see changes in elevation across the survey area and places where additional survey points could be used were identified.

Lab Nine: Survey Grade GPS Data Collection

Introduction

This lab is an introduction to high accuracy survey grade GPS techniques. With the use of a Topcon survey grade GPS, points were collected, classified and imported as a feature class in ArcMap. These points were then used to create a cartographically correct map of each different feature type that was collected.

Methods

Map one. Survey grade GPS study area. 

Figure one. Dr. Hupy (right) instructs students how to use a Topcon survey grade GPS unit.

The collection of points for this survey were conducted over a period of about two hours via pairs of students. Each pair were instructed in the use of the GPS unit, collected and classified points (figure one). All points were collected in the area around Phillips Hall, an instructional building located on the University of Wisconsin Eau Claire campus (Map one). This GPS unit is unlike the other ones used in this course as is not a single piece of equipment. It is a collection of individual parts which work together to record locational data. The satellite signal receiver is yellow and grey apparatus at the top of the black pole in figure one. The screen in the center of the image is the central computer, but it could be replaced with a modern mobile device with the correct software installed as all communication between the computing unit (screen) and the GPS receiver is done through a Bluetooth connection.

When collecting points the GPS unit had several features that were calibrated by Dr. Hupy before students were allowed to collect points. First off was how many position readings were collected for each point. There were two settings for this: Quick reading and precise reading. With quick reading 20 GPS positions are recorded then averaged for each individual survey point. Precise readings go through the same process as quick, the only differences being that 30 readings are averaged for each survey point and that all attribute data must be entered for an observation before another point can be recorded. The amount of readings for each point in both quick and precise reading modes can be specified by the GPS operator. The GPS operator is also able to specify the length of the pole to compensate for the receiver’s height above the ground. In this case the pole was two meters tall.

Another notable feature of a survey grade GPS unit is its accuracy when compared to other lower cost/quality GPS units. For this Topcon unit there were three accuracy settings.

1. Auto: Accurate to 5 meters of actual location. This accuracy rating is comparable to what one might get from a handheld GPS device not unlike the ones used for earlier labs in this course. This reading was found most often when the GPS unit was close to a building’s wall.

2. Floating: the accuracy of flouting about 2.5 meters, a bit better than Auto. There is still much room for improvement.

3. Fixed: The GPS has a good single allowing it to be accurate to less than five millimeters.

Figure two. Creating a new feature class using XY coordinate data from a table. 

 Once all points were collected they were exported from the GPS computing unit as a text file which was  then imported to ESRI ArcMap as a new feature class by using the data tables XY coordinate values (Figure two). The projection was defined as WGS84 UTM zone 16 before the points were imported. From here, it was a simple process to symbolize and map the data.

Results

Map two. Symbolized results of high accuracy GPS survey. 

The resulting points displayed a high level of accuracy. One can note this when observing the patterning of the light poles across Phillips Parking Lot. They form a gentle arc that conforms to the shape of the lot. With a lower accuracy GPS unit one would expect the poles to hardly be linear in their patterning or consistent in their spacing. Unfortunately, this is one of the only consistency between this base map and the survey points. In other places, such as the smaller parking lot just north of Phillips Lot, the accuracy is not as good. The GPS cannot be faulted for this; its accuracy was well under a detectable amount on this map. The problem is the generously generalized base map used for this display of data. High-resolution aerial imagery would have been a better choice for accuracy but would have resulted in confusion of data at this scale, as the GPS recorded features would be overlaid on their visible real world counterparts.

Conclusion

This project entailed using a GPS receiver to collect points at a high level of accuracy. There were several different accuracy levels and many customizable settings on the GPS unit for data collection. The data was then collated and distributed as a text file including each survey point’s coordinates and feature classification. This data was then digitized to create a map that symbolized all the collected data. In all this assignment reinforced the student’s understanding of GPS data collection techniques by introducing them to using a survey grade GPS unit. 

Lab Eight: Azimuth and Distance Survey Technique

Introduction:

The focus of this lab is how to collect spatial data points with the use of an azimuth compass and distance measuring devices. Tree locations and species were surveyed on the University of Wisconsin Eau Claire Campus.

Study Area:

All measurements were taken Northwest of Phillips Hall on the University of Wisconsin Eau Claire Campus. The weather on this particular day (April 5, 2016 at ~4pm) was rainy and 2-4 degrees Celsius.

Methods:

Figure one. Field notebook used to collect tree data. Author’s Note: “Waterproof paper is not worth much without a waterproof pen.”
Field notebook for scale. 

The first step to this exercise was to collect data, a Truepulse 360 B was used to measure distance, a Sonic Combo Pro was used to measure both distance and azimuth and finally a simple compass was used to compare the results of the other tools. Tree trunk diameter was recorded with the use of a tape measure wrapped around the trunk a little over a meter from the ground. All information was recorded by hand via pen/pencil to paper (figure one) and later transferred to an Excel spread sheet for integration to ESRI ArcMap.

Each point was collected by teams of four people in operation of equipment, two would go to each individual tree, measure its diameter and position, and place the range finding device on the tree’s trunk. The other two people would work the corresponding distance measuring equipment and record trunk diameter and tree species as dictated by Dr. J. Hupy, course instructor.

The final and probably the most important part of this operation was to collect a principle measurement point via a high accuracy GPS device. This point symbolized where the distance and azimuth measurements were taken from and was used as the starting point for the following operations. 

Figure two. Tool flow model for tree location data. 

After measurements and observations were taken in the field they were transposed into an excel spread sheet and added to ArcMap using the principal point coordinates. From here a series of tools were run to determine where the trees were located (see figure two).

Figure three. Bearing distance to line tool settings. 

The first tool was a bearing distance to line transformation (Figure three). By using the principle point as a starting location, the azimuth reading for line direction, and the distance measurement, a straight line from the starting point to each individual tree could be calculated (Figure four).

Figure four. Resulting line feature class from bearing distance to line tool. 

Figure five. Feature Vertices To Points tool used to create endpoints that show the locations of trees. 

It was then possible to use this new line feature class to create another point which represents the locations for each tree. A feature vertices to points tool was used for this operation. Note in figure five that “END” was selected for the point type option, meaning that only the terminal end of the line will have a point created.

Figure six. Join field tool used to add data that was dropped from earlier operations. 

 One minor setback about using this tool process is that non-essential data fields appear to have been dropped from the attribute tables of the dataset. This issue was rectified by using the join field tool as seen in figure six. This added both diameter and species observations to each tree point.

Results: 

Map one. Tree location results derived from azimuth and distance measurements. 

This survey technique is low tech and moderately accurate in relation to the starting point. These assets can also be considered the main issues with this survey technique, it is not going to be accurate in relation to the real word if the principle measurement point is not precise. But, for this exercise, real world accuracy of each point is not as important as the accuracy relative to other points. For example, one may notice how in Map one there are tree location points that appear to be inside of Phillips Hall. In reality these trees are not in Phillips Hall. What matters are the spatial relationships between each tree.

In earlier lab reports a GPS device has been used record survey points. This is easier, takes fewer people, is faster, and often more accurate to the real world than when compared to an azimuthal survey technique. Unfortunately using a GPS device to record data can also be quite frustrating as there are typically a plethora of technical issues associated with the technology. A great deal of these issues, such as problems acquiring satellites and inflexible domain settings, do not accompany azimuth distance surveying.

Conclusion:

As mentioned in Results, the real world accuracy of the tree location points are questionable when compared to the real world, as several of the trees are placed within a building. It would seem as if all the trees have been displaced Southward by five to ten meters. The tree locations are fairly accurate in relation to one another. With only a few exceptions, such as the two White Birch trees next to the river on the East side of the map. These two trees actually grew close enough together to have a merged main trunk, meaning that the two points should be nearly on top of one another. According to the map they are displaced by an approximate 5 meters. It is undetermined what the cause of this discrepancy could be, but as these trees were amongst the last to be surveyed, it would be no surprise if a lapse of user technique was the cause.

Lab Seven: Motorcycle Geospatial Question

Introduction:

Across the University of Wisconsin Eau Clarie campus there are several parking lots. Within these parking lots are areas designated to motorcycle parking. As the weather has been abnormally warm mild thus far this spring, there has been an increase in the presence of motorcycles in the area relative to the past months. This study will attempt to answer which parking area around campus has the highest average CC size (engine displacement in cubic centimeters) of motorcycle and what is the most common type of motorcycle in said area? In order to answer this question the ArcCollector mobile app was used in conjunction with a normalized feature class using numerous domains for data collection.

Study Area:

Figure 1. UWEC and surrounding areas study area.The study focused on the area around the University of Wisconsin Eau Claire (UWEC) Campus. 

This area has been further divided into several of the most used UWEC motorcycle lots (Phillips, Schneider, Towers), off campus street parking (between State Street and Park Ave from Roosevelt to Garfield), and the main Chippewa Valley Technical College (CVTC) motorcycle parking lot area. There are other areas around campus where motorcycles are parked but these are not as commonly used as the areas listed above. Figure one is a map symbolizing these areas.

Methods:

Figure two: Screen shot of choosing motorcycle classification with Arc Collector mobile app. 

The first step to this project was the creation of a feature class to aid in the collection of data to answer the geospatial question proposed above. There were three main fields and several secondary fields included for the point feature class:

• Main Fields: 
• MTRCL_Type: a text data type useing coded value domain for: Cruiser/Touring, Standard/Sport, Enduro/Motard, Scooter/Moped and Custom/vintage. This filed is to record the classification of each motorcycle observed (see Figure two). 
• CC: short integer data type with a range domain of 40-2300. This is used to record the engine capacity in cubic centimeters for each motorcycle observed. 
• Notes: a text data type with no domains other than a limit of characters. This is used to record observations or options not listed in the table. 
• Secondary Fields: 
• Make: a text data type used to record the manufacture of a particular machine. 
• Model: a text data type used to record the model of motorcycle 
• Year: a short integer data type with a range domain of 1930 to 2017. This field is use to record the approximate year of manufacture for each motorcycle. 
• Engine_config: a text data type with a coded value domain used to record the configuration of the motorcycle’s engine 
• Cyl: a short integer data type with a range domain used to record the number of cylinders for each motorcycle. 
• MAINT: a text data type with coded values used to estimate the mechanical and cosmetic condition of each motorcycle. 
• P_pass: a text data type with coded values used to record whether or not the observed motorcycle has a valid parking pass. 

After creating this feature class it was a streamlined process to export it to ArcGIS Online, create a map, download it onto a mobile device, and begin data collection. Data was collected over a period of five days with a start time for each survey. Each observation collection session took between 30 and 40 minutes to complete and was a complete survey off all areas.

Results:

Figure three. Embedded map of results for motorcycle survey.  

Figure four. Results of motorcycle survey via lot name. Note that “<Null>” represents no survey results.

Figure three is an embedded map symbolizing the locations of each individual motorcycle observation by its classification. It also shows the location of the parking areas and their average engine size. The accuracy of the GPS unit was under 15 meters for each point. Table one shows the compiled results of each individual motorcycle observation for the different parking areas. This was derived vie a spatial join between polygons and the points within them.

As a whole there were 16 total observations taken during 5 different survey sessions. 

Conclusions:

From figure three and figure four It can be concluded that Phillips lot has the highest average CC engine size (~756cc) of motorcycles on campus with touring/cruiser type bikes being the most common in the parking area. The CVTC lot ran a close second with an average engine size of 700cc. Towers lot is at the bottom of the list with a average CC size of only 450. Off street parking, located on lower campus had zero results.

Figure five. The distribution of CC size for observed motorcycles. 

The design of the feature class for this project was good. The domains and data type selection were thought out, which helped to prevent the omission of data while collecting. Despite this, there was one major issue with collecting data on motorcycles. Aside from a few die-hard riders, people do not seem to like riding in cold rainy weather. Aside from the first day of data collection the weather during this study was quite rainy and between 30 and 45 degrees Fahrenheit. Figure five is a distribution for the CC size of the dataset. While it may resemble a normal distribution, it is far to leptokurtic (peaked) and does not follow the central limit theorem. This means that this dataset is not a viable answer to the question of this survey. Put simply more observations need to be made, preferably during a time when the weather is more suitable for motorcyclists.

Lab Six: Parcel Forum

 This post, unlike the others in this series, is not about a lab. Instead, it focuses on a Parcel Mapping forum the occurred at UWEC on Tuesday March 15, 2016. Attending this forum was an opportunity for students to rub shoulders with people in the PLSS parcel mapping and land records industry. I arrived at this session as it was moving into its second half at ~12:45pm, meaning that I was able to listen to four people presenting on their work and even participate in a breakout session where groups attempted to solve complex issues relating to the field.

Jason Poser of Buffalo County, Wisconsin gave the first presentation I attended. As with many of the presentations I attended during this forum, his was ripe with jargon, explanations of how the advance of technology has drastically changed, and how land records are kept and how surveys are conducted. An example of how the advance of technology has changed and streamlined land records was that in Buffalo County there was as much data digitized in 2015 and early in 2016 as was digitized in from 2006 to 2014. This was due to a shift from using Autocad to ESRI’s parcel fabric. Poser describes this as being functionally better, more accurate, and easier to change parcels. He also said that it was quite convenient to use historical data to see changes made to parcels in the past.

Two other presentations were much akin the first. Brett Budrow of St. Croix County and Dan Pleoger of Sawyer County spoke briefly on how GIS has been implemented within their districts. Budrow even included a little history as to how parcel mapping was originally done with ink and mylar before Autocad or the use of designated GIS software products.

The presentation by Mark Netterlund of Barron County was by far my favorite and most informative. As someone who has not worked with parcel mapping as a career, I appreciated Netterslund’s avoidance of jargon and interjection of advice that can be useful for all different career types. He spoke about how the turnover rate for the surveyors in the county was low, which allowed a community-like atmosphere to develop. He also mentioned how important it is to create relationships with other departments as favors are often exchanged. Finally, he suggested to just talk with people and let them know what it is that you are up to, often times they will be quite interested and willing to help with the job.

The final part of the forum was where the people present formed breakout groups. As this forum was designed for PLSS and county surveyor people, I was able to observe (but not contribute to) one of these breakout groups. I wasn’t prepared to contribute much as much of the material went right over my head, even having spent a few hours doing simple parcel mapping in a GIS class.

During this breakout session, the idea of group polarity kept coming to mind. This is where the ideas of likeminded individuals are amplified when they are put in a group setting. The main ideas that I got from a good majority of people was how they needed more funding to complete what surveying and mapping projects they want to. Despite this, there were few who see the value in educating the people who write their checks the reason why they would like the money. One cannot blame them too much for this, as interpersonal communication has typically been a shortcoming for those in this type of discipline.

As a whole, I would have to comment that this was a fairly exclusive, yet indecisive group; with exceptions of course. They seemed to be bonded via the knowledge of common industry jargon and similar work tasks. Despite this they were usually unable to agree on ways to solve a problem as everyone who had a solution seemed to believe that their solution was best and did not wish to entertain other’s ideas. Yet from this mass of ideas, I was able to pick up some sound advice about ways to work with others instead of against them. 

Lab Five: Introduction to Arc Collector

Objective:

To use arc collector to collect data with the use of a mobile device running on IOS or Android. This lab introduces students to the creation of a web map to be shared across ArcGIS with Arc Collector. Students then are tasked to collect more micro climate with the use of Kestral wind gauges in the same fashion as the previous lab.

Methods: 

Figure 1. Classmate Audrey Bottolfson poses with an Arc Collector enabled smart phone and Kestrel wind gauge.

For the first step of this project the same feature class used in the previous lab was made available for use in this one. No changes were made to domains or coded index items as it was warmer this day (March 8^th, 2016) and adjustments were unnecessary.

This feature class was then deployed to ArcGIS Online with an accompanying base map, making it ready for collecting data with a mobile device through ESRI’s Arc Collector mobile application. As with a great deal of equipment initially it didn’t work, but after intensive troubleshooting by the course instructor it became functional.

As like the previous lab, microclimate measurements were taken with the use of a Kestral wind gauge. These readings were then entered as data points into the feature class via the use of a personal mobile device. After about 45 minutes of data collection the entire class reconvened to upload the data into a shared drive for all to access.

Results: 

Figure 2. Compiled temperature readings across the UWEC campus (five category Jenks natural breaks classification). 

While collecting data it was obvious that the locations recorded by the mobile devices were not as accurate as those recorded by the dedicated GPS devices used for the previous lab. There were several instances where the recorded point was off by tens of meters even after allowing the device to focus in on a stationary location. Data entry on the mobile device was far easier to use than with the dedicated Juno GPS due to the far more intuitive and user friendly design of the smartphone and the collector app. In my opinion the relative poor accuracy of my mobile device compared to the Juno was not a bad compromise for the ease of data entry. There is also the advantage of the data being uploaded automatically to ArcGIS Online as data connections become available, meaning that data is rarely lost and available for viewing and editing from remote locations almost immediately. 

Figure 3. Screen shot of user interface of Arc Collector. 

This lack of accuracy led to several issues when combining data with the other members of my class. For instance, not all mobile devices are built using the same GPS receiver meaning that some are far more accurate than others. The GPS receiver in my mobile device was accurate within about 20 meters. In figure three this lack of positional accuracy was apparent while collecting data. The GPS said that it was located at the blue and white arrow dot in the center of the image. My real location was actually at the northern green dot to the left of the blue one. A classmate’s device was newer and was more accurate than mine; within about 15 meters. This difference in spatial accuracy across devices does not help the data’s integrity as a whole, but for this project it isn’t very important as locational accuracy is not the focus for the exercise.

Conclusion:

Microclimate data collection via Arc Collector is far superior to using a GPS in how it is convenient and able to compute and update data quickly. The time it takes most mobile devices to acquire satellite signals is far quicker than a good majority of commercial GPS units. The drawbacks for this tool in collecting data is that the locational accuracy is not quite comparable to a modern designated GPS unit and some mobile devices use better GPS receivers in their construction than others which makes them more accurate than their counterparts. But this is often not a problem for data collection that doesn’t require specific precision.

Lab Four: Geodatabases and Domains

Introduction:

This lab was centered on creating a geodatabase with custom defined domains for different data types. This database was then downloaded to a handheld GPS unit and used to collect weather observations around the UWEC campus mall.

Methods:

This geodatabase needed to store domain information for a feature class with custom made fields to store microclimate information; for example, temperature, dew point, relative humidity, wind speed, and wind direction are all examples of the data types collected. For each attribute it was necessary to set a domain or subtype to normalize the data across multiple surveyors. An example of a domain would include setting a range for the wind speed field; it would be outrageous to accept a wind speed of -2 or 6666666, so the range was set to only accept values between 0 and 60, inclusive. The other type of domain setting was for a coded index domain, essentially an attribute field where there are a few predefined choices available for use.

Field Name	Data Type	Domain Type
Group Number	Text	Coded Values
Point Number	Short Integer	Range: 1 to 100
Temperature	Float	Range: 15 to 60
Dew Point	Float	Range: -20 to 100
Relative Humidity	Float	Range: 1 to 100
Wind Speed	Float	Range: 0 to 60
Cardinal Wind Direction	Text	Coded Values
Azimuthal Wind Direction	Short Integer	Range: 0 to 360

Table 1. All domains and associated data types used to create this geodatabase. Note that Dew point has a range that was changed to a negative value.

 An example of this was from the group number attribute; allowing the user to only select from a pre-defined list of options. After entering domains, we tested the functionality of the geodatabase and feature data set described earlier by importing it to a Trimble Juno GPS unit set up to run ESRI’s ArcPad GPS program. Observations were collected from the UWEC Campus Mall, an area roughly 500 by 200 feet in the center of the UWEC campus located in Wisconsin. Using the GPS unit and a Kestrel 3000 wind meter, a handheld weather monitoring device capable of collecting all required data, a couple points were created and the attribute table was populated.

Results/Discussion

When entering data points to the feature dataset it became apparent that several adjustments were needed; for example, the dew point attribute domain needed to have an extended range as several dew point measurements had negative values. Because of this the correct value wasn’t able to be entered for the points in the dew point field. All other fields worked well with the domains being set up for the weather present on that particular day. Unfortunately there were only two points collected for the time that was allotted for data collection. This was due to the fact that the GPS unit I was assigned failed to acquire a strong enough signal from satellites meaning that what points I did collect were recorded as being several kilometers south of where they should have been. 

 

Conclusions:

During this lab the processes required to create a geodatabase with relevant range and coded value domains were completed. From here it was a relatively simple process to export this geodatabase with its point feature class into a Trimble Juno GPS unit for a short test. During the test it was apparent that the domain range for the dew point attribute needed to be extended and that one cannot always trust their equipment to work correctly once in the field.