Conducting a Distance Azimuth Survey

Introduction: In the age of technology, surveying and mapping surface features has become more time efficient and digital. With the use of a GPS or rangefinder, a user can map an acre plot and record attribute data in one setting, however, as will all technology and field work, situations are unpredictable. If the technology were to fail, a user would then have to survey the plot by means of other techniques, one being measuring distance and azimuth to the selected features. Standing in a single spot and surveying using a compass and tape measure or paces, one can map the general arrangement of the features in question. When using a compass, it is important to remember that magnetic fields can affect the readings and true angles. The magnetic declination of a location needs to be accounted for and applied to the azimuth readings. This makes data processing more time consuming in the long run but when technology fails you, a little math can go a long ways.

Figure 1: Owen Park in the City of Eau Claire, WI is nestled next to the Chippewa River. The past site of an electrical plant, the 11-acre park now offers an assortment of recreational opportunities to the local community. 

Survey Location – Owen Park, Eau Claire, Wisconsin, USA: In 1913, lumber baron John S. Owen donated 11-acres of land to the city of Eau Claire as a hope to create a city of parks. Today, those 11-acres are still used as a park that bares Owen’s name. A band shell, state bike trail, numerous benches, tennis courts, and large trees cover the area. Owen Park is a community favorite location for festivals and gatherings throughout the year, from Blues concerts to the Eau Claire Marathon. Topography of the park is relatively flat, and is commonly flooded by Chippewa River.       

Methods:

Figure 2: The TruPulse 360B laser ranger finder was mounted on a tripod for stability. From the single location shown, 98 survey points were measured recording slope distance in meters, Azimuth in degrees, a point characteristic, and a count ID.

Collecting Survey Points: For the survey, we used a TruPulse 360 B Series laser rangefinder to record the slope distance and azimuth to the chosen points. The rangefinder was mounted on a tripod to limit movement to prevent discrepancies in the azimuth readings from point to point since only one survey location was chosen (Figure 2). The single location allowed all points surveyed to be assigned the same starting coordinates, removing the possibility of error. The location chosen as the survey point was the northeast corner of the tennis courts. This point is easily identifiable from an aerial photograph which is crucial to assigning starting coordinate values. Using the rangefinder, 98 features from trees to benches were surveyed with their distance, azimuth from the start, description of the point feature, and the count. 

Figure 3: The XY coordinate position for the survey location was assigned using aerial images and Google Earth. The position recorded by Google Earth was presented in Degrees Minutes Seconds and needed to be converted to decimal degrees before being assigned to each survey point. 

Finding Coordinates for the Starting Location(s): The coordinate position of the starting survey point needed to be assigned. Using Google Earth, a place marker was set at the northeast corner of the tennis courts in Owen Park (Figure 3). The assigned location was 44°48’9.1’N, 91°30’0.67”W. This location needed to be converted to decimal degrees to allow ArcGIS to read the coordinates. By multiplying the seconds by 60, they are converted into minutes. Adding the converted seconds to the minutes and dividing the new total minutes by 60 will result in degrees. This value is then added to the observed degrees to finish the conversion of the location to decimal degrees. This process is repeated for both the latitude and longitudinal coordinates resulting in the location 44.802531N, 91.500183W. In the conversion, make sure to hold at least 6 decimal values to allow proper and more accurate results in ArcGIS.

Figure 4: The data recorded from the laser range finder was organized into a Microsoft Excel data table. Each attribute or value was arranged into its own column for each point surveyed. 

Creating the Data Table:  The data collected from the survey was organized into a Microsoft Excel table to allow ArcGIS to import it. For the table, each attribute was listed in its own column: Number ID or count, Type, Slope Distance (SD) in meters, Azimuth in degrees, X (Longitude of shooting location), and Y (Latitude of shooting location) (Figure 4). The distances, and Azimuth values need to be extended to 6 decimals in the table to match those of the X,Y values of the shooting location calculated in the previous step.   

Importing the Table: A geodatabase was created to hold the files created in this exercise to help organize the project. To create a new geodatabase, right click a folder in ArcCatalog where you wish to store the database. From the menu select “New” – “File Geodatabase”. A new database will be displayed in the selected folder. Properties can be assigned to this new database by right clicking the database, selecting properties from the menu, then adding the desired information, but for this exercise the properties will be kept at the defaults. Once the database was created, the document work environments were set to add all new files to the database.

Figure 5: The map document properties window located under the file tab, allows the user to set a default geodatabase to store the output from tools or other processes. The properties window also provides a setting for ArcGIS to store the relative pathnames of data sources. This prevents information from being lost if moved from one folder to the next.  

By clicking File – Map Document Properties, a window opens allowing you to choose the default workspace. Under the default geodatabase section, click the folder and browse to set the workspace to the new geodatabase. Then before closing the map document properties window, check the box “store relative pathnames to data sources” (Figure 5). This allows the program to follow locations of files if they are moved from one folder to the next without having to reassign the pathways. To verify that the workspace was set to the new geodatabase, right click in ArcToolbox and select “Environments”. Under the workspace tab, two boxes should appear: “Current Workspace” and “Scratch Workspace”. Make sure that both read the location of the new geodatabase, if not browse and assign that location to the workspace. All output features from tools will now be added and stored in that location instead of a default “catch all” database. Now with the environments set and the database created, the data table can be imported directly to the geodatabase. Right click the geodatabase and select “Import – Table (single)”. The Table-to-Table tool will open. Browse and assign the excel data table as the “Input Rows” field. The output location is automatically set to the geodatabase so the table just needs to be named. After the table’s named, click “ok” to run the tool. The table should then be added to the database as well as the document.

Figure 6: The output feature class of the Bearing Distance to Line tool in ArcGIS results in a series of lines based on the distance and azimuth from the starting location to the survey point. The features were displayed over an aerial photograph for comparison. 

Displaying the Survey Points: Now that the table is in the geodatabase, the survey features can be displayed on a map. Because the actual coordinate positions aren’t known for the features, the bearing distance to line tool will be used first to locate the features (Figure 6). Open ArcToolbox and select Data Management – Features – Bearing Distance to Line. Set the “Input Table” to the table containing the survey feature information. Since the table is in the geodatabase and in the document, it should be listed in the dropdown box. Then name the output feature class that will be added to the geodatabase. Set the “X” and “Y” field to the respective columns of the table. Make sure the Longitude coordinate is listed as X and Latitude as Y. Next the “Distance field” is the slope distance and “Bearing field” is the Azimuth both are the recorded values from the rangefinder. Before running the tool, make sure that the spatial reference is set to the same as the map document. This exercise used the default projection WGS84 Web Mercator (Auxiliary Sphere).

Figure 7: The Feature Vertices to Points tool in ArcGIS uses the vertex locations of a line or polygon feature to assign a point feature. By using this tool, the survey points could be added independently of the assigned lines from the Bearing Distance to Line tool.  

The “Feature Vertices to Points” tool will convert the line feature class created by the “Bearing Distance to Line” tool to a point feature class. Again in ArcToolbox – Data Management – Features, select “Feature Vertices to Points”. Set the output feature class from the bearing distance tool to the “input features” and name the new output feature class that will be added to the geodatabase. Before running the tool, switch the “point type” to “END” instead of “ALL”. This will instruct the tool to only place a point feature at the end or last vertex of each line instead of the start and end (Figure 7). This prevents the output feature class from having doubles for each survey feature. Once everything is set, click “Ok” to run the tool.

Displaying the Survey Locations (optional):  For cartographic purposes, the user could add a feature class that contains the locations of the stationary survey points. A table with the X,Y coordinates could be added then displayed using the method discussed in previous exercises, but since only 1 location was used, a point was created in a new feature class to mark the location with the help of aerial photographs. Add an aerial image base map from the “add data” button on the tool bar and locate the study area, Owen Park in Eau Claire, WI in this exercise. Then, right click on the geodatabase and select New – Feature Class. In the window, name the feature class and select “point” from the dropdown menu in the Type Field section. Click next to set the coordinate system, again, making sure that it is the same as the map document. Continue accepting the XY tolerance and storage configuration defaults. Click finish and the feature class will be added to the geodatabase and the map document. Right click on the new feature class and select “Edit Features – Start Editing”. To add a point in the survey location, click the “create features” button on the editor toolbar. From the Creature Features window, select the new point feature class then using the aerial image, click the position of the survey location, the northeast corner of the tennis courts. This will add the point to the feature class. Click the dropdown menu on the editor toolbar, “save the edits” then “stop editing” the feature class. Now the stationary survey location can be displayed as a reference on the map.  

Discussion:Using the laser rangefinder instead of a standard compass had several advantages. The electronic technology built into the rangefinder automatically accounts for variation in viewing pitch which allowed the azimuth to be measured at any angle. If a standard compass had been used, it would have to be held level to allow the needle to rotate. This can influence angles if survey points were being measured from a slope. Also declination from true north occurs more with the use of standard compasses rather than the electronic counterparts. Declination or deviation from true north occurs at different severities based on location on the Earth’s surface. Terrain, development or even metal objects can skew the reading of a compass. The declination in Eau Claire, WI is about 1°4’48”W or about 1.867°W from true north. This means that 1.867° should be added to any compass readings recorded to correct the value to reflect the true direction. The electronic rangefinder wasn’t a perfect recording tool and presented other complications however. As distance from the survey location to point increased, the accuracy of its real world location decreased (Figure 7). The trees and poles surveyed across the park ended up being located in the Chippewa River. It was also observed that the laser’s focus was less at greater distance. Several points were measured through openings between tree branches and the laser typically reported the distance to the object trying to survey through rather than the intended point. This limited the surveyed points to direct line of sight which won’t always possible in field situations. Another problem was encountered when assigning the starting survey location to the measured points.

Figure 8: A comparison of starting locations and results of the bearing distance to line tool from locations assigned by ArcGIS (yellow) and Google Earth (blue).

A location was assigned using values obtained from ArcGIS and values using Google Earth. The ArcGIS location when mapped with the tools described resulted in the points being plotted several meters incorrectly, but the Google Earth location converted from degrees minutes seconds to decimal degrees used the correct starting location as a reference for the measured points (Figure 8). It could be possible that the accuracy of the images used to assign the point location to the starting point was less than those of Google Earth and since all other points were measured from that single starting location, any variance in that location would cause a chain reaction in the rest. Sometimes technology isn’t as reliable as the “old school” methods of a compass and pace count or tape measure. Having the capability to me versatile and adapt to what is needed in the field will help provide the best quality information to a client.

 

Unmanned Aerial System (UAS) Scenario Evaluations

This is one of the many forms of an unmanned aerial system. Multicopters (pictured) can be used in situations that require precision in tight spaces. Vertical takeoffs and landings make multicopters one of the most versatile UAS.

Introduction: With the fast paced development of geospatial technology, applications of the new information systems are growing exponentially. The perfection of remote sensors in functionality and size has expanded the range of what vehicles or systems can be fitted to carry and capture information. Tasks that were once only capable on foot or in person are now able to be done via satellite or unmanned aerial systems (UAS). Unmanned systems are both used by the military as well as by civilian contractors and are capable of unpiloted flight. Several types of UAS exist and each can provide different sets of benefits to evaluate tasks quicker and, in most cases, provide a less costly alternative.   

Fixed-wing unmanned aerial vehicles require a strip to takeoff and land. The ability to glide and carry large payloads offers the consultant a wide range of options for sensors and other hardware to complete a survey.  

Fixed-Wing Craft: Fixed wing vehicles are the most familiar or “search image” for what people think a UAV or UAS is. These crafts have two fixed wings as the name suggest that allow gliding and stable flight with greater payloads and longer flight times as compared to copters. The wings allow the craft to glide in an event of power failure which copters are not able to do providing a better peace of mind for the pilot; however, the fixed wings don’t allow the craft the ability to hover in a single spot or allow for a vertical takeoff. Like any standard gliders or airplane, a fixed wing craft requires a runway for takeoff and landing. Fixed wing crafts are available in gasoline or electric power. The gas powered flights can last about 10 hours and electric about 45 minutes to an hour depending on the payload.

Helicopters, with the ability to hover, provide the user to focus on a specific area of interest or reach areas quickly and precisely. Since they takeoff and land vertically, a helicopter can be launched in a confined space that a fixed wing craft would be unable to. 

Helicopter: Helicopters are operated by a single lifting motor with two or more blades. The lifting motor allows for a vertical takeoff and landing allowing the system to be launched in a limited area. The blades are able to be pitched to compensate for winds and directional change. For an operator, it is safer for the motor and radio equipment on board to have separate power sources. This not only helps to return the craft but also increases safety when working or performing maintenance on the system. Copters are fast and efficient but can create vibrations that can reduce quality of video or information. The location and mounting systems can limit the vibration allowing for accurate data. Helicopters are powered by either gasoline or electricity. Gas powered systems have a flight time of 4 to 5 hours while electric powered systems are limited to 20 to 90 minutes. Flight times are limited by payload as well as configuration of the system.

Multicopters provide yet an even more stable and controlled response in a UAS. With multiple arms, balance in winds and flight allow a multicopter to be the most versatile and reliable UAS. 

Multicopter: Multicopters utilize several rotating blades to achieve lift and control. Each blade set is controlled by a separate motor allowing the system to balance, provide lift, and have precise directional controls when hovering. The multiple arms allow for a higher degree of safety when performing maintenance tasks by operating at smaller radii than a helicopter. The multiple blades or arms of the system also require more powerful computer processors without which the craft can’t fly; however, the varying designs for multicopters offer a degree of flexibility unavailable to helicopters or fixed wing crafts in the location of sensors and other operating equipment. This variation also enables the multicopter to be more stable in high wind conditions compared to other aerial vehicles but with more arms on the multicopter, the shorter the flight time. The stability is great for beginner operators as well as those who need the stability to focus on an area in great detail.  

What do you need: There are several pieces of equipment that are required for every UAS, some more obvious than others. You will need to consider what you are trying to survey then determine the type of vehicle, autopilot, planning software, and payload the vehicle will be carrying. The vehicle, fixed wing, helicopter, or multicopter will determine which of the systems will be needed as well as determine the size of the payload. APM 2.6 autopilot hardware provides processors, gyroscopes, accelerometers, pressure sensors, and a GPS to any vehicle (ArduPlane for fixed wing ArduCopter for rotary wing) with a reasonable cost at $179.00. The ArduPlane or Copter software allows the user to pre-plan a flight pattern into the system allowing the pilot to monitor the system rather than solely focusing on flying the vehicle. The planning software also ensures that the necessary overlaps for images as well as insuring precise points of interest are located quickly and accurately. In the event of the vehicle flying out of the range of the manual controls or equipment failure, the software provides a failsafe by automatically returning the aircraft to the point of take off. The payload will vary with vehicle and duration of the flight. Longer durations of flight will require smaller payloads since more operating power will be needed. High definition cameras, thermal scanners, infrared scanners, or LiDAR are just a few possible sensors that are capable of being installed on a UAS. Again, which sensors are necessary will depend solely on what the subject of the survey will be.     

Scenarios: There are many real world applications of UAS. The following list provides several examples of those scenarios and provides possible solutions on which UAS and technology would be most helpful to meet the goals of the individual client.

The desert tortoise is a threatened species in the United States. Threatened species are not in danger of extinction but population sizes are small and if not monitored closely may become endangered. With this threatened title, the tortoise habitat is protected under the Endangered Species Act. 

Scenario 1: A military testing range is having problems engaging in conducting its training exercises due to the presence of desert tortoises. They currently spend millions of dollars doing ground based surveys to find their burrows. They want to know if you, as the geographer can find a better solution with UAS.

        Questions: What do we know about the tortoises? What can we use to locate burrows? How large is the testing range? How often are exercises conducted? What is the budget?

        Background: Desert tortoises are herbivores that feed on grasses, wildflowers, and cactus pads that are available in the deserts. Tortoises live in burrows to escape the extreme heat of the environment and lay dormant from November to February. 97% of tortoise burrows are associated with shrub vegetation allowing the tortoise to eat in the spring. The soil needed to create the burrows is also particular to the tortoise. Moisture and texture has to be perfect that it crumbles during digging but is firm not to collapse. Commonly burrows are found in sandy loam soils with gravel and clay not areas with a lot of sand.

        Solutions: With military operations possessing fixed wing aircraft, a possible solution to locate the burrows could be outfitting a UAV or manned aircraft with a few new sensors. The fixed wing UAS has a stable flight with the capability of long flight durations. With the burrows being typically located near dense vegetation, a near infrared scanner would be able to locate those vegetated areas. Next, the suitable soil could be located using variations in moisture content measurable by a short-wave infrared sensor. Areas with higher moisture content will display weaker in the sensor readings due to the absorbance by the water rather than reflectance by the dry sands. Once paired with the vegetation and suitable soil locations, a thermal imager may be capable of locating the entrances of the burrows. The cooler soil would stand out from the stark hot surface of the surrounding areas. An overlay analysis could be performed and those areas that are most suitable for burrows can be mapped providing a guide for later surveys. The use of these sensors can’t be 100% accurate so foot inspections may still be needed but much less intensive than without the UAS.

For years, utility companies have hired helicopters to fly along power lines to perform inspections. This is a very dangerous and therefore costly process that can be reduced with the help of a UAS. 

Scenario 2: A power line company spends lots of money on a helicopter company monitoring and fixing problems on their line. One of the biggest costs is the helicopter having to fly up to these things just to see if there is a problem with the tower. Another issue is the cost of just figuring how to get to the things from the closest airport.

        Questions: How often do problems occur along the power lines?  How much does it cost to get the helicopter out to inspect the line?  Are the lines accessible to ground crews to get close enough quickly enough?  What type of area are the power lines in, populated or rural?

        Solutions: A multicopter would be a good option to perform this task.  A ground crew with the multicopter would get close enough to the power lines that they wish to examine in order to use a multicopter. The UAV multicopter should be installed with a high quality camera and should be able to operate at a relatively low noise level to prevent disturbing wildlife or cattle that may be in the area.  It would also be advantageous to have some way for the copter to detect changes in the electrical field around it so it could detect anything that may be wrong with the power lines. This multicopter has high payload capacity (3kg) for high definition imagery, rotors covered for safety when flying near power lines, is equipped with excellent crash/accident avoidance technology such as its “coming home” function, very stable when holding position for excellent imagery, and comes with some of the best pre-mission programming.  It will cost around $30,000 per multicopter but this will pay off in the long run when considering the company cost paying a helicopter to go out to check the power lines repeatedly.  It’s not an inexpensive start up cost but will pay off in the end, as this is a very reliable model. It may also be a good idea to have a UAV helicopter along as well. If a power line needs to be looked at rapidly due to an emergency, the helicopter can be launched from further away due to its longer range.  The same considerations regarding weight, noise level, image quality, and maneuverability would need to be taken into consideration for the helicopter as for the multicopter.  

Pineapples are a short bush like plant that is commonly grown in tropical areas. As the fruit matures, it rises out of the center of the plant and gradually begins to turn a bright yellow. 

Scenario 3: A pineapple plantation has about 8000 acres, and they want you to give them an idea of where they have vegetation that is not healthy, as well as help them out with when might be a good time to harvest.

        Questions: What type of area is the pineapple plantation in? What is currently being done to assess these problems?

        Background: Pineapple plants are read to harvest when they are in the late stages of their development.  This means that they have over 1/3 of their peel as a yellow color but they haven’t lost all of their green yet. Near infrared reflection has been tested to see whether or not it can be used to detect ripeness of a plant, as the plant gets riper, the NIR tends to decrease.  This fact combined with the fact that when pineapples are ready to harvest they shouldn’t be ripe yet and shouldn’t be harvested until their skin is 1/3 yellow and 2/3 green can help determine the best time to harvest the pineapple.  Using a multi-spectral approach with both visible wavelengths to try and determine color and NIR wavelengths to determine health and whether the plant is ready to harvest.

        Solutions: A gas powered UAV helicopter equipped with a near infrared (NIR) camera sensor which will detect higher reflectance of healthy vegetation will help locate the ripened fruit.  This UAS would require pre-mission software that would allow it to track and cover the whole field recording data spatially.  Ideally this would be done during the day when there is the highest amount of NIR reflection. A gas helicopter would provide the focus and maneuverability needed to survey the pineapples. The size of the field would limit a multicopter to short flights where the helicopter could fly for longer durations. The NIR camera has a high range of reflectance allowing the distinction of healthy and unhealthy vegetation. Healthy vegetation reflects significantly more NIR waves allowing the healthy plants to be displayed more prominently than the unhealthy vegetation.

The Niger River delta is one of the largest in the world and is also one of the most polluted. Stretching thousands of miles across western Africa, the river passes through several oil rich lands increasing the pollution risk. 

Scenario 4: An oil pipeline running through the Niger River delta is showing some signs of leaking. This is impacting both agriculture and loss of revenue to the company.

        Questions: What is the range of possible leaking points in the pipe? Will there be any restricted areas to avoid? Are there fire risks? Will the UAS be flying over populated areas? What time of year will surveying occur?

        Solutions: The survey area is large requiring a UAS with a large antenna which will allow for a longer flight range from the computer base. The best type of craft for this mission would be a gas powered fixed-wing craft. This would allow the payload capacity needed and provide the stability required. The main sensor required on the plane would be a small thermal imaging camera to record the thermal readings of the ground along the pipeline, just after sunset, from which one can deduce the thermal heat capacity of the ground. If there is a leak, the oil will produce a thermal reading that differs from the heat capacity of the surrounding materials.

Surface mines are those at which layers of earth are scraped and dugout then processed to separate the precious metals or stones. The earth is removed until the ore is depleted and operation is no longer profitable by the company. This Russian diamond mine is over 600 feet deep. 

Scenario 5: A mining company wants to get a better idea of the volume they remove each week. They don’t have the money for LiDAR, but want to engage in 3D analysis (Hint: look up point cloud)

        Questions: What is the project budget? How expansive is the mining operation? Is this an open pit mine?

        Solutions: Aerial photos need to be taken of the mine area with a high percent of overlap (>60%). Use of a multicopter or helicopter would provide the control and stability to capture quality images, however if the budget is constrained, the use of a kite may also serve well to photograph the area. From online servers such as Photosynth, the photographs can be uploaded to create a point cloud of the area. A point cloud is a set or fabric of dense points that hold elevation as well as coordinate information. From the point cloud, a digital surface model can be created and be capable of calculating changes in volume along that surface. Like the servers that create the point clouds, there are open source products such as Meshlab that allow the surface calculations to be done. If the budget allows, ArcMap with the 3D analyst extension will also allow the creation of the surface model from the point cloud. With use of the SurfaceVolume tool, ArcMap can also calculate the difference in the surface from a set reference plane allowing the changes in volume to be tracked over time.   

Information complied by: Jake Burandt, Tim Condon, and Brielle Cummings

Visualizing and Refining a Terrain Survey

Introduction:  The development of geospatial information systems (GIS) has allowed new avenues and research methods to be created that help evaluate spatial questions and problems. Digital Surface modeling is one improved method for those evaluations. In a GIS, there are many ways in which a dataset of elevation values can be reviewed to create a digital representation of a real world surface. This exercise explores several options on how to create a continuous surface from a point feature class using ArcGIS software.

Inverse Distance Weighted Interpolation (IDW): Inverse distance weighted interpolation (IDW) uses the concept of areas that are closer together have more in common than those that are a farther distance apart. To help model a surface from a set of elevation data points, IDW will use a defined search area, use all the elevation points that fall within the area, apply a weight based on distance from the area being modeled, and assign a value to the surface based on the average in that search area. If the dataset used in the interpolation has a large number of survey points at a fine resolution, IDW is a great method to use; however, because there is no standard and uses average values, the results can’t be conclusive that the surface represents the real world features correctly.

Natural Neighbor: A natural neighbor interpolation of a surface uses surrounding survey points similar to IDW but instead of distance, uses a proportion of areas of overlap. The creation of Thiessen Polygons around the survey points is the first step this process runs, then using the point where the surface is being calculated, another Thiessen polygon is created overlapping the originals. Using the proportion of the overlapped original polygons to the newly created polygon, weights are applied to the surrounding survey values and used to assign a surface value at the point in question. This is better suited for datasets that have fewer survey points that are low density of the study area. 

Spline: Using a spline to create a continuous surface from a point feature class is capable from a statistical equation figured from the dataset. The surface created tries to minimize the variation among values resulting in a very smooth surface model. This is great for areas of little variance, but for formations that have drastic changes in elevation, features may be represented falsely in the digital model. Spline is a “connect the dots” form of interpolation where each elevation point is plotted at its observed value and the created digital surface passes through each recorded point like a sheet to form a continuous model.

Kriging: Kriging, unlike the spline or IDW modes of interpolation, uses a statistical model created from the survey data and applies it to the model to create a surface. Since a statistical model is used to create the surface, a degree of certainty can be calculated allowing for an accuracy of a surface model to be evaluated more directly as compared to the surface estimations created by a IDW or spline. Like IDW, Kriging interpolation weights surrounding values and applies the results to an area that doesn't have surveyed values. The weight assigned to each value is based on distance to the point in question as well as the overall arrangement of all survey points in a study area. This takes the whole area into consideration but still allows the “closer areas are more similar” idea to persist in the model. Like an IDW, Kriging interpolation is more accurate with higher numbers of original survey data points; however, Kriging interpolation can also be a reliable method for surface modeling with few data points if the survey points are accurate and reflected accordingly in the statistical model generated for that dataset.

Triangular Irregular Network(TIN): A triangular irregular network (TIN) is created from a point feature class by connecting neighboring survey points. The connection of these points creates triangles that allows the surface to shape and follow the grades of the terrain. A TIN can be used to create a 3D representation of a surface to provide the GIS user a better visual representation of a study area. When using a TIN, areas that have more variance in elevation should have more reference points to allow the triangles to better conform to the feature. If an area is surveyed at a coarse scale, subtle changes in topography may be lost by the generalization of the triangles created.

Methods:

Figure 1: Formatted in the previous exercise, the XYZ coordinates for the terrain were imported into a Microsoft Excel table. To remove the negative Z values and keep the scale of the surface true, a value of 12 was added to all X, Y, and Z recordings. 

Importing the Elevation Data: Before any terrains can be created in ArcMap, the XYZ table of values created in the last exercise needs to be imported into ArcGIS (Figure 1). The table should be saved in a geodatabase or file accessible by ArcMap, if not available, the table won’t be visible to the program. Once the table is saved in the appropriate space, right click “Layers” under the table of contents. From the menu, click “Add Data”. Browse the folders in the window to find the saved table of coordinates then click and add the file to the map. At this point, the chart is only a table and is not a feature class in ArcMap. To display the coordinates in the map view, right click the table in the table of contents and from the drop down menu select “display XY Data”. Make sure that the X, Y, and Z values are assigned correctly to correctly display the coordinates. Now the coordinates are displayed as points on the map viewer. Switch the table of contents to the “list by drawing order” view to hide the original table. The point feature class should be the only item displayed in the table of contents. The feature class now needs to be exported. It is important to EXPORT the XY display to a new feature class and not just save. By exporting the features, ArcGIS saves all the information and the relative associations by assigning object IDs to each set of coordinates. These associations are what the tools will read and use as references in later processes. Right click the point feature class and from the menu select “Data-Export Data” and follow the prompts to save the new feature class to a geodatabase. The saved feature class will then be added to the table of contents and is now able to be used by other ArcGIS tools.    

Figure 2: ArcGIS offers an extensive list of tools and operations. For the surface interpolations done in this exercise, all the tools with the exception of a TIN are located in the 3D Analyst Tools toolbox under the Raster Interpolation toolset. 

IDW Interpolation: The inverse distance weighted (IDW) interpolation is performed by running the IDW tool from the ArcGIS toolbox (Figure 2). To run the tool, open ArcToolbox and expand the “3D analysis tools” toolset. From the “Raster Interpolation” toolset, select “IDW”. The IDW tool window will open. Use the coordinate point feature class as the “input features”. Make sure to select the Z field as the “Z value field” to make sure the tool uses the elevation values in the interpolation. Name the output raster, accept the defaults, and select OK to run the interpolation.

Natural Neighbor Interpolation: The natural neighbor interpolation is performed by running the “natural neighbor” tool from the ArcGIS toolbox (Figure 2). To run the tool, open ArcToolbox and expand the “3D analysis tools” toolset. From the “Raster Interpolation” toolset, select “natural neighbor”. The tool window will open. Use the coordinate point feature class as the “input features”. Make sure to select the Z field as the “Z value field” to make sure the tool uses the elevation values in the interpolation. Name the output raster, accept the default cell size, and select OK to run the interpolation.

Kriging Interpolation: The Kriging interpolation is performed by running the “Kriging” tool from the ArcGIS toolbox (Figure 2). To run the tool, open ArcToolbox and expand the “3D analysis tools” toolset. From the “Raster Interpolation” toolset, select “Kriging”. The tool window will open. Use the coordinate point feature class as the “input features”. Make sure to select the Z field as the “Z value field” to make sure the tool uses the elevation values in the interpolation. Name the output raster, accept the all the default values, and select OK to run the interpolation.

Spline Interpolation: The spline interpolation is performed by running the “spline” tool from the ArcGIS toolbox (Figure 2). To run the tool, open ArcToolbox and expand the “3D analysis tools” toolset. From the “Raster Interpolation” toolset, select “Spline”. The tool window will open. Use the coordinate point feature class as the “input features”. Make sure to select the Z field as the “Z value field” to make sure the tool uses the elevation values in the interpolation. Name the output raster, accept the all the default values, and select OK to run the interpolation.

Triangular Irregular Network(TIN): A triangular irregular network or TIN is created by using ArcToolbox. From ArcToolbox, expand the “3D analysis tools” toolset. From “Data Management”, expand the “TIN” toolset. Select “create TIN” and the tool window will open. Use the coordinate point feature class as the “input features” and the Z field as the “Height Field” to make sure the tool uses the elevation values. Name the output TIN and select OK to run the tool. 

Figure 3: Under the layer properties window in ArcScene, elevation values are able to be applied to a surface to allow 3D displays. By right clicking the layer in the table of contents then selecting properties, this window will appear and allow the user to assign these elevation values under the Base Heights tab. 

ArcScene Evaluation of Interpolation Rasters: ArcScene is capable of displaying the interpolations as 3D models. Open a new scene and add the new IDW, natural neighbor, spline, Kriging, and TIN rasters. The elevation values need to be set for each raster to have the display model the terrain. From the table of contents, right click a raster and select “Properties” from the drop down menu. Click the “BaseHeights” tab in the layer properties window (Figure 3). Under the “elevation from surfaces” box, click the “floating custom surface” bubble. In the drop down box, make sure the layer you are working with is selected, and then click apply and ok to close the window. The layer selected will now be displayed using the elevation values and model the surface. The same process can be done for each raster to display the elevation curves. Toggle the layers on and off to compare the interpolations.
   

Discussion/Revisit: Several interesting observations were made using the interpolation models, the first being the arrangement of the features. In all the models, the locations of the features were mirrored perfectly. The elevation values were correct the horizontal location of each was opposite of the real world terrain. Coordinate values were reviewed but were correctly entered into the system so the issue remains unresolved.

Figure 4: The Kriging interpolation of the terrain was the most realistic and representative of the snow surface. The left image is an aerial view of the interpolation results, while the right is a 3D representation of those results. Boundaries of features are smooth and gradual in transition to the surrounding elements mimicking the snow. 

Of the five types of interpolation methods, the Kriging method returned a model that best resembled the real world snow terrain in the planter box (Figure 4). Feature boundaries were smooth and gradual such as those in the terrain. The edges of the valley were not defined as well as the real world terrain but were represented correctly by the assigned values. The next best model was the spline interpolation (Figure 5). Like the Kriging, the feature boundaries molded into the plain but in the spline, small imperfections in the plain and along the edge of the valley differ from the snow terrain model.

Figure 5: Results of the different interpolations. The TIN (top left) surface is accurate in location of features but not in the structure of the features. Spline results (top right) display over the plain peaks where the survey points were located created a washboard effect over what is really a smooth surface. This is very similar to the IDW (bottom right) display that small peaks are shown in the model. Also the boundaries of the features are varied and not as continuous as the snow counter parts. The natural neighbor model (bottom left) is middle ground between the IDW and Spline. Feature edges are continuous but the surface is wash boarded from the imperfections in the interpolation. 

The TIN was accurate in its representation of the feature locations and general shapes, but the resolution of survey points was coarse (Figure 5). To create a more realistic model using a TIN, more survey points may have been taken around and of the snow features. The increased number of points usable by the tool would allow smaller triangles to shape the real world features more realistically. Natural neighbor interpolation was the worst in representing the real terrain (Figure 5). Along the flat plain surface, small peaks were created in the model where the survey points were. This created a washboard appearance to the model that was unrealistic. The ridge was less defined in elevation change from the plain as well as the boundaries of the other features. Edges of the features were not gradual and even throughout the model, falsely representing the snow surface.

        As in all field work, the field provided a few big challenges in this exercise. Wisconsin, after all, isn't known for gentle pleasant winters. With air temperatures hovering around 10 degrees Fahrenheit with wind chills below zero, dexterity became an issue. Not being able to feel or move your fingers makes any job harder especially one that requires patients and accuracy in measuring features you sculpted out of snow. Also with the temperatures being so cold, the moisture content of the snow was very low. For an example, try forming a ball from powder sugar. The light, fine, and fluffy snow made forming the ridges and peaks tricky. The features were soft and not as rigid as ones that could have been made from heavier snow. This made measuring difficult since the surfaces of the features weren't solid. We had to be careful not to alter the surfaces when moving the meter stick from cell to cell. Using the string allowed the measurements to be made without changing the faces of the features drastically since the string was able to mold around the surface but still hold true to location.

Conclusion: Field work is unpredictable. Each day, especially in Wisconsin, is its own identity in terms of weather and you have to plan for everything. Weather changes along with the availability of tools and supplies. Being able to solve problems with limited resources is a skill that is great in any situation not just field work. Having the persistence and the ability to work through and problem solve successfully in these types of situations demonstrates commitment and responsibility to an employer.

Creation of a Digital Elevation Surface using Critical Thinking Skills and Improvised Survey Techniques

Introduction: Digital surface models are a new, extremely useful piece of information a geographer has available to help study the Earth. The ability to generate a digital version away from the field allows researchers to review and test hypotheses on a study area without having to travel in and out of the site. Taking a real world feature or landscape and transforming it to a model on the computer screen can be achieved several ways one which will be described in the following sections. The goal of this exercise was to create and survey a terrain field of interest.

Methods: For an introduction to surface modeling, a snow-filled planter box measuring 112cm X 232cm served as the study area. The snow in the box was leveled off by sliding a board along the top edges of the box to create an even starting point for the terrain. From this point, the terrain was shaped by hand resulting in the formation of a depression, ridge, two peaks, valley, and plain.

The snow of the planter box was first leveled with the help of a board. The terrain was then sculpted by hand and featured two peaks, valley, plain, ridge, and depression.

The use of string allowed the grid to follow the contours of the terrain without compromising the line of the grid itself.

Each grid line was placed a 8cm intervals from one end of the box to the other. By the string being draped over the terrain rather than taught from end to end, allowed the grid lines to mold to the surface without damaging the soft snow structures. 

The completed coordinate grid of 14x29 8cm cells. 

  Once the formations were completed, a string grid was placed across the formations. Each string was placed 8cm apart and stretched from one end of the box to the other and from one side to the other. The 8cm spacing was marked on the edges of the box using meter sticks and black pens. Since the peaks and ridge formations reached above the extent of the top of the box, the strings were draped across the surface to prevent damage to the formations.

Using Microsoft Excel, a spread sheet was created to help record the elevation values with the horizontal and vertical positions of cells in the string grid as well as their corresponding elevations. Surface elevation was then measured using a meter stick and the string grid as a zero reference. The grid formed by the string allowed the most extreme value (highest point if above zero or lowest point if below) of each grid square to be recorded with the coordinate position of the cell. To help keep the elevation values true, each string was pulled tight, if possible, to keep the zero constant with the top of the box.

After all grid cells were assigned elevation values, a new zero was assigned to the lowest elevation value. The difference or value of the lowest elevation was added to all cell elevations to prevent negative values. For example, the lowest recorded elevation was 12cm below the top of the box (-12cm), therefore, using that point as the new zero, an addition of 12 was added to the recorded elevations preventing those negative values. This addition of 12 cm was added to all cell values including the X and Y cells to keep the scale true to the real features.         

Discussion: Several challenges arose throughout the duration of this exercise; the first being where was zero.  The group decided that using the sturdy frame of the planter box would work well as a starting zero but wouldn't prevent negative values for elevations. It was then discussed and decided that the later relative adjustment based on the lowest recorded value would augment the elevations relative to the top edge of the box but not distort the shape and location of the formations. The second challenge was how to grid the survey area. If the strings were taught and secured on the box edges, they would cut into the soft snow formations that rose above the box. To remove the risk of damage or alterations, the strings were pulled tight then gently laid over the formations so that the line stayed true but didn’t alter the snow. The third challenge was the surface elevation measurements. The survey method chosen was a single point maximum where the point measured in the cell was either the highest (above 0) or lowest (below 0). Average point measurements for each cell may have been a more accurate survey technique but with the meter stick to measure with and the resolution of the grid, there wouldn’t be much variation in values. The measurements of those formations that exceeded the height of the box were tricky to record also. Because the surface was uneven in the cells, the meter stick often stuck in the snow about a centimeter. This variation may also skew the results slightly, leading to a model that isn’t as accurate.

Conclusion: The method of using a grid system is common when surveying terrains. Different elements that can be present in models and in real world settings pose different challenges when trying to replicate. Modeling and field work in general requires on to think on their feet and problem solve with limited resources available. Not every idea’s going to work, so the process of trial and error continues until the best option is found. All field skills are extremely important and can be applied to any instance in a person’s life.