Field Activity #4: Conducting a Distance Azimuth Survey.

Introduction

These days there are many ways to survey an area, but sometimes the most convenient technological method will be either unavailable or unpractical. In these situations it is critical for a field technician to understand and utilize basic survey techniques that relay less on technology or expensive clunky equipment and more on geographic knowledge and ingenuity. For this exercise, students will work in groups of 3 to survey an area of their choosing with a TruPulse 360/B range finder. This piece of equipment will let students record slope distance and azimuth for a number of locations in their chosen area.

Figure 1: TruPulse 360/B. Measures distances, heights, and azimuth. Can integrate with GPS and has Bluetooth capabilities. LTI 360 degrees TruVector Compass Technology. lasertech.com

Along with recording slope distance and azimuth, students will also record an attribute of each point (height, diameter, type, etc.).By carefully choosing survey points and using either GPS technology or aerial image interpretation, students will be able to also include a starting survey point for each point collected. Once all the data is entered into an excel spreadsheet, it can be imported into ArcMap and models can then be made and discussed.


Methods/Discussion

Firstly, when working with azimuth in surveying it is important to know the magnetic declination from true north. Magnetic declination is the angle between magnetic north (magnetic pole) and true north (geographic pole). It is important to know that magnetic declination changes by location and also over time. By utilizing this website set up by NOAA, anyone can calculate their magnetic declination. In Eau Claire, WI the magnetic declination is 1 degree 4 minutes and 53 seconds W (-1.08139 W in decimal degrees).

Figure 2: The magnetic declination for Eau Claire, WI generated by NOAA at this website. The magnetic declination will be added to the azimuth measurements taken in the field to increase the accuracy of the final results.

In determining whether to add or subtract this declination from the azimuth data collected, the mannerism "East is least and West is best" can be used. When dealing with a declination to the east, the value should be subtracted from the azimuth. When the declination is to the west, the value should be added to the azimuth. Therefore, for Eau Claire, WI, the declination of 1.08139 should be added to each azimuth recorded.

The study area chosen was Wilson Park, both due to its size and proximity to campus/individuals housing. The park takes up one city block near downtown Eau Claire and has multiple trees, benches, tables, electrical boxes, signs, statues, and electrical posts. The attribute collected for each point reflects the type of object found at that location (tree, sign, etc.). The goal of this survey was to measure each feature inside the one block area of Wilson Park. Three different survey positions were used and the locations of each were chosen near a permanent structure that would be easy to find on an aerial image, for when coordinates were needed later.

Figure 3: Survey position A on the corner of Emery and Farwell. This spot was chosen due to the electrical pole that would be easy to find on aerial imagery allowing the coordinates to be determined without using GPS technology.

 

 

 

Figure 4: The view from survey position A. Trees blocked the line of sight to a good portion of the park so multiple survey positions became mandatory.

 

 

 

Figure 5: Survey position B on the corner of Earl and Farwell. This spot was chosen due to the fire hydrate that would be easy to spot on aerial imagery. This spot was originally going to be located near an electrical pole, like the first, however a bus spot was located at that position and buses keep stopping for us to get on.

 

 

Figure 6: The view from survey position B. The area hidden behind trees at survey position A is now visible and the features in this portion of Wilson Park can be measured.

 

 

Figure 7: The view from survey position C on the corner of Earl and Barstow. This spot was chosen due to an electrical pole that would be easy to find in aerial imagery. The last remaining features of Wilson Park could be measured from this position and as such, position C was the last survey position used.

 

 

 

With the data collected, it could then be entered in an Excel spreadsheet in the proper format. The resulting table had 6 fields: "pnt_numb" indicating the number of each point as it was taken, "pnt_type" indicating which kind of feature is found there, "sdist" indicating the slope distance, "azi" indicating the azimuth, and finally "X" and "Y" indicating the x,y coordinates in decimal degrees to 6 decimal places (suggested format). 

Figure 8: A sample from the excel spreadsheet. The degrees have 6 decimal places by suggestion of the professor. Last time the class was taught, students tried using less decimal places and ArcMap would error when the Bearing Distance To Line tool was used.

The above screen capture of the final spreadsheet had to be modified multiple times before it arrived at its current state and could be used in ArcMap. The original table I used did not have azimuth data corrected for magnetic declination, had slightly wrong survey positions (X,Y values), and also had one row at the bottom with null values. Simply adding 1.08139 to the azimuth values corrected the first problem. For the second problem, at first, the survey positions were found via Google Earth and converted into decimal degrees. When the data was imported into ArcMap the survey positions were just slightly off. So, ArcMap itself was used to collect the survey position values by adding an aerial imagery base map and going to [Customize - ArcMap Options... - Data View]. Here the units displayed on screen could be set to decimal degrees with 6 decimal places. The final problem was most frustrating to figure out. The null values in the table made ArcMap error when the Bearing Distance To Line tool was used. It took awhile to figure this out but once the data was entered into a new clean spreadsheet with no null values the Bearing Distance to Line tool worked perfectly.

 

A geodatabase was created in ArcMap to hold all the files pertaining to this lab. This was done by rightclicking the desired folder in the Catalog window, hovering over New..., and choosing File Geodatabase. Once the geodatabase was created, the excel table was imported into it by rightclicking the geodatabase, hovering over Import..., and selecting Table (single). Next the ArcToolbox contains the two tools needed to take the data collected from the field and display it on a computer: Bearing Distance To Line and Feature Vertices To Points.

 

 

Figure 9: The ArcToolbox with relevant tools highlighted in red.

 

 

 

Figure 10: The Bearing Distance To Line dialog box. Here the table imported into the geodatabase is selected, its output location is selected, and its different fields are placed in the appropriate sections.

 

 

 

Figure 11: This is the result after the Bearing Distance To Line tool ran successfully. A line shapefile was created that illustrates where each feature measured relates to its survey position.

 

 

 

Figure 12: Next the Feature Vertices To Points tool was run. The line shapefile create earlier is selected and all vertices receive a point, creating a point shapefile for each of the features measured.

 

 

 

Figure 13: This is the result after the Feature Vertices To Points tool ran successfully. Both new shapefiles are present and aerial imagery can now be placed underneath to give some context to what the feature are.

Figure 14: With aerial imagery of Wilson Park underneath the shapefiles, the effectiveness of the survey method can be analyzed.

Conclusions

 

Upon examination of the final product (Figure 14), it can be determined that several things went wrong during data collection. The goal of this survey was to measure each individual feature inside the one block area of Wilson Park. That being established, there are multiple points that are found outside of Wilson Park. The points that are one or more blocks away were most likely caused by faulty distance readings when the data was collected. Instead of bouncing the laser off the intended feature, the laser continued out until it found a feature beyond our study area. The points that are in the streets around Wilson Park may have been caused by faulty azimuth readings when data was collected. Most of the points do seem to correlate with the aerial imagery, although its difficult to compare the aerial imagery available to current conditions due to seasonality. This suggests that if more care was taken in collecting measurements this survey technique would be quite accurate. Using the TruPulse 360 was incredibly easy, involving nothing more then pointing the device at the feature and clicking a button. The ease of use, small size, and accuracy of the TruPulse 360 make it a quick and easy way to survey an area without the need for large survey equipment such as a total station and tripod.

 

 

 

 



Activity #3: Unmanned Aerial System Mission Planning

Introduction

Mission planning is of the utmost importance when preparing for a project, especially when Unmanned Arial Systems are involved. Unmanned Aerial Systems (UAS) can cost thousands of dollars and because many of the parts that make up a UAS are specialized, it can be difficult to maintain or replace. When planning for an outing that will use UAS technology it is important to choose the right type of Unmanned Ariel Vehicle (UAV), proper sensors, and to choose the right time of day and year to execute the mission. UAS along with remote sensing technology can be useful for many applications, such as: terrain modelling, topographic surveys, inspection work, monitoring deforestation and vegetation health, mapping structural attributes like biomass and basal area, and the list goes on. The field of UAS has been and continues to grow quickly and the need for people that understand how to properly plan and execute UAS missions has become steadily more prevalent.

Students were given five different scenarios in which UAS could possibly be used but currently was not. For each scenario, the students were to decide which type of UAS to use, which kind of sensors to use, and to think of any limiting factors that could hinder the mission from any point during its process. Given only the scenarios and a couple websites to get started, students could use any means necessary to find solutions to each scenario that involves the use of UAS.

Figures

Figure 1: The APM Copter offered by 3D Robotics. An example of a rotary craft. This is the basic ready-to-fly version with a base cost of $749.99 with configuration options available for added cost. Upgraded models of this UAV are available ranging in base price up to $1,350. Just the copter frame can be purchased along with all the indivdual electronic parts to construct your own UAV. 3D Robotics (use side panel to explore parts options)

Figure 2: The APM Plane offered by 3D Robotics. An example of a fixed wing craft. This is the basic ready-to-fly version with a base cost of $569.99. Configuration options are availabe for added cost. 3D Robotics

Figure 3: The Canon S-95. Captures visual and near-infrared (NIR) light.

Specifications

Resolution	Field of View	Weight	Spectral Bands
3264 x 2448 pixels	50 x 39 degrees	250 g (0.55 lbs)	RGB, NIR

 

Figure 4: The ICI 7640 Thermal Camera. Captures thermal long wave infrared (LWIR) light.

Specifications

Resolution	Field of View	Weight	Spectral Bands
640x480 pixels	48x37 degrees	127.6 grams (4.5 oz)	7-14 microns

 

 

 

 

Figure 5: SOC710-GX Hyperspectal Imager

SOC710-GX System Specifications

Spectral Coverage: 400-1000nm

Spectral Resolution: 4.2nm

Bands: 120

Pixels per line: 640

Speed: 90 lines/second

Focal Length: Configurable

Lens Type: C-Mount

Weight: 1.25 Kg*

Dimensions (DL): 10.3cm x 20.0cm*

Power: 12-VDC / 10 Watts

Figure 6: Fixed Wing vs Rotary

Fixed Wing

Flight Time	Longer
Speed	Fast
Structure	Simple
Best Use	Aerial Mapping, Terrain Modelling larger areas (mine sites, stockpiles), Topographic surveys
Flight ability	One-Way, circle pattern
Limitations	Need takeoff/runway, can’t carry all types of payloads, no hover capability

Rotary

Flight Time	Shorter
Speed	Slow
Structure	Complex
Best Use	Inspection work, hard to reach areas (pipelines, bridges, power lines, rail tracks)
Flight Ability	Every direction horizontally and vertically, hover
Limitations	Short flight time and complex maintenance

Scenarios

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.

 

The best type of UAV to use for this case is a fixed wing craft. With a fixed wing craft, large areas of desert can be covered and analyzed. The desert landscape is perfect for construction of runways which are needed for take-off and landing of fixed wing UAVs (figure 6). The desert tortoise lives in burrows which need to be kept at temperatures lower than 85 degrees Fahrenheit or the tortoise will suffer brain damage. The resulting tempertures of these burrows is lower than the surface temperature in the deserts of the southwestern United States. The ICI 7640 Thermal camera can be utilized to see this temperatures differences between the desert tortoise’s burrows and the surrounding surface temperature. By viewing this temperate data, the burrows of the desert tortoise can be found and avoided.

 

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.

In this case, a rotary craft would be ideal. Complications in transporting crafts from the airport and the great cost of flying helicopters would no longer be a problem. The small size of a rotary UAV like the APM copter in Figure 1 is perfect for easy transport. The maneuverability and hover capability of a rotary UAV (figure 6) will make inspections of the power cables quick and easy. By fixing a Canon S-95 (figure 3) to the craft, quality pictures in the visible spectrum can be captured and analyzed for problems in the tower. The combined cost of a rotary UAV and Canon S-95 (figure 3) is far lower than even just one flight in a helicopter and the setup will provide quick quality results.
 

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.

 

With such a large area to cover, a fixed wing craft would suit this scenario best. The longer flight time and faster speed of a fixed wing craft compared to a rotary craft (figure 6) will allow for a more extensive area to be covered. By utilizing NIR and hyperspectral images, photosynthetic processes can be monitored which include both vegetation health and growing-season length. The Canon S-95 (figure 3) and/or (depending on payload capabilities) the SOC710-GX Hyperspectal Imager (figure 5) can be used to capture the images needed for analysis. 

 

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.

 

Because the oil pipeline in the Niger River delta is hard to get to by person, a rotary craft would be best to perform the needed inspection. Great maneuverability is needed to follow the pipeline and detect leaks which a rotary craft would be able to handle (figure 6). Using the Canon S-95 (figure 3) or the SOC710-GX Hyperspectal Imager (figure 5), would enable oil sheens and stressed vegetation health to be seen and monitored. Cloud penetration is lacking for these kind of sensors, so a clear cloudless day is needed.

 

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.

 

Depending on the size of the mine, either a fixed wing or a rotary craft could be used. By using both the Canon S-95 (figure 3) and the ICI 7640 Thermal Camera (figure 5), three sets of data (RGB, NIR, and Thermal) can be combined and utilized to create an accurate 3D model of the mine each week. A cloudless day is needed for quality images.

 

Sources

http://www.infraredcamerasinc.com/infrared-camera-7640.html

http://aggieair.usu.edu/aircraft

http://surfaceoptics.com/products/hyperspectral-imaging/710-gx/

http://diydrones.com/

http://planner2.ardupilot.com/

http://planner.ardupilot.com/

http://www.usa.canon.com/cusa/support/consumer/digital_cameras/powershot_s_series/powershot_s95
http://www.mdpi.com/2072-4292/3/9/1847

http://www.ciesin.org/TG/RS/sarsens.html

http://150.197.2.88:8080/MultiData/A0156/265/12.pdf

http://www.whrc.org/education/rwanda/pdf/Walker_SAR_Veg_Mapping.pdf

http://www.militaryaerospace.com/articles/print/volume-24/issue-10/technology-focus/electro-optical-sensor-payloads-for-small-uavs.html

http://hse-uav.com/fixed_wing_vs_rotary_wing_uav.htm

http://www.popularmechanics.com/technology/military/planes-uavs/hyperspectral-sensors-the-flying-eyes-that-see-the-invisible

Field Activity #2: Visualizing and Refining Your Terrain Survey

INTRODUCTION

This is the second part of a two week exercise in which groups of students were to make landscapes in the snow, collect elevation data, and make 3D models of the terrain. First, the landscape needed to be made and coordinate system set up. With that accomplished, elevation data could be collected using meter sticks, string, and a little ingenuity. At this point, part 1 of the exercise was completed. A more in-depth view of these procedures can be seen in my last blog post, Field Activity #1: Creation of a Digital Elevation Surface Using Critical Thinking Skills and Improvised Survey Techniques.

In this part of the exercise, the elevation data taken in part one will be entered into excel, imported into ArcMap or ArcScene, and used to create 3D models by utilizing different interpolation methods. Five different surfaces will be compared and the best method will be chosen.

METHODS

Now that the elevation data was collected, it could be entered into an excel spreadsheet with column headers of “x”, “y”, and “z”. The lowest number collected was -12 and as such 12 was added to each elevation sample to eliminate any negative numbers and make the data set easier to work with.

Figure 1: The excel file that holes all the x,y,z data for the landscape. After importing the sheet into ArcScene the data was displayed as XY data and converted into a shape file to be used for interpolation methods.

With the excel sheet in the proper format, it could then be imported into either ArcMap or ArcScene to experiment with different interpolation methods. The excel file was imported into ArcScene and converted into a shape file by first displaying the XY data and then exporting the data as a point shape file.

Figure 2: This is what the XY data and the resulting shape file looked like from an overhead veiw in ArcMap or Arc Scene. In ArcScene the Z data also comes into play and the points are shown at their relative heights.

With the data in the proper format in ArcScene, different interpolation methods could be used to visually and spatial analyze the landscapes elevation.  Five different methods were used: IDW, Kriging, Natural neighbor, Spine, and TIN. These can be preformed by navigating to the ArcToolbox > 3D Analyst Tools > Raster Interpolation. To make a TIN navigate to the ArcToolbox > 3D Analyst Tools > Data Management > TIN. An interpolation method uses sample data points, in this case elevation points of the landscape, and creates a raster in which it predicts what values the adjacent cells would have based on different parameters. There are two types of interpolation methods, deterministic and geostatistical. Deterministic methods base their predictions on measured values and mathematical formulas, while geostatistical methods use statistical models to predict surfaces.

IDW (Inverse Distance Weighted)

This interpolation method estimates values by averaging the data in groups centered on a processing cell. As such, it is a deterministic method. A group of points and their accompanying processing cell is called a neighborhood. More weight is given to points closer to the center of the processing cell. Points farther away from the processing cell are assumed to have less influence. The manner in which IDW interpolates can be altered by changing the number of processing cells or by specifying the radius the neighborhood will take.

Figure 3: The surface interpolated using the IDW method. This method fit the survey the poorest. The surface has many peaks and depressions that are not present in the actual landscape.

Kriging

The kringing method is somewhat complex and uses statistics to predict surfaces, making it a geostatistical method, with a level of accuracy and certainty that is not possible with deterministic interpolation methods. This method assumes there is a spatial correlation between points based on their distance and direction to each other to create a surface. This method is best used when a distance or directional bias is known or assumed about the data. This method is best used for soil science and geology.

Figure 4: The surface interpolated using the Kriging method. This method fit the surface well. The mountain peaks are curved and the river bed is continuous and drops in elevation properly.

Natural Neighbors

This interpolation method uses area to apply varying weights to sample data that surround a query point. The surface created will pass through all sample points and will not produce any peak, pit, ridge, or valley that is not explicitly present in the data used. Unlike IDW, the weights given are based on overlapping area and not distance. However, like IDW, Natural Neighbors is a deterministic method.

Figure 5: The surface interpolated using the Natural Neighbors method. This method fit the survey well but not as well as others. The mountain peaks are a bit to pointed and the ridge is lacking in elevation.

Spline

Being another deterministic method, the Spline method uses a mathematical function to create a surface that passes through each point minimizing surface curvature for the entire area. The surface created will be smooth and will pass through each sample point exactly. There are two types of Spline interpolation, regularized and tension. Regularized creates a smooth surface with values that could extent past the data range, while tension creates a less smooth surface but will have values that fall within the data range. This method is best used for elevation, water table heights, and pollution concentrations.

Figure 6 (1): The surface interpolated using the Spline method. This method fit the survey the best. By running the surface through each point taken, the resulting model represents what the snow landscape actually looked like.

TIN (Triangulated Irregular Network)

A TIN surface is constructed by placing vertices at the sample points and connecting these vertices with edges to form triangles. The resulting network of contiguous triangles models the values between points without changing the position of the sample data. This method is best used for areas with a high degree of irregularity.

Figure 7: The surface interpolated using the TIN method. This method fit the survey but is less visually pleasing then other methods. The snow landscape did not have straight edges however the look of the surface is simply a side effect of the method used. By creating a network of triangles, the majority of curvature in the landscape was lost.

DISCUSSION

Of all five interpolation methods, IDW (Figure 3) fit the survey the poorest. Each point dips down or protrudes up in a fashion that does not correlate with what the landscape actually looked like. The mountain peaks consist of multiple peaks instead of just one and the river valley is littered with recesses that make the surface resemble Swiss cheese. The Nearest Neighbor (Figure 5) method fit the survey better but consists of too many rigid and sharp edges. The mountain peaks, ridge top, and river valley were not as pointed as this method would suggest.

The TIN (Figure 7) method for the landscape does fit the survey but in a more abstract fashion. The rigidness of the features is somewhat visually displeasing but is simply a side effect of the method. The mountains and ridges fit better than the river valley and lake depression. The area around the river valley is very blocky with straight lines that do not correlate with the actual landscape. More samples of elevation points are needed to make this method more viable.

The Kriging (Figure 4) method fit the survey very well with the only exception being the ridge top. The surface has multiple tips on the ridge when in reality the ridge has a smooth top. The river valley has a smoother and more continuous bottom and the mountain peaks are more curved when compared to the IDW (Figure 3), TIN (Figure 7), and Nearest Neighbor (Figure 5) methods, which correlate better to the actual landscape.

The Spline (Figure 6) method fit the survey best. The mountain peaks are curved yet still reach the proper height and retain their proper shape. The ridge top is curved and continuous. All the features are portrayed with the proper amount of curve and look the way the landscape should. This method also captured the slight ridge around the river valley better than the other methods did. The only problems are slight depressions around the ridge and in the river valley. This may be caused by improper sampling technique or it may be realistic to the actual surface. The area appears flat in person, however the data suggests otherwise and it can be difficult to tell minute changes in elevation when starring at all white snow.

Figure 6 (2): The surface interpolated using the Spline method. This surface represented the actual landscape the best of all five methods used. The mountains and ridge are peaked yet curved, the river valley curves and drops in elevation like it should, and overall the model fits the survey well.

CONCLUSION

After experimenting with different interpolation methods and analyzing the surfaces created, it was decided that the Spline (Figure 6) method represented the landscape best. This would make sense given the fact that the spline method creates a surface that passes through each sample point and gives the surface a smooth appearance. Using snow as the construction medium resulted in smooth curved features which was represented best with the Spline (Figure 6) method, decently with the Nearest Neighbor (Figure 5) and Kriging (Figure 4) methods, and poorly with the IDW (Figure 3) and TIN (Figure 7) methods. Taking more sample points would result in better surfaces for all methods, although each method except for IDW (Figure 3) produced surfaces that did represent the landscape well. Due to time and schedule restraints a re-survey was not possible. Had a re-survey been done the results would be incomparable due to changes in the landscape from more snow falling and wind caused snow drifts.

Using the top of the box as sea level and using a slightly distorted coordinate system was not in itself a major contributor to error in the surfaces created. More sample points in areas with swift changes in elevation were needed to create better representations of the landscape. Though the measurement method was slightly flawed, the models made in ArcScene do correlate with the landscape and look pretty neat.

Group #2 did an excellent job of working together to get part one of the exercise complete. However, it was harder to come together for part two. We were unable to arrange a re-survey which would have improved our results and a lack of communication was to blame. I learned quite a bit about the different interpolation methods offered in ArcMap and ArcScene from this exercise, as well as how to take data from the field and create useful models. The group’s inability to come together at the end of the exercise is a good reminder to start working early and communicate often. Despite the horridly frigid weather, this exercise was enjoyable, thought provoking, and open ended enough to feel like the students were not being simply walked through another assignment.