Distance Azimuth Survey

Introduction:

Technology can fail, and often does right when it is needed most. This can have devastating effects on field collection of data, if one is not prepared for it. This lab was intended to teach us a method to for collecting data, when left with only the most rudimentary equipment. This method, distance-azimuth, requires only a compass and some means of measuring distance, like a measuring tape. This allows it to be a backup in situations when other equipment is rendered inoperable or unavailable. When standing at a known point, one can determine the location of a second point, by determining the compass heading needed to travel between the two, and by determining the distance between the two points. This allows for trigonometric calculation of the second point's location on the earth's surface.

My group chose to survey the campus commons, as the area had a large number of possible data collection points (Figure 1). As it is a relatively open area, points are easily identified on aerial images. This will allow for my group to easily compare the accuracy of our collected data to existing aerial imagery.

Figure 1: One view of the survey area.

Methods:

The class was divided into groups of two. We were to use a TruPulse laser distance finder to determine the distance between our location and the desired point. The TruPulse had a built-in compass, so we used the azimuth information it determined. We recorded the locations of rocks and trees by recording the Object ID, Distance, Azimuth, and Type in an Excel spreadsheet. We used a Trimble Juno 3b to record the coordinates of our survey site.

Figure 2: Myself 'firing' the laser at rocks and trees
 to determine distance and azimuth.

Figure 3: Ally entering points into the spreadsheet in the field.

We conducted our survey between 2:52 PM and 3:38 PM on September 30, 2015. The weather was clear and sunny, and the campus commons was very busy. We recorded points from a fixed location with myself leaning against a small tree to stabilize the laser, and with Ally recording the measurements and types into a laptop (Figures 2,3). After we finished collecting the points, we obtained the geographic coordinates from the Juno and added the coordinates to X and Y labelled columns in the excel spreadsheet (Figure 4).

Figure 4: A portion of our excel spreadsheet.
Note the adjusted labels of the X and Y columns.

Next, we imported our adjusted spreadsheet into a file geodatabase using the table to table (single) import tool in Esri ArcCatalog. After the table was imported, we used the Bearing Distance to Line tool in Esri ArcMap to convert our feature table to lines (Figure 5). Our first attempts to convert the feature table to lines resulted in extremely skewed data, at which point we realized our X and Y coordinates were accidentally swapped (Figure 4). After correcting this user error, the results displayed properly (Figure 6).

Figure 5: The 'Bearing Distance to Line' tool.
If your X-Y data comes from a GPS, make sure your spatial reference is WGS84!

Figure 6: The output of 'Bearing Distance to Line'

Next, the 'Vertices to Points' tool was used to convert the endpoints of the generated lines into an independent feature class (Figure 7). By selecting the 'Point Type' 'END', the output feature class will contain only the endpoints of the bearing lines (Figure 8).

Figure 7: The 'Vertices to Points' tool.
Make sure you set the 'Point Type' to 'END' to only get the endpoints.

Before comparing the points to the aerial imagery, I projected the points to the Eau Claire County Coordinate System (WKID: 103417) so the aerial images and points would be in the same coordinate system (Figure 9).

Figure 8: Output of the 'Vertices to Points' tool.

Figure 9: The 'Project' tool with Eau Claire's County coordinate system selected.

Figure 10: The complete data flow model in ArcGIS Model Builder.

After the point locations were generated by the 'Vertices to Points' tool, we were able to truly compare the results of the survey to the aerial image. The initial results were quite inaccurate for the entire surveyed area (Figure 11). We believed this was due to the inaccuracies of the GPS we used. In order to compensate for this inaccuracy, we compared the aerial image in ArcMap to a more updated image in Google Maps that featured the newly planted trees on campus, and added a new point where we thought the origin should be (Figure 10). After bringing the new spreadsheet into my geodatabase, I created a new map with the new origin point. The new origin point brought the points closer to their original locations, but I thought the accuracy could be further increased by compensating for the magnetic declination. The declination for our location is around 1 degree 6 minutes West, according to the NOAA Geomagnetic Calculator. I added a column to calculate new azimuth values by adding the existing values with the geomagnetic declination of -1.1 degrees. After adding the new spreadsheet to my geodatabase and mapping the points, it delivered higher accuracy than any of the previous points (Figure 11).

Figure 11: Points from the original GPS point.

Figure 12: Original azimuth values from the new origin point.

Figure 13: Adjusted azimuth values from the new origin point.

Discussion:

'Firing' the laser from a sitting position allowed the point collector to maintain the same position for the entire length of the survey without the use of a tripod. It did however have one major drawback, as some rocks became obscured by the slight decline in elevation between the data collection site and the Little Niagra Creek (Figure 14). The accuracy of the point locations seems to decline as the distance between the observer and the surveyed point increases. This inaccuracy might have been caused by the equipment, or possibly the slight tremor in my hands causing the laser to be fired off-target.

Figure 14: The red polygon shows the area viewed in the two images.
The top image was taken at the same location and perspective from which the laser was 'fired'.
The bottom image was taken at the same X-Y location as the top image, but was captured while standing.

If I were to do this lab again, I would set the laser on a tripod to eliminate any possibility of shaking hands affecting the recorded data, and to eliminate inaccuracy caused by the low surveying perspective. I would've also moved locations several times to shorten the distance between the survey location and the points surveyed to compensate for any mechanical inaccuracy. The tripod's positions would be measured using distance-azimuth from the corners of buildings, so a GPS wouldn't be needed. I would also make sure that the points I'm surveying have actually been captured on an aerial image or another form of control.

Conclusion: 

The technique allowed my team to collect points over a larger area in a substantially faster time than other survey techniques would've allowed. This technique could be used when collecting points at a distance when survey-grade accuracy isn't necessary, like if one were to map out all of the locations of "Wall Drug" billboards along US-90 in South Dakota. For surveying applications, this technique has been replaced by the 'total station' which uses the same principles, but with substantially higher accuracy. The survey points with the new origin point and azimuth adjusted for the geomagnetic declination had sub-meter accuracy in some locations, but had irregular patterns of inaccuracy. I believe this may have been a result of electromagnetic interference (EMI), but as my partner and I didn't have a second compass to verify the strength of EMI at our survey location, we can't be sure.


Sources:
http://www.ngdc.noaa.gov/geomag-web/