Volumetric Analysis with UAS Data

Introduction

Volumetric analysis is a type of analysis often used in conjunction with UAS data to determine the amount of soil/material added to or removed from an area. It is an analysis of the volumes of a specified area. Volumetric analysis can be performed for many different purposes. When moving soil away from one location, it can be used to determine how much soil has been moved in a certain time frame to help give accurate time assessments for future projects. It can be used in mines for determining how quickly work is happening. Two methods utilized in this class to perform volumetric analysis are using Pix4D volume calculations and ArcMap volume calculations. UAS data is very valuable for these types of analyses because of the ease of access to data, and the high quality of the data. The higher quality, the more accurate the calculations can be. The easier it is to acquire data, the more often the data can be taken.

Methods

Pix4D Wolfpaving Calculations

Volumetric calculations can be quickly done in Pix4D for a fairly accurate assessment of an area, so we started there with the Wolfpaving data set. Utilizing the volumes tool and tracing out an area to calculate a volume for, Pix4D create volumetric assessments for each of our piles. The calculation areas for each pile is shown in Figures 1 and 2.

Figure 1: Pile A (right) and Pile B (left) calculation areas

Figure 2: Pile C calculation area

Pix4D calculated the volumes of these piles based on the area we outlined, and the base elevation that we gave it. Ensuring the base elevation is correct is very important for this step because the calculations take the volume of everything above that elevation in the selected area. Volumes for each of the piles are shown in Figures 3, 4, and 5 below.

Figure 3: Pile A volume

Figure 4: Pile B volume

Figure 5: Pile C volume

ArcMap Wolfpaving Calculations

Volumetric calculations in ArcMap are a bit more difficult than in Pix4D, but can be more accurate as well. In ArcMap, we utilized the Litchfield data set. The first step is to create a feature class, and create polygon areas in which we are trying to calculate volumes for. The polygons created are shown in Figure 6 as an outline around the feature.

Figure 6: Pile A, B, and C polygons

After the polygons have been created, the next step is to the Extract by Mask tool to extract each of these features for calculations. After the mask extraction is completed, calculations can be done. Just as in Pix4D, setting the base elevation as a reference is very important. The volumes calculated in ArcMap are shown in Figures 7, 8, and 9.

Figure 7: Pile A volume

Figure 8: Pile B volume

Figure 9: Pile C volume

The differences between the volumetric calculations in ArcMap and Pix4D are quite large. All of the calculated volumes are much larger in ArcMap. Figure 10 is a table below that compares the two sets of calculations, and shows the difference in volume between them.

Figure 10: Pile comparisons (in m^3)

Figure 11 shows a map of the area calculations and a reference to where the survey was done.

Figure 11: Wolfcreek Piles Volumetric Data map

ArcMap Litchfield Calculations

The same process used for the Wolfpaving dataset volumetric calculations will be used for the Litchfield calculations. However, one additional step will be completed for this new data set: resampling. The data will be resampled at 10cm prior to calculating volumes to ensure consistency across the timeline. Considering that the calculations will be done from surveys that are months apart (July 22, August 27, September 30), we need to ensure some form of consistency so the data is properly comparable.

Figure 12: July 22nd clip

Figure 12 shows the clip from July 22, with an area of 41,816.80 cubic meters.

Figure 13: August 27th clip

Figure 13 shows the clip from August 27, with an area of 83,810.11 cubic meters

Figure 14: September 30th clip

Figure 14 shows the clip from September 30, with an area of 51,635.62 cubic meters. There was a slight difference in volumes between the initial sampling and the resample at 10cm. For example, the data for July 22 decreases volume by about 11 cubic feet after the resampling. However, with figures upwards of 40,000 cubic meters, 11 cubic feet is a nominal difference. For smaller data sets, it would be very necessary to resample to ensure the most precise numbers are found. In a large area like this one, resampling is helpful and does provide a more accurate assessment, but 11 cubic meters makes little to no difference when we're talking about 40,000+ cubic meters.

Discussion

Figure 15: Litchfield Survey Area Volumetric Changes map

Over the course of the three months this dredge operation was surveyed, we were able to calculate the change in volume over the course of that timeline.  The changes throughout the surveying were evident, with August being roughly twice the volume as July, and September being about twenty-five percent more than July. Resampling of the data to 10cm made a very minor difference in the volumes as was discussed a bit above. However, with volumes this large, the resample made very little difference. Two things that would have made very large changes to the numbers would be selecting a different base elevation, and clipping a different area. As has been discussed, the base elevation's accuracy is very important. Without the base elevation being accurate, the volumes would be much higher or lower than they should be and would not properly reflect the amount of material being moved to and from the site. Clipping the area was done by hand, so there was room for human error in this component. If the area had been changed slightly, the numbers could be skewed greatly. I tried my best to keep certain areas, like the southwest side consistent in my clipping because they stayed very consistent throughout the surveys. However, as with September 30, a large area has been removed and a call could have been made to include or not include different areas of that date.

Conclusion

Volumetric analysis is a crucial component of UAS data analysis and can allow for the tracking of an area's volume over time. This can be very helpful to know how quickly work is being done, and how efficient the workers are. If a certain amount is supposed to be moved under contract, or if the contract amount is based on how much volume was moved, volumetric analysis can be a great way to keep everyone involved in the operation honest. UAS data is a precise enough way to calculate these volumes, and is more more consumer-friendly and less time consuming than other methods that have been used in the past.