Hadleyville Cemetery Mapping - Week 3

Introduction

The biggest problem that the Hadleyville cemetery faces is the loss of all records and maps that detail who is buried in the cemetery and where in the cemetery they are buried. This cemetery has been in use since 1865 and many of the tombstones may be broken or illegible. The data must now be generated from scratch due to the fact that there is no background material to reference. A GIS will allow the creation of an attribute table that will be attached to a feature class that represents the individual plots located at the cemetery. Due to the fact that the data will be entered into a geodatabase that can be viewed, analyzed and updated makes this more than just a simple map project. A GPS unit, field notebooks, and a UAS drone will be used the gather the necessary data. The overall objectives include identifying as many stones and burials as possible and entering gathered data into a spreadsheet as accurately as possible for the use of creating the GIS map. The GIS will include a basemap and plotted grave marker points which will include attribute data such as the first and last name, year of birth and death, and a picture of the plot just to name a few.

Study Area

The study area is located in Eleva, WI located in the northwestern part of Trempealeau County (see figure 1 below). Located on the south side of County Road HH, it is out in the countryside surrounded by fields and trees.

Figure 1: Eleva Locator Map

The data was collected in early September about two weeks before the official start of the fall season. 

Methods

The class used a combination of a survey-grade GPS unit, a UAS drone, cameras, and field notebooks to conduct the survey. The GPS unit was used to plot the grave markers, the UAS drone gathered aerial imagery of the cemetery to be used as a basemap, cameras were used to take close-up pictures of the grave markers, and the field notebooks were used to gather observed attribute data off of the grave markers. Due to time constraints in the field while gathering data, data was collected in a quicker pace that may have led to possible data entry errors. This was evident when comparing recorded attribute data with the pictures that were taken of the grave markers. The attribute data from the grave markers were recorded through observation in field notebooks and in an organized fashion. The graveyard was "mapped" out in grid form where students designated grave markers with a letter for the row followed by the number that the grave marker fell within that row (refer to figure 2 below for clarity). The recorded data was then transferred over onto an online spreadsheet that all students were able to access and edit which then provided each student with the same, normalized table. The final, standardized attributes that the class agreed upon were:

•  PointID (a letter and number associated with each grave marker)
•  Notes
•  Joint tombstone (whether or not a grave marker was shared between two or more people)
•  Legible
• First name
• Last name
• Middle initial
• Year of birth
• Year of death
• Standing (if the headstone was upright or not)
• Marker type (type of material marker was made from), and
• Occupancy number (number of people that shared the same stone)

Figure 2: Cemetery Grid Map

Some issues that occurred when transferring data was making sure everyone was labeling his/her grave markers with the proper PointID, students having different terminology in the same field (for instance, some may have entered 'Partially' as opposed to 'Somewhat' in the 'Legible' field), and some confusion about what some students meant exactly in the 'Notes' field with what they wrote in. Overall, however, the data normalization process went smoothly considering how much effort was needed to collaborate with each other on how to proceed with the data entry and normalization process.

Using the UAS imagery in the GIS and the survey data in conjunction with the grid map in figure 2, it was possible to then plot each grave marker in the GIS and join the attribute table to the plotted points. The UAS provided high resolution imagery that made it possible to clearly see each individual grave marker which allowed for relatively easy plotting. 

Results/Discussion

The final result featured a grave marker attribute table (see figure 3 below) and a GIS map with locator maps included (see figure 4 below).

Figure 3: Graves Attribute Table (partial)

Figure 4: Hadleyville Cemetery GIS Map

The time creating the GIS itself was much longer than the time spent collecting the data. The survey GPS only collected a small portion of grave marker plot points, so that method ended up not being utilized. Instead, students used the grid map along with their recorded data to plot (as best as possible) each grave marker. The time spent in the field collecting data was approximately two hours, while the transferring of data and the creation of the GIS took several hours and included transferring the recorded data into a spreadsheet, plotting each individual grave marker and attaching attributes and photos to each one, and creating the final map. Possible sources of error include data being inaccurately recorded and transferred, attaching attribute data and photos to the incorrect grave markers, and not precisely plotting the grave markers (as some were located underneath trees that cannot be seen through in the UAS imagery). 

Having more time or using a different type of GPS unit for plotting the grave markers would have allowed for more thorough and faster data point collection and subsequent plotting in the GIS, and having created the cemetery grid along with a normalized spreadsheet prior to recording the attribute data would have allowed all students to be on the same page and made collecting data much more efficient. 

Conclusion

Overall, the methods transferred to the objectives quite well, although it could have been done more efficiently if students had known how to handle the collection of data beforehand, but such is the way of data collection in the field. While the UAS provided accurate imagery for the basemap, the recorded data was prone to human error in recording and transferring of data, and this became evident when comparing the data in the attribute table to the grave marker photos. The sources of error are not negligible and may only have a possibility of being acceptable if a grave marker was difficult to read. Otherwise the attribute data should have been recorded and transferred accurately. Although it was not 100% accurate, this GIS project does provide something better than the original situation, which was no records of any kind whatsoever. The survey was successful overall as it provided a GIS map with plotted grave markers that included attribute data and photos that can now be updated and edited in the future. 

Sources

Eleva locator map (figure 1): https://en.wikipedia.org/wiki/Eleva,_Wisconsin

Cemetery grid map (figure 2): Marcus Sessler