Mapping the old stone walls of Northborough
A proposal for the creation of an interactive, interpretive historical map of Northborough's earliest organizational structures

swim.jpg

Please note that this document is a non-final draft proposal and is not yet complete.

Overview

I propose to map Northborough's old stone walls using Geographic Information Systems (GIS). We can work together to create the Northborough Stone Walls Interpretive Map using a combination of GPS equipment, GIS software, LIDAR imagery, and satellite photos. The map and its accompanying dataset can be published alongside a new essay on how the stone walls relate to the history of the town; I propose also to research and write this essay over the course of the project, and to share it with the public along with an interactive zoomable JavaScript version of the map. I aim to release the dataset and its associated files under a Creative Commons license, which would allow the townspeople to take ownership of the data, improve upon it, and produce derivative works for the benefit of both our own historical knowledge and that of future generations.

Approach

I have looked at similar efforts in order to define a workable strategy that can be completed in a reasonable time-frame. Perhaps simplest in approach, there is apparently a one-man project to map stone walls in the town of Royalston using GPS. At the other end of the spectrum there is a bold initiative in New Hampshire aiming to map every stone wall in that state; LIDAR data are algorithmically processed to identify candidate features, which are then verified by volunteers on the ground.

A GPS-only solution seems impractical, and it is likely not feasible to measure every last wall in perfectly precise fashion. Therefore I have decided to use a hybrid approach and combine multiple data sources, working with a few volunteers in coordination with local officials. The process should have several stages, which can to at least some extent be carried out in parallel:

  1. Superimpose LIDAR data over a geo-referenced 1887 map of Northborough with all major historical roads, and identify an initial set of candidates by tracing LIDAR data along the roads. As I understand it, the vast majority of stone walls were completed before 1887 and so this procedure should suffice to identify many of the historical roadside wall candidates. (Some walls are not wide enough to show up in the available LIDAR imagery, so this will not be a perfect strategy.)
  2. Walk all the major roads that existed in 1887 with an accurate GPS unit and record the positions of roadside walls, fixing good candidates, removing bad candidates, and finding any additional walls that were not identified as candidates.
  3. Visually identify a set of property-line candidate features in LIDAR data, primarily looking for clear lines that coincide with tax parcel boundaries.
  4. Use GPS to map whichever property line candidates as can be accessed with written permission from property owners.
  5. Map hidden walls in the forest using GPS. I know that Edmund Hill Woods and Mount Pisgah have stone walls, and more may exist elsewhere. Researching the historical circumstances of why a given property is no longer developed could add substantial interest to these portions of the map.
  6. Use satellite photos and/or LIDAR to "soft-verify" any remaining candidates for which this is possible. These can be marked on the final map in a different color or line style in order to distinguish them from GPS-verified walls.
  7. Hide all the remaining candidates that could not be verified.

Timeline

I estimate that we could map over 85% of the town's walls in the span of approximately two years. While the resulting map might be incomplete, the project could still be of enormous worth, because we can add rich historical context in the form of numbered map points which correspond to clickable galleries of photos and informational sidebars. We can also include scientific information about the geology of the stones themselves, and of the ancient glaciation processes which deposited them in the earth for the settlers to find.

Once a "skeleton" of verified walls can be traced on the map along major roads and parcels, a preliminary interactive version of the map can be published online, showing verified walls and unverified candidates in different colors; the site could include an invitation to property owners, who could volunteer to help complete the project by allowing their walls to be measured on-site. I estimate that this preliminary version of the map could be produced in 6 to 8 months. I plan to use OpenLayers, a permissibly-licensed free JavaScript interactive mapping application, so that the town can hold the map in perpetuity without the fees associated with commercial software. Industry-standard GIS formats such as Shapefiles and GeoPackages will be provided as well, so that the raw data can be improved further in the future.

Techniques

A detailed discussion of the techniques and equipment involved in mapping old stone walls can be found at this link. In order to prepare for the project, I've secured permission to practice these techniques and test out the equipment by bringing a small team to map approximately 4,000 feet of old stone walls on an 80 acre private property in Royalston in mid-May of 2023.

One additional way to enrich the map would be to introduce points of interest identifying other notable stones such as cairns, the "horse block" next to the Unitarian Church, and the hitching posts on Whitney Street near Country Candle Lane. I certainly would regard the massive sphere in the cemetery (across from Zeh School) as a notable stone; the grave of the Reverend John Martyn (1706-1767) would seem itself to be no less impressive. But I leave it to others more knowledgeable than me to decide whether it is proper to include gravestones in this list.

I believe there is an old stone cellar remnant somewhere in the woods between Summer and South streets which I saw as a child. This should be visible on LIDAR, and it might be interesting to use machine learning to identify other stone cellars or even walls.

Resources

I have already purchased a ruggedized field recorder and a standalone GPS receiver of reasonable precision (GPS Coordinates within: +/- 2.5m CEP). Although this device alone would not be precise enough for walls whose average thickness is about 1 meter, I believe we can use the fact that most walls coincide with roads or property boundaries as a corrective factor, combining LIDAR, satellite photos, and field observations for a given location in order to produce a result that is sufficiently accurate and which "looks right" when superimposed on other map layers.

As I will be providing my own transportation and equipment, my proposal contains no request for any funding.

What I would need to ask for is permission to visit any stone walls that might be in restricted areas owned by the town of Northborough. I might also ask for copies of certain information, maps, or historical records in order to help establish some context around the data and to help flesh out the accompanying essay.

About me

My name is David O'Toole and I have lived on Whitney Street for most of my life. I attended Worcester State University and received my bachelor's degree in computer science in 2001. I currently work as an IT support specialist for a small consulting firm based in Templeton. My friend Mike Leonard is a lifelong resident and we would like to work together on the project.

I can be contacted via email if you wish to discuss the project. Thank you for reading.