Over the past few days I have been looking at historical US/BLM land survey documents (for Florida). These were done between 1827 and 1857 (with a small number after that). These are the original surveys for Florida. The older ones are titled ‘Territory of Florida’, while the later ones ‘State of Florida’.
These were primarily done to establish township/range boundaries (6 miles x 6-miles), and section (1 mile square) lines. As the team was doing these measurements, they came across various features, either directly on, or next to a line they were measuring (i.e. close enough that they could see it without going off into the wilderness).
Some of the things that have been seen: natural wells, native-american burial mounds, forts, military roads, native-american trails, vegetation notes, etc. These items were (for the most part) static and would not have changed location. Another interesting discovery is that some currently mapped features, had names that were spelled differently 160-190 years ago.
Is there value in entering some of this historical data ?
N.B. I found one survey this morning, that defines a location for a settlement called Wacahootie, with a small notation ‘Bowlegs Old Plantation’. That particular document is scaled 40 chains : 1 inch, and contains accurate northing and easting measurements.
Yeah OHM is at an early stage, but certainly this sort of info belongs better there than on OSM itself. There is no problem creating combined visualisations, for instance by using the OHM Overpass instance with an OSM background map or with umap.
Your use case would seem particularly suitable as most of the items you are mapping are fairly discrete and you can establish a date when they existed. Of course these might also engender interest in trying to find them again: at which point they can simply be added to OSM as well. In fact we are increasingly coming across OSMers, such as yourself, who have interesting historical data they want to map alongside stuff which still exists (at least 3 papers at the recent SotM-Fr conference alluded to this problem).
The basic rules are that start_date and end_date should exist on any tagged element. You may also want to things like earliest_start_date, latest_end_date, and in principle other values depending on the fuzziness of the date. There is also an epoch tag, something like epoch=roman. The first two are essential for the rendering mechanism.
As for other tags you are of course free to use what you like. In general OSM tags work fine, but avoid using historic=castle for something which has just been built! In many cases (like this) it highlight slight issues with the OSM tags: it will take longer to work out how to refine these simply because we have fewer contributors and they are less likely to encounter each others work.
Personally, I also like to add a snapshot date (most stuff I have added post 1800 is from maps dated at a particular time. If unsure about tagging you can create personal tags which can be changed once a more widespread tag is identified. So I use my username SK53:feature=xxx.
Adding source tags is really important for OHM, particularly if we ever want academic historians to appreciate its value.
On this same subject, one of the things that popped up in my mind last night, was “what datum was in use when these historical surveys were done ?”. The answer appears to be none, other than for the fact that they had to establish a latitude/longitude for the starting point (Tallahassee Meridian, 1824), likely by solar methods. I am assuming that the township corners in the field were established similarly, then networked back as the lines were surveyed.