[This is incomplete. Pictures and more text still to come. Emrys 2017-06-26.]
Emrys Williams surveyed the churchyard for a new map in 2017, ably assisted at various times by James Nott, Robin Smith and Anna Blomfield. In fact, it took something like 40 visits to complete, and 30 sessions of data entry back home. Perhaps 120 hours in total.
The intention was to create a map that would allow people to find individual monuments easily, and show how the churchyard was changing over time. Tombstones move about surprisingly often, falling over and being resited, or moved for convenience. A map accurate to maybe half a metre would be nice.
Emrys first tried a mixture of photos – google earth at various times, and the aerial photos that were used to make the Parish Path Map, and photos taken from both ground level and the top of the church tower. It turns out that satellite photos don’t really have the resolution needed. All overhead photos suffer from obscuration by the trees. Ground photos are OK, but the bushes get in the way, and converting ground photos into a map is not easy. Photos from the church tower are great, but the perspective adjustment is extreme. By the time the photos has been dragged enough to match a flat map, the accuracy is lost.
Emrys then though to use GPS. He actually owns and managed to a find a survey-capable GPS unit, 2004 vintage, which uses postprocessed correction to give GPS readings accurate to half a metre. The GPS still works! But it needs a windows XP computer to run the correction program. He actually managed to find one! And it still boots up! And the program still runs! But… Since 2004, the standard for publishing the GPS correction information has moved on a version. The 2004 program will not work with 2017 correction data. And the program was never updated, because the 2004 GPS unit is obsolete. It might be possible to make it all work with a modern open source program, but that’s a lot of effort in itself. And the GPS isn’t great, anyway, even half a metre of error is enough to get two gravestones in the wrong order on the ground.
Which leaves actual real surveying, like surveyors do. Emrys made two attempts to buy a theodolite on EBay, but was outbid each time – he’s being cheapskate in these straitened times. So, how to survey?
String and tape measures. Offset surveying marks out a straight line, a datum, and measures disances along the line and at right angles to the line for each plotted point. String makes a good straight line, with each end of the string defining the datum on the map. A 30m tape measure runs alongside the string to measure distance from one end. An 8m tape measure works at right angles.
But how to measure the right angle accurately? Emrys made a plane table out of a camera tripod, a bit of plywood and some steel wire pins, with a bit of fishing line to hold up a plumb bob. By sighting through the steel pins both along the datum string and at right angles, it’s possible to locate everything along the string to within a few centimetres, even for objects 10 m to the side.
Using this procedure, surveying looks like:
- Choose your datum line. There must be no obstructions (gravestones) for the string, and the ends of the string must be well-defined and repeatable. For example, the corner of the church, or a gatepost, or a measured offset from something solid like that.
- Set out the datum line and check it carefully. Get your datum wrong and every measurement will be wrong. (He did. It was. Once.)
- Divide the survey into targets. For example, tombstones, or graves. For each target, fix one place on the target to define the target’s location; for example, the south-west corner of every tombstone.
- Use the plane table, sighting along and across the datum line, to fix the distance of the target along the datum line. Measure the distance along the datum with the tape measure running alongside it. Measure the distance at right angles to the datum with another tape measure, making sure it crosses the datum under the plane table. (For targets very close to the datum, less than a metre, probably doing it all by eye, without the plane table, is good enough. But the same ideas hold.) Note the distance along and across in a table.
- For each target, note enough other measurements to be able to draw it on th emap. For example, for a simple rectangular tombstone, note the width and depth and angle to the datum, measured with a steel tape and a protractor. For more complicated targets, like a kerbed grave with a cross and a plaque, a sketch is needed, with lots of measurements marked out. make sure that it’s clear which location, measured at step 4 above, goes with which target sketch, found at step 5.
- Keep going until you run out of targets. Or get too thirsty. Or it starts raining. Or you trip over your plane table and break it. Or you get distracted by someone asking about the stainded glass. Or whatever.
- Enter the data onto the map, back home, after each survey session.
Make sure you know what data goes with which datum. If the position of a datum turns out to be wrong, you may have to move the datum on the map, and thus move all the associated targets too, and you need to know which target goes with which datum. Emrys found that a wall was about 1.3 m out of place midway through the survey. Re-surveying the wall and moving it on the map meant that trees and hedges had to move too.
It would be good to keep all these data and sketches in a nice bound surveyor’s notebook, with page numbers, and dates, and notes of who did what, with what equipment. Emrys did none of that, but it would be nice.