In trying to determine the answer to this, I’ve found a lot of stuff talking about how they used to hand-draw them, that the art of shading is dying out, etc. But I guess more what I’m wondering is how they knew in such detail what the elevation was of every little bit of land. And then how would that be represented in such a way as to provide the information for the artist to draw accurately?
They took aerial photographs with two cameras separated by a known distance. The photos were examined with a viewer that superimposed the two images and allowed the operator to move a dot of light which was adjusted to appear at the same position and altitude as a point on the surface. The three dimensional coordinates were either recorded, or perhaps the operator traced out contours manually by adjusting the spot to be at a fixed altitude, then scanned it in x and y to position it on the surface.
It has been many years since I heard this described, so I only remember the essential detail of the adjustable spot of light viewed by a human being.
And before then they used trigonometry:
The Great Trigonometric Survey
This was the survey that nailed Mount Everest’s height to within 30 or so feet of the measurement based on the most recent technology.
The device used is a stereoplotter. When I were a lad I was shown one, and shown shown how to use it. I was amazed at both the mechanical complexity, and the quality of manufacture. (This was a fully mechanical version that directly cut contour maps.) But I couldn’t get the light spot to sit on the terrain.
I thought the OP was talking about before any technology was available. And I imagine he also meant “European”… or not as the case may be.
I have it from travel writers that it was done, often by the militaries of various countries, (occasionally by explorers) who would be exploring areas previously “unknown”, unknown that is to them, by simply using the eye and estimating, penciling in various landmarks, etc.
Many travellers who used these maps would often be in for surprises.
I was under the impression that (from the 19th century at least) they had human surveyors actually walk the terrain and measure the elevations all over the place (as in this movie). And, at the same time, setting official milestones that could be used for locating and sizing lots for legal / title purposes. Is that mostly a fairy tale?
No, I did mean in the kind of mid-20th century era. I don’t fully understand the technique described with the point of light (or how they could know exactly where the planes were), but it sounds pretty amazing. A lot easier to just bounce lasers off everything (which is what I assume they do now).
They would already have surveyed the major landmarks, e.g., road intersections and major buildings, and it would be relatively easy to locate those landmarks on the aerial photographs, then interpolate the position of everything in between.
Even now I’m not sure they do know all elevations so accurately: this mountain was re-measured to settle debates. Bear in mind that this is a mountain that has hundreds of people going up it every year and the Ordnance Survey, who make some of the most accurate maps in the world, weren’t sure exactly how high it was - I have a map from the 1960s that shows the height as being 2980ft, 8ft lower than the later survey got it. Think of mountains in even more remote areas where knowing the precise height isn’t important to anyone. Also bear in mind that the bits in between summits may not be measured so accurately - every so often you hear a hiker saying that they thought a map showed a crag as being too high relative to a nearby col or something.
In the US, the National Geodetic Survey had a series of benchmarks all over the country. There were brass markers that gave the exact elevation at that point. Other locations could be determined by this, and the NGS would put out topographic maps using these as . . . benchmarks.
But I’m talking more about the maps you could get for hikers, that were way off in the middle of the wildnerness away from all that stuff. Yet you could get ones that were super “zoomed” in and incredibly detailed. Hard for me to understand how they did it, or for that matter *why *they would do it if it takes that much work and it’s just for the very occasional hiker.
Lots of things get surveyed, you only need a couple of known locations to get your scale. From that scale you can superimpose aerial photos and draw out what you want. The stereoscopes make a 3D image and because you know the scale you can figure out how high something is and make topo maps that way. It’s not that difficult and they’ve been doing it for 70+ years.
Maps are not as accurate as people seem to think. Who cares if something is off by 20 feet. They don’t get super critical until it becomes a chart and is used for navigation. Even then they have symbols for items that are close or probable.
They sometimes cut corners: compare
old map
with
new map
(with the new maps you’ll have to zoom out to 1: 50 000).
Or how about this? Look at the top that’s called Toman Coinnich in the new map:
old map 2
new map 2
935m is 3067ft. In the old map the top is only just above the 2750ft contour, so it is wrong by about 250ft. I read about this error in this book.
These are in the comparatively heavily populated Scottish Highlands - I would guess that in the vast uninhabited expanses of North America and other continents, such errors on maps are not unusual.
They weren’t very accurate, it’s as simple at that. Even today maps often show trails going in the wrong location, or bridges in the wrong spot, and elevations off by a fairly significant number of feet.
By the way the old maps in my example are from the 1920s, but the same topography was still shown on maps in the 1970s, perhaps later.
And yes, Telemark, I have frequently found the depiction of paths on maps in Scotland to be a work of fiction. Maps in England are not so bad, because of the system of public footpaths.
If the OP is in the US, I’m assuming he means the 7.5’ USGS topo series. Sometimes trails and structures and things are out of date because they often don’t bother updating them in wilderness areas, but the actual elevations are extremely accurate. The bulk of these were made in the 1940’s through late 1970’s and though they were aided by aerial photos, most of the actual data on the maps comes from good old fashioned boots-on-the-ground surveying.
The term you want to look up is photogrammetry. Wiki has a decent article so no need for me to to outline the technique here. For information on an older technique look up “plane table and alidade”.
Remember maps are always a “best fit” situation and so will always have inherent inaccuracies. However, as equipment and techniques have become more sophisticated resolutions have increased.
Because land masses change with time, map data is updated periodically sometimes with significant changes in both horizontal and vertical positioning. When comparing differing maps (and perhaps even more important when using GPS) one should be certain that the same datums and epochs are being used or adjusted for.
A plane takes aerial photos, clicking (for example) every 2000 feet for photos that cover 5000 feet across. As a result, you have two photos, taken from different vantage points, of every section of terrain. Operators look through an eyepiece at side-by-side photos and their brains create an image with the illusion of depth, such that they can move a light dot around on areas that appear to be the same distance below them. (This takes a skill similar to being able to see those MagicEye pictures). These curves made by moving the light dot become contours, which are one way to represent topography. These contours are interpolated and matched up with known elevations, with a fair amount of artistic license used to create a complete, realistic picture. I don’t have any examples at hand, but quadrangles done as part of one mapping project (covering one national forest, for example) probably would not match up perfectly with quads done for another project at a different time.
The vast majority of USGS topos have contour lines done this way. Hypsography based on Digital Elevation Models or LIDAR has only become possible in the last 20 years.
In the 80’s and 90’s the trails of the White Mountains in NH were completely mapped with GPS and they made lots and lots of corrections. The new maps were significantly more accurate, but not all areas in the US have been remapped. Even elevations were corrected in many places, not necessarily of summits but of trail junctions and sub-peaks.
The older maps were plenty serviceable, and were pretty accurate. But the new technology allows them to be even more accurate.
Very interesting info, thank you! I understand now about the “magic eye” thing.
It still blows my mind that the USGS did so much work in the mid-20th century to make all those maps. I wonder: if they knew the technology that was coming a few decades later, would they have been tempted to say “ah, screw it…let’s just leave it for our kids or grandkids to figure out with their fancy gizmos”?