Loopy Loops
So you have happily surveyed away, squirrelling data into compass after every dive and then anxiously plotting it. Never satiated, always wanting more, you need that fix; watching the red plot lines branch and reach out ever further and then ….you close a loop and its obviously wrong. The cave didn’t really look like that. Some azimuths are clearly not right. What happened?
Well what happened is that compass tried its best to fix your loop. If compass detects a loop and finds it to be within some defined percentage of error, it will automatically fix it for you, rather than give you an error. You even have a choice of a few fix algorithms, but do you really want the fix?
I asked this question to Jim Coke, famous cave cartographer and directer of the QRSS. He advised me to never cook the data as it will inevitably come back to haunt me. Loop closure errors are of course to be expected, especially consideringthe environment that we are surveying in.
Double Trouble
At some point of course we must close the loop, to make the map either by letting computer software such as compass redistribute the errors or by manually tweaking the data a little bit here and there. So how do we do that without losing the raw data? Well we can’t, at least not easily. We need to keep two sets of data, the raw data and the cooked data which will be used as the foundation to the cartography. Although its a bit annoying and tedious, compass has quite nice features for moving individual surveys between projects.
So now its clear that we should not cook the data, the next question is what is acceptable accuracy. I asked this question on the Cave Diver Forum (CDF) and got some interesting answers. The answer I liked best came from Andy Pitkin, I posted the response here :
“Good advice from Jim, who may be the most experienced underwater cave cartographer around.
As Kelly says, it depends on the size of the loop, but also a couple of other things, particularly how much effort it took to get the survey in the first place. If you don’t need your map to be highly accurate to (for example) give divers a good idea of the cave layout, then you can leave a larger error than if you are surveying to get a precise location (e.g. of a sinkhole in the jungle that you need to cut a trail to). If you had a miserable dive thousands of feet into the system through horribly unstable passage to get the survey, you may decide that the additional accuracy you might get with resurveying isn’t worth it.
As an aside, errors in azimuth usually contribute more to loop closure errors than errors in distance, so a large error will not usually because you used knotted line rather than tape. One single incorrectly recorded digit in an azimuth, or a couple of digits transposed, can often make a huge difference. It is worth temporarily trying to close the loops in Compass (you can set up a temporary survey for this purpose that you delete later) and then running the blunder detection capability in Compass as sometimes that will show you that one single shot is much more likely to be the cause of the closure error than others. With some practice it can tell you where you may have gone wrong and therefore you can go back and resurvey a few selected areas rather than the whole loop.”
Article posted by AlanF
