General surveying questions

[FW]

In the past, I have either surveyed small caves, or sumps. Sumps are usually part of larger caves, and someone else always did the maps, and I just turned in the data. Small caves are easy, since there is usually only one surveyor.

Now I have a big project on my hands, and I want to know how most surveyors number stations, connect different parts of a cave, tie into the surface, GPS, etc.

Specific questions include:

Do you start numbering on the surface?
Do you start with station 0, 1, or something different (like A1)?
If you are working on a multi-entrance cave (like Peacock) and have what you thought were two different caves that end up connecting, how do you tie them together, especially if they don't have unique station numbers?

You probably noticed I have started using Walls software, but I suspect any other software would have the same type issues.

[SuPrBuGmAn]

1, 2, 3, 4... for main branch(which is often a guess anyway), number from the surface.

A1, A2, A3, A4....B1, B2, B3, B4... ect for branching passage.

[mpoucher]

I start numbering at the surface, usually 0, or 000 or have a prefix MAD0 or MAD000 if I know there will be multiple entrances. I use a letter in front for each survey A1, A2... , B1, B2.... Then use two letter combinations for offshoots off of those surveys such as AA1, AB1, etc. Sometimes I use descriptive lettering such AQ1, AQ2 for the Aqueduct survey. I've also used a system like AB1a, AB1b, AB1c for short offshoots off a particular station. It gets complicated if you have a large complex system like Peacock. If you have the Taming of the Slough, you can look at the data in the back for how the stations were named.

If your planning to put in markers so your survey stations match what is in the field, you may just want to use seqential numbers. This is what we did in Peacock. You can add prefixes if helps to organize the data. I no longer put markers in for my surveys unless I am planning to come back and do sketching in the cave on a subsequent dive.

To tie two surveys together, you usually have a common station that identifies the tie point. Each survey will have a station called A1 for instance that is the tie point. In Compass, you can have duplicate names in different surveys and you have to create a file that lists tie points so the program knows to close the loops. I don't know how Walls handles loop closure. It may just assume that two stations named the same are a tie point and require that you don't have duplicate names. I that is the case, your station numbering convention may change. ( I have gone in a renumbered entire surveys for this reason.)

Walls may also allow you create fixed stations that are tied to a GPS point. This allows you to fix the entrances of several cave relative to one another to view in the software or export out to some other program for viewing and plotting.

[FW]

Thanks Mike, that is what the "dry" surveyors seem to do. I am hoping to get the numbering right the first time, to avoid having to re-number later. So far, I only have a couple dozen stations, so it won't be hard right now. Later it might be much harder.

FWIW, this system may end up connecting a lot of sinkholes, like Falmouth/Cathedral, or even Leon Sinks. The viz is only good a few weeks out of the year, so it may be a long time before we get it all surveyed. Now if I could only get Barbara interested in helping with this system :roll:

I guess you don't use On Station anymore?

[mpoucher]

I guess you don't use On Station anymore?

I'm trying to convert to Compass and anything new I do there. OnStation is easy to get started with, but it has some quirks - the worst being that sidewall information always exports in meters, which makes exporting maps challenging. The work around is to set everything in meters, then do some fancy scaling when you get it into Autocad. The nice thing about Compass is it will export into GoogleEarth and several CAD or GIS formats. The only downside I've found is it doesn't do loop closure between two fixed points, (or I can't figure out how to do it) so something like Peacock, where you have mulitple entrance with good GPS points, it won't correct the survey.

[jlillest]

I don't think OnStation is supported anymore and will probably be unusable to anyone on Vista, besides Walls and Compass are more feature rich.

As for labeling and numbering stations, it doesn't really matter how you label stations. The best reason for using a consistent theme to how you do them is for readability and usability.

I personally despise the labeling scheme that Mike mentioned above for offshoots (AA1 off the main line of "A", etc.) because I think it clutters the lineplot which I use a lot during the exploration stage of mapping. I prefer to use a different and sequential prefixes for each survey day or section of cave depending on how large the cave is. Because Mike's system works for him and lots of others, that's all that matters as long as it leads to a completed map.

Most important of all is the usability of survey stations. We survey "dud" caves all the time, small caves that require only a handful of shots, so we don't worry about recovering stations. In a potentially large cave system like you've mentioned, you need to worry about recovering stations when you connect survey/dive lines and connect entrances to each other. With the low viz you've mentioned, you may need to leave marked stations all over the place unless you're absolutely positive when reached a "T" in the passage where you suspect a connection may occur.

I'd suggest finding some sort of system now for labeling and leaving some sort of markers so that you'll have less trouble tying survey legs together months or years later. Cookies or arrows with some sort of unique label on them should be ok as long as you leave them at a known station.

Also, remember we're doing the same exact type of surveying that the dry cavers do. They (we?) just have more accurate tools and more time to goof off!

-Jon

[FW]

One problem with this system is it is prone to violent floods. Nothing like in Florida. When there is a big rainstorm, the sinkholes in the area gush water. I suspect short of drilling bolt holes, no markers are going to stay put. I have seen TAG caves pile up gravel at the entrance 15' high. Can you imagine the force of water needed to do that? I was outside a system, just a few miles north of this one during a heavy rain, and it looked like open fire hydrants at the entrance. We jumped in a truck, and tried to outrun the flash flood, and ended up pulling up out of the stream bed, and waiting until the following day to get out of the area.

One mistake I made on out first trip in, was using a log as a survey station. I suspect it has moved hundreds of feet, if not out of the cave, taking most of our line with it :(

[RN]

What I've done with systems that require multiple dives and/or multiple divers is transfer all data to an excel worksheet before inputting into Compass. This way I can use the original numbering system used in the water and then create another column in my spreadsheet with station numbers to enter into my mapping software. This keeps a record of the station names regardless of what they end up being in the final project.

[Webmaster]

Usually start with A0 or A1 at the dripline/surface and go from there.

We usually classify offshoots as to whether it looks major or minor. If it looks minor we try not to waste a letter on it and will probably do something like AB1. But I have been fooled before. Or for real short stuff just use stations in the main trend and pick it back up after the 2-3 shots into the side passage. They don't necessarily have to be entirely consecutive in labeling.

Saving the single letter designations for the major passage trends helps you keep track of things a bit better I think, it shouldn't really matter to the software what you call it.

I've seen people take it to extremes, but I don't think you would hardly ever need to go past 3 letters.

From what I've been reading on Walls, I believe you can have duplicate station designations, even within the same cave if they are defined in separate .SRV files. Probably not a good idea as a regular practice since that could get confusing, but it would eliminate any software issues from connecting two caves with duplicate station labels.

One thing to keep in mind, especially in large surveys is which letter designations you've already used. Printing out a list of all the unique letter trends to carry with you can come in handy. It's not really a software problem, you can change it, but it's a PITA to get back and realize you have 10 pages of sketches and data with the same station labels as previous work. Makes it confusing.

[apitkin]

I personally despise the labeling scheme that Mike mentioned above for offshoots (AA1 off the main line of "A", etc.) because I think it clutters the lineplot which I use a lot during the exploration stage of mapping. I prefer to use a different and sequential prefixes for each survey day or section of cave depending on how large the cave is.

I don't understand what you mean by this. How does it clutter the line plot? In Compass you can turn station labeling off anyway, but I don't see how you are using your different and sequential prefixes. Can you give an example?

Andy

[FW]

I don't understand what you mean by this. How does it clutter the line plot? In Compass you can turn station labeling off anyway, but I don't see how you are using your different and sequential prefixes. Can you give an example?

Andy
You can in turn labels off in Walls too.

[mpoucher]

One problem with this system is it is prone to violent floods. Nothing like in Florida. When there is a big rainstorm, the sinkholes in the area gush water. I suspect short of drilling bolt holes, no markers are going to stay put. (

Forrest: I don't know how complicated your system might get, but I've had pretty good success with identifying survey stations without using markers, or coming back after re-lining. Often the stations are obvious, at worst you may have to survey a few stations on each side of the connection to determine where it fits. Gary Walten told me thats how they did it in Mexico and it doesn't get any more complicated than what they have down there. You may want to talk to some surveyors in Mexico to see what else they might do. Another thought might be to establish critical stations at cave features that won't move, or can be marked so they don't wash away in a flood.

[huxley]

would be nice to have some sort of reference to come back and measure the distance to the walls left to right and top to bottom for a 3d model.

[FW]

would be nice to have some sort of reference to come back and measure the distance to the walls left to right and top to bottom for a 3d model.
Yes, we should have taken LRUDs while the visibility was good. We can still get rough measurements with a handheld sonar.

[Neptuno]

Hi forrest, now there is a soft called ariane http://www.arianesline.com/ that Sebastian is making in mexico, is 10 times better that all the other stuff available and very visual.

to identify the stations is enough taking the angles where you tied and the angles and distance to the near ones, also can automatically add prefix to the surveys and colors, for ex. A001, or cave01, cave02, etc.

there is a video tutorial on the website, they are using it to survey all the new caves in mexico, right now we are exploring what is so far the longest cave in mexico with a single entrance (over 25000 feet) and is all mapped in ariane and works very well, also you can work over a server and all the teams can add information on the server

[huxley]

We might need to see what software we want to use for this system, I would assume that we soon will need several people entering data. I looked at a couple other ones
I downloaded therion and wanted to play around a little
http://therion.speleo.sk/download.php
what else is out there and really worth to focus on
walls, compass, wincarst, arianes, onstation ......
FW should we look into the capabilities of other software?

[Neptuno]

well, therion lasted very short in my computer... so damn complicated... :)

[huxley]

i started to read into the capabilities and yes it sounds complicated but they say it's capable of large and complicated systems
http://therion.speleo.sk/downloads/tharticle-vercors-2008.pdf

[Neptuno]

please try ariane, if you need another feature, just ask sebastian to add it..... the only missing thing is that is for stick maps (and 3d computer generated map), but you can import/export to compass and do the rest on inkscape I guess in case is needed.

[apitkin]

We might need to see what software we want to use for this system, I would assume that we soon will need several people entering data. I looked at a couple other ones
I downloaded therion and wanted to play around a little
http://therion.speleo.sk/download.php
what else is out there and really worth to focus on
walls, compass, wincarst, arianes, onstation ......
FW should we look into the capabilities of other software?

The really neat thing about Compass is its ability to export kml files that you can view in Google Earth. That way you can share with your fellow explorers the map overlaid on terrain as it develops. The kml files are just text and can be easily edited. For me that is the killer feature that Compass has over the others. Essentially they do much the same thing, with greater or lesser amounts of bugginess and varying user interfaces. If it could output kml files directly I would probably switch to Walls.

Andy

[jlillest]

We might need to see what software we want to use for this system, I would assume that we soon will need several people entering data. I looked at a couple other ones
I downloaded therion and wanted to play around a little
http://therion.speleo.sk/download.php
what else is out there and really worth to focus on
walls, compass, wincarst, arianes, onstation ......
FW should we look into the capabilities of other software?

It should be noted that Therion is a map drawing program and does not crunch data by itself. Therion crunches data externally through cave survey programs like Survex. Survex is another nice program, but more complex than Walls & Compass without the nice map display features. If you're a programmer you should be able to pick up on Survex fairly quick as it's entirely non-GUI based.


The really neat thing about Compass is its ability to export kml files that you can view in Google Earth. That way you can share with your fellow explorers the map overlaid on terrain as it develops. The kml files are just text and can be easily edited. For me that is the killer feature that Compass has over the others. Essentially they do much the same thing, with greater or lesser amounts of bugginess and varying user interfaces. If it could output kml files directly I would probably switch to Walls.


I'm not worried about exporting to kml as I can do better maps with GIS applications, still share shapefiles and data and not have to worry about Google making my private data available to everyone without my knowledge (which they've done once or twice).

-Jon

[FW]

Lets see, Fords are better than Chevrolets, unless you try to integrate GPS from the passenger seat? So far, I like Walls better that anything else I have seen, but I admit other survey programs have some nice features.

[rchrds]

1. Do you start numbering on the surface?
2. Do you start with station 0, 1, or something different (like A1)?
3. If you are working on a multi-entrance cave (like Peacock) and have what you thought were two different caves that end up connecting, how do you tie them together, especially if they don't have unique station numbers?

You probably noticed I have started using Walls software, but I suspect any other software would have the same type issues.

1. Yes and No. Yes, your survey has to be tied to the surface. (unless you know of a USGS survey marker that is underwater) No, your numbers do not have to sequentially tie to the surface. Only your survey software has to make that connection. There are two things at work here. The first is how accurately do you wish to locate your entrance? If your survey only consists of underwater survey, with no stations or ability to surface connect further into the cave, than GPS is probably going to be good enough to establish a datum. (Fixed location from which your survey is referenced.) If your cave has 35 miles of passage, and the furthest reaches are 10-12 linear miles from your entrance, GPS is NOT going to be good enough. In either case, you have to have the ability to either get a good GPS fix (i.e. sky not partially obscured by a bluff wall or mountain on one side) or the ability to surface survey to a surveyed reference. (USGS marker or similar.) This will likely require at least one, possibly more shots from the surface of the water out into an open area or to a marker.

Example 1: My short underwater cave, 30 stations, 1 mile of passage. I want to overlay it on Google Earth to see if there are any other nearby entrances. Go ahead and use the GPS. It will get you close enough for what you want.

Example 2: My 37 mile cave has three entrances, and I want to find out which valley to dig in 13 miles away to try to open a fourth entrance. In this case, you will find that for every foot or so of error you have in the GPS position, the end of the passage 13 miles away might swing 100 feet. (Obviously geometry matters here.) In this case, you need to be able to surface survey not only to a minimum error reference station, (preferably the same one your maps are referenced to) but you also need to be able to surface survey between the entrances to close the loop and mitigate error.

2. It doesn't matter. For my first dive or trip into a cave, I always label my GPS station as 0Datum. This is how it is labeled in walls. Then I count up. I might later have to renumber stations if i survey out, and forget, but hey, that is easy. Some pencil and erasor time on the original survey book sheets, and then change them up in Walls. If I know for a fact that there are multiple nearby entrances that might or might not be connected, then I use something more descriptive, like Turtle0, Scary0, etc shortening the name of the cave. However, I only use this for the zero datum stations, the rest of the numbers are generic. This is because I can use the #prefix statement in walls. (later question.) All of these datum stations will eventually be brought together in a single georeference file in walls, so they can be easily managed.

3. If this happens, and you have used generic labels throughout the separate caves, use the #prefix directive (directly from the Walls manual):

--------------------------
If our project area contained many caves connected, for example, by surface surveys or GPS positions, it would be nice if we could compile the complete data set without having to revise the original station names simply to make them globally unique. Also, if we wanted to insure that some arbitrary collection of SRV files is truly self-contained, we would need some way to say whether or not station S1 in file A and station S1 in file B, for example, are physically the same station. To satisfy such requirements, the SRV format provides the following directive for creating qualifying name prefixes:

#Prefix <name>

or

#Prefix

The first form assigns an implied name prefix to all stations following it in the file whose prefix is not explicitly overridden. The second form reassigns the default empty prefix. For example, PEP:A123, BRINCO:A123, and :A123 are station names with explicit prefixes. These stations are assumed to be physically different, although they all could have originally appeared in survey books as simply A123. A colon is used to separate the prefix from the original name portion. (The override :A123 is equivalent to a non-prefixed A123.) Notice that in the Hackberry Cave illustration (see SRV Files - Overview), all stations are assigned prefix "HACK", except for the surface tie-in, SUR:S1.
------------------------------

Ok, and for some separate points brought up later:

Numbering stations: After dealing with the 37 mile long cave I am currently drafting, I have to agree that the Prefix letter method, exactly as Mike described, has worked quite well. In a manner, it allows you to know where things are in the cave, particularly when they are quite out of hand with maziness. I have seen this used in a number of places in Mexico, and it works well there also. I see Jon's point, and perhaps this might work on a project where the day locations were found might be important (as seems to be with some sorts of projects) However the survey tends to be disjointed, particularly in mazy stuff, and it can be difficult for someone not involved with the project to intuitively find their way around using this method. Keep in mind, that you are limited to 8 characters with walls (not including the #prefix!)

Survey Markers: I also agree with Mike to an extent in this regard. I do not leave station markers except at junctions, important locations (bottle drops) or leads (which will become junctions, or be removed.) There is just no need to leave stations in almost any other circumstance. The only exception would be in a very long passage where your stations might be very difficult to discern due to uniformity of passage depth and shape, and here, you could leave a station. However, as you commented, stations get blown out. This is why it is so important to make stations on permanent locations where possible, and then TAKE A NOTE AT THE STATION when you survey- that way, you can find the station again, regardless of the location of the line. I cannot tell you how important I find this particular piece of information that almost nobody does. It takes only an extra second or two to write <small rock on right> or <floor loop before dome> and with this information and the depth, 90 percent of the time, you can return to the same station again if the line was well placed, as you would likely use it again anyway.

If you have the time and stability to make custom station markers as Mike did in Peacock, then by all means. But I have found that using the line arrows that are rough on the back, along with using a survey pencil that is not too sharp, will allow you to write the survey station on the back (or front) of the marker in the water, and it wont disappear. I have placed survey stations in Jackson Blue that are now 10 years old that I can still read. If your station changes, just write a new one, and take the old one out and scrub it clean for re-use.

Data storage and management:This is one thing that cave divers to badly. Writing your survey data on a slate, copying it down into the computer and then erasing the slate is bad practice. Photocopying the slate, and then erasing it is a step in the right direction. The best answer is keep your survey data until you are done with it, and then, turn it in to the NSS (or publish, as Mike did an excellent job of doing in Peacock.) If you merely copy your data into the computer and do not keep the original data, there is a very real possibility that some time later (when trying to close a loop) you may realize that you made a "blunder" in your data. Of course, you will then go back to your old data and try to figure out what happened. But where was the blunder? In the transposing from the slate to the computer? Or was it on the original data? Or did you write it onto the slate incorrectly? Who knows? The best answer is to use an actual survey book, with removable pages that you can keep for the duration. Some of you have seen the survey books I use and the underwater cave survey paper that I sell at cost. (Much cheaper than trying to make a little order from Rite in the Rain.)
I cannot tell you how much work I have been saved by people who had "finished with a project" and then turned their data into their local caving organization (in this case the QRSS) who was then able to share it so we could continue to add to the data. Don't be a data hog. You will die some day, and someone else will find a new passage in your cave. Don't make them resurvey the whole cave.

Therion: AAAAGGGGHHHH!! It looks cool, and the one (1) guy who really pushes it loves it, but it absolutely has its limitations. Most of which you will probably never hit. (Roundtripping.)

Ariane: Haven't seen it yet. But 25,000 feet is not a large cave in the mexico (or US) realms. I was just sent a map of a dry cave system in QRoo with (gulP) 100 kilometers!! of passage in an area approximately 7km by 5km. It started with 14 separate cave systems. It's all being done in Walls. Also- if it cant export a stick map, you are wasting your time. You need to be able to get your data out. Walls will export to a shape file, and a KML export is being developed, however you can make the conversion easily in many different programs. You need the stick map export to be able to draw the cave around it. I suppose you could draw within the program? But is that as flexible as using inkscape or AI?

Car comparisons: It's not really the same at all. Those cars all do the same thing. These programs do very different things (except maybe compass/walls) but once you get into the details, and start having the hard questions (how do I tie in my dry surveys and my wet surveys into one map? How do I add survey to a map I have already drawn without redrawing the whole thing? How do I get around long station names?) Only Walls so far has held up to the really difficult projects, and in addition, is still being supported by the developer! It does have a down side, and that is the interface, but as with any computer thing, it seems, the more it does, the more you have to learn about it.

Jason

[FW]

Thanks Jason! Now if it would only clear up enough to do some more surveying.

[jlillest]

Example 2: My 37 mile cave has three entrances.....

Three? What happened to the fourth?

Great stuff Jason, though I'd like to add that over several thousand feet of overland survey GPS coords slowly become a better loss of accuracy (ie more accurate) than a Suunto survey. Also, a GPS point gives a single point and has no effect on the "swing" of a cave passage. Only the actual cave survey can cause the rotational errors that you describe. Even better than GPS points, though, is picking points off of DOQQs if you can find 1m res for your project area. This is very possible for Florida and the rest of the US but may take a while for Mex.

-Jon

[rchrds]

Three? What happened to the fourth?

Great stuff Jason, though I'd like to add that over several thousand feet of overland survey GPS coords slowly become a better loss of accuracy (ie more accurate) than a Suunto survey. Also, a GPS point gives a single point and has no effect on the "swing" of a cave passage. Only the actual cave survey can cause the rotational errors that you describe. Even better than GPS points, though, is picking points off of DOQQs if you can find 1m res for your project area. This is very possible for Florida and the rest of the US but may take a while for Mex.

-Jon

Example. So, three :)
The DOQQ suggestion is valid, if the entrance is obvious.
The swing described happens when the figure or error is used in plotting the GPS position. Figure the circle of error for a final position at the end- that is magnified by the circle of error inherent in each individual position. Draw the circle around each reported position, then take the worst case location in that circle and use that as a point to find the far waypoint. (ArcGIS can do this automatically, though I haven't figured out how) and check out how large a swing you create. Of course this requires more than one entrance point to pivot about.

Also- Jon, if we do a 3 day (Friday-Sunday in cave camp) survey trip at BSC, you interested? (non diving related, sorry.)

[jlillest]

Also- Jon, if we do a 3 day (Friday-Sunday in cave camp) survey trip at BSC, you interested?

Sure, when did you have in mind? I might even be able to do longer if the objectives justify it.

-Jon