Jump to content

Thoughts on tools


Recommended Posts

I started to write something up on this about a year ago, but didn't post it. One thing as a surveyor, you learn is that knowing where you are vertically can be of assistance in finding a BM. I am talking about the vertical control BM's and not so much tri-stations or horizontal control points.


For example, if you know the height of the horizontal plane that the benchmark is supposed to be in, this provides a valuable additional dimension that can be useful in minimizing a search area. Well sometimes anyway. Examples I see are BM's along roads with steep banks or hilly roads, BM's on bridge abutments or other structures, etc.


The trouble is, or course, that you have to get closer to survey skills and work to establish that plane reasonably accurately. Sometimes just knowing within a foot or two can be useful, but even that level of accuracy is not trivial to obtain.


More advanced or dedicated amoung you might have already tried some of this, or acquired something like a rod and level. Those tools could be used in a number of ways to get a close elevation into a BM location. However since BM's are frequently spaced at some interval of distance, it might mean running a level line 1/4 to 1/2 mile or more. In some cases you might consider it though, just another extension of dedication to the search that some might enjoy trying. A homemade rod and hand level might work well enough in some cases.


However the idea I was playing with was barometric leveling. This used to be used for reconnaisance surveying for roads, etc. The equipment was pretty expensive still and consisted of a precise aneroid barometer/altimeter and a recording altimeter. The latter would be set as a fixed point and the former taken for a ride, intermediate points where elevations were desired would be read and recorded along with the time. The results of the roving altimeter would then be compared to the stationary one for the same time and a differential elevation or corrected elevation could be determined for the point.


That type of equipment is still available, and a lot of it comes up surplus also on Ebay. The company that made such equipment was named American Paulin.


What caught my thought process, though was if more modern and less expensive digital altimeters could be used in the same mode. I had acquired a hiking class digital altimeter for about $150 at REI and started testing it as a stand alone device. I have never had a GPS with a built in altimeter, and that would be another extensive topic of discussion as to viability and testing. In this case, it appears to me that it probably only has a usable accuracy of 1-2 meters, even though it can read out to even feet. Changes in location often show jumps of 2-3 feet which indicates to me a possible conversion from integer meters to feet, but it is hard to tell how the thing really works. I have used it to go from a known elevation point next to my driveway to a BM a mile or so away and be within a few feet.


What you start to see is that barometric pressure is in a constant state of flux so that even a few minutes of time the altitude set to a BM elevation can change significantly. By the time you travel off to some search area and get back you may have 10-30 feet difference. Now you will have to use some record keeping. There are two approaches I can postulate. One would be time based rate of change, so that you prorate the changes of pressure over the loop run based on time. That would be only as accurate as the assumption that the change in pressure was more or less linear change with time. The alternate would be to have two such devices and do the relative positions somehow.


The next step was to look for something akin to the recording altimeter for the fixed station. My search led me to an inexpenive module someone has been building for the model airplane and rocketry folks that can record altitimeter readings at user set intervals and records them in internal memory and can then be dumped to a PC file. The particular module I found was called ZLog. I acquired two of them and the device needed to interface to it. For use in this application it is a bit of a kluge, because you have to rig up a small battery pack for each module.


I have not tried this in BM applications, nor evaluated how well it works at this point, so I am just throwing out ideas and wondering if anyone else has been down this road at all.


So the way it would work, is that you would start up each module with an approximate or known elevation. Leave one, and take the other on a drive that encorporates maybe one known BM as a check and one point in the vicinity of your search location noting the time since it was started, or maybe several points can be occupied, then return to base. Download the data and do some differential computations.


The interesting thing about the ZLog is that it is very sensitve. If you set it on your desktop and initiate it, it will change to a foot lower after being lowered half a foot off the desk. Since it only reads out to the foot, that is about as good as you can get. I can't recall if the digital data is more accurate or not.


The next thing is that you notice even more how rapidly pressure changes and results in different altitude reading. Differential comparisons seem like the only way it could be accurate for correct elevation. Another factor that seems clear to me is that how far geographically you try to extend a level run would deteriorate the results the farther you go since weather fronts, etc. are not generally constant over a wide area. If you have a stable high or low, you might be able to go farther distances.


Just some thoughts for today.


- jerry wahl

Link to comment

Follow up to my first post.


GPS units such as Garmin's with built in digital altimeters could be good or bad. I have heard that there is a way they can be set so that the GPS reading is used to update the barometric altitude. If you knew how this worked it could be more accurate or more unpredictable.


Raw GPS elevation is just pretty sloppy, +-10-20 feet is about as good as I think you can reproducably get, but that is just my intuition based on my own experiments. I hear reports of very unhappy users of altitude enabled GPS units. But since I don't have one at present I cannot say. There might be a method to smooth (average over time) the two device readings and get a good result. Whether this is done on not I have no idea.


Knowing what elevation a 'candidate' mark is at can be useful. I think you probably see it here almost every day. If you know the metal plug on the bridge abutment is 2.5 feet lower than the BM you are searching for 1) you can start to discount it. 2) you can search at points higher and maybe find it.


Datum would ultimately also be an issue and if you do any type of leveling you need to be sure you are on the same Datum as the marks are published in.


I just think it is an interesting possibility, as yet unproven, that inexpensive technology may be available today to replace expensive analog devices costing $1000's of dollars in the barometric levelling area. This type of application would be moot for surveying applications since RTK or static GPS would be able to give accuracies to 1-2 cm, but for BM search applications it might be viable.


I am skeptical of the pressure/altitude conversion in the ZLog, and therefor would want to do testing if there were large variations of altitude between the base and target areas. For example 500-5000 feet for example.


Another concern is temperature compensation. Should the base be in the same temperature environment (outdoors) as the roving unit.


- jlw

Link to comment

The first and foremost thing that comes to my mind is we all have to be on the same plane so to speak.

On the level...

A "Defined level datum".


What it was then might not be what is now.

But that has also proven questionable.

I have one I use from the 1800's that has a published data that my GPS matches almost everytime.


I have found much bigger errors on restamps of elevation than just a few meters (or feet).

I have found in many cases that the GPS elevation was closer than the 10-20 feet.

There are also other programs I run on the laptop that average the reading to finer numbers as well.


Now there are a few of us who can look at terrain or topography and tell the lines of near elevation,I use that tecnique often and it pays of well finding them.


As an interest in doing things like this I now own a good transit and have applied and been accepted to be a part of the MSPS (Missouri Society of Professional Surveyors) so I can properly learn how to use the older equiptment and maybe head on for my Survey licence.


As a tester we had to take barometric pressure and temperature at every location so it has relevance.

Link to comment

jwahl -


I figured I'd throw in my 2¢ on this interesting and useful topic. As we have seen, for finding some USGS marks that have little or no remaining landmarks from which to measure, elevation is about all there is to use.


It seems to me that using 2 barometric altimeters would be a good plan, but I think that, for best results, you will need 2 people - one to do the roaming and the other to provide new readings at 'base' as needed. So, the plan would include 2 altimeters and 2 cell phones.


I would recommend calibrating any technique by using two bench marks of good order that are sufficiently far apart to model real use. Of course a day of 'changeable weather' would require aborting the test - no use collecting bad data.


I certainly agree with you that a regular GPS receiver is not useful for measuring altitude. To remove observational bias, simply put your GPSr next to a bench mark of known altitude and watch the GPSr's readings for 20 or 30 minutes. I did that with a local bench mark and it was like riding an elevator in a 3 story building with 3 sub-basements! Up and down, up and down, for a half hour. I almost got motion sickness. :huh:


As you know, the idea of averaging data with a wide range will get nowhere. The higher the standard deviation of the data, the lower the confidence value of any average you might get.

Edited by Black Dog Trackers
Link to comment

Operationally, my approach would be to use the altimeters to transfer elevation from an official benchmark to my own temporary benchmark point in the vicinity of the mark I'm searching for. The TBM could be a square that I chiseled, or a wooden stake driven solidly, a manhole cover, or whatever seemed handy.


This measurement could be done several times for averaging, if there was variability, and this is the advantage over roaming. Then I'd use conventional optical leveling to work in the neighborhood.


For short range optical leveling, I think an amateur could take a very cheap rifle scope, attach the best hardware store spirit level tube you could afford, and mount it on a camera tripod. Learn the 2-peg test to get it into adjustment. Most of the small levels I've looked at have a repeatability of no better than 6 minutes of arc or 1 in 500. So this would get you an error of less than a foot within a reasonable sighting distance of the TBM. Maybe better if you took the average of several re-levelings of the instrument.


The other way to make your own optical instrument would be to take a good quality carpenter's 3 foot level (maybe 1:1000) and attach a sighting mechanism.


Regarding averages of things with a large standard deviation: if the measurement is unbiased (roughly meaning errors equally likely to be above or below the true value), then averaging long enough will get you good results. The catch is that "long enough" is an extremely long time if the standard deviation starts out large. To reduce the standard deviation to 1% of the raw data stddev, you have to average 20,000 time constants. So if the instrument takes 5 seconds to significantly catch up to a true change that would be 28 hours of data.

Link to comment

Thanks for the replies. Yes I think you could or should essentially establish you own set of temporary BM's near each search area. They could just be a level surface that you document in your own notes, or something marked. The differential barometric levelling could be run a few times to check or possibly refine values for the TBM's.


Since the zlog modules (about the size of a postage stamp), do record data, one could be left alone at a base location if it is safe, like in your back yard. The procedure in a run could include one or more existing BM's in the area, or on the same line as the ones you are searching for, and a TBM in your back yard could then be derived and then used for future work.


- jlw

Link to comment

I guess as a pilot, this has gotten my interest. No doubt calibration on a vertcon BM is a good way to go, but could the "altimeter setting" we use in aviation altimeters be used? Those settings are updated often (don't remember the details of how often), broadcast on ATIS frequencies, given by control towers & approach controls. A simple VHF receiver will get you them pretty much anywhere near any civilization. And, are they possibly (?) available on the internet? If set properly, the result is altitude above mean Sea Level (MSL). I'm thinking they will be pretty accurate. How does MSL relate to Ellipsoid height? Close enough for our purposes? Dunno....


I browsed altimeters on Ebay, but nothing grabbed me, yet.... I sure would like one of the aircraft types, but the prices are up there.... Also probably 28VDC or 115VAC, 400Hz power....

Link to comment

>how does MSL relate to Ellipsoid height


Klemmer, I think we are after height above the geoid (orthometric height). That's the basic idea of MSL and what the main number on the data sheet gives.


Ellipsoidal height is what is measured by GPS and must be converted using local gravity. The conversion or offset value (Geoid Height relative to the ellipsoid) is given on data sheets and may be 10's of meters. The GEOID99 or later tools give this conversion, with an expected accuracy of maybe less than a foot? I don't remember exactly.

Link to comment

The real operational method using is differential elevation differences, and so depends entirely on what controlling elevation you decide to use to 'calibrate' or set up on.


I do not have access to my equipment right now or else I could upload a few pictures and examples. As it is, it may be a few months before I get a chance.


- jlw

Link to comment

I think many of us tend to ignore THE MOST IMPORTANT piece of data on the sheet for a bench mark: namely THE ELEVATION!


So it's good to see this discussion. And although it's a fact that we have little ability to come up with absolute elevation in the field (our consumer grade tools are just too inaccurate), elevation differences from nearby stations are another story.


Case 1: about a week ago I recovered a tidal benchmark near a building in the Bronx New York (the Riverdale Yacht Club) . Although the piece of concrete it was mounted on looked slightly out of level (see my log: KU0932), it was certainly within an inch of the true elvation. Not good for surveyors, but good for me. Why? There is a mark on the other side of the same building that has been not found several times. The fact is (from the datasheets) that it is just about 1 inch lower than my recovery. If I get back there (and get permission to do a little prowling) I can certainly find the spot on the building and hopefully find the mark. Heck, getting to within an inch is like heaven for us amateurs. If I hadn't checked the elevation, I'd be all day there checking the corner of the building.


Case 2: there is a mark (a chiseled cross if I recall correctly) on the Queensboro Bridge in Manhattan that is on the side of a stone bridge pier, just where a loading dock has been built. The problem is I couldn't be sure if the mark might be just above the loading dock, and hopefully findable, or down behind the dock and lost forever. So I had an idea to do some "poor-man's leveling": from a known elevation at a bench mark on the bridge on the other side of First Avenue, I used features of the bridge itself, such as water tables and similar lines, which I assumed must be horizontal, and translated the elevation across First Avenue and down the block to my missing mark. Where the elevation of the bridge feature I was following had to be abandoned for one reason or another, I measured with a tape to another feature, higher or lower as needed, and continued till I got to my mark. When I finally got there, I found my missing mark was about 5 feet below and behind the loading dock. Case closed: the mark is out of reach. Read my log (KU1417). You might be amused. Here's one photo of my poor man's level line:


Note we are almost at the loading dock and we are still (at 57') about 9 feet above the missing mark.



Edited by Papa-Bear-NYC
Link to comment


Yeah, thanks. Geod vs Ellipsoid.... I should have checked my terms. I know better. Brain gear slippage. What I didn't know was that aviation defined MSL altitude is essentially geod elevation. Different industries, different terminology. So for our purposes, MSL altitude = Geod elevation. Good to know. Thanks.

Edited by Klemmer & TeddyBearMama
Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

  • Create New...