NGS Monuments/Data Sheet Question

how bad is bad?

can you tell how these were run from the data sheets? same series should match better (well they should all match, but apparently they are not)

did the NGS GPS any of them?

how is the new geoid in your neck of the woods? perhaps you can try an older model just to see. there was some article published in the past year or two where a new model was busted by 0.5'.

stability issues? big stability issues? ever look at the velocities of the nearby CORS to see how things are moving there? if there is a good shift going on, you can date the establishment of elevations and calculate a distortion. of course, they only realized the velocity recently, so you are bound to assume things were constant over the years.

someone else can go on about the GPS and elevations, i have to go play on eBay for SWMBO.

Two questions. How much did the marks not check by? Where are these stations located?
Much of southern Mississippi is an active subsidence zone. You might want to check in with Denis Riordan, NGS Mississippi state geodetic advisor -- denis.riordan@noaa.gov

Thanks fr the replies.

I will check with Mr. Riordan tomorrow.

I have an OPUS solution on EH2557. The OPUS elevation is 198.37 feet. The published elevation is 195.17. My crew is very well versed about measure-ups, so I am very confident in their observations.

I processed some of these stations using both Geoid12A, and Geoid 09. The differences were within a few hundreths.

I have held each of the monuments as control, and the elevations are all over the place. Nothing consistent without double checking my notes at the office.

These observations were taken on three different days over about a week or so.

I have never run into anything like this before. I am generally fairly confident in my abilities in regards to post processing, but this one has me baffelled.

Thanks,
Jimmy

This is a total shot in the dark but maybe try starting a brand new GNSS Solutions Project and process it again. I have noticed with various brands of processing softwares sometimes something gets fubared in there and starting over will fix the problem.

Also only hold one NGS point fixed and see if maybe you can find a bad one by comparing minimally constrained adjusted coordinates with published coordinates. There could be one bad one that is throwing least squares into a giant tizzy. I've seen that too. Even one bad antenna height can blow the whole network up. It's amazing how one bad height can fubar a whole network as LS tries to spread the blunder all around. Could be a transposed number. Those things can be hard to see sometimes. I've looked at something 15 times before I finally saw it then said, "Why couldn't I see that before?"

What is a few feet between marks? The 1800's.

If you skip the CORS and run control with approximate N/E and hold the published elevation, how do the other marks check?

I looked at MSB5, a CORS in Bobo, MS. It is showing a subsidence of -0.0063 m/yr (NAVD). That is screaming at 0.02' a year! The data sheet for EH2557 states the elevation was established in 1991, 22 years ago. This would translate to a drop of 0.45' which does not help your 3.2' difference. If geology is the cause, this CORS may not reflect local conditions of your monument.

| NAD_83 (2011) VELOCITY |
| Transformed from IGS08 velocity in Aug 2011. |
| VX = 0.0011 m/yr northward = 0.0009 m/yr |
| VY = 0.0057 m/yr eastward = 0.0010 m/yr |
| VZ = -0.0028 m/yr upward = -0.0063 m/yr

Have you tried running through CORS a second time? Perhaps you punched in the wrong rod height when uploading? (It happens to the best of us).

Like the tree said to the lumberjack, I'm stumped.

Thanks to everyone for the replies.

I am going to look at this again in the morning, when I get back to the office.

I will also send the NGS adviser an email as well.

Paul in PA ran some observations for us. We have observed 5 different NGS benchmarks at varying times. Paul ran the first day's observations and created a VRS, and ran the information through the CORS.

This project has me stumped. We are racing against time at this point. The river is rising, and this project has to get out the door. We are considering returning to the site, and running about 5 or 6 miles of levels to double check. We just need to have the warm and fuzzies before we send this one out the door.

Just A Thought

One of my earliesr GPS projects was with 3 ProMark's putting an elevation benchmark for a contractor on a Navy dock. During the day I occupied several tidal benchmarks and a tie point at a USGS benchmark. I gave my project report to the contractor who several days later calls and says "The Navy says you are wrong." I dig into my data and datasheets and note that all the data on one tidal mark does not add up. It would add up if benchmark values were tranposed. It took me a while but I finally found someone at NOAA would would pull all the benchmark data. He agreed there was in fact a transposition, so I asked "Are you going to change it." "No."

I transposed the values in Solutions and reprocessed and the results were accepted.

There may be errors that may not be corrected because they have been relied on in the past, an interesting concept, BLM, please note.

Similar errors aside levee benchmarks are in the levee. Everytime a levee gets put to the test it becomes 100% saturated leading to the possibility of subsequent severe settling. What should really prick up your ears is finding a levee benchmark higher than it should be.

A 5 mile level loop along a levee is at least less of a task and safer than leveling along highways. Hopefully the level loop regains your confidence in GPS.

Paul in PA

The 1992 date for the NAVD 88 height on the datasheet is the just the date when the adjustment was published, not when the leveling was performed. For that you need to look at the history on the datasheet which in this case shows that NGS did the leveling in 1983. Consequently the estimated height difference would be -0.0063 m * 30 = 0.189 m/0.62 ft. That being said the OPUS solution is 3' higher not lower. I'm pretty sure there hasn't been any significant uplift in this area.

I am probably not going to be a lot of help. One question I have is you said "All of the monuments are Vertical Order Class II monuments". There's first order class one and class two, second order class one and class two, third order one and two. Not sure what "order" you're referring to. However, that probably doesn't matter, it sounds like you are missing outside of any of those precisions.

Also, are you looking at "vertical differences"? Compare the measured vertical difference between two of the monuments you are measuring, and their published vertical differences. Maybe that can help shed some light. Try different combinations of points to check between.

You probably are, but make sure you are looking at the elevation differences and not the ellipsoid differences. (Maybe I am suggesting things you know better than...I'm not meaning to insult your intelligence.) But perhaps you can be overlooking something simple.

Are all of your benchmarks from the same level runs from the original vertical runs? Sometimes, a different line can not match well to another line.

One other problem I have run across, is where we found a monument in a railroad switch-box pedestal or some such concrete pedestal. We could not get an elevation to match to it for the life of us. Also the "walk-up" description seemed kind of goofy. We finally realized that at some point in time the railroad had probably dug up the whole pedestal and relocated it. It wasn't anywhere near the same vicinity of what the original elevation should have been.

Just some thoughts.

good catch, i saw that date and compared NGVD to NAVD (pretty close there) and something was kicking the back of my mind. i should have spent another minute or two to use the correct date of survey. of course, if there is another CORS that produces another subsidence (or uplift) rate, then it could be better. i will not guarantee the CORS i used was close enough to the site to be the same.

Since the OPUS difference is almost exactly 1 m (0.975) and it's higher, I going with someone recorded or entered a bad antenna height as the most likely cause.

I thought that also. He used 2 meter fixed rods for the OPUS observations

"We are considering returning to the site, and running about 5 or 6 miles of levels to double check. We just need to have the warm and fuzzies before we send this one out the door."

That's the ONLY thing that we make me feel warm and fuzzy!

Mystery Solved!

Jimmy and I made a run to the site today to try to figure out the problem before the water got too high. When we went to the first marker, Jimmy looked at the monument box and called out a designation number. I didn't know the boxes were even marked (the marks were covered with mud), and the number he called out was not what I expected to hear.

Here's what happened: I was identifying the marks by the mile markers next to them, which were pointed out in the station descriptions on the datasheets. Well, guess what! The mile markers were all of by 1 mile for at least 20 miles down the levee, maybe more. This means that the marks I thought I was occupying were actually 1 mile away.

We ran into someone from the levee board, who told us that they knew the marks were wrong and they were to be corrected in a few weeks.

I should have figured it out right away from the horizontal coordinates, but I was only concerned with vertical, and I just didn't notice they were different. Well, pardon my ignorance!

Anyway, after we figured out which marks we were using and discovered a processing error which popped up for the second time in about 2 years, everything from our previous observations checked very well, within a few hundredths. It only took three extra trips to the site to figure it out. Sheeeesh!

Now there is a great lesson, Check and check until you get it right! Smart effort to track down the source of the error and doing it before you start pointing fingers!!

Glad that you found the problem with the designations. I have to ask, though. What do you mean when you say Paul created a VRS?

Mystery Solved!

> Here's what happened: I was identifying the marks by the mile markers next to them, which were pointed out in the station descriptions on the datasheets. Well, guess what! The mile markers were all of by 1 mile

This is the reason we do station rubbings and photos for campaign projects. Misidentifying a mark is a sure-fire way to screw up a network!

Mystery Solved!

You could have told him that a thousand times before this happened, and explanations may have been ignored. I have a feeling he has learned this lesson now though, and he will never let this happen again. (I don't mean him specifically, I mean anyone). I am just saying that the bunder is probably the best teacher. Checking against the published horizontal coordinates is probably a good lesson too...

Check Your Work As I Do

On almost all GPS projects I take the time to download an ortho photo and register it to SPC. I then bring in my GPS points to verify they look right. Most times everything is within 0.5 foot horizontal. I guess it comes from having to do some photo control for aerial topograpy in the past. First choices for those GPS points then are not predetermined by existing traverse work or benchmarks are 90° curb corners, ends of parking space's stripes, anything that will stand out in the photo. Sometimes that is all I use the orthophoto for, but usually it migrates to my working drawing for also checking the field work, overlaying deed plots etc.

Paul in PA