Notifications
Clear all

Pushing the limits of GPS accuracy using GPS Post processing

39 Posts
8 Users
25 Likes
954 Views
(@pfirmst)
Posts: 90
Member Member
Topic starter
 

Any GPS experts on the forum?   I've been mulling over a few questions.

In Australia, AUSPOS provides accurate post processed GPS positions, it only processes observations from GPS satellites.

AUSPOS allows processing up to 7 full days of RINEX data.

Australia is one of the worlds fastest moving continents, moving approx. 7cm NE each year, 0.19mm / day or 1.35mm per week.

I have seen people process 24 hour data intervals over long periods and after converting to GDA2020 coordinates, average the result, which makes a lot of sense.

You have to wait a couple of weeks before processing, if you want the most accurate results.

The reality is, at this level of accuracy, the Australian plate is moving quickly, everything is changing, the ground might also be subsiding, uplifting, you get the picture, it's just a snapshot in time.

Now my antenna's don't have ground planes, usually it takes my antenna's about 7 days to produce the same level of confidence result as a geodetic antenna fitted with a ground plane can in 24 hours.   

This does have me thinking about the improvements that modern Geodetic antenna's based on the choke ring design, however it's my understanding that choke ring antenna's may not capture as signals from satellites over the horizon so well.

To quote AUSPOS:

Question 3.1 - How does AUSPOS work?

All computations are undertaken using the Bernese Software System. Bernese is a high precision orbit and geodetic parameter determination software system. For more information see Bernese GNS Software.

The International GPS Service (IGS) product range is used in the computation process. Precise orbit parameters, Earth orientation parameters and coordinate solution IGS products are used.

NOTE: The IGS final orbit product is not available until approximately two weeks after the observation day. The rapid orbit product is available two days after observation. If both the final and rapid orbit products are unavailable, then the IGS ultra-rapid orbit product will be used.

Once a user has submitted a RINEX file, the nearest 15 IGS and APREF stations are used as the reference stations for the processing. That data is retrieved from Geoscience Australia's GNSS Data Archive. A precise solution using a 'double difference' technique is then computed using these stations. The coordinates of the IGS stations are constrained with uncertainties of 1mm for horizontal and 2mm for the vertical.

 

From AUSPOS reports:

 

 
Posted : March 19, 2023 4:36 am
(@pfirmst)
Posts: 90
Member Member
Topic starter
 

I guess Einstein was right when he said everything is relative.   So I guess, if you have some really accurate observations, at intervals over time, you'll at least understand how a patch of dirt is moving relative to the rest of the country and this might work in nicely with an annual level traverse, performed over a number of years.

Even deep seated rods going down to Permian bedrock may be affected by mining and blasting, dewatering, or underlying tuff's or faults.   In areas of underground mining, the surface undulates.

Clients are already pretty happy with AUSPOS results, and I usually only do 24 hour static observations, but it does make me wonder about newer antenna designs, I understand that if they're set up for multiple GNSS frequencies, they might not be as good as an antenna specifically set up for GPS frequencies.   Has anyone got any thoughts or recommendation's around Geodetic antenna's?

Might be time for an Aspirin.

Cheers,

Peter.

 
Posted : March 19, 2023 4:56 am
Jitterboogie
(@jitterboogie)
Posts: 4112
Supporter Supporter
 

if you build an LDP for the area you're working in, save the front end time, it's always where it is.

I was spoiled to have such a system when I was at Las  Cruces, now I'm back in the real world.

Chasing millimeters is a futile thing for most work, unless you're building the world's tallest buildings and or the longest and deeper tunnels, millimeters don't mean much until they're centimeters and so forth and so on...

it really just depends...

 

 
Posted : March 19, 2023 4:25 pm
CV Nevada reacted

Carlson BRx7 GNSS Receiver

(@lukenz)
Posts: 434
Member Member
 

@jitterboogie 

And if you really are after mm it's usually over a short distance and you need quality total station and fixed stave with level.

 
Posted : March 19, 2023 11:50 pm
(@pfirmst)
Posts: 90
Member Member
Topic starter
 

@lukenz @jitterboogie

Installing deep seated survey rods, no other accurate references (a mine in the middle of nowhere), nearest mark has an estimated location accuracy of 1ft, derived from AGM66 a few km away, hasn't been measured for decades, other marks on site have been destroyed, other than a permanent GPS base station permanently installed on site.   We usually do 24 hour static observations, but have done as long as a week to get an accurate height (don't need to in this case).   I noticed the improvement a good geodetic antenna with a ground plane provides, can in 24 hours, provide similar accuracy as a normal antenna would take a week to achieve.   We'll be tying it in using a level and total station with least squares (including GPS vectors), our level reference is the base station antenna, doesn't have a prism under the antenna, so we'll be reliant on bearings, zenith angles and level to tie it in, we're not allowed up there to measure the pole diameter.   Level shot is 130m distance, with 32x zoom 45mm lens with 0.1mm micrometre, with 10mm range, going to use the micrometre to adjust the level plane into alignment with the antenna base plane, then transfer it to a level rod close to the instrument.

All distances measurements with TS will be less than a km, so yes short distances.

We get down to 4mm easting and northings at 95% confidence in 24 hours, but elevation is usually around 10 - 11mm, the base station on site has a base plane, but it's not a choke ring antenna, has an elevation window within 6mm from 20 days of static post processing.

It would be big and bulky, but if it provides a big improvement in height confidence, a choke ring antenna might not be a bad option, for the odd job where it's required.

Fully understand that this isn't a normal requirement.

Cheers,

Peter.

 
Posted : March 20, 2023 1:33 am
Jitterboogie reacted
(@lukenz)
Posts: 434
Member Member
 

@pfirmst 

Wow that sounds like a cool job!

 

When your doing week long static do you have mains supply to the receiver?

 

I know the Trimble Zephyr Geodetic antennas were what was being used for research when I was at Otago school of surveying. Can you not just ask Geosciences team what they use themselves?

 

Also if you are in the middle of nowhere why does the absolute position need to be so precise? Would have though one point to within 0.05m-0.1m would be fine and then hang a local control network of that which is precise within itself.

 
Posted : March 20, 2023 2:28 am
pfirmst reacted

Carlson BRx7 GNSS Receiver

(@pfirmst)
Posts: 90
Member Member
Topic starter
 

@lukenz Yep, they're using a Zephyr geodetic 2.   I had figured the choke ring was an improvement and some of the later designs have improved capability to receive signals from satellites around the horizon, while avoiding ground signal reflections.

To establish a reliable datum, against subsidence from underground mining.  Realistically, the GPS base station on site (every day is recorded), is the only reliable long term datum, so we'll be networking our ground references back to it.

 
Posted : March 20, 2023 4:47 am
jflamm
(@jflamm)
Posts: 266
Member Member
 

Posted by: @lukenz

 

Also if you are in the middle of nowhere why does the absolute position need to be so precise? Would have though one point to within 0.05m-0.1m would be fine and then hang a local control network of that which is precise within itself.

This is where my mind was at.

 

 
Posted : March 20, 2023 4:51 am
(@pfirmst)
Posts: 90
Member Member
Topic starter
 

@lukenz Yes mains supply for anything longer than 24 hours, although it would also be possible to swap out batteries daily, for a week.

 

 
Posted : March 20, 2023 5:12 am

Carlson BRx7 GNSS Receiver

Jitterboogie
(@jitterboogie)
Posts: 4112
Supporter Supporter
 

what you need to provide the data you want is a full up gravity survey.

I got started in gravity, and it's used to detect and define subsidence and position of the oil fields in the Kingdom.

They are essentially are injecting sea water to surround and keep the fields under pressure with all of the massive extraction.

Company I worked for did a year long gravity survey for them to utilize for exactly what you're describing, except not for mining. It's also not published from what I remember, and that's what the client wanted.

Yeah, the Zephyr series of antennas are supposed to be superior to the old choke ring set ups, we used them for everything we reduced as our base antennas.

Sounds like interesting work, but don't reinvent the wheel if you don't have to.  Just add some paint and fringe.  😎 

 
Posted : March 20, 2023 5:14 am
pfirmst reacted
Jitterboogie
(@jitterboogie)
Posts: 4112
Supporter Supporter
 

@lukenz 

and staff trained and certified to actually do that work consistently with the same in the gear quality.

yeah, mm is like making the watch or the nuclear reactor jobs, and some bldg and bridges and all tunnels I suppose.

super specific training for the highest quality work.

thank God for the schlebs and the construction staking Industry we can all survive....

lol

😆

 

 
Posted : March 20, 2023 5:19 am
lukenz reacted
(@pfirmst)
Posts: 90
Member Member
Topic starter
 

@jflamm That's close to the tolerance they use in the pits for mining.   They want it tighter around the plant, it doesn't cost much to let the GPS sit there for 24 hours, and it gives us a good vector back to the GPS base station on site (more important than absolute position), that's within total station measurement tolerances.

Cheers,

Peter.

 
Posted : March 20, 2023 5:19 am
Jitterboogie and jflamm reacted

Carlson BRx7 GNSS Receiver

MI-Other-Left
(@michigan-left)
Posts: 373
Member Member
 

You spent quite a bit of text explaining how gps/gnss positioning works to a bunch of people that are pretty well versed in said technology.

Then you indicate that you are literally chasing millimeters in horizontal and vertical axis.

Then you cite numerous examples of how unstable your area is, including the fact that there will be pit mining operations in the vicinity.

and then you say:

Posted by: @pfirmst

a permanent GPS base station permanently installed on site.

and

Posted by: @pfirmst

we're not allowed up there to measure the pole diameter.

Up where? On top of a pole mounted on a building?

What is the configuration of your permanent gps base station?

If it's not an off-site deep-drilled and braced monument like this:

Then you won't get even close to the type of results you're aspiring to.

 

After reviewing your initial two posts, the only questions you actually asked were:

Posted by: @pfirmst

Any GPS experts on the forum?

and

Posted by: @pfirmst

 Has anyone got any thoughts or recommendation's around Geodetic antenna's?

Citing specifications and using all of the best gear on the planet is a means to an end. Obviously you desire optimum precision/accuracy. We all do. To one degree or another.

What are you really trying to accomplish?

It sounds like what you need is a high precision control network/system, but you never come out and say it.

And if your goal is a high precision control system, you're not really asking the right questions on how to plan and implement a strategy to use all that increased precision/accuracy. (That's probably out of reach, given the site conditions.)

Do you have a surveying background?

 
Posted : March 20, 2023 9:17 am
Jitterboogie and Rover83 reacted
Jim Frame
(@jim-frame)
Posts: 7121
Member Member
 

Posted by: @jitterboogie

Zephyr series of antennas are supposed to be superior to the old choke ring

Fun fact:  a few years ago I bought 5 each Zephyr Geodetic antennas to replace the Compact L1/L2 and Microcentered antennas I had been using with my 4000SSi receivers.  I paid $35.00 a piece plus shipping.  🙂  They were well-used but are still working great.

 
Posted : March 20, 2023 12:10 pm
pfirmst and Jitterboogie reacted
Jitterboogie
(@jitterboogie)
Posts: 4112
Supporter Supporter
 

Posted by: @jim-frame

Posted by: @jitterboogie

Zephyr series of antennas are supposed to be superior to the old choke ring

Fun fact:  a few years ago I bought 5 each Zephyr Geodetic antennas to replace the Compact L1/L2 and Microcentered antennas I had been using with my 4000SSi receivers.  I paid $35.00 a piece plus shipping.  🙂  They were well-used but are still working great.

 

That's amazing!

retail was a nightmare for those.

 

I have 2 zephyr antennas my old employer said no longer worked.

 

they're working just fine.

 

 
Posted : March 20, 2023 1:12 pm
pfirmst reacted

Carlson BRx7 GNSS Receiver

Page 1 / 3