Say I want to establish a baseline by running static at each end - and I know the horizontal uncertainty of the coordinates (@95%) is going to be +/-10mm for each point - And I wish the azimuth uncertainty of the baseline to be +/-20"....how long would the baseline need to be?
Pointing me toward some pdf on the subject would be appreciated, or a formula. I know I did this in college but it's gone through the sieve of my brain.
Is it the propagation of errors equation?
if you use 2 static receivers, simultaneously, in clear areas, (no obstructions),?ÿ your accuracy for establishing directional control, will be much better, than if you perform 2 OPUS observations, with only 1 receiver.
Relative positioning can be real good, with 2 units.
I used to run a cluster of 6 L1 post processed gps units.
I got a good "feel" for it. Alas! I've fallen into the big dark hole of rtk!?ÿ
🙂 Kent MC could run the numbers on the back of his hand!
N
Yes I would rather process the baseline than do 2 opus's, but since moving to Windows 10 I've had no luck getting GNSS Solutions to work - I think it's dead...Someone wants me to calibrate the heading sensors on an 8m boat - trying to think how to do it. I even had a look for Polaris last night. It's a real neck breaker at lat 48 degrees!
Maybe you could:
Put 2 stakes in the ground, intervisable. GPS on these. 1/4 MI apart.
Then, use total station to transfer it to the boat.
Or to the dock. Some mechanism where he can re calibrate anytime, as desired.
Way back when I took a few flying lessons, we'd calibrate the gyro to the compass, and confirm it with the runway direction.
He may not be able to benefit from greater than 1/2 degree accuracy.
Dunno.
N
It's a commercial boat on Lake Superior - I have asked what the accuracy spec is they require but then I'm asking myself "how do I evaluate the accuracy of my azimuth anyway?" How do I turn the coordinate errors I will know into an azimuth error??ÿ
Just draw a thin triangle.
Sideways error = Distance * tangent(azimuth error)?ÿ?ÿ
where both errors are at the same confidence level standard error=1 sigma, 90%, 95%, or whatever).
If you do independent observations good to 10 mm the sideways error estimate will be 14 mm (square root sum of squares of individual errors).?ÿ?ÿ For 20" then we get 0.014 meter / tan(20") = 145 meters.
As noted above, the relative position error for simultaneous observations over moderate distances will be much better than the absolute errors.
@cf-67 Question: How accurate is the heading sensor on this 8 metre boat?
JOHN NOLTON
Good question
I believe 0.015 degrees or about 1'
I even had a look for Polaris last night. It's a real neck breaker at lat 48 degrees!
Ouch. I find it hard enough at 42 latitude.
Just quietly, our bathymetric survey kit has an external GPS+GLONASS antenna with a heading sensor in it. Here it is on the far right of the photo (taken with my 0.8MP phone camera), and it's the cat's meow. The antenna is NMEA 2000 linked to the sounder/chartplotter which is used to navigate when doing the sounding runs. The antenna heading sensor is calibrated via the settings in the chartplotter, and involves doing circles until it says she's ok.
Looks like a nice office!
This calibration is quite involved - I would be shooting the fore and aft sensors with the boat on land to give them a starting azimuth to dial in - then the boat goes in the water, tied to the dock - I shoot them again - compare with the heading sensor - then the boat is turned around and repeat...I'm not scared of anchor bolts, but this job has me a bit worried 😛 ?ÿ
It was, and it came with lunch (on all of the 7 days):
The hype for that particular antenna (Lowrance Point-1) says:
- Integrated electronic compass provides 10Hz heading for chart stabilization and radar overlay, plus accurate course over ground even at low speeds.
So I suspect that is what the running circles calibration is about.
@cf-67 Thanks for the link above on the sensor. Besides doing something like Nate the Surveyor said in his post above about
laying out control I would think that NGS has a lot of control around Lake Superior. Give us Lat/Long of about where the boat will be on land and we can search and see what they have. Come off their control and check your GPS then do a pre-analysis in some software
(star*net is used by most surveyors on this board) and go from there.
JOHN NOLTON
@cf-67 Bring the boat down to a reasonable latitude.
48-21-47 N
89-13-45 W
DMS
In Canada. There is a CBN Pillar and azimuth marks on the other side of the city, but around 10km away.
I'm thinking I may get a demo of Carlson's baseline processor - do that, and maybe do a polar for a check and for the fun of it.
@cf-67 I did a search of NGS data base for Lat of 48 to 49 and Long. 89 to 90
6 points came up. If I had a topo map I could see where I really need to search. That will be up to you.
pid SZ0062 designation is 008740082; SZ0060 BLAKE; SZ0057 MACGREGOR ;SZ0061 MCKAY; SZ0064 WARE
and SZ0063 WHITEFISH
This is where a Triangulation Diagram or the Geodetic control Diagram (which was a more detailed map) of the area
would be handy. I understand you can plot NGS points on a topo map on the computer but I have never done that.
JOHN NOLTON
Thanks for looking into this - I can definitely plot those coords into google earth and see where they are.
Some of those are dated 1964!
For me. they all come up in an NGS search as "Outside NGS publication area" with no data. How did you see them?
As of about 2000 when Geocaching took their snapshot of the NGS data base, SZ0063 was assigned to a point in Minnesota.?ÿ The others don't come up in Geocaching, so weren't active at that time.