As a trial run, I set up my new PM3 receivers on 3 know bench marks. (solving for elevations only) The bench marks are spaced approx. 1 mile apart in a E-W direction. I collected data at 5 sec intervals for 20 minutes. 10 sattelites in view, PDOP less than 2, 16 mile observation range. In processing the data, I held the published elev of the middle benchmark. The other 2 benchmarks missed by 0.27 & 0.58 ft. Would I get better results by collecting data longer? What is the rule of thumb for occupation time in determining elevation? Thanks.
Roy
Good rule of thumb is double a positional occupation time for a good elevation. For elevations you also get better results with better orbits. Go to http://igscb.jpl.nasa.gov/components/prods_cb.html
and get the ultrarapid or rapid orbit. the rapid orbit is generally available about noon EST the next day. If you are psotprocessing the timing is ususally not sifnificant.
Unzip the file once it is downloaded and put it somewhere that Solutions can find it. The bin folder is one spot, that way once you tell Solutions where it is it will look there for it on the next project. To keep from clluttering up my bin folder I put the orbits in my project folder, then on each project I have to tell Solutions where the precise orbits are. It always looks there for that particular project, so whatever you do, be consistent.
Consider the accuracy of the original leveling before thinking your GPS elevation is poor. Also consider how accurate the geoid is in your area.
If elevation is critical on a project the chance are you will be on the site more than minimum time anyway. Take some of that time to tie down you elevation. Use at least three offset vertical control points, holding the highest order and checking against th others. From point to point within your project you can stay within 0.02' on long static observations.
Have fun.
Paul in PA
From point to point within your project you can stay within 0.02' on long static observations.
Vertical accuracy tends to degrade considerably with baseline length. Even with dual-frequency receivers and multiple-hour observations, getting 1cm repeatability on vectors over 5km can be a challenge.
Were all three monuments put in by the same agency?
The differences(from record) you are talking about are very substantial.
Within a project I can expect 0.02' on at least half of my observations. The projects are small enough that I can also expect 0.02' on proper field work. I may not always carry my verticals to that precision and will adjust field work to my long static GPS observations. Obviously thgis is not a PLSS state and projects are under a mile, most under 0.5 mile. Point to point within a project, degrades as there are more points and farther apart, maybe I should have said within close points in a project.
I do not know what to expect of the geoid in Alaska. Roy may already be in the ballpark.
Paul in PA
Modified By Lawrence Paul Lopresti on 1/6/2006 at 11:45 AM
Roy,
For good vertical positioning it is best to have 4 satellites above 60° horizon. I do not know if you can achieve that in your area. With lower satellites the geometry for vertical positioning falls off in a hurry.
Paul in PA
At what point would the available sats all be at lower latitudes? Alaska's pretty much at 60 degrees latitude.
Here, we get sats to the north. I would think he doesn't.
I did a planning for Anchorage. Setting a mask of 50° I get 5 visible from 22-24:00 UTC today. that is much better than any other 2 hour period. That is during the middle of the day up there, so planning is going to be essential for vertical GPS work. Anchorage is above the latitude of the GPS orbits, but the orbits of satellites just 30° East or West of his position give suitable positions. There are satellites that will cross 2 quadrants in a NS/SN line. Getting the full duration of that passing time should tighten up verticals considerably.
Being that far North he actually gets satellites due North of his position, which is great for horizontal control.
Paul in PA
Modified By Lawrence Paul Lopresti on 1/6/2006 at 12:13 PM
I've done a bit of work in Alaska....up near Fairbanks. I recall before I left people told me it would be more difficult up there (GPS-wise), but it wasn't. If there were any issues, it was due to terrain....they have a few mountains.
Sat
Accurate differential levels were run between the bench marks for a previous project. At the time I collected the data, only 2 satelites were visable above 50°. What is considered a long observation time? I only observed for 20 minutes.
Roy
For the 2 hour period I mentioned the satellites are each in 2 quadrants. 2 hours for vertical control at that time is probably better than the 6 hours any other time of day. If I were there I would look for at least 3 above 50° for 1 hour, for vertical work. If mountains are not a problem your horizontal work can be done in 20-30 minutes. On most projects vertical control points are usually much fewer than horizontal control points.
Paul in PA
Thanks for the help
Roy
If the vectors were "fixed" and not "float" or "partial" you should be getting better than a half a foot vertical on such short lines.
I believe I understood you to say that the observation timer indicated you should have had a fixed solution in that time?
To answer your question, yes, you will see improved results with more time on site. You will also be able to measure farther with more time on site. But there is a limit to where you begin to see diminishing returns. If you know someone else who is an experienced GPS user with GNSS Solutions, perhaps you could ask them to look at your data and see if they can do better...
I've never seen those sorts of errors.
Ever..
Deral Paulk, PLS OK
I would bet you have it Deral.
GNSS Studio defaults to ellipsoid heights.
Although for a mile apart, you wouldn't think the Geoid delta would be that great.
There might be some value in a little light reading.
http://www.ngs.noaa.gov/PUBS_LIB/NGS-58.html
If you ever get a chance to go to one of Dave Zilkoski's seminars it is worth your time.
http://www.ngs.noaa.gov/PC_PROD/WorkShops/
Modified By Phil athome on 1/8/2006 at 1:08 AM