Thanks for your advice! We're not doing any vertical work for this job. We didn't occupy the GPS points twice at different times. What exactly is a static session? We just shoot a GPS baseline point with our rover for 33 epochs and then move on.
Poor Field Techniques Or Equipment, Start Over.
I want to give you credit for asking these questions. They are good ones. I hope you will ask them of the LS in responsible charge of your work.
> Thanks for your advice! We set our epochs to 33. Sometimes we get an "uh-oh" (bad PDOP?) from our FC-200 just prior to shooting a point. I realize that means we're about to float, but we don't have all day. Poor excuse, right?
By the time you get a warning that you are about to loose fix you have really already lost it. Some of these receivers will extrapolate solutions for a short period but they degrade quickly. Yes, it's a poor excuse.
>Does poor satellite geometry mean the satellites are grouped mostly in one quadrant?
In a manner of speaking, yes. In practice - now that we have a fully deployed constellation of both GPS and GLONASS satellites up - it usually means that the signals from satellites that are positioned in some a part of the sky are being blocked by obstructions - trees, buildings, etc.
>I have no idea what a multipath is.
Mutipath occurs is when the satellite signal is received after being reflected off some surface. The receiver also receives the same signal directly from the satellite. That is a signal sent by the satellite at a given moment is recorded by the receiver more than once and at different times. Obviously this distorts your results. Anything that reflects light will also reflect radio signals. So if you can see it, it is a multipath source. The receiver will just ignore any signals it receives from below a user defined masking angle, typically 10° or 15° above the horizon, so the ground and very low objects aren't a problem. And signals that reflect of distant objects are fairly easy for the software to detect and filter out. So if you are collecting RTK near trees, buildings, etc. they are going to block out some of the satellites altogether and fuzz up the data from the ones that aren't blocked.
>I once heard someone in the office mention post processing, but I've never seen it done.
Post processing - resolving the positions back in the office using software designed for the purpose - isn't going to do a whole lot to improve on poor field practices. And it takes time.
>Maybe I should download OPUS and learn how to do it myself.
OPUS isn't a program you download and run. It's a web service you send your data to for resolution. You will have to occupy each point for a bare minimum of 7.5 minutes and preferably much longer - several hours perhaps- to use OPUS. It may be appropriate to have your RTK base collect raw receiver data while you rove through the day for submission to OPUS. You could then shift your whole data set onto the national datum, which is a useful thing to do, although it's full usefulness may not become apparent until sometime in the future.
I hope that was a joke!
No joke! I'm trying to learn the correct way to do this so when I have my own crew, it'll get done right.
Go To College, You Have Too Much To Learn
And this is not the place for you now.
If you do not know if you have good satellite geometry you are a hazard to the profession and the public.
Paul in PA
Go To College, You Have Too Much To Learn
He won't likely get it in college, Paul. I would be surprised in universities spend the requisite time making sure graduates know all they need to know about GNSS positioning. By the way, I put in 3 semesters at my local junior college, none of which were surveying, and I don't have a degree. I use GPS daily, effectively.
No need for all of that, Paul.
I need to see if our collector is recording the vectors. We don't use a base station, so can I get the job done reliably without it? We use L-Net ( http://www.lengemann.us/l-netftpdatasite.aspx). In another post you mentioned only the RTK base position gets held when using STAR*NET. How does that apply to me?
Poor Field Techniques Or Equipment, Start Over.
Why 33 epochs? What's the magic there?
One thing to keep in mind. GPS will, in general, save you a ton of time. So it's okay to reinvest some of that time in longer occupations and redundancy. Also, as you noted, sometimes, it's just time to pull out the gun. Even then, using a pair of points to run conventional traverse from saves on the traverse time that would usually be required. So, spend a few minutes on the point. Shoot those pair points twice and average.
Also, GNSS positioning is 3D - automatically. It just comes that way. So when you do your checks with the total station, check the vertical difference too. In this way, you can check 2 dimensions of a 3D vector - distance and elevation difference. The only one left is direction. But if 2/3 checks well, you can reasonably figure the third checks well too.
Thanks for your advice! Can you recommend any useful GPS books for field crews?
yes. Make sure you are comparing apples to apples. You're in Florida, so probably not as much of an issue, but still worth knowing. Data collectors take GNSS data that is sent to the data collector from the RTK head which is in latitude and longitude format. The data collector then projects this on a Grid (such as State Plane). Then the distances between point A and B reported by the collector are along this Grid, which isn't necessarily coincident with the surface of the Earth. The total station reports a slope distance corrected for vertical angle (which is determined from the Geoid). So you have two different distances, one is Grid, the other is parallel to the Geoid, aka horizontal. Applying the right scale factors in the right way makes is the proper way to compare the two distances.
Note that network RTK is pretty good, but nowhere near as good as having a local base (I'm sure this is debatable). Considering your correction source, I can't stress enough: time on site, redundancy.
Poor Field Techniques Or Equipment, Start Over.
>The receiver will just ignore any signals it receives from below a user defined masking angle, typically 10° or 15° above the horizon, so the ground and very low objects aren't a problem.
Not exactly true... There is no mechanical filter that rejects signals below the mask angle. The mask angle is a software setting that filters out signals from satellites based on the computed elevation of the satellite from the observer. So a signal from a satellite at an elevation of 30° could bounce off the ground to the receiver and the receiver would still receive that reflected signal even with a 20° mask. A satellite at 5° though could bounce a signal off an overhead tree limb and the signal would be disregarded.
The ground is the number one source of multipath, but receiver hardware causes the signal to be delayed from ground reflections, hopefully allowing the direct signal to arrive first and allowing the receiver to determine the true signal. Even choke ring antennas don't completely eliminate ground multipath, it simply causes the signal to go up and down each baffle, delaying its arrival to the antenna. This is, at least, what I've been told. It's all kind of "above my pay-grade" so to speak.
Poor Field Techniques Or Equipment, Start Over.
No magic to 33 epochs. I think my party chief likes that number because it's 3 more than we're required to shoot.
Go To College, You Have Too Much To Learn
This is unfortunately very true. I took a university course on GPS surveying a decade ago and ended up teaching half of it.
It's good to ask questions Field Dog. The scariest people out there surveying couldn't be bothered to understand what they're doing.
Just because I'm paranoid, doesn't mean they aren't out to get me.
This is a real good place to start....
> In another post you mentioned only the RTK base position gets held when using STAR*NET. How does that apply to me?
Not at all if you aren't adjusting your measurements in StarNet or some other LS package. Also, using StarNet doesn't mean you have to hold the base position, and only the base position, that is just the way I prefer to do it.
But, a) you have to hold something, somewhere, and b) the coordinates your RTK is giving you are the product of measurements from some base, virtual or actual. So I recommend that the base position be held and the (multiple) measurements from it be adjusted. After all, if you have multiple measurement between the same 2 points what else is there to do? It wouldn't make much sense to hold one of the rover positions and adjust the base, would it?
In your case, if you don't have access to an LS adjustment package, or need results quick in the field, you could simply average the coordinates of any number of RTK shots on your control points. That's not as good as a rigorous adjustment but better than just cherry picking a result you like.
Norman, you consistently give such good advice that I wonder whether it's time to ask Wendell to update your user name to Norman Oregon or at least Norman Route66.
Hello
Don't use a range pole. Set up the GPS on a tribrach on a tripod, make sure you are levelled up exactly on the point. The bubble level on a range pole is inaccurate, even when it is dead centre the pole is not vertical over the point. If the pole is old and has been banged around the back of the vehicle it may be far out of adjustment. Collect 3 minutes of data every second. Measure every point twice with at least 1 hour between the setups. Finally setup on a known high order control point as close to the survey as possible, compare you results to the published co-ordinates.
John
Thanks! This is excellent!
Thanks again for your help!
Thanks for your advice! We're using a range pole and a bipod.