Last night at our chapter meeting one of our more intelligent young Engineer members (took the LS this year) asked me about doing topo and transferring Elev.s for Elevation Certificates with the PM3s.

I advised him that if he was really serious, the PM3s are not the ideal choice for that. I'd much rather have dual freek gear. I could do topo but it takes 15 seconds minimum to get a shot and you never know what you have until you process.

RTK is much better for production work.

On the other hand, what I use them for is breaking down sections, mainly horizontal and they are ideal for that especially considering the low acquisition cost.

I told him to be very careful doing Elevations Certificates with GPS. The old-fashioned level run may take more time but it may just be the best choice.
Modified By David Karoly on 5/5/2006 at 12:37 PM

Mr. Karoly,

Not to be contrary, however, the results I have had with elevations from the PM2 units indicate that there are some very appropriate times to use them for transfering elevations (with checks, you can see the results are reliable).

Elevation Discussion

Also, I have had some good results with topo maps using the PM2 units. It did take a while to understand the procedures and conditions required. They could make for a very economical way to topo. Consider that with four units and three people, you could actually be getting three shots every 15 seconds. I personally prefer 30 second observations with the PM2 units.

Conventional and PM2 Comparison


Photo Intensive Post of Comparisons

I certainly have gotten very good agreement between my PM3 surveys and Total Station checks, I just get nervous about it.

I've got a topo to do up near Red Bluff. I was considering running one base and two receivers. The last time I used it for topo I got a lot of fixed but red vectors which was fine for topoing fence but a little rough for contour mapping.
Modified By David Karoly on 5/5/2006 at 8:12 PM

If you are getting red vectors with Stop and Go, there are several things to consider.

The first is that, based on my own personal experiment, a fixed solution with GPS will get you less than 0.15' horizontal and less than 0.20' vertically nearly 100% of the time, epoch by epoch, regardless of the duration of the point occupation. In other words, a rolling kinematic session (single epoch per point) will be almost as accurate as if I had done a stop and go survey of each point. You can verify this yourself by setting up your rover unit on a point and log a kinematic file. Then note the difference of each point from the mean. As you will see, the thing to look for in GPS is fixed vs. float solution. If you're fixed, you're golden.

But this brings up point two. If you don't have enough data solutions does a poor job of estimating the errors of the points in a stop and go or kinematic survey. It appears that the break over point is thirty second occupations. With thirty seconds, the statistics become more realistic. Less than thirty and the software becomes very pessimistic of the results. But again, we can ignore the statistics if we concentrate of fixed/float and have enough redundancy built in to look for obvious boogers.

Lastly, red vectors with fixed solutions can be easily changed to green vectors by simply allowing for a more relaxed error "Desired Project Accuracy" setting. This doesn't make your data any better but you can allow bigger error tolerances and see what allows the points to pass. If you have float solutions, work on the data to get it fixed. Personally, I would not trust ANY float solutions nor any points within a minute or so before and after float, unless you reinitialized immediately after the float.

Best thing is to learn through some experimenting just what you are getting. I highly recommend the kinematic test on a static point. You will see how the software computes the errors and be able to compare that to what the errors actually were.

Shawn

I was using 5 second epochs. Typically my static sessions will be 5 second epochs or I will do 5 minute S&G on control with 5 second epochs.

With 15 second shots it might help to use 2 second epochs.

I recommend a 2 second epoch interval and a 20 second stop-and-go occupation. that gets you 10 shots to average.

I agree with Mike! Set up your S&G work for 2 second epochs and collect for at least 20 seconds (I still will use 30 seconds occassionally). You need enough data for the s/w to compute good positional averages. Collecting only 3 epochs is bordering on the edge.