with a local base.

Has anyone habitually forced the use of PM2(or PM3), units in situations where lock-loss is constant? Using the rover rod, external antenna and postprocessing to the local(less than 500 feet), control.

What kinds of errors . . . or better yet, what kinds of accuracy is usually achieved?
Modified By John Francis on 10/18/2006 at 10:00 AM

That would be a sign of a mental defect. Tried it once, 3 PM2s and a Z-12. Max range was 1,000'. Neither GNSS Solutions nor Astech Solutions satisfied me. My biggest mistake was in having the Z-12 under cover also. Had it been out in the open like the PM2 bases, I might have been satisfied with a best 2 out of 3 position. However the ±3m gave me an idea of what was right. It was rock rows to be added to a map.

I have done as well with a compass and pacing. Shoulda had a disto, I could have triangulated better.

It took a long time to wrangle through the multiflora rose with a 2 meter rod. Were I to do it again, I would use a light weight adjustable pole. For what I got, a PM3 without an external antenna would also have been easier.

Paul in PA

Hi John,

I have done at least 50 test jobs under extreme cover and have some customers who live by the PM3 under really heavy, tall timber.
Here are my rules:

“1.5 hour occupation in morning, 1.5 hour occupation in afternoon, set elevation mask to 25 deg in Process Options, don’t screw with mapping out any ‘bad’ SV’s (unless I need to remove one (and only one) to fix.) Two fixed bases in as clear areas as I can possibly find.”

Then I like to see the same (1 tenth foot) solution morning and afternoon, with either base processed individually or as a network.

I have a test configuration at my home that has pretty bad canopy (and multi-path). I have shot it with RTK and then taped into the trees. So I have a known solution there. I have done lots of testing with the antennas shoved right up against the trunk of the trees. Good solutions abound.

In the real world uses, the traverses would be a really-really big deal, so I don’t have anyway to check.

I have never been able to get a fix using ZMax’s in these heavy cover situations. I have waited the full 1.5 hours. I have run ZMax static (dual freq) side by side with PM3 and am not able to form a conclusion as to if ZMax static generates better results than PM3.

I have found that the morning and afternoon occupations typically match very well. This gives me confidence that I have a reasonable solution. Often I will run a third occupation a couple of days later and generate even more confidence.

In my experience I will always get a fixed solution with 1.5 to 2 hours of occupation. The issue is confidence that the solution is reasonable. The confidence intervals reported by GSolutions under heavy cover are not dependable, they are very optimistic. So I reoccupy and hope for close a close match between sessions. I use fixed locations for the two bases (I know this violates Phil’s rule but I will typically have days of static on the bases and as they are in the open, I can process the bases against a network of 3 CORS stations.)

I have been playing around with a ‘trick’ recently: I run a 6 hour occupation, then in process options I mask out the second 4 hours and process; then mask out only the first 4 hours and process; then process the whole occupation. I am thinking that this is just like two occupations, but I can just leave the receivers (connected to a big battery) and come back at the end of the day.

Most of the time I don’t encounter heavy canopy in Utah. My issues are typically very deep canyons and low SV counts. (Saturday I had 11+ SV’s on mountain tops and 4-5 in gulch bottoms with hard rock vertical faces, ZMax would fix given 40+ minutes and 95% confidence. With 99.9% confidence my ZMax’s will not ever fix in those conditions. I know PM3’s would dependably generate fixed solutions under the same conditions with 30 minute occupations.)

I know that there are completely different issues in the east and south.
I have a couple of PM3 customers who are doing big boundary surveys in heavy timber in Idaho who use approximately the same rule sets. The main issue is confidence in a solution and we develop that with multiple occupations.

To answer the question: "tenth foot repeatability, accuracy unknown."

What have been your experiences?

Mark

I've never seen good results with my PM3's in canopy. But I can get repeatability with my HiPers XT's and glonass using 3+ hour occupations under heavy deciduous canopy on a dry day. You need to know your equipment and the conditions to estimate the occupation times. Also we use 5 to seven receivers at time a to create a strong network with many vectors simulations. I put my Promarks in open areas only. Just two receivers will not cut it here. We often set coordinate pairs in the conditions then traverse between them with total station. Most often checking horizontal within 0.06'. Vertical never works with these conditions.

Lee
Modified By Lee Green on 10/18/2006 at 2:29 PM

for me is about locating grave markers in a cemetery.

WAAS averaging, base & rover, autonomous, or total station for about 2 feet accuracy?

I will not do post processed kinematic in very heavy canopy.
I will not do post processed kinematic in very heavy canopy.
I will not do post processed kinematic in very heavy canopy.
I will not do post processed kinematic in very heavy canopy.
I will not do post processed kinematic in very heavy canopy.

You can try using the mapping mode, post processing, with external antenna. Run a few points with your total station and see how close you can get with the PM3 in mapping mode.

Honestly, John, this is a total station job, as is any job that is described as Very Heavy Foliage.
Modified By Mike Margolis_ on 10/18/2006 at 8:32 PM

I have had good luck with the using a similar method to Mr. Silver with the PM2's. Longer occupation times and more SV’s are always better in hostile conditions. I seem to have better luck with recording at a 2sec rate then 5 or 10 for some reason.

I have set point pairs and static both then traverse between just as a check, theory being different multipath environments at each point. If my traverse distance agrees with the static distance gives me some confidence my coordinates are right. I have also set up for 4 hours and then process the data in different time intervals and record the coordinates. I have chopped up a 4 hours vector into 1/2 hours times, some processed and some did not.

Also had some extention cables made to get the antenna up out of swamp brush. This only works if there is no wind. I have been trying to figure out a way to raise an antenna 60 feet and have it stable and over a point. Have not come up with anything yet.

Charley

I just did a test last week using the PM3 in MM mode with 2 PM2's as base stations. I had a concrete monument in heavy canopy I ran traverse to a week before. But one of my static reference stations didn't get recored. GERRR! So as I was running a static session to locate it I walked through the heave woods to the conc mon with the PM3. I then set the PM3 on top of the mon, and logged 15 minutes in MM mode as a point. I processed the PM2's in static last week to locate the mon. I got time last night to post process in MM Office using the PM2's as reference stations. At 5,200 feet from my maine reference station I was within 0.25' of the traversed location. The only problem was I had to do a lat lon inverse to get the correct azimuth, but the distance was right on time. I need to play around with cleaning my processing azimuths and distances down.

MM mode ROCKS!

M2C's

jjp

Ok! I answered the CONVERGENCE of

0 15 56.00

&

SCALE of

0.99997627

thus giving me 0.12' or difference!!

Yea John I'd do a MM job!

Jules J.

Please do not let me get in the way of any method that works.

There is no substitute for success!

Thanks Phil!

You have been part of the success!