Using the Locus receivers, on very short lines, short in GPS terms, I'm talking about lines of 1 to 2 miles, how long should you observe? Especially if the satellites are down to only 6 ?

I know that time spent on a point is somewhat relative, the once you are there theory, but I wonder about sitting for 35 to 40 minutes waiting for the first blink on the occupation timer when the lines are only one mile or so long.

On a related note, now that the leaves are off the trees, how about occupation times in the woods? Same distances as in the previous question...

Sitting here in Limbo...
Tom Bryant PLS ST Louis MO

the truth be known, i usually get the first blink at 20 minutes or so with my Locus...at 35-40 minutes, it usually shows that i'm good for 10Km/6 miles or so...

chances are, if it's taking that long to get the first blink, there's a reason...

with L1/L2 receivers, i'll cheat a little on a short baseline, say eight minutes for less than 5km/3miles...with the Locus, i'm a bit more conservative and wait until the the occupation timer indicates that i've been on a point long enough for my calculated distance...

For L1 only, I've always gone by....

20 min for 5km
30 min for 10km
45 min for 15km
60 min for 20km

Why not observe that 3km baseline until your Locus blinks, then take it into Solutions and start trimming time until you see where you lose your fixed solution.

I'd still stay there for the 20 minutes minimum even with a one mile baseline.

To me, if it's serious business (and I take all observations serious), it's worth the wait to stay at least long enough for the receiver to tell me sufficient data has been collected for a 10k line if I'm only looking for 5k. It seems much better to have too much than not enough and have to make a return trip. Been there, done that.

jd

Tom, J.D.'s got the idea. You can shorten the time from factory recommendations, which are admittedly conservative. The question is "What will it cost to remeasure if it doesn't work?"

Tom,

One thing that you may consider is the sv geometry during those seemingly endless occuptaions. It is my understanding (you guys correct me if I'm wrong) is that Ashtech software is capable of identifying sv's that are "clumped" together that could cause a weakening of the overall sv geometry. I know that I have occupied points for over an hour while seeing 6 sv's being logged just to get a 10k solution. The same session took 45 minutes to hit 5k. It's not just a matter of how many sv's, but the geometric quality.

J.D.

What i have found many times is that when it takes 30-40-50 minutes to get the 5k solution, and only 5-10 more minutes to get the 10k, just sometimes not always.

I always wait for at least the appropriate number of blinks for the estimated/calculated distance. I hate going back. Not only inefficeint and costly but twice as boring !!

Jimbo

in my above post about sv geometry I meant to say that the onboard software in the Locus could analyze the sv geometry through the broadcast ephemeris for determining quality of geometry. didn't mean to sound as if it were something special that the post processing software could do this "remarkable feat".

Tom

I agree.. I always wait for the light and most of my work is within 2 miles.. I hate going back

In the trees,in winter, is another matter.. It works but don't go by the lights. I like to have at least one of my three rec's in the wide open .. then with a full SV complement .. I'll let it run at least 1.5 hours, more if you've only got 5-6 SV's. good luck..

Steve G

Using yellow 4600's here. I have been doing alot of occupations in the woods lately and been really impressed with the baseline solutions I'm getting. I usually wait a minimum of 45 minutes and that has worked pretty well. I haven't had a baseline not fix on me yet. I will usually go in and trim the data on the lower elevation sv's or raise the elev. mask to 20 or 25 deg. Even if it blinks at 20 minutes, I will stay there 45. If it doesn't blink until 30 or 40 minutes I may stay there an hour, depending on the situation (tree density, ground topography, space weather, etc.)

What length lines? What receivers? Accuracy vs. occupation time is a complex equation with the variables:
Length of line measured
DOP (satellite geometry)
Observables (L1, L1/L2) recorded
# of SV's observed
Obstructions (station specific)
Multipath environment

lines less than a mile
3 4600LS's
PDOP <6
L1 only
between 5 & 8 sv's
Obstructions being trees with no leaves, density varies.
I would say fairly high mutipath conditions on one or two of the receivers, I try to keep one of them in an open area.

There really are two types of multipath..

Multipath (stationary blockage, such as building)...
The multipath will tend to balance out over a period of time. Since the SV's are on the move, then the multipath will also move (increase and decrease) during the occupation. The easiest way to work with these types of obstructions in to run the mission plan and occupy the point when you have the greatest amount of SV's in the air and the lowest PDOP for the day. For advanced users, then I prefer to look at the actual track of the SV's during the session and can usually predict which would be greatest affected by the occupation.. Even if the SV data looks clean, with enough birds then you can mask a probably source out...and of course extend the occupation time from the norm...Some software will allow you to put in the obstruction (height-bearing-width) and it will plot the outages and best times for you.

Multipath- Kinematic (such as trucks and leaves)
These are much more problematic. Having a point very close to a road or other possible cause of blockages generally leads more to cycle slips than multipath, but it can creep into the data.
Same for canopy (trees), and these are probably the hardest to deal with since with the various possibilities of geometry of SV's coming through the leaves, then the multipath is constantly changing, and again cycle slips can be a frustrating problem..There's nothing guaranteed about a baseline that 'fixed' when in trees. It can still be suspect and the ONLY way to tell for sure is to have redundant measurements (here is one place where the non-trivial will come into play)...
For example, you set up two units in the open and the third in canopy...After you process, then both lines to the canopy point may (and I say may) show a good closure, while actually both lines are wrong and your position is in error. You'll never be able to catch it and can only rely on the processing statistics (which by the way, can sometimes be very good, but very wrong)..Again, much longer occupation times will tend to produce accurate results and also using the times where the PDOP is the lowest and the most birds are in view..

Tree multipath is also intensified by moisture on the leaves. Attempt these when possible during the driest (early afternoon) time of the day for best results.

Leaves being off the trees is better, but the branches will also cause slips in the data, so more is better allowing you to work with what you got when post-processing...

TM

And of course the worst possible multipath that I ever encountered was in East Texas working along the Sabine...Worked much too late (dark thirty) and wandered off the cut line (which is very easy to do)..Spent six hours attempting to find the truck, or a road, or a house, ect....Talk about multipaths...

Okay...I'm not quite through yet...While we aren't talking RTK, it applies as well...When you watch the RMS and stats, these sometimes are in error as they look for consistent data, which can be a wrong fix..I think that is one of the main reasons that sometimes a 'fixed' point, just jumps out and is really an outlier...
Redundant-Redundant-Redundant
Probably learned this best from my grandfather while helping him build cabinets...
Measure twice..Cut once..Good advice for surveyors it also seems to me...

Sorry for the verbosity...

Trimble

How bout.. "We do it right, cause we do it twice" my carpenter brothers favorite saying..

Steve G

Position Dilution of Precision (PDOP) is generally one of the better indicators on how a session will end up...No amount of time will make up for bad DOP's...
To expound on JD's thoughts...

You can have eight SV's (with all of them in the same quadrant of the sky and have a very poor PDOP)and a poor position with hours of data..You can also have 6 pretty well scattered, one in each quadrant and one nearly directly above and have excellent DOP's and some redundancy...

Here's how I picture PDOP...

Stand in the middle of a square room...

Attached 6 ropes to one wall...You can swing the ropes in a wide arc when facing the wall...
Now...
Attach a rope to each corner of the room and one directly overhead...Can't budge it except straight up (and this is the reason that the vertical component can never be as good as the horizontal...No ties below to 'lock' the vertical component.


While on the subject, there's something else to look at if you don't need the most accurate vertical position. Instead of PDOP (which is a combination factor) look at the HDOP (Horizontal Dilution of Precision)..Many times you can get a good horizontal with a poor vertical..Why might you do this you say?
If you are in a bad place and need to get the best available horizontal, then your low HDOP period may be longer than the PDOP period(and most likely it will)..This might give you a little more occupation time at a rough spot in the woods.

TM

One of my favorite cartoon's shows Dagwood with a truck load of lumber in the back of his truck. When asked what he was going to build...he replied "A bird house"...Sorta sums up my carpenter skills....
TM (not a woodworker)