I have had some very pleasing results on recent kinematic surveys where I have lost lock and reintialised on-the-fly after 5 mins of stationary recording.

We carried out a stop and go pipeline survey over 130 kms - run down the centre of the adjacent road and booking offset dists to the pipe features - shots approx every 400 metres. Approx 2 weeks later, we ran kinematic up and back in 20 km sections using 2 control stations approx 20 km apart - travelling at 90k/hour.

With light tree canopy, we regularly would lose lock and regain by simply stopping and recording for 5mins (4 sec rec interval)and carrying on.

Comparing the verticals(for average measts each 400m) to the 2 kinematic runs, even after lock lost, their was typically only 0.5 metres difference between the 3 over the entire 130 km extent.

These units definately reintialise OTF.

Frank, how are your kidneys? I sure wouldn't want to follow a pipeline at 90k/hr. Now 9k/hr. I could handle on a quad.

When were you on the road and when were you on the pipeline, or is this road specific for the pipeline and of a better quality than most pipeline access roads?

Why 4 second intervals? You lose a lot of ability to cross check information without having increments of 5 or 10 seconds. At 4 seconds you get only 1 comparable epoch on the majority of CORS. i.e. if it starts at 0:01:00 and the next reading is 0:01:04.

Just wondering?

Paul in PA

Yes, of course they initialize OTF. The only question is....how long does that initialization take? Because of the constant changes in the satellite constellation, we recommend 20 minutes (ie a New Point initialization). Can you get by with less?

Maybe yes, maybe no. But you'll never know until you're back in the office.
Modified By Mike Margolis_ on 6/18/2003 at 7:35 AM

Clarification of Terms...

OTF (On the Fly) Initialization is term associated with Dual-Frequency GPS receivers, not L1 only receivers.

If the L1/L2 receiver loses lock, your not required to re-Init the receiver, simply keep moving along the kinematic trajectory. Dual Frequency GPS observables provide the Post-Processing S/w with enough data to recover momentary loss of SV lock, Dual freq receivers data collection is more robust, more observables.

L1 receivers like PM2 require Kinematic Initialization, This can be accomplished by using the TN-Ashtech Init-Bar (20 Cm Baseline) for the default time observation of 300 Seconds, or 5 minutes.

Alternatively, a known control point can be used if the Init-Bar is not available. The initial kinematic Initialization should collect 300 seconds / 5-minutes.

During PM2 kinematic data collection if
the kinematic alarm goes off (less than 5 SV's), the field observer can perform additional kinematic re-initializations
with as little as 2-Epochs, more data is better.

Call your first Initialization "INI1",
then INI2, INI3, etc...

My technical opinion, a 20-minute kinematic Initialization is not required.
Modified By Bob LeMoine on 6/18/2003 at 8:11 PM

Sorry about the delay in reply.

Paul, I am in Western Australia.While we have some rugged pipeline routes, this main is a just a water supply main adjacent a public road - sealed bitumen.
The survey was mainly to measure the hydraulic profile and fix the line features ie valves, scours, changes in pipe size etc. The topography is mainly long rugular undulating type hill country - wheat, sheep fields.

The accuracy required being sub metre for H & V.

The control units were also recording at 4 sec intervals - so the epochs matched.
The net result was meast about every 50m (at 90/km/hr).

As I reported, I went through the majority of the survey after processing and simply comparing verts of my pick up survey shots (which averaged about ea 400m) to the nearest kinematic shot( ea approx 50 m)it was typically within 0.5m - never more than a metre.

When I checked your profile it said USA.

I was curious as to the general method you were describing.

Assuming you where in the USA, I based my epoch question on this area. We have a good, and getter better, network of CORS stations. A good number are now on 1 second epochs and more are changed to 1 second as time goes on. The rest are at 5, 10, 15 and 30 seconds. To get a good selection of CORS in order to determine the best solution to the location equation, a minimum 5 second epoch is most appropriate. Does the government or private industry provide CORS equivalent information in Australia?

The 4 seconds seemed an odd choice for epoch rate.

Paul in PA

Paul,
Yes I assume by CORS (what does the letters stand for?)you refer to stations with continuous SV meast being carried out for which the data is available.

We do have broadcast data for real time differential but as far as I know it is only used for metre accuracy - for real time positioning.

I assume for "CORS" stations you can obtain the precise data for post processing - we dont have such a network Aust wide though I expect such stations exist in the capital cities - I work in the country.

In West Oz we use a Govt network of SSM's - "State Survey Marks" for which a precise position is available - in most areas I work they are usually within 5 to 10 kms.

FT