GPS Accuracy

[JerryBiedenbender]

I am having some difficulty getting a straight answer about this subject.

One person tells me that (ArcPad/GPS Analyst extension/GPS correct/geoXH6000) will get be 4 inch accuracy for vertical and horizontal.

Someone else tells me that there is no way ArcPad is capable of collecting accurate Vertical accuracy's. They say the only combination That will produce 4 inch accuracy on both vertical and horizontal is (TerraSync/Pathfinder Office/geoXH6000)

Does anyone have experience on this area and can give me some advice?


Thanks,
Jerry

[DaveChamberlin]

Jerry,
        The accuracy of the vertical with any gps unit isnt going to be entirely accurate.  You have to look at the horizontal, and can pretty much believe that the vertical is going to be twice as bad.  Its based off a single geoid that is preloaded onto the unit.  so if you go out and take a point, and read its elevation, then build a large structure, stand atop it, and take a point, the elevation will not change.  Plus too, Trimble is pushing and pushing their new floodlight stuff, when most base stations around the US cannot even post process it.

Dave

[ericbock]

Jerry,

I have done significant testing with all the Trimble MGIS units and can tell you that the GeoXH 6000 IS capable of 4inch accruacy in the horizontal and the vertical, but there are certain things to consider.

1) Will you be collecting data in Real-time or will you be post processing?  Real-time would require VRS or single base corrections.  Post processing requires GPS Analyst or Pathfinder Office software.  You will need to watch how far you are from the base stations....see FAQs link http://trl.trimble.com/docushare/dsweb/Get/Document-528897/GeoExplorer6000Series_CustomerFAQs_0411.p...

2) ArcPad Elevations

[INDENT]-ArcPad displays and stores elevations in Height Above Ellipsoid (HAE) by default. If you want to view or store mean sea level (MSL) then you would have to enter a Geoid Separation value to calculate MSL. Dave mentions this in his post and I agree with him about the Geoid Separation not being accurate because its based on a single position.  Also, you have to go out to a website and enter a location to get this separation value, then enter it in the GPS settings.

If you will be using ArcPad, my recomendation is to collect in HAE and calculate the MSL values in ArcGIS using the Geoid Separation values for each location.  The Geoid may or may not change that much over a small area, but I imagine this changes depending on where you are located.  I have a recommendation under the ArcPad Ideas to add Geoid model capability.  The Geoid model is similar to the Geoid separation, but it covers all of the US.  Go and vote fort this in ArcPad Ideas!  This may be why someone told you that TerraSyc/PFO may be a better option, because it handles MSL better.[/INDENT]

3) Always use an antenna and pole when collecting elevation data, just holding the GPS in your hand defeats the purpose of accuracy.  Make sure to measure the pole  height, even though it may have markings on it...they arent always correct!

Also, one thing to consider is if 4inch vertical accuracy is good enough for the type of data you are collecting.  things like manhole inverts, topo surveys, etc. may need greater accuracy, which would require survey grade GPS or other means of data collection.


The 2008 GeoXH was able to collect decimeter (4inch) accuracy in the horizontal and vertical as well and it doesnt have Floodlight technology. THis technology allows you collect more data in tougher environments due to increased satellite coverage (GLONASS), advanced algorithms, barometer, and anything else I may have left out.

Hope this helps, eric

[SumantMallavaram]

Eric,

Thanks for an excellent and concise post. I wish I had seen your post 2 weeks ago when I was in a similar situation and had to figure out the best way to get elevations out of a Trimble GeoXH 2008. Having never done that before, I went through a lot of trial and error and came to the same conclusion as you did i.e, capture the HAE (GPS.Z property) and then compute the geoid99 separation online for every XY coordinate during post-processing. I then calculate my ground elevation (NAVD88) as GPS.Z - Geoid99 separation - Antenna height. We are using a range pole and the unit is fixed to it at a constant height to make it a bit easier.

Having done all this, I now want to know if GPSAnalyst is capable of outputting a vertical accuracy number similar to the horizontal accuracy it provides as "Average accuracy" and/or "Worst accuracy." I understand we can't get survey grade acuracy with a GeoXH2008 but I would still like to be able to quantify the accuracy in some way. I am currently trying to shoot some points at the same locations as that of control points in my study area. Preliminary comparisons between the surveyed elevation and GeoXH 2008elevations are all over the place with some points within 0.5ft of surveyed elevation and some within 3ft of surveyed elevation.

I might consider shooting multiple points at the same location and compare their elevations to see how precise my unit is. I would appreciate some information about a more definite and tried and tested approach.

Sumant