Replies

1. 
   PaulBinCT | Jan 05, 2006 04:26pm | #1

   I thinkkkkk that's in excess of the error signal that's built into civilian GPS units...

2. 
   junkhound | Jan 05, 2006 05:31pm | #2

   re:  relatively inexpensive

   Simply, NO.

   To get to 1 foot non-military you need to locally install a few additional benchmark surveyed local on-the ground beacons that gives a local signal, then you can get to +/- 1 foot when you hae a receiver that includes those local benchmark signals .

   Using maritime local beacon type GPS receivers, you can get to within about 10 feet most of the time.

   $100 handhelds typically locate you to about 25 feet elevation.   E.g. My $150 Magellan Meridian unit will read anywhere between 370 ft elevation and 420 ft elevation in the driveway (actual 396 ft elevation) depending on time of day and number of satellites in view.  You can hold this unit at your knee and raise it to above your head and it will tell you that you moved 3 or so feet, but the accuracy of the real altitude is not correct.

1. 
   Catskinner | Jan 05, 2006 09:27pm | #5

   Thanks, Junk. So much for that idea. <G>

2. 
   JohnSprung | Jan 05, 2006 11:06pm | #8

   What you describe sounds like differential GPS.  The idea is that if you put a second GPS at a nearby fixed location, any position changes it shows over time are simply bogus.  So, you can subtract off those changes from what the first GPS tells you, thereby eliminating the old "Selective Availability" and atmospherics.  (IIRC, SA was turned off in the late 1990's.)

   That'll get you much better relative positions, but doesn't help with getting the absolute position right.  That leads me to wonder, why would anyone really need the absolute lat/long?  Could you get it by averaging GPS over a long period of time?  Could you get it from things like Google Earth, which goes down to hundredths of an arc second?  Is Google Earth really that accurate?  0.01" of arc is only like a foot or so.   

    

   -- J.S.

    

1. 
   CapnMac | Jan 06, 2006 12:08am | #9

   Could you get it from things like Google Earth, which goes down to hundredths of an arc second?  Is Google Earth really that accurate?  0.01" of arc is only like a foot or so.

   Hmm, I want to remember reading somewhere that they were using fractal math to generate the reported values that small (reported values not being the same as scaled boundaries).

   But that's just remembering, too.Occupational hazard of my occupation not being around (sorry Bubba)

3. 
   gordsco | Jan 05, 2006 08:42pm | #3

   Try this link. Scroll down for reviews etc.

   http://www.geocaching.com/about/buying.aspx

   geocashing.com also has discussion forums. They are the GPS masters.

   Gord

1. 
   Catskinner | Jan 05, 2006 09:27pm | #6

   Thanks.

4. 
   dgbldr | Jan 05, 2006 09:19pm | #4

   No GPS, but you can use a precision altimeter.  You calibrate it at the nearest small airport.  Runway elevation is always known within one foot.

   DG/Builder

    

1. 
   Catskinner | Jan 05, 2006 09:29pm | #7

   Good idea. I think I'll look into that.

5. 
   JonE | Jan 06, 2006 12:51am | #10

   In a word, no.  I use a Magellan SporTrak Pro which is good to about ten feet, for informal observations, and a $7000 Ashtech ProMark system for survey-grade GPS.  A crew I was working with today (from another firm) was using a brand-new Leica system with radio-frequency positioning.  Accurate to a fraction of an inch, but ridiculously expensive.  I can get to within 0.01 ft of true horizontal and vertical coordinates (and relative to other points) with my Ashtech system, but it requires some time and effort  - not a simple two-minute observation from the hood of your car.

   You cannot get that kind of accuracy with any handheld, you generally need (2) units working together, a firm reference point, and an external antenna.

    

1. 
   PaulBinCT | Jan 06, 2006 01:21am | #11

   Just curious Jon... I have heard about algorithms for taking into account the error signal that's broadcast with civilian GPS signals.  Do you know if that's how the real high end units work?

   PaulB

1. 
   JonE | Jan 06, 2006 03:58pm | #13

   As far as I know, the "error" that was broadcast was discontinued by the feds a few years ago.   The limitation with handheld GPS is the lack of an external antenna, a good secondary reference point for a cross-check, and the limitations of the software itself.   High-end units use post-processing software, either in the unit itself, or more often in an office computer, to tell you where you've been with great accuracy.  With all the high-end units, you need at least two of them, one of which must be on a known coordinate reference point, to get the accuracy they are reputed to have.  There are also CORS (continuously operating reference stations) around the continent that can act as the second or third unit - generally you need to be within range of one.   If you set up two units without a reference, they will be moderately accurate with reference to the geodetic datum you are in, but still maintain a very high level of accuracy with regard to each other.  This is good for localised survey work where a tie to a state plane or national coordinate system is not necessary. 

   OK, so you say, well, why can't I use two handheld units?  Because they can't "talk" to each other and there is no post-processing software available.  Even the high-end units are not capable of extreme accuracy when used solo. 

1. 
   PaulBinCT | Jan 06, 2006 04:00pm | #14

   Very ineresting Jon, thaks for the update!

   PaulB

2. 
   Catskinner | Jan 06, 2006 09:46am | #12

   Good to know. Thanks.I already feel better about paying a surveyor. <G>