[time-nuts] nuts about position

[Dana Whitlow]

It is true that most handheld GPS receivers have WAAS capability these
days; however the accuracy is more like 3 to 4 meters even with several
minutes of averaging.  I've always been puzzled by why it is so much
worse than good professional equipment can apparently achieve with
similar "features".  Perhaps it has something to do with the simple antennas
that handheld units use.

A day or two ago I had related the story of our survey at the Arecibo
observatory done with professional equipment.  Soon after that was
done, I took my handheld Garmin GPS60CX up on the roof and used it
to make the same measurement with WAAS corrections turned on and
about 10 min of averaging.  I don't remember the exact numbers for
the discrepancies between the two measurements, but they were in the
neighbor hood of 4 meters.  In fairness to the handheld, I should mention
that the position involved did not offer complete sky visibility- it was in
sort of a canyon such that the sky was visible only down to about 30
or 40 deg elevation to the east and west but with excellent low-elevation
visibility to the north and south.

Incidentally, Puerto Rico does have a WAAS ground station fairly close
to the observatory site, just off the San Juan airport roughly 50 miles
away from the observatory.

Dana


On Wed, Apr 25, 2018 at 8:21 PM, jimlux <jimlux at earthlink.net> wrote:

> On 4/25/18 11:18 AM, Brooke Clarke wrote:
>
>> Hi Tom:
>>
>> As part of a FireWise community mapping process I'd like to get GPS
>> coordinates of the fire hydrants (Lat, Lon, Ele). Is there a civilian GPS
>> receiver that makes use of WAAS and/or DGPS corrections?
>>
>
> I think almost all handheld receivers these days use WAAS for improved
> performance.  WAAS should give you 1 meter kind of accuracy, particularly
> if you compare a known location in the area.
>
> DGPS was a thing back in the 90s (I fooled with a Trimble Scout with a pod
> that received corrections over FM broadcast SCA) - I'm not sure it's widely
> used today.
>
> The USCG DGPS transmits corrections on MF beacons, but is being
> decommissioned.  Most of the inland stations have been shutdown.
>
>
> If you're a surveyor, you get corrections from a network (CORS or
> something similar), or you're primarily interested in "relative" position -
> you set up your base station and your RTK rover tells you where it is
> within 1 mm + 1ppm of distance from base.
>
> http://www.xyht.com/ has regular features on the latest GPS survey gear.
>
> You might be able to convince the local survey equipment rental house to
> come out and demo the gear (or give you a good price on a rental)
>
> Or, why not just do the survey optically (!) - none of this new fangled
> GPS stuff.  Rod, level, theodolite.  If Everest could do it in the 19th
> century in India, you can do it too.
>
> If you can find a couple benchmarks to work from, you can get accuracy of
> 1 part in 1000 with a decent 200 foot tape measure and something to sight
> with (a cheap laser level at night works pretty good to keep your line
> straight).  You're doing a series of triangles - SSS completely defines it,
> so no angles need be measured.
>
> With decent survey gear 1 part in 10,000 or so is straightforward.
>
> 1 ppm is hot stuff with conventional optical gear - you're going to be
> making multiple measurements, compensating for refraction, etc. It's like
> GPS at 10cm accuracy - lots of things cause errors of that magnitude.
>
> A nice theodolite (like a Wild T2) is readable to 1 second of arc. That's
> about 5 microradian.  At 100 meters, the horizontal uncertainty would be
> 0.5mm.   Yeah, not quite 1ppm, although you could probably do multiple
> setups and average in on 1ppm.
>
> Of course, you'll then need to go out and get a decent tripod, a rod and
> target, and a rod person to wave the rod, etc.
>
> But another poster did comment on "why not use the telescope" you could
> precision point to a series of stars and calculate using celestial nav
> where you are.  Although, that might be painful to the 1 meter sort of
> accuracy - the "tables" probably don't really account for deviations from
> ellipsoid and so forth.
>
>
> _______________________________________________
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/m
> ailman/listinfo/time-nuts
> and follow the instructions there.
>