EDM Calibration Baselines

[+GEO*Trailblazer 1]

Well I finally got out to the Springfield EDM Calibration baseline.

AC7212 EDM BL 0M

 

MO-DNR EDM'S

 

These specifications and guidelines are intended for control surveys using GPS technology. They were developed to provide the information and guidance necessary to achieve horizontal and 3-D second-order relative positional accuracy for lines of less than a few tens of kilometres in length.

 

I was thrilled but not surprised at my findings of the 2 GPS calibrations.

RINO 120 AND 110 WITH WAAS ENABLED.

 

I started with the Controlled data at the 0-METER benchmark.

I then took a 20-30 minute set on each of the 4 marks and marked each reading.

Then measured back to each from the next.

 

If a longer set period would have been taken the average of the mean would have been much more accurate.

Like the LC signature discs.

That is If you were to do a 3 day 5 hour session each day and take the mean of the whole.

For our purposes we need not be that precise but you can see that the GPS is highly accurate in our intended purpose.

 

AC7212 BM O METER

37 15.30600 93 10.20500

GPS

37 15.30691 93 10.20467

+.00091 - .00033

 

GPS 2

37 15.33938 93 10.11191 ACTUAL 150 METERS

149.720 METERS -.028, 28 MM OR 1.1 INCH.

GPS 3

37 15.39444 93 09.95729 ACTUAL 400 METERS

400.059 METERS +.059,59 MM OR 2.32283 INCHES.

 

GPS 4

37 15.58071 93 09.41471 ACTUAL 1275 METERS

1273.042 METERS - 1.958 MM OR 77 INCHES.

 

ALL ELEVATIONS WERE WITHIN LESS THAN 5 FEET.

LETS SEE WHAT YOU CAN FIND ON YOURS.

 

Maybe this will help to satisfy your minds on how accurate your handheld really is.

I find that it is within way less than the standard disclaimer of 3-5 meters.

[tosborn]

I'm suprised that your results are that good with consumer grade receivers. Was the satellite constellation particularly good when you were taking these averages?

Edited August 29, 2006 by tosborn

[+GEO*Trailblazer 1]

This particular EDM has a clear view of the birds in all directions.

Yes the Ephemeris must have been favorable.

 

I get the daily reports on the birds as well.

 

But I did not do an advance search.

 

I wanted to show how on a avereage day with and average GPS you can get good results.

I have always known the accuracy is well beyond all that is stated.

It was the main reason for Selective Availability (S.A.)way back when.

They found the system to be more accurate than ever planned.

 

I also have some secrets to the Initial start-up as well.

I initialized mine at a NGS Triangulation Station and also use that for a control point within my GPS.

 

Not that it makes any difference but I think it does.

 

I will almost bet you if you do a test at a calibration point you will find it to be more accurate than you realize.

 

I am a measurment Tech too.

I use to calibrate mercury meters.

So I may be a little more of a perfectionist than most.

[+GEO*Trailblazer 1]

Got another EDM Calibration done.

 

I am much more surprised at this one.

 

All marks were well within 5mm.

 

I did take more time and set up a little longer than the first one.

This one is not listed as of yet so I logged it on the closeset recoverable benchmark.

 

HE0191 K 28

[MarkDuster]

Got another EDM Calibration done.

 

I am much more surprised at this one.

 

All marks were well within 5mm.

 

I did take more time and set up a little longer than the first one.

This one is not listed as of yet so I logged it on the closeset recoverable benchmark.

 

HE0191 K 28

 

So your GPS reads out to 1/100,000 of a minute of arc?

Or 6/10,000 of a second? Which is 1.9 cm in lat and 1.5 cm in lon so tell me how you got mm accuracy again. Are you just lucky or did I miss something about significant figures?

[ArtMan]

Like tosborn, I am surprised at these results.

 

The other day I downloaded SA Watch, a program that graphically plots readings sent from a GPSr every two seconds.

 

Even over several hours, the mean location was still a couple of meters off. And there were plenty of short periods when the location reported was significantly off base.

 

Admittedly, I had only limited satellite coverage (5-6 birds, typically) in my preliminary test, but I am quite amazed that a user can get sub-centimeter accuracy with an inexpensive handheld unit, even under the conditions described.

 

-ArtMan-

[+PFF]

I posted some test results in the GPS forum. I can read only to Seconds with one decimal point, but the 60cx gave me the same reading as the published coordinates of a benchmark (likewise rounded to Seconds with one decimal place).

 

Accuracy of Garmin 60cx.

 

This is my first experience with a GPS receiver which produces an Averaged reading. Very impressive!

 

-Paul-

[+Klemmer]

I really don't want to start a long discussion on this. It has been done before on the GC.com GPS forum, and it got a little ugly. Lets try to keep this simple (and civil). For background, if you are interested, here is a good read on Accuracy vs Precision. Taking long-time averages using a civilian GPS is interesting (I've done it also), and you can achieve a decent accuracy with enough good reception, software and time. But averaging a signal with (at best) 3 meters of accuracy can never achieve precision. The result is not valid. You cannot use it for anything meaningful, such as surveying or safe high quality navigation. If you could just do long term averages and achieve valid (accurate and precise) results, there would be no need for BILLIONS of dollars of higher grade GPS orbital and ground equipment (two / three frequencies, classified / military P-code receivers, survey grade post processing, etc) that surveyors and other use daily. It's fun, but please don't read too much into it.

[+Harry Dolphin]

Uh oh. Someone hit my buzzwords: Precision versus accuracy. With a scientific training, this was drummed into me. Precison means coming up with the same answer. Accuracy means coming up with the right answer. How's that for simplicity?

I work in exporting. When we build a box, we measure to the nearest inch. So, if I build you a box that is 57x27x65", the cube is 57.891. Everyone insists that they need three decimal points! But the box could be 56.5x26.5x64.5. Cube = 55.887. Or is might be 57.5x27.5x65.5. Cube = 59.937. Wonderfully precise numbers! But meaningless. I've beaten my head against the wall explaining that 58 is a good an answer as any. Nope. We must use three decimal points. I give up.

I don't have the math in front of me, but... NGS tells me that KV1228 is at N 40 53 00 W 074 33 24, plus or minus 6 seconds. Geocaching tells me that it is at N 40° 53.000 W 074° 33.383. Delightful precision. Meaningless accuracy. Thus my discussion in another thread concerning KV4159.

KV4159 ___________________________________________________________________

KV4159* NAD 83(1996)- 40 45 55.48385(N) 074 09 26.93611(W) ADJUSTED

KV4159* NAVD 88 - 1. (meters) 3. (feet) SCALED

KV4159 ___________________________________________________________________

KV4159 LAPLACE CORR- 5.92 (seconds) DEFLEC99

KV4159 GEOID HEIGHT- -32.24 (meters) GEOID03

KV4159

KV4159 HORZ ORDER - THIRD

KV4159

KV4159.The horizontal coordinates were established by classical geodetic methods

KV4159.and adjusted by the National Geodetic Survey in September 1999..

KV4159

KV4159.The orthometric height was scaled from a topographic map.

KV4159

KV4159.The Laplace correction was computed from DEFLEC99 derived deflections.

KV4159

KV4159.The geoid height was determined by GEOID03.

 

Wonderfully precise numbers, but do they actually mean anything?

 

KV4159 STATION DESCRIPTION

KV4159

KV4159'DESCRIBED BY COAST AND GEODETIC SURVEY 1913 (SF)

KV4159'STATION IS ON THE E SIDE OF PASSAIC RIVER, ON A FILLED IN DOCK

KV4159'BELONGING TO THE NAIM LINOLEUM COMPANY WHOSE WORKS ARE 5/16

KV4159'MILE BELOW.

KV4159'

KV4159'THE STATION IS 6.05 FEET FROM THE NW CORNER OF THE DOCK AND IS

KV4159'MARKED BY A C. AND G.S. STANDARD BRASS DISK CEMENTED IN THE TOP

KV4159'OF A 4-INCH TILE PIPE AND SUNK TO THE SURFACE LEVEL.

KV4159'

KV4159'THE SUBSURFACE MARK IS A BURIED BOTTLE SET IN CONCRETE.

KV4159'

KV4159'TWO REFERENCE MARKS, EACH A COPPER NAIL IN CONCRETE CONTAINED

KV4159'IN A 4-INCH TILE PIPE AND SUNK TO THE SURFACE LEVEL, ARE PLACED

KV4159'RESPECTIVELY 43.4 FEET S 51 DEG E AND 43.9 FEET S 49 DEG W FROM

KV4159'THE STATION.

 

I'm told that I don't understand surveying. I'm sure that that is true. We have a dock on the Passaic River, belonging to the Naim Linolem Company, in 1913. Station is 6.05 feet from the NW corner, in a clay pipe sunk to ground level.

 

KV4159

KV4159 STATION RECOVERY (1933)

KV4159

KV4159'RECOVERY NOTE BY COAST AND GEODETIC SURVEY 1933 (RWW)

KV4159'STATION REPORTED LOST.

 

Dock is gone. Clay tile pipes are gone. Naim Linoleum Company is gone. It's now a park in a residential area. A hundred, or so, of these clay tile pipes set along the river banks in 1913 were logged as "STATION REPORTED LOST" in 1933. But we can now locate where they used to be to within five decimal points of a second? Is there any actual accuracy to 40 45 55.48385(N) 074 09 26.93611(W)? Or is this precision for the sake of precision?

Yes, I know. I don't understand surveying, and should just keep my snout shut. Or is someone willing to explain this to me?

[DaveD]

There is accuracy. Note "KV4159 HORZ ORDER - THIRD." Third-order is an Federally adopted standard that defines the location of this station as one part in 10,000 (1:10,000) relative to the local network of other horizontal control points. This means you could have a potential positional uncertainty of 1 unit in 10,000 units such as 1 meter in 10,000 meters. NGS uses 5 decimal places in latitude and longitude to resolve the relationship between individual stations at not worse than 1 mm. It does not imply positional accuracy - that's defined by the order of accuracy A, B, 1st, 2nd, or 3rd. You can find info relating to these orders of accuracy at the Federal Geodetic Control Subcommittee --

 

"Standards and Specification for Geodetic Control Networks"

and

"Geometric Geodetic Accruacy Standards and Specification for Using Relative Positioning GPS Techniques"

 

at http://www.ngs.noaa.gov/FGCS/tech_pub/

 

You will also find "Geospatial Positioning Accuracy Standards." Following the completion of NGSs national readjustment of the North American Datum of 1983, schedule for Febrary, 2007, these porportional accuracy statements will replaced with positional accuracy (e.g. 1 cm, 5 cm etc.)

[+GEO*Trailblazer 1]

Got another EDM Calibration done.

 

I am much more surprised at this one.

 

All marks were well within 5mm.

 

I did take more time and set up a little longer than the first one.

This one is not listed as of yet so I logged it on the closeset recoverable benchmark.

 

HE0191 K 28

 

So your GPS reads out to 1/100,000 of a minute of arc?

Or 6/10,000 of a second? Which is 1.9 cm in lat and 1.5 cm in lon so tell me how you got mm accuracy again. Are you just lucky or did I miss something about significant figures?

 

I did not mention that I downloaded the results into the GPSTrackmaker which then converts it to 7-10 or so decimal places.

 

I do wait to get all the birds I can locked on and reading D that is 3-D.

I take my time.

 

Yes you are correct in the precision vs. accuracy statement.

 

And I can not answer the last querry.

 

Looks like a good subjest to talk to DaveD about the next time I get to meet with him.

But only the Maker knows when that will be again.

 

I did get a bunch of other questions answered and we spoke of this very topic.

The degradation is set at 10 meters for civilian handhelds and he was surprised that I could get that good of results as well.

 

Lucky happenstance as he would say.

I said I like them lucky happenstances.

 

Great to meet with you all again.

A truly remarkeable 2-1/2 year journey.

 

OH yeah I'm back.

[+Black Dog Trackers]

Following the completion of NGSs national readjustment of the North American Datum of 1983, schedule for Febrary, 2007, these porportional accuracy statements will replaced with positional accuracy (e.g. 1 cm, 5 cm etc.)

I'm looking forward to seeing this. This will relate much better to our situation of using GPS receivers to find locations. Even though the numbers will be more accurate than our instruments can indicate, we'll be able to relate to them better than the relative accuracy specifications.

[Bill93]

I think I understand the relative accuracy concept in general, but there must be limits to its applicability.

 

If a station of a given order is 1:10,000 relative to a local network with a typical spacing between stations of 3 km, then that calculates out to 0.3 meter. And if a pair of stations are only 1 km apart their relative positions are accurate to 0.1 m. Seems reasonable.

 

But I have trouble believing that if there is another station 3 meters away (rare but I think we've seen examples) the relative distance between those two is accurate to 0.3 mm.

 

So can we say something like Third order is +/- 1:10,000 +/- 1 mm?

[DaveD]

When distances between stations are small then all sorts of issues arise. If you read the documentation I noted previously you will find that it defines minimum spacing for various classifications of surveys. In general when distances between stations were short was when we would break out the calibrated invar steel tape. With careful procedures it was not unusual to get .1 to .3 mm in just a few meters.

 

"So can we say something like Third order is +/- 1:10,000 +/- 1 mm?" - NO, Third-Order is 1:10,000 period.