RE: {Disarmed} Re: {Disarmed} RE: {Disarmed} Re: [okgis] Accuracy Problem


Chronological Thread 
  • From: Tom Eckert CST <tom AT permiangeomatics.com>
  • To: <okgis AT gis.ou.edu>
  • Subject: RE: {Disarmed} Re: {Disarmed} RE: {Disarmed} Re: [okgis] Accuracy Problem
  • Date: Tue, 31 Mar 2015 16:16:13 -0500

Ditto your post from a dirt surveyor.

Thank you for your comments.

From: Gerry Creager - NOAA Affiliate
Sent: ‎3/‎31/‎2015 3:19 PM
To: okgis AT gis.ou.edu
Subject: {Disarmed} Re: {Disarmed} RE: {Disarmed} Re: [okgis] Accuracy Problem

Realizing I'm new to the group and don't really work in the field any
longer, I do have a fair bit of experience with location augmentation
systems (i.e., Differential Correction) and with registration, and datum
transformations. What I don't have (I'm an Open-Source guy and a theorist)
is a lot of ESRI experience.

In watching this thread, a couple of things come to mind. One is that there
are errors in Google Earth data that aren't necessarily apparent to the
casual user. This was something Google stated openly at the Open Geospatial
Consortium meetings when they first joined, and made the offer to
completely open KML/KMZ. Their data sources are varied and they have
serious problems with orthorectification, edge matching and scale that they
do their best to (automatically) handle but which are still present. For,
especially, large scale features and measurements, Google Earth is great,
but it can expose errors if you try to use too much precision.

It sounded like the original poster had not considered exactly what datum
was in use when gathering the data. Assuming WGS84, there's an annual
adjustment, and if the updates aren't made, error can creep in. Couple
these with different projections, between Google (cylindrical [simple]) and
another projection that might be, perhaps, state-plane and a bit
"non-standard" and your error budget grows.

Using aerial imagery, one needs to consider the possibility that it was not
registered and rectified... or if it was, that it was not done so to the
precision and accuracy for your application. Doing a 4-point registration
(7-point is considerably better) can improve your representation of points
on the ground.

Since 2 MAY 2000, when Selective Availability was turned off, we've seen
uncorrected accuracies on the order of 6 meters or better virtually all the
time. Based on work I did some years ago, I would expect differential
correction to improve that to somewhere between 60-300 cm (0.6-3.0 m)
depending on other error parameters, especialy urban canyons. Long-baseline
GPS surveying can readily achieve 2-3 cm over a 4-hour occupation time
without augmentaton (e.g., RTK) with sub-centimeter ellipsoidal accuracies
possible with extremely long occupations and good statistical review.
Vertical measurements, using carrier-phase measurements (as with
long-baseline surveying) can achieve 3-4 cm accuracies and ellipsoidal
accuracies on the same order with 4-6 hour occupations. Vertical doesn't
get much better until your occupation times reach, well, months.

As a rule of thumb, I expect differential correction to improve autonomous,
or even averaged, position determinations by 90% (6.0m -> 60 cm) and
multi-frequency (L1/L2/L5 combination) long-baseline/long-occupation
geodetic accuracies to be on the order of 2-4 cm. Real time kinematic
surveys can achieve 3-5 cm accuracies although often, anything below 5 cm
should be considered survey-quality. Of course, post-processing and
statistical verification need to be done to assure QC.

On Tue, Mar 31, 2015 at 9:15 AM, Brenda Fennel <bfennel AT paristexas.gov>
wrote:

>  I also wanted to add that I worked with someone who was collecting data
> for underground utilities with a Locator unit in GCS-WGS 1984 and then
> adjusting the points to the Google Earth Pro imagery… then when they would
> go back to the data they would find that the image and the points they had
> so carefully “adjusted” to the image no longer lined up…my SWAG theory is
> that as Google adds imagery updates the existing images may shift somewhat
> due to the web Mercator projection and other ground control (or lack
> thereof) issues.  Perhaps the solution may be one of the  Cartography tools
> to make the maps look right and still retain the accuracy of the GPS for
> future field crews.
>
>
>
> *Brenda*
>
> Brenda Fennel, GISP  |  GIS Coordinator  |  The City of Paris
> Texas-Engineering, Planning & Development
>
> bfennel AT paristexas.gov  |  903.784.9212 x1212  |  P.O. Box 9037  Paris,
> TX 75461-9037
>
>
>
> *From:* Brenda Fennel
> *Sent:* Tuesday, March 31, 2015 8:45 AM
> *To:* okgis AT gis.ou.edu
> *Subject:* {Disarmed} RE: {Disarmed} Re: [okgis] Accuracy Problem
>
>
>
> I recently had a similar issue and after working with WDS we discovered
> that the difference, +/- 2.8 feet, was due to the differential correction
> parameters and the base station choice. My solution was: In the
> Differential Correction Wizard>Change the REFERENCE POSITION setting to
> “Use reference position from base files”, this setting retrieves the
> “Correction tables” from the base station rather than from Pathfinder
> Office and results in a more accurate correction.
>
> NOTE that the older settings used a base station 64 miles from my location
> but which had the correction tables, since then a closer base, 1 mile away,
> was updated to provide the correction files.  I reprocessed the same set of
> raw data from the receiver and now have a difference of 6 inches….good luck!
>
>
>
> *Brenda*
>
> Brenda Fennel, GISP  |  GIS Coordinator  |  The City of Paris
> Texas-Engineering, Planning & Development
>
> bfennel AT paristexas.gov  |  903.784.9212 x1212  |  P.O. Box 9037  Paris,
> TX 75461-9037
>
>
>
> *From:* Bob Pendergrass [mailto:bobpende AT gmail.com <bobpende AT gmail.com>]
> *Sent:* Monday, March 30, 2015 6:45 PM
> *To:* okgis AT gis.ou.edu
> *Subject:* {Disarmed} Re: [okgis] Accuracy Problem
>
>
>
> To all the above I would add that you are comparing the GOOGLE map and
> your GIS map at very different scales. Your JPG clearly shows that you have
> zoomed in way too far so that the image is blurred and pixilated. I would
> lay odds that if you zoomed the GOOGLE map to the level as you did for the
> GIS map, the points won't line up either.
>
>
>   *Bob Pendergrass, GISP*
> email: *MailScanner has detected a possible fraud attempt from
> "rwpendergass AT tallgrassgis.com" claiming to be* *MailScanner has detected
> a possible fraud attempt from "rwpendergass AT tallgrassgis.com" claiming to
> be* *MailScanner has detected a possible fraud attempt from
> "rwpendergass AT tallgrassgis.com" claiming to be* *MailScanner has detected
> a possible fraud attempt from "rwpendergass AT tallgrassgis.com" claiming to
> be* *MailScanner has detected a possible fraud attempt from
> "rwpendergass AT tallgrassgis.com" claiming to be* *MailScanner has detected
> a possible fraud attempt from "rwpendergass AT tallgrassgis.com" claiming to
> be* RWPendergrass AT tallgrassgis.com <http://rwpendergass AT tallgrassgis.com/>
>
>       robert.pendergrass AT tulsacc.edu  <robert.pendergrass AT tulsacc.edu>
> cell phone: (918) 706-1523
>
>
> *"Many eyes go through the meadow, but few see the flowers in it." *
> *Emerson*
>
>
>
>
>
> On Mon, Mar 30, 2015 at 4:38 PM, Scopel, Tracy <Tracy.Scopel AT owrb.ok.gov>
> wrote:
>
> If the datum of your points is WGS84, and you’re overlaying them on a
> NAD83 aerial, you need to apply a data frame transformation for the points
> to display accurately. Have you done that? If not, see image below:
>
>
>
>
>
>
>
>
>
>
>
>
>
>
> *Tracy Scopel *GIS Manager
>
> [image: Oklahoma Water Resources Board logo]
>
> [image: http://optima:8180/graphics/TWP_Oklahoma_Portrait_2014_AW.jpg]
>
>
> *Oklahoma Water Resources Board *
>
> *Phone: 405.530.8883 <405.530.8883> | Fax: 405.530.8900 <405.530.8900>
> 3800 N Classen Blvd, Oklahoma City, OK 73118 www.owrb.ok.gov
> <http://www.owrb.ok.gov/> | @OKWaterBoard
> <https://twitter.com/OKWaterBoard>*
>
>
>
>
>
> *From:* Miles Layne [mailto:miles.layne AT gmail.com]
> *Sent:* Monday, March 30, 2015 4:24 PM
> *To:* okgis AT orion.csa.ou.edu
> *Subject:* Re: [okgis] Accuracy Problem
>
>
>
>
>
>
>
> On Mon, Mar 30, 2015 at 4:19 PM, <miles.layne AT gmail.com> wrote:
>
> delema....I gps manholes and overlaid them on the aerial photos which is
> standard procedure for my CIP's; however, when I entered the lat & long
> from my
> gps point into Google Earth, the point was more accurate in locating and
> marking the manhole than what was overlaid on the aerials.  Is there a way
> to
> be more accurate when overlaying gps points to an aerial photo?  I ask
> because
> the cities and towns use the maps we create of their infrastructure for
> locating the various lines and appurtances we locate, gps and map for them.
>
> Thanks
>
>
> --
> This message has been scanned for viruses and
> dangerous content by MailScanner, and is
> believed to be clean.
>
>
> ------------------------------------
> To manage your subscription go to http://gis.ou.edu/sympa/info/okgis
> To unsubscribe email okgis-unsubscribe AT gis.ou.edu
>
>
>
>
> --
> This message has been scanned for viruses and
> dangerous content by *MailScanner* <http://www.mailscanner.info/>, and is
> believed to be clean.
>
>
> --
> This message has been scanned for viruses and
> dangerous content by *MailScanner* <http://www.mailscanner.info/>, and is
> believed to be clean.
>
>
>
>
> --
> This message has been scanned for viruses and
> dangerous content by *MailScanner* <http://www.mailscanner.info/>, and is
> believed to be clean.
>
>
> --
> This message has been scanned for viruses and
> dangerous content by *MailScanner* <http://www.mailscanner.info/>, and is
> believed to be clean.
>
> --

[The entire original message is not included.]

--
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.


Archive powered by MHonArc 2.6.16.

§