As I was afraid of, the ORGI2 results of L1/CORS was not of the greatest quality. My nearest CORS station some 65 miles away has been off line for over a month now, but having gone back into the old original Test 1-3 files I found it would not have made anywhere near this difference.

Following are misclosures only:

Ave of Orig. Test 1-3 rel. to ORGI2

-0.239 North Orgi2 95% Err 0.425
-0.019 East Orgi2 95% Err 0.428
-0.244 Ortho Orgi2 95% Err 0.324

Ave of Opus Test 1-2 rel to ORGI2

-0.240 North
-0.009 East
-0.191 Ortho

My guess is poor satellite geometry over Stumpwater. Given the warning signs in the Ashtech Mission Planning software, I would probably pick a different time for such a session.

I still have no idea how to actually quantify/qualify the (personal opinion following) excellent results, especially horizontal, from the first three test sessions.

I do believe that if there had been another unit on fixed NAD83 coords within 30-40 miles, the preliminary Stumpwater results would be better.

J.D.

p.s. Jimbo, explain what I need to do to send my Locii data to site and I'll try. Or, better yet I'll convert to Rinex format and forward to you and you can put it on ftp site. I still think I'm ftp challenged.


pss We did run 3 units. The comparison of coordinates of three local points computed alone, relative to independent CORS processed coords of same 3 points was substantial. "Local" line lengths were 69.315' 3191.378' and 3233.498'. Elevation differences (point to point)were -6.599', -0.956' and -7.071' for "Local" Lines. The independent CORS coordinates resulted in lines of 69.288', 3191.580' and 3233.622 '. Cors derived elevation difference (point to point) were -6.810', -0.478' and -7.288'.


Modified By J.D. Billings on 2/5/2002 at 12:00 AM

With the limited sv constellation on 02.034, this is the best I can do to determine the location of the tower at the Crawfish Kitchen in downtown Louisana

Lat N 30°12'41.41425"
Long W 093°12'57.34567"
Ellipsoid -13.296

NAD 83 SPC Louisiana South Zone

N 627294.214 95%err 0.284
E 2686222.078 95%err 0.310
El 46.128 (Ortho) 95%err 0.221

Actually, this was just processing all vectors from CORS-CORS and CORS-JIMBO (and CORS-POST), holding fixed all CORS stations.
ARL5 Arlington TX
HOUS Houston TX
BEA5 Beaumont TX
DQUA DeQueen AR
WNFL Winfield LA
ENG1 NewOrleans LA (I think)


J.D., taking a wild shot in the dark

Touch of the flu slowed me down.
Haven't felt like making my brain work all day. Went home early and slept !!! till now.

I'll see what i get maybe later today...if the gears start spinning inside the ol'noggin.

Jimbo

Link to the FTP site.
Where ya normally type http://www.xxxxx.xxx
instead type
ftp://orgi:orgi@jimbo1.d2g.com
Then open my computer and go to your files.
Then mark your files, right click and pick copy from the menu.
Then go back to the FTP window and right click on jd-tx, and click paste from the pull-down menu.

Should work.

If it doesn't email them to me and I'll post'm.

Jimbo

3 CORS 10 degree mask
only edit is to set time to 0400-1500UTC
BEA5
WNFL
ENG1 - Baton Rouge


JIMBO - 95%
N 627293.755 ±0.402
E 2686219.481 ±0.448
Ortho 47.857 ±0.448

All vectors processed.
All CORS fixed.

Jimbo

Jimbo,

Didn't you run ties to the Tower from the local HARN point, or have you had the chance to do that yet? That would give us a preliminary "target".

JD


ps re-installed the max-ftp software and fixed problem. Now only problem is aol logging me off during a file transfer

Modified By J.D. Billings on 2/6/2002 at 10:38 AM

don't wannnabe a lightning rod.

Holding off on the HARN tie till clearer sky.

Jimbo

PS What happens when ya drop ARL out of your mix. When i tried the extra long on ORGI 1 ARL seemed to throw in a really big movement.

Jimbo

I gave your site a 4 character designation for my file - JMBO.

I re-ran the adjustment, excluding ARL5 as you suggested.

NAD83 LA-SOUTH ZONE

N 627294.249 95%err 0.305
E 2686222.116 95%err 0.331
Ele 46.144 (ortho) 95%err 0.241

We should also consider exclusion of DQUA as it's extended beyond WNFL. But, the software appears to be weighting the data well since you notice very little difference in the inclusion/exclusion of ARL5.

Now, to get to the point of what we discussed doing back in December with this "multi user L1" thing, I ran it again excluding ARL5 and putting "POST" in as a constraint point (used the OPUS derived "Precise" file coords). Here's another set of numbers:

NAD83 LA-SOUTH ZONE

N 627294.243 95%err 0.310
E 2686222.130 95%err 0.336
Ele 46.127 (ortho) 95%err 0.244


Keep in mind, I have a different mask angle for every vector, and in a few cases exclude sv's, segments of data, and occassionally restrict the use of a sv as a reference satellite. Also, some vectors can be made "fixed solution" rather than "partial solution" but will sacrifice quite a bit in quality as noted by the vector length error estimate (95% level). The "Yellow Box" guys need to remember there is a difference in the way Ashtech handles QA as compared to Trimble. If the question comes up, it's back in the archives, or Brian, Mike, Phil or someone more knowledgeable than myself will need to explain the process again.


J.D.

it was ARP not ARL that I tinkered with way back when.

Yeah, I'm gonna add Galveston and POST but I haven't started tinkering with the data. that's what I'm heading off for now.

Jimbo

get you a bit more "Southerly" coverage wouldn't it? Have you been able to spend the time tweaking the mask angles and cutting some of the really high resid. data out yet?

I'm fixin' to start a preliminary run on one of TM'S points. The man is blessed with CORS sites for the "box geometry".

JD


ps guess I'll have to wait to do TM's point as the UFCORS site is in a bit of disarray tonight and I can't download the files I want. I'm sure there's a work around to get 'em, I just don't have the time or energy right now. Maybe the easy stuff will be back up later.

Modified By J.D. Billings on 2/6/2002 at 8:58 PM

Started working thru the mask angles...say BEA5 to JMBO

I get the best with a 39° mask...by a lot ! The length errorstd. res. is 0.379 for vector length 305049.257'


Jimbo

PS: Have not started scrubbing the residuals yet.
Modified By James Webb on 2/6/2002 at 10:21 PM

You probably are using all the time on the receiver and downloaded CORS to extend front and end of session. I'm using CORS download time 0400-1500 and trimming the Locus times accordingly. That may be some of the minor differences we may see.

On the same vector, BEA5 to JMBO if I mask at 39° the std err is 0.472(USF). If I mask as 22°, and eliminate sv 25 (which made very little difference) the std err is 0.446(USF), and the vector length is 305048.633 (USF).

I still think we're "numerically challenged" as far as the constellation allowed during the session.

J.D.

at home now but after looking again later, I think I had SV 1 cut-out for a hour or so. Had been playing a little with the high masks and think I accidentLLY LEFT THE one tweak in place.

I'm only using the 0400-1500 UTC time block. Anyways...

Gotta go hunt up some real work. Things are dead around here...been talking with 5 other surveying co.'s and nobody is doing anything. Almost to that scary stage its so slow....and not just surveying.

Jimbo

my easting matches yours now....
don't know when or how but after I had pasted in the lat/long of all the CORS...they got changed somehow....maybe I didn't have them fixed but.....

anyhow I'm at about .4H and .35V but I haven't added in HOUS or GAL1 yet.

Jimbo

Some HARN derived coordinates would be nice for a comparison. Have you noticed the loonnngggg periods during the session where some vectors had only 2 sv's available? Still makes me think the O in O.R.G.I. may be outa place on this session. Maybe I'm just a bit pessimistic though.

I looked ahead at mission planning and found a couple of nights in the next 2-3 weeks that look ideal from constellation standpoint. Seems 02.034 was just not cut out for ORGI, in Northeast Texas anyway.

JD