OPUS report question

The report output is determined by whether or not one is processing rapid static or static data. The report for static data returns # FIXED AMB. The report for fast static returns QUALITY IND.

http://www.ngs.noaa.gov/CORS/Articles/Snayetal-OPUSPt3-AS2011Vol8No7.pdf
try the link

> ? average of W-ratio (separation between candidate sets of ambiguities) of last 3 epochs

In order to calculate the position of a receiver, a baseline processor has to determine the number of integer wavelengths to each of the SVs it observes at a given epoch (point in time). However, because of uncertainties introduced by the variety of error sources inherent in the system, there's more than one set of integer estimates, and a confidence value associated with each set. As I understand it -- and I'm inferring quite a bit from information I've picked up here and there -- for each of the last 3 observed epochs in a submitted data set, OPUS-RS takes the 2 sets of integer estimates with the highest confidence values and divides the highest by the second-highest to derive a W-ratio. It then averages the W-ratios for the last 3 epochs. It does this for the confidence values derived in both network mode and rover mode, and publishes these values as the quality indicators.

Thus a quality indicator reported as 21.91/64.08 means that in network mode the best integer solution set was almost 22 times as strong as the second-best set, and in rover mode the best was over 64 times as strong as the second-best set. Values less than 2 are considered suspect.

Unfortunately, this link doesn't seem to work. At least not today on my systems.

In any event, take a look at this discussion http://www.citg.tudelft.nl/over-faculteit/afdelingen/geoscience-and-remote-sensing/research-themes/gps/lambda-method/

The ratio is basically a confidence indicator based on iterative solutions. Fast convergence leads to larger values.

HTH,

DMM

my link is to the American Surveyor magazine

I checked the link this morning. IT WORKS JUST FINE ON MY COMPUTER. I would highly recommend finding the American Surveyor article and reading it. I really appreciated the post by GEEODDMIKE, as the NGS article did not provide an explanation of the LAMBDA TEST.

Thanks for the replies all. I have read them and need to let some of that sink in. Jim your straight-forward wording was very helpful. The articles are good, but require extended study.

My wife hijacked me this weekend and I was not able to thank you all sooner. But at least I have lots to chew on this week with all the info.

> OPUS-RS takes the 2 sets of integer estimates with the highest confidence values and divides the highest by the second-highest to derive a W-ratio.

A little more investigation reveals that I oversimplified the derivation of the W-ratio. Here's a slide from Getting There Faster – The RSGPS program and OPUS-RS:

Similar concept, but a more sophisticated implementation.

Link to the NGS OPUS published articles list

Dane's link works fine for me too. For anyone having trouble, here's the link to the list of articles published by the NGS on OPUS. The article is the seventh in the list: "How Precise is Opus? Part 3"

Melita

Since I am not the mathematician my son is and he isn't handy at the moment. What is the VAR in the denominator portion of the EQ. Is that just short for variable? If so, it seems a bit redundant.
The Student's t distribution has a good write up in WikipediA.

The critical information to me is whether I have a fair or great solution. This indicator shows me the Networks & Rover "Health" of the solution. The value to hope for is greater than "3" for both where a higher value means the solution was derived quickly. At least that is what I am getting out of this so far. Thanks to all for your inputs! This has helped a lot.