OPUS-RS - FAQs


	Online Positioning User Service


OPUS Upload \| What is OPUS-S \| Using OPUS-S \| Faqs OPUS-S \| Recent Solutions \| What is OPUS-RS \| Using OPUS-RS \| Faqs OPUS-RS \| Privacy Policy \| OPUS Policies \| Contact OPUS

FAQs

1) Will OPUS-RS take L1 data?: OPUS-RS takes only L1/L2 data. OPUS-RS requires that at least four four observation types be present in the user's file. These may be L1, L2, P1,and P2, or L1, L2, C1, and P2. A mapping grade version of OPUS, called OPUS Mapper, is being developed, and it will be able to take L1 data.
2) I saw your 92% and 97% confidence message on my printout. How can I be more sure of my results?: These messages have been removed. Better accuracy messages are now being provided.
3) What is the maximum time period for a session, and could the same file be put through OPUS-RS that is put through OPUS?: OPUS-RS is set to not run and return a message to users for datasets over 4 hours. Also, OPUS suggests a minimum of 2 hours, so it is possible to send datasets between 2 hours and 4 hours to both utilities.
4) OPUS-RS has been in a testing or "Prototype" phase for over one year. Is this version finished, or set in stone?: OPUS-RS has undergone thousands of tests, so we have officially released it for operational use by the public. However, it is still being refined so users may notice some changes in the future.
5) What are the quality indicators on the data sheet?: A quality indicator is the average value of the w-ratio at the last three epochs of data. The w-ratio is a measure of how well the LAMBDA algorithm was able to determine the correct integer ambiguities. There are two of them because OPUS-RS uses the RSGPS program twice. The network mode run uses only data from the reference stations; its purpose is to determine the ionospheric and tropospheric refraction for the time and location of the user's data set. The rover mode run uses the information from the rover run to find the user's position. Values of the quality indicators below 1.0 should be taken as warning signs. However, low quality indicators by themselves don't tell the whole picture - it is possible to have low quality indicators along with low standard deviations (located to the right of each position component) and still have a good solution.
6) What is the normalized RMS?: This quantity is also known as the reference standard deviation (the square root of the reference variance), the standard deviation of the observation of unit weight, and other names. It is a unitless quantity, and its expected value is 1.0. Values of the normalized RMS greater than 1.0 mean that the weights assigned to some or all of the observations were too large. This can happen if the user's data set or one or more of the reference stations contains particularly noisy data. Most runs of OPUS-RS produce a normalized RMS of 1.0 or less, meaning that the noise in the data was actually less than would be indicated by the assigned weights.
7) How are the weights assigned in OPUS-RS?: In OPUS-RS, the weight of an observation is the inverse of the square of the observation's standard deviation. For one way phases observations, the standard observation of an L1 observation is 0.019 meter/sin(elevation angle), and the standard observation of an L2 observation is 0.0244 meter/sin(elevation angle). For one way range observations, the standard deviation is initially set at 0.05 meter/sin(elevation angle) for P1 and 0.064 meter/sin(elevation angle) for P2. However, the range observations are filtered to detect noisy data and multipath. Noisy range data may be downweighted by assigning a larger standard deviation. In addition, OPUS-RS contains a priori constraints on all the parameters. The assigned standard deviations are 0.4 meters for double difference ionospheric refraction delays, 0.025 meters for the correction to the nominal tropospheric delay, 1.44 meters for the correction to a priori value of an L1 double difference ambiguity, and 2.3 meters for the correction to a priori value of an L2 double difference ambiguity.
8) What happened to the peak-to-peak errors?: OPUS-RS uses a simultaneous least squares adjustment of all the available data (reference stations plus the users data set), while regular OPUS uses three single baseline adjustments. The simultaneous solution is the more theoretically correct approach, but it doesn't produce the peak to peak errors. It does produce standard deviations for the coordinates of the user's receiver but these are almost always too optimistic to be useful. For those users who desire it, the covariance matrix of the user's receiver from the adjustment is available in the extended output.
9) How soon should a GNSS user submit his/her observations to OPUS-RS for processing?: Roughly speaking, OPUS-RS should provide reasonably accurate coordinates for data-collection points located in the 48 coterminous states if the user waits at least 2 hours after the completion of his/her GNSS observing session before submitting the collected data to OPUS-RS for processing. OPUS-RS may provide more accurate coordinates, if the user waits longer to submit their data. Little or no increase in accuracy should be obtained by waiting more than 28 hours.
More precisely speaking, to process a submitted GNSS dataset, OPUS-RS needs (1) corresponding GNSS data from at least three CORS and (2) corresponding GNSS satellite orbits. Moreover, OPUS-RS uses CORS data that spans a time interval which extends the time span of the user's data by as much as 26.4 minutes in each direction. OPUS-RS uses this extra span of data to accurately estimate atmospheric parameters that characterize the refraction delays experienced by the user's GNSS data. Now, most CORS data are collected hourly within 30 minutes after the turn of the hour. However, some CORS data are collected only daily within 4 hours after midnight at Greenwich, England. Hence, NGS recommends that users do NOT submit their GNSS data to OPUS-RS until at least 30 minutes after the turn of the first hour which occurs at least 26.4 minutes after the end of his/her observing session. The chances of having GNSS data available from more CORS are improved, if the user submits his/her GNSS data to OPUS-RS at least 4 hours after the first Greenwich midnight following the end of the user's observing session. NGS studies reveal that the accuracy of OPUS-RS results generally improves with the use of data from additional CORS. Hence, OPUS-RS will use data from as many as nine CORS, if these data are available (and if these CORS are within 250 km of the user's data-collection point). Also, the accuracy of OPUS-RS results should be better if this utility is able to use the `postfit' satellite orbits, as opposed to the `predicted' satellite orbits, generated by the IGS. The IGS releases new orbits four times per day at 3:00, 9:00, 15:00, and 21:00 relative to Greenwich time. Moreover, each release contains postfit orbits for up to 3 hours before the release time. Hence, a user never has to wait more than 9 hours (and he/she may only have to wait as little as 3 hours) to have OPUS-RS process his/her GNSS data with postfit orbits. Note that accuracy differences between postfit and predicted IGS orbits continue to shrink as understanding of the underlying physics has improved over the years.
19) What does "D.N.E." mean in the output?: D.N.E. means "Does Not Exist". It is a flag used to indicate that insufficient information exists to compute the field in question. For example, when D.N.E. is listed under orthometric height accuracy, this indicates that there is not enough information about the relationship between the hybrid geoid, the geometric reference frame and the local vertical datum in this region to estimate the accuracy of the orthometric height.
20) What does "No official datum supported" mean in the output?: This flag is used when there is not enough information for a GPS based solution to yield access to the official vertical datum of a region, if it exists. Specifically, this means that the hybrid geoid for this region does not contain information about the GPS derived ellipsoid heights on benchmarks which have leveling-derived orthometric heights in the official vertical datum of the region.
21) What does "No NGS Geoid Model Available" mean in the output?: This flag is used in regions where NGS has not computed a high resolution model of the geoid (such as Guam or American Samoa) or where such a model may exist, but is not supported such as the portion of the CONUS geoid which overlaps with non-USA territories.
22) OPUS-RS does not have as high a chance of a successful solution as OPUS because of the added constraints of OPUS-RS. Can I find out where in the U.S. I should consider occupying the site for at least 2 hours, so if it fails in OPUS-RS I can still use OPUS?: OPUS-RS map shows the estimated vertical standard errors achievable with 15 minutes of GPS data when using OPUS-RS in the coterminous U.S. These standard errors were computed as a function of the IDOP and RMSD values provided by the CORS network as of May 2008. The areas of high vertical standard errors also correspond with a higher than normal percentage of failed solutions.

National Geodetic Survey - OPUS Team
Modified 21 March 2008
http://www.ngs.noaa.gov/OPUS/Faqs-RS.html