Traditional PPP solution needs 24 hours to achieve or reach the centimeter level. Also, it needs less than 5 minutes to reach the meter level.

what do you mean when talking about the resulting quality?

if you mean it as the standard deviation of the solution, this number expresses the quality of the least squares adjustment. and after two hours, it is expected to be fewer than 5. at any way, this is number does not represent the accuracy of the output coordinates.

Best Regards

Mohamed 

Dear Mohamed,

The quality number is a term used by RTKlib to indicate the mode of the output solution. For PPP, it should be 6. Although, I put all the options for PPP and gave the ultra-rapid corrections from IGS, the log still says that the results have quality of 5 which is a standard point positioning.

My data is static, so it shouldn't take that long to be in sub-meter ( or decimeter) accuracy. I am not talking about few centimeters but at least it should reach sub-meter level within the two hours for such a static scenario.

Using GIPSY PPP with 2 h data you obtain about 5 mm for horizontal, and about 20 mm for vertical coordinates (Refer to Hayal and Sanli 2016)! This is with IGS final orbits. I don't think the result will change much with ultra rapid ones but in any case I would also use final orbits to see whether you gain any improvement. Cheers. D. Ugur Sanli

Thanks Dr. Ugur for your reply. I have two questions for you:

First: is GYPSY PPP is a free software or not? 

Second: In my work, I need to do the PPP code myself as I will integrate it later with other sensors. I have problems in the accuracy as I still feel that I am not getting the benefit from the precise data. Can GYPSY or any other software provides me with the corrected satellite position and clock that I can directly use in my codes (so it will do the interpolation, Sagnac effect, relativistic terms, ..etc) or they only give the final solution?

Dear Mohamed,

it is  possible that ultra-rapid igs products can cause this result. as they have a long sample interval about 15 min. i suggest using final igs products because they have 30 second (sampling interval) in this link http://www.igs.org/products.

yes Mohamed, but the orbits are still every 15 min. I think my main problem is in the satellite position. I am just hoping that some might did this on his own like me so we can revise the steps that should be done to see what I am missing

Would you try a different data set at a different location and time ? Just in case!

Dear Mohammed 

I talk about the Igs final clock file which you used. the ultra-rapid file has 15 min while the final satellite clock has (30 sec or 5 min).

there is a big difference when using the final satellite clock product with interval of (30 sec or even 5 min).  in this link, ftp://cddis.gsfc.nasa.gov/gps/products/  , 1)enter the week of your data, 2) download the two files igswwww.sp3 and igswwww.clk. use these files in your processing.

Best Regards

Still I can't see superiority of the precise data to the regular corrections from Broadcast ephemeris. May be I should account for the other errors. So, does anyone know how to read the satellite antenna phase center from the ANTEX file and how to implement its correction?

I found the answer to my question. RTKlib didn't give a solution quality of 6 unless I gave it the final IGS products with all possible precise files available (ex: DCB, ATX)

I like the answers given so far, is there any Matlab scripts which can be used to read both the sp3 (15 minutes) and clk (30 seocnds or 5 minutes) of the IGS final product. Currently am able to read the sp3 file with precise satellite position, however i am reaching accuracy in meters upto 10 meter plus . I guess this is due to the wrong estimation of the clock errors . Planning to integrate the clock bias into my solution using the CLK file.

Please i am processing with RKTLIB on PPP static mode. I have been able to get the results out but i need help in able to identify the correct coordinates of the points and their corresponding standard errors after processing for 24 hours.

Thanks

Dear Mohamed,

I have included few online PPP services link. You can check it with your data. mm level accuracy receive with all below online services with final orbit.

AUSPOS : http://www.ga.gov.au/bin/gps.pl

OPUS: https://www.ngs.noaa.gov/OPUS/

NRCAN: https://webapp.geod.nrcan.gc.ca/geod/tools-outils/ppp.php?locale=en

Hope it will help you.

With best regards,

Somnath Mahato

Dear Mohamed Elsheikh ,

Where did you find that information about RTKLIB and final orbits ?

Also, do you know if short measurement time (1 hour) will give good results ?

Best Regards,

Bogdan

Franklin Onyeagoro The correct coordinates of IGS stations you can find in the SINEX file which is available with IGS products. (The SINEX format is defined in here" https://www.iers.org/IERS/EN/Organization/AnalysisCoordinator/SinexFormat/sinex.html )

or alternatively, some online tools can give you the station position http://sopac.ucsd.edu/sector.shtml

If your static point is not an IGS station but it is a new testing point, you can collect long static data (I suggest not less than 24 hours and the more the better) and then process it using one of the online tools such as the ones Somnath Mahato provided in his answer and they will give you their estimation of your point. Then, you can compare your solution with theirs.

Somnath Mahato Thanks for your answer. The problem of this question was about the RTKlib itself and I put the solution in one of my answers. However, the tools you mentioned are useful for people who want to compare their solutions to what we can say more trusted solutions.

Bogdan Narusavicius

The RTKlib software has its own manual that comes with the codes and has lots of detailed explanations to its algorithms. The final orbits you can know about from IGS website and its analysis centers.

For static data, the convergence time might take 20 to 40 min, and then you should be able to see a precise static solution less than 10 cm.

Apart from precise IGS products (ephemeris, antex and clock) input for ppp, RTKLib has one another option which is 'ionosphere'. One may use 'ionofree LC combination' instead of 'broadcast (default option)' along with the IGS products for obtaining ppp using RTKLib.

Regards

And yes, of course one should have dual frequency RINEX observables for calculating postprocessed ppp on ionofree LC using RTKLib