BDS-3/GPS/Galileo OSB Estimation and PPP-AR Positioning Analysis of Different Positioning Models

Round 1

Reviewer 1 Report

Please see the attachment.

Comments for author File: Comments.pdf

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Reviewer 2 Report

The paper deals with an interesting topic of Observation Specific Biases (OSB) , however, it provides hardly any new or innovative information, since are routinely generated and made available by nearly all IGS Analysis centers (ACs). Of which 3 AC OSB products were used for verification (note not the CNES but GFZ OSBs were used). The English of the paper is good, except for a few terminology words outlined below. The main problem I have with the paper is that it shows a significantly better precision for single epoch kinematic PPP positions than for the 4-h static ones. This is not possible, as in general static PPP results are significantly better than the kinematic ones, as the static PPP precision should be akin to the precision of all the single epoch kinematic PPP position solutions within the respective 4-h period. This needs to be addressed and corrected.

 

After this is addressed and corrected as well as the detail suggestiins/corrections/comments listed below, the paper may be acceptable for a publication 

 

-------- detail corrections/suggestions/comments --------

p.2, 

- ... from float to fixed ambiguity PPP ...

- LAMBDA (reference)

- replace profoundly with e.g. thoroughly : 68.3%, 34.71% with 68%, 35% etc. (and elsewhere, 1 decimal for %improvements are not necessary, I think) : clients? (replace e.g., with approaches?

- narrow-lane FCB

 

p2. and 3  OSB has been adopted by IGS and all  the IGS ACs!

                hardware parts => receiver hardware

                phase wrap => phase wind-up

 

p.3 : "is the projection function of the trop.." => the mapping function of the trop.

 

Eq. 3 delta_DCB and delta_b need to be explained in the text below

 

Eqs 4, 5 as well as in Eqs 7 and 8; alpha+beta=1 is not needed here

 

p. 5 GPS needs to be spelled out here, possibly with a reference

 

 Eq. 17, the 2nd right hand side (r.h.s.) portion is incorrect, consequently the IF lane is wrong, for GPS it should be 10.7 cm , not 6.3 mm!

 

Eq. 24 r.h.s. need to distinguish Br and Bs

 

Eq. 27: what is b_C1, it needs to be defined here

 

Fig. 1 is it necessary? (the relevant details are in the text)

 

p.9:  IGG3 anti-variance method (reference)

       Note Galileo OSBs and to a smaller effect BDS3  ones are more stable than for GPS. Galileo OSBs has been shown to be stable over days or even months!

 

      "Unlike the OSB-solved stations,  22..." => different 22 

 

Table 1; note for Galileo CODE is not using IGS14.atx PCO/PCV (only for BDS and GPS), they are using a different atx file. Was IGS14.atx used for Galileo, if so the Galileo results are negatively affected . The Galileo results in Fig 4 and Tab 2 should be even better ..

- the Galileo E14 and E18 (in an eccentric orbit) were excluded here (in spite that CODE include them in their orbits/clocks and OSBs) , it should be mentioned ..

 

-  constrained station coordinates (Tab 1 needs to specify to which coordinates were constrained/fixed. if IGS solutions were used, which one?)

 

p.10 "the self-programmed software from .." => the software developed by

- "The timing diagrams .." => the plots ..

 

p.12   CNES??? you mean GeoForschungsZentrum's (GFZ)? Also correct elsewhere, e.g. abstract, Tab 3 etc. PLEASE CORRECT EVERYWHERE! CNES does not use BDS, only GPS, GLONASS and Galileo.

 

Tab. 3 CODE is making their OSBs available, please correct here

 

p.13 how are defined 'fixed epochs", it needs to be specified here 

 

- GPS Float or AR PPP (single epoch) kinematic E precision (the best= 1 cm) cannot be better than the 4h static one (~1.6 cm)! this need to be explained/corrected

 

Tab. 4 should clearly state it is for GPS-only PPP's, What about E, C and GEC PPP??? (like in the Section 3.4) and for DOY 300-306

 

p.14: "However, the convergence time was slightly higher compared to the other two prod-ucts due to the longer convergence time of their PPP solutions" 

Not sure what is meant here, that  CASM has a slightly longer (worse) convergence time for both static and kinematic PPP? (note the float and AR convergence also depend on the AC orbits/clock quality

 

p.16: ".. single-GPS performed worse than the first three .."  Please rephrase, this is potentially ambiguous and misleading as GPS-only PPP's are usually better that Galileo-only or BDS-only PPP's

 

Figs 12, 13 and Tabs. 5, 6. As already mentioned  above, the 4-h static PPP precision cannot be worse than single epoch kinematic PPP precision. Something is wrong here?

 

p. 20: "The GPS fixed-ambiguity PPP solution shows the most significant improvement among all the PPP solutions. The possible reason for this could be that the PPP solu-tion is less accurate than the dual- and multi-systems, resulting in a more significant fixed-ambiguity PPP solution."

 

Note that this is obvious and also equally true for all single GNSS PPP's, like Galielo-only, BDS-only PPP's, etc. Please rephrase or improve or shorten

 

Acknowledgments: All authors gratefully acknowledge CODE, WUM, CNES ???

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf

Round 2

Reviewer 2 Report

The paper still needs some minor modifications/improvements, even though the authors considered or implemented nearly all of my suggestions/comments. Except for clarifying what is meant by a fixed epoch (are the ambiguities of min. 3, 4 or all the satellites of the epoch fixed?) And how and where from were obtained the fixed station positions which were used for the OSB generation and/or as a reference for the PPP testing. This is essential information in particular for static PPP's. Directly related to this is the use of 'accuracy', as even  the most precise reference (station positions) does not justify the use of accuracy. precision or repeatability would be more appropriate here. Also, not sure what is meant by mean errors used in the text and Figs/Tabs legends. Are these STD or RMS, this should be corrected/clarified throughout the text and in all Figs/Tabs legends. Also in order to prevent any confusion, it should be clearly stated in the abstract and in the introduction that CNES OSB products are generated using GFZ (GBM) orbits/clocks. since this is highly unusual that AC would not be using its own orbits/clocks and instead be using different AC orbits/clocks.

 

Below listed are a few more details, which should be considered/corrected

 

----------- 

 p. 2 - replace profoundly with thoroughly

      - replace FCB benchmark with FCB datum

 

p. 3  - OSB products by CNES (based on GFZ Rapid orbit/clock      solutions)

      - tilted path => slant path 

 

p.8 and p.9 - benchmark => datum  

 

p. 11  - the OSB estimation since the number of

 

Figs 3-7 and Tab 2 legends should say it is for DOY 301, 2021

 

Figs 7-8 legends - seven days => DOY 300-306, 2021

 

p.13 - by CNES (based on GFZ Rapid orbit/clock solutions)

 

p. 14 - fixed epochs? (are the ambiguities of min. 3, 4 or all the satellites of the epoch fixed?)

Figs 12-15 are old, need to be updated/corrected

 

Figs 14, 15 legends - IF model (top) and the UDUC model (bottom)

 

In all Figs/Tabs, please replace 'statistics' and 'average positioning errors' with either positioning repeatability, or standard deviations of position differences or RMS of position differences, whichever is applicable

 

 

 

Author Response

Please see the attachment.

Author Response File: Author Response.pdf