Appendix D. Instruction on DOCB in WUM Rapid products - PrideLab/PRIDE-PPPAR GitHub Wiki

This guide is intended for PPP users who want to apply WUM0MGXRAP products (including orbits, clocks, and biases) to PPP software other than PRIDE PPP-AR. Users must be able to modify their software accordingly.

Compared to conventional IGS products, the aligned product offers additional orbit and clock estimates for each GPS, Galileo, and BDS satellite at the second midnight epoch (24:00:00) in the SP3 file and CLK file, respectively. The continuity index at day boundaries, termed as DOCB (Discontinuity of Orbits, Clocks, and Biases), is added to the SINEX-Bias file. To ensure optimal continuous data processing, users’ PPP software must:

  1. Read the estimates for the 24:00:00 epoch from the previous day's product and the 00:00:00 epoch from the next day's product. Prevent orbit and clock interpolations across day boundaries (i.e., 24:00:00 or 00:00:00). This mechanism is illustrated in the following diagram. Taking orbit interpolation as an example, conventional methods (left) tend to place the interpolation point in the middle of the interpolation interval. However, in our proposed method (right), when the interpolation point falls on the previous day, the epoch 24:00:00 of the previous day’s products should be used; when the interpolation point falls on the following day, the epoch 00:00:00 of the following day’s products should be used instead. Because the interpolation methods for orbit and clock differ, the compatibility of their interpolation results will be compromised by the discontinuities in orbits and clocks. However, this problem can be avoided by interpolating them within a day.

image97

  1. Read the code and phase DOCB values from the SINEX-Bias file, then compute the DOCB values according to specified signal choices for each satellite. If the DOCB value exceed a predefined threshold (e.g., 0.10 cycle), the rapid products for this satellite cannot support continuous processing at the day boundary, and a reset of its corresponding ambiguities is necessary. The code and phase DOCB values in the SINEX-Bias file is defined as follows:

    image98

    The columns are similar to those of “OSB.” From left to right, they represent the abbreviation “DOCB”, the SVN and PRN number of the satellite, the observation type label, the time of the day boundary (the two time labels are exactly the same and is intended to emphasize the “day boundary” concept), the unit of DOCB, the DOCB value, and the standard deviation of DOCB.

    For example, when a user selects the observations “L1W”, “L2W”, “C1W” and “C2W” for the satellite G02 and forms a dual-frequency ionosphere-free combination, the DOCB values of the corresponding observations, $d_{L1W}$, $d_{L2W}$, $d_{C1W}$, and $d_{C2W}$ should be read and formed as the wide-/narrow-lane combinations,

$$ d_{WL}=(f_1\cdot d_{L1W}-f_2\cdot d_{L2W})-\frac{f_1-f_2}{f_1+f_2}\cdot(f_1\cdot d_{C1W}+f_2\cdot d_{C2W})\ d_{NL}=\frac{f_1^2\cdot d_{L1W}+f_2^2\cdot d_{L2W}}{f_1-f_2} $$

where $f_1$ and $f_2$ are the frequencies of the selected dual-frequency observables, $d_{L1W}$, $d_{L2W}$, $d_{C1W}$, and $d_{C2W}$ are in units of seconds, and $d_{WL}$, $d_{NL}$ are in units of cycles.

Then, it is determined whether $d_{WL}$ and $d_{NL}$ exceed a predefined threshold (e.g., 0.10 cycle). If it does, the ambiguity parameter of the satellite is regarded as having a “cycle slip” at that day boundary, and it should to be reset to avoid degraded PPP results.