DSLWP-B has now been for more than a month in lunar orbit, since the orbital injection was made on May 25. Scott Tilley VE7TIL has sent me his latest batch of S-band Doppler measurements, including data for all this first lunar month. Having a complete lunar month of data is interesting for orbit determination purposes, since it gives observability of the orbit from all possible right ascension angles.
I have run my orbit determination with the new data.
Here is a comparison of the new orbital parameters with the ones obtained before.
% 20180610 DSLWP_B.SMA = 8762.40279943 DSLWP_B.ECC = 0.764697135746 DSLWP_B.INC = 18.6101083906 DSLWP_B.RAAN = 297.248156986 DSLWP_B.AOP = 130.40460851 DSLWP_B.TA = 178.09494681 % 20180623 DSLWP_B.SMA = 8765.95638789 DSLWP_B.ECC = 0.764479041563 DSLWP_B.INC = 23.0301858287 DSLWP_B.RAAN = 313.64185464 DSLWP_B.AOP = 113.462338342 DSLWP_B.TA = 178.5519212
The epoch is UTCModJulian 28264.5 in both cases. We see that there are noticeable changes to the inclination, RAAN, and argument of periapsis. The figure below shows an overview of all the Doppler data measured by Scott, as well as the comparison of both orbital elements. In green we have measurements considered for orbit determination; in red we have measurements excluded; and in orange we have points that might indicate a ground lock, where a groundstation transmits through the spacecraft transponder in order to perform a two-way Doppler measurement.
The figure below shows the residuals for the new and old sets of elements. We see that the older set doesn’t give a good match after MJD 28280. In contrast, the newer elements do not match the data before 28272 as well as the older elements. This effect is interesting and perhaps it indicates some factor that is not being modelled correctly in the orbit propagation. It would be interested to compare these residuals with residuals obtained from the tracking files published by Wei Mingchuan BG2BHC in dslwp_dev.
The two sets of elements give quite similar Doppler values. To examine the differences more closely, we show the data by segments of a few days. In the first days of data, we see that there are slight differences in the Doppler curves, and the older elements match the measurements more closely.
In the middle section of the data the Doppler curves are very similar and the measurements fit equally well both sets of elements.
In the final section we see noticeable differences between both Doppler curves. The measurements match the newer data better, indicating that overall the newer elements are more accurate.
Still, the main problem with these orbit determinations is the dependence on the local oscillator of the spacecraft. It has been estimated that the S-band beacon is 3400Hz high in frequency, so this value is subtracted to Scott’s measurements. However, this estimate is far from being accurate and the frequency might drift over time. Using two-way Doppler measurements would be much better, since the dependence on the spacecraft’s local oscillator is eliminated.
Scott’s data includes some frequency jumps that seem to indicate ground locks where a two-way Doppler measurement is being performed by the Chinese Deep Space Network. DSLWP-B uses a standard 240/221 turn-around ratio, as it is the standard in S-band DSN. Therefore, the uplink frequency is 2095.1MHz.
So far I have been unable to find the location of the transmitting groundstation. Without knowledge of this, the two-way Doppler measurements are pretty useless. The Chinese Deep Space Network has tracking stations in China as well as overseas in Neuquén, Argentina, Swakopmund, Namibia, and perhaps others. This is complicated further by the fact that China has also a fleet of satellite tracking ships, so the location of the transmitter could literally be anywhere.
The fact that the spacecraft needs to be visible from the groundstation at the times when the ground locks where measured indicates that the groundstation longitude is between Oceania and America, and discards all the groundstations in China. Other than this, I haven’t been able to find anything else.
The GMAT script and supporting Jupyter notebook for the calculations in this post have been updated here.