PDS_VERSION_ID = PDS3 RECORD_TYPE = STREAM OBJECT = TEXT PUBLICATION_DATE = 2007-07-27 NOTE = " E-mail and other messages concerning MGS Radio Science on or about 1998/035." END_OBJECT = TEXT END From trish@rodan.jpl.nasa.gov Wed Feb 4 00:18:40 1998 Received: from rodan.Jpl.Nasa.Gov (rodan.jpl.nasa.gov [128.149.43.24]) by magellan.Stanford.EDU (8.8.5/8.8.3) with ESMTP id AAA08636 for ; Wed, 4 Feb 1998 00:18:39 -0800 (PST) Received: from 128.149.43.27 (trish-mac.jpl.nasa.gov [128.149.43.27]) by rodan.Jpl.Nasa.Gov (8.8.5/8.6.6) with SMTP id AAA18902; Wed, 4 Feb 1998 00:18:37 -0800 (PST) Message-Id: <199802040818.AAA18902@rodan.Jpl.Nasa.Gov> Date: 04 Feb 98 00:25:27 -0800 From: Trish Priest Subject: Occultation #13 To: David Hinson , Richard Simpson , Joe Twicken , Len Tyler CC: Salvadore Abbate , Sami Asmar , Jennifer Caetta , Paula Eshe X-Mailer: QuickMail Pro 1.0.2d1 X-Priority: 3 Reply-To: Trish Priest Content-Type: text/plain; charset="iso-8859-1" Content-Length: 440 Status: R Hello, Occultation #13 went smoothly. The signal disappeared at the expected time. Record Times: 035-07:35:00 to 035-07:50:00 One-way Times: 035-07:36:05 to 035-07:49:28 Expected Surface Begin: 035-07:47:27 SOE: p117a.01.soe SPK: spk_i_980202_OD116_126_V1 Script: 98035.45.a116.scr There were two attenuation changes. The second occurred at 07:38:35. The next occultation occurs at 6:04 PM PST Wednesday over DSS 45. -Trish ------------------------------------------------------------------- From hinson@nimbus.Stanford.EDU Tue Feb 10 13:30:49 1998 Received: from nimbus.Stanford.EDU (nimbus.Stanford.EDU [36.10.0.96]) by magellan.Stanford.EDU (8.8.5/8.8.3) with ESMTP id NAA06688 for ; Tue, 10 Feb 1998 13:30:44 -0800 (PST) Received: (from hinson@localhost) by nimbus.Stanford.EDU (8.8.4/8.8.3) id NAA26524; Tue, 10 Feb 1998 13:30:31 -0800 (PST) Date: Tue, 10 Feb 1998 13:30:31 -0800 (PST) From: Dave Hinson Message-Id: <199802102130.NAA26524@nimbus.Stanford.EDU> To: trish@rodan.jpl.nasa.gov Subject: Problem with occultation #13 Cc: joe@neptune.Stanford.EDU, rsimpson@magellan.Stanford.EDU, len@nova.Stanford.EDU X-Sun-Charset: US-ASCII Content-Length: 617 Status: RO Trish, Occultation #13 has the same type of problem that appeared in #1 and #16. After steering the data, frequency estimates prior to the atmospheric occultation have a sawtooth shape when they should be nearly constant. The problem was more subtle than on the other occasions, so we did not notice at first. We've seen it now at 15, 45, and 65. It's clearly a problem at the DSN (not the spacecraft) and it appears that the POCA is involved. We've never seen this type of problem before. Please relay this information to Sal. Fixing this problem should be a very high priority. Thanks, Dave -------------------------------------------------------------------- From joe@neptune.Stanford.EDU Thu Feb 12 11:42:41 1998 Received: from nova.stanford.edu (nova.Stanford.EDU [36.10.0.123]) by magellan.Stanford.EDU (8.8.5/8.8.3) with ESMTP id LAA03936 for ; Thu, 12 Feb 1998 11:42:40 -0800 (PST) Received: from neptune.Stanford.EDU (neptune.Stanford.EDU [36.10.0.149]) by nova.stanford.edu (8.8.8/8.8.4) with ESMTP id LAA02277; Thu, 12 Feb 1998 11:42:39 -0800 (PST) Received: (from joe@localhost) by neptune.Stanford.EDU (8.8.5/8.8.3) id LAA16726; Thu, 12 Feb 1998 11:39:00 -0800 (PST) Date: Thu, 12 Feb 1998 11:39:00 -0800 (PST) From: Joe Twicken Message-Id: <199802121939.LAA16726@neptune.Stanford.EDU> To: hinson@nova.Stanford.EDU, len@nova.Stanford.EDU, rsimpson@nova.Stanford.EDU, trish@rodan.jpl.nasa.gov Subject: Occultations and Synchronization Problem Content-Length: 3334 Status: RO Hello all- As you know, we have seen three open-loop recordings recently where the synchronization appears to have been lost between the raw data samples and the POCA. This loss of sync produces a sawtooth pattern in the carrier frequency estimates from the steered data. The problem has occurred as follows: Occ #01 1998/024 DSS15 Occ #13 1998/035 DSS45 Occ #16 1998/037 DSS65 I have spent quite a bit of time looking at the occultation #16 data. I first modified our steering coefficient maker (SCM) to account for all deterministic frequency drifts *except* for that due to the changing POCA; the program simply assumed a single POCA ramp for the entire data set. After steering the data with the new coefficients, all of the POCA ramp rate changes were obvious in the carrier frequency estimates (the frequency estimates were piecewise continuous with changing slope). Sad to say, there was (approximately) a 5-second discrepancy between the time that the POCA changes were actually observed in the data and the time that the POCA ramp rate changes were reported in the ODR headers. The observations were as follows: Change POCA Time Observed Time of Change of Change 1 53350 ~53355 2 53500 ~53505 3 53640 ~53645 4 53770 ~53775 The five second-discrepancy coincides with the duration of the leading edge of the sawtooth which we saw following our normal processing. Dave Hinson observed this and mentioned it yesterday at our weekly meeting. So, it is possible that the POCA times are being reported five seconds early. It is also possible (and my contention) that the data time tags are approximately 5 seconds too late. I ran my experiment geometry (EXPGEOM) program with a reconstructed SP-kernel for the given orbit to estimate the time of the occultation. The estimated occultation time was 53896.2 sec past midnight. According to our original data processing, the occultation appears to have occurred at 53905.6 sec (some 9.4 seconds after the predict). Dave Hinson also observed this problem which led his software to conclude that the planetary radius was a few km smaller than expected (thereby causing the late occultation). After discussing this situation with Dave, I modified our PREPMO program to subtract 5 seconds from all of the *data* time tags, leaving the POCA alone. I then reprocessed the occultation #16 ODR from start to finish. Because the actual time discrepancy is not exactly 5 seconds, there is still a bit of a sawtooth in the steered frequency residuals, However, the magnitude of the sawtooth is one order of magnitude smaller (~0.3 Hz versus the original ~3.2 Hz). The occultation now occurs 5 seconds closer to the predicted time (which is actually somewhat early itself for ingress occultations because I do not account for bending of the ray in the martian atmosphere). The signal is also properly steered to the center of the band using our usual USO frequency model. I think it is pretty clear that the data time tags are on the order of 5 seconds too late. I do not know how the data acquisition system works at the DSN or how this might occur. We can go back and look at the other two problematic occultations to see what the time discrepancy is in those cases. I will keep you all posted on my findings. Joe