Major Events and Software changes since launch
Friday 6/10:
The Delta-II rocket carrying the Aquarius/SAC-D spacecraft lifted off on schedule at 14:20:13 UTC this morning. The initial ascent was nominal (that's NASA-speak for perfect) and the second stage and spacecraft are coasting to the orbit altitude. The final burn and spacecraft separation will occur in ~30 minutes.
The first real-time HKT (HKTMRT) file from the Malindi was detected and ingested successfully by ADPS. The data showed that the AOCS mode transitioned from Standby to Safehold to Survival. The solar array currents indicated deployment of the arrays during the pass.
The HKTMRT file from the first Svalbard pass showed that the GPS receivers were active and generating realistic orbit vectors.
The HKTMRT files from the Alaska and McMurdo stations were very noisy and had many frame drops. The Svalbard passes continued to be normal quality.
During the first ETC pass (21:37 UTC), the X-band transmitter was observed to power on in the real-time HKT. The first stored HKT (HKTMST) file was not received until early the following morning.
Fred performed an analysis of the AOCS pointing control using one of the HKTMRT files. This showed that the -Y (solar array normal) axis was Sun pointing, and the spacecraft was rotating around the Sun line at about 2 RPO, which (we learned later) is the norm for AOCS survival mode.
Saturday 6/11:
The first X-band HKTMST file (CGSS_20110611_092100_10020110611083306_SACD_HKTMST.bin) was received and ingested early in the morning, containing about 2 hours of data.
Fred performed analysis on this downlink, showing the satellite ground track based on the GPS data and the solar array currents as the spacecraft passed through eclipse.
Joel reported that the estimated attitude quaternion in the HKT AOCS block was all zeros. Fred identified another quaternion, based on raw star tracker data, that was valid except when the tracker viewed the Earth as the spacecraft pitched in Survival mode.
Fred reported that the AOCS transitioned from Survival to Safehold (nadir pointing) at 22:13 UTC, using HSTMRT data.
Sunday, 6/12:
Fred reported that the AOCS transitioned back to Survival between two Svalbard passes at 00:12 and 01:47 UTC. The time was later determined to be 01:22 using the HKTMST file.
CONAE reported that they would provide all of the HKTMST files starting from separation.
Monday 6/13:
A report was received from ULA on the injection orbit, showing errors well below the 3-sigma estimates. Later that day, CONAE reported that the injection altitude was about 800 m above nominal.
Fred provided an initial list of AOCS fields to display on the SAC-D HKT web site for Liang and Norman to implement in the analysis and display tools.
CONAE delivered an HKTMST file containing about 42 hours of data starting from spacecraft separation.
Wednesday 6/15:
Fred generated a plot of the AOCS mode and eclipse periods, showing that the transitions from Safehold to Survival mode occurred at the end of every eclipse that started in Safehold. Later that day, Fred performed analysis of the -Y to Sun angle during the eclipses, showing that the angle increased significantly during the eclipses, well above the 45-degree limit that triggers the transition from Safehold to Survival. Subsequently CONAE would report that this accumulation of control error resulted from transitory behavior in the Coarse Sun Sensors (CSSs) at eclipse entry that caused a drift in the roll and yaw angles.
Thursday 6/16:
Fred performed analysis of the AOCS telemetry showing that the reaction wheels are not commanded in Safehold mode during eclipses. This corroborated findings in a report delivered by CONAE yesterday afternoon.
Friday 6/17:
Joel delivered hktmerge_aquarius to generate a merged HKT file from multiple input files.
Fred reviewed a block diagram of the AOCS Safehold control algorithm, and reported that the Triad algorithm used with the Sun and magnetic field vectors would not work in eclipse. The diagram also showed the gyro data as input, but apparently this was not implemented. INVAP subsequently confirmed this.
CONAE successfully tested the disabling of the reaction wheel control by command prior to the start of an eclipse, which prevented the accumulation of Sun angle error and the return to Survival mode.
At Gary Lagerloef's request, Fred performed analysis of the equator crossing times on this day compared with the first orbit after launch, and showed that the time was about 100 seconds later, indicating the deviation from the weekly repeat cycle. After this Fred continued sending weekly updates of the crossing times.
Norman completed the implementation of the AOCS fields on the telemetry analysis web site.
Saturday 6/18:
Fred reported that CONAE maintained the AOCS in Safehold for about 10 hours by disabling the reaction wheel control each orbit.
Monday 6/20:
John implemented the hktmerge_aquarius processing in the ADPS.
Tuesday 6/21:
Fred performed analysis showing that the reaction wheel commands are determined almost entirely from the attitude error derivative (dot_qds) values, and are mostly uncorrelated with the error (qds) values; and also that the dot_qds values are noisy, since they are determined from angle sensor measurements with no gyro input. These results support one of CONAE's proposed solutions to the eclipse mode transitions, which is to reduce the control gains.
Based on responses from Dalia, an effort by Liang and Norman to display information about the Aquarius software messages was suspended.
Fred started reviewing an ongoing E-mail exchange between CONAE and FDF on the CODS ephemeris data comparisons with the FDF ephemeris.
Wednesday 6/22:
Jose Relloso of INVAP responsed to Fred's analysis, indicating that he would prefer other solutions than reducing the gains to resolve the eclipse problem.
Fred and Gene participated in an AOCS walk-through by NASA scheduled by Dankai Liu.
Friday 6/24:
Fred delivered a report showing that one proposed solution to the eclipse problem, increasing the threshold on the CSS to cause an earlier transition to eclipse, would probably not be sufficient to resolve the problem.
Fred and Gene participated in a discussion led by Marcelo of their proposed solutions to resolve the AOCS eclipse problem.
Monday 6/27:
CONAE performed a test of increasing the CSS eclipse threshold with control disabled.
In response to requests from the Science Team, Joel staged a set of sample quicklook files on samoa.
Gene sent a query to Cristian Filici about when CONAE would start delivering CODS products.
Tuesday 6/28:
CONAE performed a test of increasing the CSS eclipse threshold with control enabled.
Fred delivered a report in which he derived the AOCS control gains in Safehold mode and presented further arguments for reducing the gains. Jose Relloso responded with a confirmation of the results in the report, and re-iterated that he recommended other solutions than reducing the gains.
Wednesday 6/29:
A number of AOCS fields in the HKT data were corrupted starting at about 09:00. CONAE declared a spacecraft emergency. Fred analyzed the HKT data and showed that the problem was a 2-byte shift in the middle of the AOCS block. It was determined that the time corresponded to a passage of the South Atlantic Anomaly (SAA).
In response to the delays in the completion of the S/P commissioning phase, Gene requested an extension of the NEN support from Lynn Myers.
Thursday 6/30:
Fred provided a proposed set of AOCS field limits to Norman for the HKT web page.
Friday 7/1:
Norman implemented the limits provided by Fred, who made some minor changes based on the results of the limit checking.
Saturday, 7/2:
The HKT telemetry corruption was corrected by CONAE at about 17:00.
Monday, 7/4:
Gene and Fred participated in a telecon to discuss the path forward to resolve the AOCS Safehold eclipse problem and proceed with S/P commissioning.
Tuesday, 7/5:
Gene and Fred participated in a telecon wit JPL to develop NASA's position on the AOCS path forward. Fred sent comments on the position document, but later learned it had already been delivered to CONAE. The decision was to raise the Sun angle threshold to 65 degrees and make no other changes (other than the previous change to the CSS eclipse threshold).
Wednesday, 7/6:
Gene queried CONAE about the status for routine delivery of the HKTMST files, which was still taking longer than had been predicted before launch.
Thursday, 7/7:
Daniel Caruso delivered a schedule with Aquarius commissioning starting on 10 August.
A status report from Shen included a proposal to have JPL analyze the GPS data using the GIPSY/OASIS software. In response to questions from Gene, Daniel Caruso clarified this.
Friday, 7/8:
CONAE commanded the Sun angle threshold to 65 degrees and the AOCS mode to Safehold, where it remained stable.
Fred began work on software to process the GPS data in the HKTMST files and generate ephemeris products in the CODS format, based on the SeaWiFS navigation software. He sent a proposal to Gene for the completion of this effort.
Dalia re-initiated discussion of the AOCS configuration for Aquarius antenna deployment.
Sunday, 7/10:
CONAE commanded the AOCS to Science mode at about 17:30. The attitude control stabilized very quickly.
Monday, 7/11:
CONAE commanded the AOCS to yaw steering mode for the first time at about 09:30.
CONAE performed the first pitch maneuver test at about 18:00. The spacecraft was successfully maneuvered to -45 degrees pitch and back to nadir.
Norman implemented data aggregation for the multi-day plots on the HKT analysis web site.
Fred generated the first set of ephemeris products with the adapted SeaWiFS software.
Tuesday, 7/12:
CONAE began routine delivery of the CODS ephemeris products. The definitive orbit products were found to contain about two days of data, starting near the middle of a day, with about one day of overlap between successive deliveries.
Daniel Caruso distributed a proposed schedule for the orbit adjust maneuvers, which spanned more than three weeks and included an exercise of the CARA (conjunction assessment) engagement process. The proposed adjustments were a reduction in altitude of 700m and in inclination of 0.0017 degree. This schedule also showed Aquarius commissioning starting 11 August.
In response to Daniel's schedule, Gary started a discussion of the requirements for the mission orbit. Fred provided analysis in support of this.
Thursday, 7/14:
The propulsion system CSRR was held by CONAE.
The Science Team began a discussion of the requirement for geodetic pointing during science data collection, in response to Fred's report that the current control was geocentric.
Friday, 7/15:
Gene re-iterated the GSFC requirements for the handover from JPL, as part of the planning of the PLAR and handover review.
Sunday, 7/17:
CONAE performed a test of a -180 degree pitch maneuver starting at about 18:57. During the return maneuver, in eclipse, the AOCS transitioned to Safehold mode at about 17:17, and recovered nadir pointing after exiting eclipse. CONAE commanded the AOCS back to Science mode at about 23:00.
Monday, 7/18:
Review of the data during Sunday's maneuver showed that the star tracker stopped tracking stars as expected during the maneuver, when the tracker began to view the Earth. However, during the return maneuver the tracker erroneously indicated that it resumed tracking while still Earth pointing, and produced an erroneous quaternion, which caused the Safehold transition.
CONAE successfully performed a -180 degree pitch maneuver starting at about 21:22.
Tuesday, 7/19:
CONAE began delivering the 72-hour HKTMST files.
Fred delivered the orbit processing software to the Subversion repository.
Susie initiated discussion with CONAE of the need to use the Scheduler Client when working remotely (not at Goddard).
Wednesday, 7/20
Fred showed that during the CSC test on Sunday, at the time when the tracker produced the erroneous quaternion, it was pointed at nadir over the Antarctic ice sheet in darkness.
Daniel Caruso distributed a proposal to delay the orbit adjust maneuvers until after the Aquarius commissioning, in order to maintain the commissioning schedule. Gary Lagerloef called a telecon to develop a Science Team position on the orbit needed for the science commissioning phase.
Thursday, 7/21
Gary delivered a recommendation to perform an orbit altitude reduction of about 600m before the start of commissining, with the final reduction and inclination change to be performed afterward.
Friday, 7/22
CONAE performed the first test of AOCS Propulsion mode at about 12:30. This was also seen to cause the GPS HKT fields to become invalid. CONAE responded that in this mode, certain AOCS values were sampled more frequently, replacing the GPS values.
During the test, the AOCS mode transitioned back to Science and almost immediately to Safehold. This was determined to result from a timing issue in the mode change commands.
A pitch maneuver of -180 degrees was successfully performed in Science mode at 19:40.
Sunday, 7/24
CONAE successfully tested the propulsion system, with a 4-second burn at 16:50.
Monday, 7/25
Fred performed analysis that verified the effect of the test burn on the orbit propagation.
Liang started a discussion with CONAE of the HKT time tags, specifically deviations from 8-second sampling in the AOCS and GPS time tags.
Tuesday, 7/26
CONAE performed two orbit raising calibration burns.
Fred delivered source code updates for the orbit processing software.
Wednesday, 7/27
CONAE performed a -180 degree pitch maneuver in Propulsion mode, to verify the ability to perform orbit reduction burns.
Thursday, 7/28
CONAE performed two orbit raising calibration burns.
Saturday, 7/30
CONAE performed two orbit reduction calibration burns, at -180 degrees pitch.
Monday, 8/1
The AOCS was in Safehold mode from about 11:20 to 23:20 (need to remember why).
Joel installed a new galaxy reflection table delivered by Frank Wentz.
Tuesday, 8/2
During a scheduled propulsion maneuver at -180 degrees pitch, the AOCS transitioned to Safehold. The mode was commanded to Science at 20:11.
Thursday, 8/4
A -180 degree pitch maneuver was performed in Propulsion mode, verifying the ability to perform the orbit reduction burns.
Friday, 8/5
CONAE performed the first major orbit reduction burn at 10:19, reducing the altitude by about 160m.
Joel modified l0gen_aquarius to produce output L0 records when the Aquarius DPU is not powered on and tested it with an ATLO file provided by Liang.
Adam Freedman raised a question about the delays in the ADPS acquisition of the HKTMST files. A report from John showed that the data transfer times were short (2-3 minutes through 27 July, increased sporadically through 2 August, and were consistently long (up to 3 hours) after that. Paul began an effort to trace the connections between the CUSS server and the ADPS acquisition nodes. Liang also started working on modifications to the HKT analysis tools to reduce their processing time, althought this was not a factor in the file transfer and availability.
Saturday, 8/6
The last eclipse occurred late on this day.
CONAE commanded the AOCS pointing into geodetic mode at the end of the day.
Sunday, 8/7
CONAE performed the second major orbit reduction burn, reducing the altitude by about 450m.
Monday, 8/8
CONAE performed a test of a 90 degree yaw maneuver in propulsion mode, required to perform the inclination adjust, at about 21:15.
Fred sent a question to CONAE about possible discontinuities in the AOCS OBT. This was in response to discontinuities observed in the computed pitch angle.
John completed the installation of Fred's orbit processing program, orbgen_aquarius, and started routine processing of the daily merged HKT files.
Tuesday, 8/9
The inclination adjust calibration burn was performed at about 22:00.
After HKTMST transfer times had improved for a few days, transfers began to be long again (2-3 hours).
Wednesday, 8/10
Tests of AOCS Safehold mode were performed in preparation for Aquarius antenna deployment.
Susie was able to connect remotely to the Scheduler Client from her home, in order to be able to support Aquarius commissioning while Liang and Alicia are in Cordoba.
Thursday, 8/11
CONAE reset the CUSS server and updated the dsa-key/rsa-key on the CUSS. HKTMST file transfer times improved again.
Sunday, 8/14 via Skype:
The Aquarius ICDS was powered on at 13:32 UTC.
First downlink file arrives on CUSS: IMT_20110814_164700_18020110814163603_SACD_RawAQ.bin ODPS ingest fails because IMT prefix was not expected for downlink files. The ODPS was updated to support the IMT downlink files.
After resolving the ingest problem, the IMT_20110814_164700_18020110814163603_SACD_RawAQ.bin downlink file failed to process through l0gen_aquarius because the magic word, "deadbeef" was not found.
Joel delivered a new version of l0gen_aquarius that handles byte-shifted downlink files. It was installed in the ODPS, and that allowed the downlink file to be processed to L0. Previously l0gen_aquarius assumed that the synch word was 4-byte aligned within the downlink file.
Monday 8/15:
The l1agen_aquarius program initially failed to process the L0 file. This was traced to duplicate ephemeris records in the temp-ephem file created for the l0gen_aquarius run. The ODPS pre-processing rule that selects predicted ephemeris data for the quicklook processing was updated so that only one predicted ephemeris file -- the one with the most-recent creation time -- is selected.
l1agen_aquarius ran successfully on the L0 files, but the l1info_aquarius could not derive any geobox information, so the L1A-metadata could not be created. Also, the analysis tool was failing because the output group parameter was not being specified. The missing geobox data was traced to bad attitude data. This was caused by the gen_aquarius_att program putting invalid records in the attitude file in cases where the HKT and ephemeris data did not align. Joel modified the gen_aquarius_att program so that it only generates attitude records for times that have both HKT and ephemeris data. Liang e-mailed a new version of the readSPHKTfromL1A.m that fixed the problem with the missing argument for the call to the analysis tool. Both of these fixes allowed the l1agen_aquarius process to complete successfully.
Some of the L2-quicklook streams failed in the l2gen_scatter program because of missing IONEX data. Joel delivered a new version of the program that resolved the problem.
Some of the L2-quicklook streams failed in the l2gen_aquarius program because the program wanted one or two additional Y-anc files. Joel delivered a new version of the program that will substitute missing Y-anc files with one of the specified Y-anc files, i.e. use Y-anc1 if Y-anc2 is missing, and use Y-anc2 if Y-anc3 is missing.
It was determined that the l1agen_aquarius program was not writing the HKT data into the L1A files in the ODPS configuration, while it was doing so in Joel's environment. This problem was traced to a different calling sequence in the ODPS that included a 5th argument, "QL" to l1agen_aquarius. The program was not accounting for this argument when deciding whether to write the HKT data to the L1A files. Joel delivered a new version of the l1agen_aquarius program, and all of the L0 files were reprocessed.
Another problem with writing the HKT data was traced to an HKT file collected during the period with the byte-shift (29 July to 2 August). This file had an anomalous time range and was being staged for every HKT merge run. The file was give a failed QC status to resolve the problem.
Analysis of all of the Aquarius downlinks showed that all of the IMT files have offsets to the first valid synch word, and none of the CGSS files do.
Tuesday 8/16:
The first phase of antenna deployment for the reflector was performed at 14:35.
The analysis tool was decoupled from the AQL1AGENQL recipe to allow the L1AQL files to be available more quickly. Once archived in the DB, the L1AQL files are scheduled for processing with a new recipe, AQL1HKT, to produce the telemetry data, and the advantage with this scheme is that those streams can run in parallel.
The AQL1AQL recipe was modified to allow failures with the L1-metadata creation to be ignored. This allows the remaining granules to be archived where before the entire stream would fail.
One L1A file failed processing due to the computed beam locations being off the Earth. This occurred during the first phase of the antenna deployment. The problem was traced to the use of the filtered quaternion, which stops updating in the AOCS Safehold mode that was commanded for the deployment. Joel designed an initial fix to issue a warning instead of failure for off-Earth locations. Fred proposed a switch in Safehold to the raw quaternion, which continues to be valid in this mode. This will happen again on 8/17 during the second deployment phase, but Safehold mode should not occur during routine operations.
Wednesday 8/17:
The second phase of antenna deployment for the boom was performed at 14:57.
Joel delivered an update the l0gen_aquarius program that includes the byte-shift value in the L0-info file.
The arbitrary 3-day delay for making the daily GSFC merged HKT files was replaced with a scheme that checks a day for full HKTMST coverage. As soon as the day is covered by HKTMST, the merge processing is scheduled.
One downlink file, CGSS_20110817_115600_10020110816234419_SACD_RawAQ.bin, failed processing yesterday. The downlink web page indicated 82,318 checksum errors. Analysis of the file showed that there were multiple data dropouts that caused the synch word to be lost partway into the file and all subseqent blocks to fail the checksum check.
Thursday 8/18:
Fixed a bug in the logic that selects HKTMST files for input to the HKT-merge process. The logic was failing to select files that completely contained the day being merged because it was looking only for those files that started or ended on the day. This only affected the merged-HKT files for 8/16 and 8/17, and those files were regenerated after the fix was made.
A second downlink file, CGSS_20110818_094500_10020110818092834_SACD_RawAQ.bin, reported a large number of checksum errors. In this case, the cause was found to be a data gap a few blocks into the file that causes the synch word misalignment. Joel delivered a new version of l0gen_aquarius that handles downlink files that have numerous checksum errors. This allowed the downlink file to be processed successfully.
A problem with the VDC wrapper for the mk_aquarius_ancillary_data program caused 30 AQL2GENQL streams to fail where there was only one Hycom Salinity file available. John updated the VDC script so that the script uses a single salinity file as the second file if only one is available.
Monday 8/22:
A gap was observed in the sequence of orbits for three consecutive downlinks on Saturday, 8/20. Data from the same four orbits on 8/19 were found in each downlink, well after they should have been overwritten. Analysis of the downlink files showed that a segment of data was left at the end of the ICDS memory, and that the new data recording was restarted at the beginning of the memory, at about 08:50 UTC. By the last downlink of the day, the new data had begun to overwrite the old data normally. It was determined that this was the time of the science patch upload, and that this was expected to reset the memory pointer.
Tuesday 8/23:
A new limits file, L1B_limits_loose-8-22-2011.txt, was installed for the l2gen_scatter program.
Wednesday 8/24:
The Aquarius commissioning was completed, with the scatterometer enabled to transmit on all three beams at 21:33 UTC.
Joel delivered new versions of l1agen_aquarius (1.04), to fix an error in the block number and frame number metadata; and and l2gen_aquarius (1.03), to add the RFI-filtered Ta fields to the output. The new versions were installed.
Thursday 8/25:
The radiometer DPU OpLUT was commanded from 0 (nominal) to 6 for the period 12:30 to 13:30 UTC.
Joel delivered an updates to l2gen_aquarius (1.04) to correct an error in the rad_samples field calculation, and (1.05) to correct and error in writing the rad_rfi_flag field. The latest version was installed and 51 L1A files were reprocessed.
Updated Aquarius HKT limits (v5) were delivered by the instrument team.
Friday 8/26:
The Aquarius Commissioning CSR was held this morning.
Saturday 8/27:
A scheduled X-band downlink at Matera was missed due to antenna control unit problems.
Monday 8/29:
CONAE conducted an AOCS test with a change to the momentum bias, which reduced the roll and pitch control errors by roughly a factor of 3.
Joel delivered an update to get_acs_times (1.01) to use the CDH OBT instead of the AOCS.
Tuesday 8/30:
A problem was identified with the NEN stations not locking on the S-band signals due to the power-on of SAC-D instruments, which were including strings of zero's in the data stream. This is the same problem that occurred immediately after launch with the service platform data.
Joel provided an explanation for the improvement in the products after 8/25 from the change in the galaxy wind table.
Thursday 9/1:
Fred observed a small pitch spike at 01:46, and the instrument team independently observed a 4-second Aquarius time tag glitch at the same time. Fred related this to a previous AOCS time tag time tag glitch on 8/16, and filed
- a Trac anomaly report.
John implemented a fix provided by Liang for the HKT analysis tool to process a specified number of most recent hours of data from the 3-day HKTMST files. This substantially reduced the delay in posting the telemetry plots from these files.
John submitted all of the Aquarius L0 files for quick-look reprocessing with the latest software updates.
Adam Freedman reported a number of NaNs in the L2 quick-look products.
Friday 9/2:
Cristian Filici distributed a proposed schedule for completing the orbit adjust propulsion maneuvers, from 12 through 30 September.
Joel delivered multiple updates to the L2 processing software (1.06), documented in Trac tickets #30, #33, #34, and #35.
Doug Vandemark and Thomas Meissner reported bad navigation fields in one L2 file for 1 September.
Frank Wentz distributed a weekly salinity map produced with a modified version of the radiometer science software.
Emmanuel Dinnat reported that the NaNs in the L2 products (see previous day) occur only at the very beginning and end of the time range, or before the radiometer power-on. This problem is a characteristic of the L2 quick-look processing, which does not remove the 10 minutes of padding at the start and end of the L1 files, and should not occur in the standard processing.
Tuesday 9/6:
Both morning ETC X-band downlinks were lost due to a ground station anomaly. All of the affected data were included in the Matera downlink. The anomaly was documented in Trac ticket #38.
Joel diagnosed the bad navigation (see 9/1) as a problem with blocks within a half-second of the day boundary and delivered a fix (Trac ticket #36). He also delivered an update to chance the missing calibration temperature values from NaN to -9999 (Trac ticket #37).
Lee Poulson delivered an update to the scripts and TLM Viewer pages.
Wednesday 9/7:
Frank Wentz delivered instructions, new routines and code fragments for implementing his latest updates to the radiometer science software. Joel opened Trac ticket #39 and began work on the implementation.
The last daily Commissioning Phase Cal/Val telecon was held. The weekly telecons will be resumed each Tuesday.
The first merged Level-1A files were generated for internal review.
Updated Aquarius HKT limits (v5a) were delivered by the instrument team.
Thursday 9/8:
The transfer time for the morning X-band downlink at ETC was observed to be much slower than usual. This was reported to CONAE.
A telecon to begin discussion of the JPL-to-GSFC transition was held by Yuhzyen Shen. It was agreed to continue these discussions at the weekly staff telecons.
In discussion with Gary Lagerloef and Simon Yueh, it was agreed to start the weekly cycle on 25 August at 00:01, the first South Pole crossing on the first day after the completion of the instrument commissioning.
Fred documented another instance of a 4-second time tag glitch in the AOCS OBT and the Aquarius block time tag.
Friday 9/9:
Joel updated the Level-1a processing code as follows: l1agen_aquarius 1.05 (Sep 9 2011 09:38:26) l1amerge_aquarius 1.01 (Sep 9 2011 09:38:38)
These include a few changes to the L1A metadata and setting the Cycle,Pass number to start on Aug 25 for the merged L1A files as described in the previous day's entry.
All the Aquarius Level-1a files for both the operational and quick-look recipes were regenerated and the routine production of the Level-1a merged files was started and the check for new orbits to merge was put in place with a repeat check every hour. These files are now available on the data archive page at:
http://oceandata.sci.gsfc.nasa.gov/Aquarius/L1/2011/
Liang opened Trac ticket #40 to update the Aquarius HKT alarm limits.
Saturday 9/10:
Joel produced the first 7-day SSS map file using the latest software updates.
Sunday 9/11:
CONAE attempted an inclination adjust maneuver, but the burn was cancelled due to unavailability of the Malindi ground station. The 90-degree yaw maneuver was performed starting at 14:26.
CONAE commanded a momentum bias change immediately after the yaw maneuver, reducing the variation in roll and pitch.
Dalia raised a concern about the temperature increase shown by the SSPA during the yaw maneuver, which brought it within a few degrees of an onboard threshold (45 degrees) that would have shut down the SSPA. She recommended that the threshold be raised to 50 degrees for any future yaw maneuvers. The inclination adjust maneuver planned for Monday was postponed to allow the threshold change command to be tested on the EM. The hex value corresponding to the new threshold was to be provided by INVAP, since this is a S/P command.
Monday 9/12:
The command to raise the SSPA temperature threshold was successfully tested on the EM. This command was added to the Activities Plan for the inclination adjust maneuvers.
Joel completed the preliminary and internal validation of the updated the Level-2 processing code based on the delivery on Wednesday and committed the code to the processing system. The details of the implementation are as follows:
Implementation of the V1.1 SSS algorithm in the Aquarius Data Processing System
- ==================================================
The coeff_loss_v3.txt, apc_matrix_v1.dat, and fd_dtb_dwin_run5.dat files were first converted to HDF5. The latter file is read by the initialize_static_data() and specified with the "RAD_DTBDWIN_FILE" parameter.
The fd_sss_vpol.f routine was added to the libaqsss.a library archive file. The wentz_dems() code was added to the fd_water_refl.f file.
The new algorithm and data files are run by specifying the following l2gen_aquarius parameters:
- rad_apc_file=$OCDATAROOT/aquarius/radiometer/apc_matrix_090711.h5 rad_dtbdwin_file=$OCDATAROOT/aquarius/radiometer/dtb_dwin_090711.h5 coeff_loss_file=$OCDATAROOT/aquarius/radiometer/coeff_loss_v3.txt sss_algorithm=VPOL_ONLY
Note the the previous SSS algorithm can also be run using:
- rad_apc_file=$OCDATAROOT/aquarius/radiometer/apc_matrix_051211.h5 rad_dtbdwin_file=NONE coeff_loss_file=$OCDATAROOT/aquarius/radiometer/coeff_loss_v2.txt sss_algorithm=PRELAUNCH
New version of l2gen_aquarius on the trunk.
- l2gen_aquarius 1.06 (Sep 9 2011 13:21:45)
This incorporates the updated V1.1 SSS algorithm delivered by Frank.
The suite parameter "V1.1" should be used.
This code has been put into production for the quick-look stream and all the level-2 data files have been reprocessed. John has implemented the optimal ancillary data wait period in the quick-look processing which means that while the level-1a files will still show up very soon after the downlinks, the processing of the level-2 files will wait until the optimal ancillary data has been found or when 3 days after data collection has been reached and then it will use the best it can find.
The newly reprocessed level-2 files and all future quick-look files processed with this version of the software are now available on the data archive page at:
http://oceandata.sci.gsfc.nasa.gov/Aquarius/L2/2011/
Tuesday 9/13:
CONAE attempted an inclination adjust maneuver, but the burn was cancelled due to lack of telemetry during the Malindi pass. The 90-degree yaw maneuver was performed starting at 15:22.
Alex Fore delivered source updates (Trac ticket #42) to correct the setting of the scatterometer RFI flags, and requested that this change be made before the standard Level-2 processing was started. Joel began work to implement and test the changes.
The Cal/Val telecons were resumed on a weekly basis. The results of the latest science software updates were reviewed at this week's telcon. A special telecon was scheduled for this Friday to make a decision about whether to being standard processing with the latest software.
Joel tested the use of the radiometer Unusual Brightness Temperature flag to mask data collected during the OpLUT change on 25 August.
Wednesday 9/14:
Joel and Fred found an error in the documentation of the radiometer Unusual Brightness Temperature flag. The original flag description was based on the expected Tb, but the software actually computes expected Ta. Frank Wentz confirmed that the test should be performed on Ta.
Thursday 9/15:
CONAE performed a 180-degree pitch maneuver starting at 10:03, in preparation for the altitude reduction maneuver planned for Friday.
Alex Fore delivered an updated scatterometer limits table (Trac ticket #43) and additional RFI flag updates (Trac ticket #44).
Friday 9/16:
Daniel Caruso provided an updated IOC Calendar showing the scheduled for the remaining orbit adjust burns and SAC-D instrument commissioning activities.
CONAE attempted an altitude reduction maneuver, but the burn was cancelled due to a ground station contingency. The 180-degree pitch maneuver was performed starting at 09:36.
A special Cal/Val telecon was held today. The Science Team approved the start of standard processing (merged L1A files and L2 files) with the latest software.
Joel delivered all of the scatterometer software updates to the ADPS for testing (Trac ticket #45). John installed the updates, and the entire group reviewed sample products. The L2 quick-look files were regenerated with the updates.
Saturday 9/17:
CONAE performed an altitude reduction manuever at 10:27 UTC today. The 180-degree pitch maneuver was performed starting at 10:06.
An Aquarius downlink was not received for a scheduled Matera pass.
Monday 9/19:
An Aquarius downlink was not received for a scheduled ETC evening pass.
Norman noted that salinity values were showing up in odd locations on 9/17. Fred determined that this resulted from the data processed during the 180-degree pitch maneuver.
Liang requested that all of the HKT files be reprocessed with all of the latest updates to the analysis tool and limits.
Tuesday 9/20:
Norman noted that a few instrument HKT fields were still hitting red limits with the new configuration. Liang followed up with the instrument team to get the correct limit settings.
Norman also noted that some Aquarius orbits with 180-degree pitch maneuvers were missing navigation information. Fred followed up with a summary of the effects of the maneuvers on navigation, and proposed an attitude flag in the L2 files.
Norman completed the implementation of the Aquarius L1 and L2 products in the ODPS Browse and Order system. The L2 products are restricted from distribution until the First Light image is announced by Gary Lagerloef.
Final plans for the Aquarius First Light announcement were made at the weekly staff telecon.
Wednesday 9/21:
Norman installed updated HKT limits from Liang for the web site.
The first meeting of the Aquarius SOCB was held.
Thursday 9/22:
Gene announced the availability of standard Level-2 products to the team, in preparation for the First Light press release.
Fred updated the Level-2 product format specification.
Joel implemented arctangent scaling in smigen (4.2.6) for the SMI salinity products (Trac ticket #49).
Monday 9/26:
CONAE provided an explanation from Matera for the missed downlinks over the weekend: network problems due to thunderstorms.
CONAE planned an inclination adjustment maneuver, but the burn was cancelled due a ground station contingecy. The 90-degree yaw maneuver was performed starting at 21:19.
Thomas Meissner reported inconsistencies in the orbit numbers stored in the products. Fred investigated and found that the orbit number was 1 too low in roughly half of the products. Joel opened Trac ticket #60 for this.
Norman installed the latest HKT alarm limits for the forward processing.
Tuesday 9/27:
CONAE performed an inclination adjustment manuever at 22:15 UTC today. The 90-degree yaw maneuver was performed starting at 21:48.
Joel delivered a fix to l2gen_aquarius (1.13) for the abnormal termination and missing navigation fields during the pitch maneuvers, reported as Trac ticket #50. John regenerated the affected L2 products from 15 and 16 September.
Fred developed a revised yaw angle calculation to handle both the 90+ degree yaw maneuvers and the -180 degree pitch maneuvers (Trac ticket #59). This update was made to both the processing software and the analysis tool.
Simon Yueh requested that quick-look processing be terminated.
Wednesday 9/28:
Joel delivered a fix to to l1amerge_aquarius (1.02) to correct the roundoff in the orbit number calculation, and to gen_aquarius_att (1.02) to implement the revised yaw angle calculation.
Thursday 9/29:
Fred performed analysis of the orbit that showed that as of Monday it was now very close to the weekly repeat cycle, within 1 second. There was still a small residual longitude drift, but the latest inclination maneuver had not yet been performed. He sent this information to CONAE.
Friday 9/30:
Fred proposed a radiometer flagging scheme for off-nominal pointing in the Level-2 data products. The scatterometer flags already include this. He wrote Trac ticket #64 to implement this.
Saturday 10/1:
CONAE performed the final inclination adjust maneuver at 22:29:30. The yaw maneuver started at 22:03.
Sunday 10/2:
A problem was reported by CONAE with the timing of the Aquarius downlink commands starting on Monday 10/3. The problem was traced to changes in the allocated contact times after the initial ARs were delivered, which resulted from the orbit adjust maneuvers. The ARs were regenerated in time and delivered to CONAE.
Monday 10/3:
Lee Poulsen approved the implementation of the updated Aquarius HKT limits in the analysis tool. All of the HKT files were submitted for reprocessing.
Tuesday 10/4:
CONAE performed a pitch/yaw maneuver as for an orbit raising burn, starting at 23:39.
At the weekly Cal/Val telecon, Frank Wentz proposed a correction to the Ta values based on the expected Ta from the HYCOM data. Fred responded the next day with a proposed implementation within ADPS.
Thursday 10/6:
CONAE planned an orbit raising maneuver, but the burn was canceled due a ground station contingency. The pitch/yaw maneuver was performed starting at 22:57.
Norman activated the automatic telemetry alarm notifications. The first alarm was issued for the canceled o
Friday 10/7:
CONAE performed an orbit raising calibration maneuver at 21:57. The pitch/yaw maneuver started at 21:48.
Saturday 10/8:
Two Aquarius downlinks from the previous day were not processed. John traced the cause to "no route to host" messages during the data acquisition. The files were subsequently acquired and processed, and the problem has not been repeated.
Tuesday 10/11:
CONAE planned an orbit reduction maneuver, but the burn was canceled due a ground station contingency. The pitch maneuver was performed starting at 09:23.
Trac ticket #40, to implement updated HKT limits, was closed. It was originally opened on 9/9.
Wednesday 10/12:
At the weekly Calibration telecon, it was decided that the science software updates would be delivered by early next week. The data will be reprocessed with the updates in a evaluation mode in preparation for the Cal/Val workshop in November.
CONAE performed an orbit reduction calibration maneuver at 18:25. The pitch maneuver started at 18:03.
Friday 10/14:
CONAE republished a number of SAC-D HKT files from launch through 19 July to fill in gaps in the previously delivered files. The files were acquired by ADPS and processed with the Analysis Tool.
Joel implemented a radiometer calibration correction developed by Jeff Piepmeier and Liang.
Saturday 10/15:
CONAE performed the final orbit adjust maneuver of the service platform commissioning phase. The orbit raising maneuver was performed at 12:25 and the pitch/yaw manever started at 12:17.
Monday 10/17:
Thomas Meissner delivered a radiometer science software update to implement the Ta bias correction. Gene opened Trac ticket #70 for the change and Thomas entered the details.
Tuesday 10/18:
In response to the announcment of the Mission Anomalies Walkthrough later this week, Fred delivered a report on the AOCS OBT anomalies to CONAE.
Alex Fore added updates to the scatterometer RFI flagging and wind retrieval algorithm to Trac ticket #70.
Wednesday 10/19:
Chris set up the Trac server for the joint CM process with CONAE.
Thursday 10/20:
An attitude transient of about 3 degrees in roll was observed in the AOCS data at about 1:00. The time corresponded to Moon interference in the star tracker.
CONAE held the Mission Anomalies Walkthrough and the Orbit Maneuvers CSR.
Joel completed the implementation of the radiometer Ta bias correction in the L2 software and delivered two test files to Thomas Meissner for verification.
Sunday 10/23:
Another 4-second OBT anomaly was observed at about 12:55.
Monday 10/24:
John produced sample Level-2 files with the radiometer Ta bias correction (V1.2) for internal examination.
ACCS and CONAE began the routine (weekly) planning cycle for the first full week.
CONAE reported that the S-band tranmitter unexpectedly powered off just after the start of the ETC pass at 21:38.
Fred reported a recurrence of the 4-second AOCS OBT glitch yesterday, and requested that Anomaly #208 be re-opened to address this. He subsequently determined that this has occurred at least 6 times since mid-July. This information was added to Trac ticket #32.
Tuesday 10/25:
Joel and Alex Fore discovered a difference in the scatterometer software output between swdev (32-bit) and his workstation (64).
Fred updated the Level-2 format specification to include the Ta bias correction attributes and the internal calibration fields added by Joel.
Wednesday 10/26:
Jeff Piepmeier presented an alternative approach to the Ta bias correction at the calibration telecon, using the internal calibration data (e.g., CND) as the basis for the correction.
Alex and Joel found two bugs in the scatterometer software that accounted for the environment differences. This was documented by an addition to Trac ticket #70.
Thursday 10/27:
Gene reported that the evaluation products for V1.2 (Ta bias correction) had been generated for orbits 1164 through 1716, as requested by the Science Team.
Friday 10/28:
CONAE announced a plan to change the service platform bus voltage from 33V to 31.5V, to be implemented the week of 7 November.
Monday 10/31:
The Aquarius Instrument Team raised several questions regarding CONAE's plan to reduce the bus voltage. It was also raised at the SOCB telecon today. Marcelo scheduled a telecon for Wednesday afternoon to discuss this.
Jeff Piepmeier reported an uncorrected gain jump in one radiometer channel on Day 272 (29 September).
Tuesday 11/1:
CONAE reported a high temperature emergency on NIRST and attempted to schedule two NEN passes. This was not successful, and the contingency procedure was executed during the first ETC pass.
Wednesday 11/2:
CONAE held a telecon to discuss the planned reduction in the SAC-D bus voltage.
Monday 11/7:
CONAE implemented the new allocated pass files, which were successfully ingested by ADPS.
Tuesday 11/8:
Thomas Meissner delivered an updated Ta bias table extending coverage to Day 2011299 (26 October) at 06:05.
Wednesday 11/9:
The Aquarius PLAR was held at JPL.
Dalia revived her prelaunch concerns about the level of checking performed on the full pass scripts by ACCS. Gene had a discussion with Simon Collins on this issue and Liang and Simon will talk about it in more detail at the upcoming Cal/Val workshop in Pasadena.
Thursday 11/10:
The V1.2 L2 and L3 products were regenerated with the latest Ta bias table.
Norman produced separate composite images for the ascending and descending data using the latest V1.2 products, showing marked differences between the two sides of the orbit and provided this to the science team.
Friday 11/11:
CONAE performed the first of two orbit maintenance maneuvers at 15:10. The pitch/yaw maneuver started at 15:01.
Tuesday 11/15:
First day of the Cal/Val workshop at JPL.
Wednesday 11/16:
Second day of the Cal/Val workshop at JPL.
CONAE performed the second of two orbit maintenance maneuvers at 11:48. The pitch/yaw maneuver started at 11:39.
Thursday 11/17:
Third day of the Cal/Val workshop at JPL.
Friday 11/18:
Thomas Meissner delivered an updated Ta bias table extending coverage to Day 2011314 (10 November) at 03:17.
Monday 11/21:
The V1.2 L2 and L3 products were regenerated with the latest Ta bias table.
Friday 11/25:
There were two short gaps in the Sun angle computed from the SHO Sun vector in the HKT data, at about 05:10 and 07:10. These were later determined by Liang to correspond to the spacecraft passages through the solar eclipse.
Monday 11/28:
The Aquarius L2 data processing stopped over the long weekend. This was caused by an interruption in the production of sea ice data by NCEP. Investigation was started of alternate sources of data.
Tuesday 11/29:
The PO.DAAC announced the availability of the Aquarius/Salinity Forum.
Wednesday 11/30:
Liang initiated the implementation of science data metrics based on the presentation by Adam Freedman and Jeff Piepmeier at the Cal/Val Workshop.
Thursday 12/1:
The Aquarius Project Office transitioned from JPL to GSFC.
Liang initiated the addition of first-order derivatives of selected HKT fields to the analysis software and web site.
Sunday 12/4:
The NCEP sea ice data availability resumed. The gap prior to this date will be filled in.
Thursday 12/8:
The NCEP sea ice data availability was halted again on 12/6 and resumed today.
Friday 12/9:
The V1.2 L2 and L3 products were regenerated with the latest Ta bias table from Thomas Meissner to Day 2011333 (29 November) at 05:37.
Sunday 12/11:
A problem occurred during an ETC X-band contact in which the SAC-D antenna was shadowed by another antenna, interrupting the downlink. There was no net loss of data.
Wednesday 12/14:
CONAE performed a collision avoidance/orbit maintenance orbit maneuver at 15:51.
- The pitch/yaw maneuver started at 15:43.
Thursday 12/15:
Liang implemented updates to the HKT analysis tool to include the AOCS OBT and also to compute differences of time tags and block numbers.
Thursday 12/22:
Joel delivered two software updates to detect and correct the 4-second time shifts: l0gen_aquarius (1.04), for the Aquarius science blocks times, and gen_aquarius_att (1.03), for the AOCS OBT. The first of these writes an entry to the log file that is also used to generate a notice on the Aquarius downlink monitor web page. This was documented in an update to Trac ticket #32.
All of the Aquarius Level-1A files affected by the 4-second shifts were reprocessed with the above software updates.
Friday 12/30:
A number of observatory components were unexpectedly powered off at about 10:00, including both transmitters, MWR and ICARE. Aquarius was not affected. CONAE began an anomaly investigation.
Sunday 1/1:
Two roll angle transients were observed that coincided with Moon interference in the star tracker. The first, starting at 01:38, showed an onboard roll of about 7 degrees, and the second, at 13:04, about 6 degrees. There were also smaller transients in pitch and yaw. In both cases, there was attitude motion lasting about 10 minutes.
Friday 1/6:
A telecon was held with CONAE to discuss the previously known time tag anomalies
- -- the 4-second shifts and the AOCS OBT fractional second -- and the recent
Moon interference AOCS control anomaly.
Thomas Meissner delivered an updated Ta bias table, covering data through day 362 (28 December 2011).
Thomas also opened Trac ticket #92 to deliver the radiometer software update for the bias determination and application.
Monday 1/9:
Alicia reported an Aquarius software message that was not defined in the FSW User's Manual. Dalia and Adam responded that it is not a problem.
Friday 1/13:
The Version 1.2 Level-3 products were made available for public distribution.
Wednesday 1/18:
The first quarterly Aquarius Telemetry Workshop was held on. The purpose of these workshops is to review the instrument performance based on the ongoing analysis of the housekeeping telemetry (HKT). The participants included the Aquarius Instrument Engineering Team at JPL and GSFC and the Aquarius Ground System Team at GSFC. In this inaugural workshop, all of the teams described their methods that have been developed for routine analysis of HKT data, displays of instrument performance statistics, and issuing alarms, and showed example results. There were also requests for and discussion of additional routine analyses. The workshop concluded with a review of the anomaly detection and reporting process.
CONAE distributed a preliminary report on the unexpected power-off of MWR and ICAR by the RTU.
Friday 1/20:
Joel produced a weekly salinity map using the implementation of Jeff Piepmeier's
- deflection-ratio based calibration correction.
Monday 1/23:
Simon Yueh delivered a software change request for the Combined Active-Passive (CAP) algorithm.
Tuesday 1/24:
A solar coronal mass ejection was predicted to produce peak radiation levels today. No effects were observed for SAC-D or Aquarius.
Sunday 1/29:
The Level-1A processing failed for one downlink. Joel determined that this was caused by a series of data dropouts in the X-band downlink that incorrectly triggered the 4-second time tag glitch logic. Fred reported the file corruption to CONAE. The downlink coincided with a S-band (real-time HKT) file that contained only two frames.
Tuesday 1/31:
Norman noted that the upper deployment mechanism (UDM) temperature, after rising monotonically since Aquarius commissioning, appeared to have leveled off at the end of January.
Wednesday 2/1:
Emmanuel Dinnat reported two problems in the L2 data products. The first was occasional Aquarius record times that did not differ by 1.44 seconds (Trac ticket #94). The second was irregularities in the pass and cycle numbers (Trac ticket #95).
Joel delivered the Level-1A processing fix for the Level-1A processing issue discovered on 1/29. The Level-0 file was processed by ADPS without errors.
A telecon was held with the instrument mechanisms engineers to determine whether changes between prelaunch and postlaunch antenna alignments could account for the observed beam offsets. The conclusion was that the uncertainly estimates provided by the engineers are not large enough to account for the offsets.
Friday 2/3:
After an inconsequential GPS glitch triggered a red limit alert, Fred proposed a relaxation of these limits (update to Trac ticket #40). These were implemented by Liang.
Saturday 2/4:
A delay in the posting of Level-1A records on the Aquarius Downlinks page was determined by John to result from delays in the file migration system due to an unrelated data processing request. He resolved the problem by raising the priority of the Aquarius report file migration.
Sunday 2/5:
Joel determined that the Aquarius L2 time issue in Trac ticket #94 resulted from a logic error for times that were almost an exact second. He delivered a fix to the Level-1A software. This will be corrected in the data products during the next reprocessing.
Monday 2/6:
Joel determined that the problem reported with the pass/cycle numbers in Trac ticket #95 had already been fixed in the Level-1A software, but the time periods reported by Emmanuel (September 2011) had not been reprocessed yet. This will be corrected in the data products during the next reprocessing.
Fred delivered a software change request for the ND-based radiometer calibration update to the SOCB.
Tuesday 2/7:
Adam Freedman sent a message summarizing a list of changes that were being implemented for the scatterometer processing.
Wednesday 2/8:
John completed the installation of the ND-based radiometer calibration correction (V1.2DR) and generated the first test products for verification.
Friday 2/10:
The complete mission data set was processed to generate evaluation products for V1.2DR. Norman Kuring produced a montage of global images for every week of the mission. Gene notified the Science Team that the products were available for review.
A special Cal/Val telecon was held to finalize the approach to the pointing correction. It was agreed to move ahead with a single roll/pitch correction for all beams. RSS will need to update the land mask table using the new pointing information.
Monday 2/13:
Joel investigated a problem with a number of orbits that failed SMI processing during the V1.2DR evaluation run and delivered a fix. The SMI processing was re-run and Gene notified the Science Team that the products were available.
Tuesday 2/14:
The Aquarius L1A processing was re-run for the entire mission to correct the orbit number and Aquarius block time errors (Trac tickets #94 and #95).
Thursday 2/16:
CONAE delivered a report on the increase in the S/P panels from the start of the mission.
Fred submitted a Trac ticket (#98) for the ND-based radiometer calibration correction (V1.2DR) and the updated roughness correction delivered by RSS (V1.3).
Friday 2/17:
David Levine submitted a request for the first cold sky calibration to be performed on March 24 (Trac ticket #99).
Tuesday 2/21:
Joel fixed a residual bug in the L1A processing of the Aquarius block times that affected one orbit. This was a follow-up to the fix made for Trac ticket #94. The L1A merged file for this orbit was regenerated.
Joel completed the implementation of the updated roughness correction and delivered a sample file to RSS.
Thursday 2/23:
Joel found and fixed a bug in the V1.2DR implementation in l2gen_aquarius (v1.17). The L2 products were regenerated.
Norman posted the first set of 7-day HKT statistics plots
http://aquarius.nasa.gov/cgi/7daystats
as well as plots of the most recent month's worth of orbit-by-orbit statistics.
http://aquarius.nasa.gov/cgi/orbistats
Friday 2/24:
CONAE raised concerns about the requested 10-minute dwell for the first CSC maneuver and proposed a series of tests to validate the behavior of the AOCS.
Monday 2/27:
Joel completed implementation of the scatterometer-based roughness correction (V1.3) and provided sample results to Thomas Meissner. The results showed frequent failures of the roughness correction. Joel and Thomas started an E-mail discussion of this.
Tuesday 2/28:
A telecon was held between ACCS and CONAE to discuss the joint CM plan and procedures.
The radiometer software update was installed in APDS as V1.2.2.
Wednesday 2/29:
A planned orbit maintance maneuver was cancelled due to problems with the McMurdo ground station. The pitch/yaw maneuver was performed starting at 17:19.
Fred reported that a 4-second glitch on 2/28 was different than all previous events, in that more than a single time tag was affected. The affected time tags spanned 80 seconds for both the Aquarius and AOCS OBT times.
Thursday 3/1:
A series of telemetry alarms was issued after NIRSt was powered off by the S/P due to an overcurrent condition. Liang determined that the relevant HKT field, countLoadsDisconn_OC, is an event counter, not an indicator that will reset. The limit was revised to stop the alarms.
Friday 3/2:
An orbit maintenance maneuver was performed. The pitch/yaw maneuver started at 18:14.
CONAE distributed the CSC validation plan for review. The plan proposed a series of test maneuvers for 12 through 16 March.
Tuesday 3/6:
The Level-2 format document was updated to reflect the new fields for V1.2.2 and distributed.
Wednesday 3/7:
The entire Aquarius mission was reprocessed to Level-3 with the V1.2.2 software and the products were made available on the web site.
Friday 3/9:
A telecon was held with ACCS, CONAE and JPL to discuss the CONAE CSC validation plan. It was decided that the Aquarius instrument team would provide positive notification to proceed after each day's test.
Monday 3/12:
The first CSC test maneuver was performed with a dwell of 1 minute. The maneuver started at 16:50:50 and was completed at ~17:22. The pitch angle overshot the target slightly, to about -183.4 degrees, and had settled to about -182 when the return maneuver started.
The first set of DR correction coefficents was delivered by Joel for forward stream processing (Trac ticket #100).
Tuesday 3/13:
The second CSC test maneuver was performed with a dwell of 2 minutes, starting at 12:43:30 and ending at ~13:16. The pitch angle settled to within 0.5 degree of the target before the return maneuver.
Wednesday, 3/14:
At 08:54 UTC, the SAC-D AOCS mode changed to safehold (SHO). This was determined to coincide with Moon interference in the star tracker. The CSC test maneuver was cancelled, and CONAE used the extra passes to monitor and command the spacecraft. The star tracker was observed to be operating normally, and the AOCS was commanded back to science mode at 15:58 and remained stable in that mode.
Aquarius temperature excursions were observed during the SHO period. All temperatures recovered to normal values after the return to science mode.
Although Aquarius continued normal operation throughout the event, the data was declared as lost due to the sigficantly off-nominal pointing.
CONAE and ACCS agreed to perform the next two CSC maneuver tests (4 and 6 minutes) on Thursday and Friday. Since this meant that the 8-minute test would not be performed, ACCS recommended to the Science Team to change the request for the March 24 CSC from 10 to 8 minutes dwell time, and this was accepted.
Thursday 3/15:
Following an extended E-mail discussion on the loss of salinity retrievals resulting from the use of the scatterometer RFI flag in the V1.2.2 roughness correction, Thomas Meissner proposed a revised implementation of the roughness correction.
The third CSC test maneuver was performed with a dwell of 4 minutes, starting at 16:35:30 and ending at ~17:10.
Friday 3/16:
The fourth CSC test maneuver was performed with a dwell of 6 minutes, starting at 15:30:30 and ending at ~16:09. The pitch converged at the target of -180 degrees after about 3 minutes.
Monday 3/19:
Joel delivered the first routine weekly set of DR coefficients (Trac ticket #100) to be installed for the forward stream.
Joel corrected a bug in the use of the navigation flag in l2bin_aquarius. The affected data products were regenerated.
Fred and Gene met with Melissa Vess, a GSFC SME who has specific flight experience with the Galileo star trackers on SAC-D, to discuss the March 14 AOCS anomaly. Vess followed up with a specific set of recommendations that were forwarded to CONAE.
Thursday 3/22:
Joel corrected a bug in the generation of the navigation flags for the radiometer (Trac ticket #104).
Saturday 3/24:
The first CSC maneuver was performed with a dwell time of 8 minutes, starting at 02:48 and ending at ~03:28.
Sunday 3/25:
Based on a report of a possible conjunction between the SAC-D spacecraft and a piece of space debris on the evening of 26 March, a series of telecons and data reviews were held on 23, 24 and 25 March between CONAE, the CARA group at NASA/Goddard and the Aquarius Project. Although the conjunction at the final review on Sunday was outside of the range where an avoidance maneuver was required, CONAE decided to take advantage of the planning that had already been done and the additional NEN support that had been provided to conduct an orbit adjust/maintenance maneuver at 2012/03/25 22:56:40. This maneuver was conducted successfully.
Monday 3/26:
First day of the Cal/Val Workshop in Santa Rosa
Tuesday 3/27:
Second day of the Cal/Val Workshop in Santa Rosa
Wednesday 3/28:
Third day of the Cal/Val Workshop in Santa Rosa
Thursday 3/29:
Thomas Meissner provided direction and code for finalizing the V1.3 implementation of the radiometer software.
Friday 3/30:
Joel proposed a reorganization of the radiometer flags to make better use of the array elements and free up some spares, in order to accommodate new flags specified by Thomas Meissner. Thomas agreed with the change. David Levine raised a concern about preserving the solar flare flag, although this is not currently set.
Thursday 4/5:
The Aquarius products for the mission were regenerated using the V1.3 science software (Trac ticket #92). V1.3 was also implemented for the forward-stream processing.
Thursday 4/19:
Joel implemented a modification to the radiometer L2 software to exclude the values from the first short accumulation (SA1) from the block averaged (Trac ticket #107). This was based on a presentation by Paolo de Matthaeis at the Cal/Val Workshop that showed that the SA1 data were biased relative to the other SAs, causing a large number of spurious RFI flags. This effect was confirmed by Jeff Piepmeier and others after the workshop.
Friday 4/20:
Joel implmented a fix for erroneous radiometer geolocation values that were being computed during the 180-degree pitch maneuvers (Trac ticket #108).
A telecon was held with CONAE to discuss AOCS issues regarding the Moon star tracker interference, the SHO-SUR transitions during eclipse and the pitch overshoot at the end of maneuvers. In preparation for the telecon, Fred sent a proposal for a change to the pitch maneuver commanding to reduce or eliminate the overshoot.
Monday 4/23:
Two Aquarius software error messages were observed. In reviewing this event, Norman discovered that these did not generate an alert, and corrected the monitoring software.
John generated a set of L2_CAL files with the SA1 exclusion.
Tuesday 4/24:
A meeting was held with Tom Jackson and Rajat Bindlish to discuss the implemenation of the Aquarius Soil Moisture processign software.
Joel generated a set of DR coefficients using the L2_CAl files excluding SA1.
Wednesday 4/25:
In preparation for the eclipse season, CONAE raised the Sun angle threshold for the SUR transition to 65 degrees.
Friday 4/27:
A telecon was held between Goddard and JPL to discuss the procedure for Aquarius power-on and begin planning for a test on the SAC-D EM.
Sandra Torrusio announced that the SAC-D PLAR had been postponed from May 15.
In response to the discussion during Tuesday's Cal/Val telecon, Thomas Meissner circulated a draft schedule for the radiometer software updates through October 1.
Saturday 4/28:
The second CSC maneuver was performed, with an 8-minute dwell starting at 02:56 UTC.
Sunday 4/29:
The SAC-D AOCS mode changed to safehold (SHO) at 05:09. This was determined to coincide with Moon interference in the star tracker. CONAE commanded the AOCS to SCI at 10:55.
Monday 4/30:
CONAE reported that X-band data was not acquired during an ETC pass. No Aquarius data were lost as a result of the missing downlink.
Tuesday 5/1:
CONAE reported that X-band data was not acquired during an ETC pass. No Aquarius data were lost as a result of the missing downlink. Gene raised a concern that this had happened twice in a 24-hour period.
Alex Fore delivered updates to the scatterometer L2 software, including the pointing angle biases.
Wednesday 5/2:
Joel generated a week of L2 files with the SA1 exclusion for analysis by the algorithm team.
Saturday 5/5:
A telemetry alarm was issued for a number of Aquarius software messages. Simon Collins later reported that these were DRAM errors in the RAD6K processor, most likely caused by radiation.
Monday 5/7:
The first eclipse of 2012 was flagged in the SAC-D HKT at 01:53.
