1. ADEOS OPERATIONS STATUS REPORT MEETING
Takao Fukuda
Associate Senior Engineer
Earth Observation Center, Japan Space Exploration Agency
The ADEOS Operations status report (OSR) meeting was held Mar. 19-20, 1998,
at the Earth Observation Center (EOC) in Hatoyama, Saitama, Japan. A total of
49 individuals attended and 21 papers were presented by 10 organizations in
this meeting.
At the opening, JAXA noted that they had anticipated having an annual meeting
for the next ten years, but due to the tragic loss of the Advanced Earth
Observing Satellite (ADEOS), this would be the first and last meeting, and the
loss of ADEOS was especially sad since many users had wanted to use its data.
The purpose of the ADEOS OSR meeting was to focus on the ADEOS routine
operations phase, rather than ADEOS-II matters which were discussed at the
ADEOS-II GDR, so participants could share operations and data processing
experiences and use these experiences as input for ADEOS-II to achieve more
reliable operation. The meeting consisted of reports from JAXA, sensor
providers, and NASA Ground Networks (NGNs) including lessons learned.
(1) ADEOS BUS OPERATIONS
(2) ADEOS MISSION OPERATIONS STATUS
(3) ADEOS CALIBRATION AND VALIDATION
(4) ADEOS SCIENCE PROGRAM
(5) NGN OPERATIONS STATUS
(6) ADEOS FAILURE
(7) OPERATIONS STATUS OF SENSOR PROVIDERS
On Mar. 19, the following presentations were provided.
<1> ADEOS BUS OPERATIONS
The JAXA Tracking and Control Center (TACC) reported on the ADEOS bus
operations, covering its history and status as well as improvement of
ADEOS-II.
JAXA explained that it had observed an attitude fluctuation in mid December
and then investigated the telemetry. The fluctuation itself did not exceed the
specification. However, the new type of fluctuation was bad for the reaction
wheel, giving it many zero crossings, which is not good for the reaction wheel
life. JAXA therefore changed the parameters on 14 May 1997 to reduce the zero
crossings.
<2> ADEOS MISSION OPERATIONS STATUS
RESTEC presented the MMO Operations Report, covering planning results,
significant events, data losses, simulations, and lessons learned. Data losses
occurred for OCTS and mid-rate AO sensors due to MDR restrictions. Some data
was also lost due to a scheduled power outage at EOC (for safety inspections)
on 5 and 12 July 1997.
The receiving and recording system, including its function and scheme,
operations history, results of Level 0' and Level 0 processing, results of
Level 0 reprocessing, anomalies and their resolution, termination of the ADEOS
recording system, and operations lessons learned were also reported.
Operations lessons learned
(1) DIR is very sensitive to dust.
(2) EOC did not have an inspection system for D1 and Level-0 data; one is now
planned.
(3) Data gaps occurred between Level-0 tapes and Near-Real-Time (NRT) data
because the EOC system could not properly reprocess NRT data or could not
easily process low-bit-rate data from ASF's or WFF's raw data.
(4) Large file sizes caused ILAS Level-0 on-line data transmissions to stop
frequently.
RESTEC made a presentation on AVNIR and OCTS, covering the operations
status, processing results, problems and corrective actions, and operations
lessons learned.
Operations lessons learned
(1) LAC requests exceeded the design, influencing processing.
(2) The three subsystems composing the ADEOS processing system (OCTS NRT data
processing, OCTS routine data processing, and AVNIR data processing) had
no back-up function. OCTS NRT data was especially affected.
RESTEC presented the EOIS operations report, consisting of NRT data
transmission results, transmission error abstracts, FTP server access results,
shipping and distribution status, and lessons learned.
Lessons learned
(1) The Data Distribution system (DDS) did not have redundancy; transmission
had to stop for maintenance. ADEOS-II will have a redundanct system.
(2) Better confirmation of shipped and received dates is needed to prevent
data being lost in shipment.
JAXA EOC's EOIS reported on the EOIS user interface software (EUS),
describing EUS functions, requests from users to make EUS more flexible and
convenient, and the schedule for releasing EUS enhancements.
<3> ADEOS CALIBRATION AND VALIDATION
JAXA EORC reported on the calibration and validation of the Ocean Color
and Temperature Sensor (OCTS), first expressing their gratitude to the OCTS
team and supporting scientists. External data used for calibration included
underflights and in-situ measurements. A large amount of data was collected,
but only a small number of match-up data sets could be produced. For example,
1347 in-situ chlorophyll (CHL)-a measurements were made, but only 17 of these
met the match-up data set criteria (cloud-free, within 24hrs, and within
800m).
For normalized water leaving radiance (nLw), 496 in-situ measurements were
made with only 11 meeting the match-up criteria. For the Sea Surface
Temperature (SST) products, the team developed a filter to eliminate stripes.
EORC concluded by noting that OCTS version 3 products were much better than
version 2 products, that CHL-a can be estimated with an accuracy of 68%, that
SST can be measured with an accuracy of 0.70K, and that geometric accuracy is
1.3 km. These products are available for scientific and application uses.
Desired future developments include a single-detector sensor (multiple
detectors cause striping), a sensor with appropriate response to high-input
radiance, and higher stability.
Also reported were activities on calibration and validation of AVNIR,
Advanced Visible and Infrared Imaging Spectrometer (AVIRIS) underflights,
destriping, image quality, and actions. As of 5 March 1997, 9,470 panchromatic
and 17,388 multispectral images had been acquired. JAXA and NASA JPL
conducted joint AVIRIS calibration and experiments on 7 November 1996, and on
28 February and 4 March 1997. AVNIR products have two types of stripes,
periodic and nonperiodic. Periodic stripes can be removed by tuning software;
nonperiodic stripes depend on target brightness and are not removed.
In conclusion, calibration using AVIRIS underflights has been completed, and
image quality is almost sufficient for applications use. However, there is
some vertical noise in both multispectral and panchromatic images, and
destriping software is being developed. Future actions include installing
AVNIR version 2 software to suppress stripes and opening the calibration
factor to the public.
<4> ADEOS SCIENCE PROGRAM
In the ADEOS Science Program, calibration and validation have been
completed for all instruments except IMG and POLDER. Algorithms have been
completed for NSCAT and TOMS. Three issues remain for AVNIR and OCTS:
Attitude retrieval from AVNIR navigation band, ADEOS attitude data
uncertainty,
and geometric accuracy assessment of OCTS data. A calibration campaign was
conducted at Edwards AFB, California; an atmosphere validation campaign was
conducted at Fairbanks and Barrow, Alaska; and an ocean validation campaign
was conducted off-shore of Sanriku in the North Pacific. Results were
presented at workshops in Tokyo 1-5 December 1997 (Arctic Ozone Workshop) and
11 March 1998 (Sanriku Campaign Workshop). Campaign data is available at the
EORC web site. New products expected from ADEOS sensors include weekly global
ocean chlorophyll distribution, aerosol distribution over the ocean,
temperature and water vapor vertical profiles, greenhouse gas distribution,
vertical distribution of ozone in the ozone hole, high-resolution (spatial and
temporal) sea surface wind vectors, UV absorptive aerosol distribution, cirrus
cloud distribution, and water vapor distribution. In addition, ADEOS sensors
provided some unexpected results such as detecting the Arctic ozone depletion
in the spring of 1997, onset of the 1997-98 El Nino, ice and vegetation
monitoring from NSCAT, nitrate retrieval from OCTS, and oceanographic study
using chlorophyll and SST from OCTS.
On the second day, the following presentations were provided.
<5> NGN OPERATIONS STATUS
Ground receiving systems are located at the Alaska Synthetic Aperture
Radar (SAR) Facility (ASF) which supports 10 passes per day and at Wallops
Flight Facility (WFF) which supports two passes per day. ASF will be able to
support ADEOS-II with a 10-meter and an 11-meter antenna; WFF, with an
11-meter antenna. Several lessons were learned. Tracking a real satellite was
useful during testing. Link tests using the spacecraft's test signal generator
are very useful for link evaluation and performance between measurements.
Exchanging compatibility test tapes is useful for confirming compatibility
across different locations. Bar coding greatly simplifies tape management
tasks. If Reliability Reports are provided more quickly, fewer tapes need be
used. The DRN/RCN exchange should be continued. Activating a carrier before
scheduled acquisition speeds acquisitions. WFF should have many blank D-1
tapes available. Email proved to be a good alternative when ADEOS network
failures occurred. Email Discrepancy Reports reduced paperwork and
facilitated database development for problem tracking. Perform end-to-end
testing as early as possible. Provide simulated spacecraft data to NGN as
early as possible.
<6> ADEOS FAILURE
JAXA EOSD reported on the ADEOS accident investigation. (Refer to the
"REPORT ON THE CAUSE INVESTIGATION FOR ADEOS ACCIDENT" for details.) On 1
July 1997, immediately after the accident, JAXA established a Task Team to
investigate the cause. The investigation lasted about three months and focused
on the solar paddle subsystem (PDL) and the attitude orbit control system
(AOCS). The solar paddle consists of 50 blankets joined by two types of
hinges,
a pin hinge and an adhesive hinge. During the second deceleration maneuver (28
August 1996), unusual temperatures applied an unexpected tensile force to the
blankets, partially breaking a pin hinge where the force was concentrated.
Subsequent thermal cycling as ADEOS circled the Earth 18 times per day caused
the hinge to break completely (17 December 1996), after which the hinge was
supported by the solder connections for electric power circuits. The solder
connections began breaking down on 23 June 1997 and were completely broken on
30 June 1997, leading to the loss of all satellite functions. The excessive
tensile force was assumed to be due to the blanket's low temperature thermal
expansion coefficient, which was twice as large as estimated during
development.
<7> OPERATIONS STATUS OF SENSOR PROVIDERS
NASA JPL made a presentation on the NSCAT status, covering the instrument,
science data processing, and lessons learned. The NSCAT instrument was very
stable and operated without any problems, providing excellent wind
measurements
and sigma0 measurements for land and ice studies. Science data will be
reprocessed using updated calibration tables and model functions from March to
May or June 1998. Lessons learned include the need for better planning of
special instrument events, more timely generation of REQQ, and shorter
turnaround of commanding cycle. The need for continuous, automatic telemetry
monitoring was stressed. In particular, it was noted that important instrument
and spacecraft parameters were not being monitored often enough or closely
enough and were not trended enough to detect when significant changes or
anomalies had occurred.
GSFC presented the TOMS status. TOMS provided the first global total ozone
image on 13 September 1996. On 7 October 1996, near-real-time TOMS data became
available on the TOMS web site. The instrument performed well with no
anomalies until the ADEOS failure in June 1997. The major concern was the many
gaps in near-real-time data (5.1% missing). Special Level 0 processing was
able to recover 112 of 167 complete orbits and 26 of 41 partial orbits.
Lessons learned include the need to begin interface definition early, to
perform system tests for all ground elements using actual and anticipated data
loads, and to develop flexible data acquisition plans.
NOAA reported that Level-2' NSCAT products were routinely sent to JPL and
JAXA and that 90% of the 1033 files sent from 16 April to 29 June 1997 were
delivered in less than 4 hours. For OCTS, 40 Level 0 files (average size 75
Mbytes) were sent from HEOC to NOAA between 1 May and 16 June 1997, and 251
raw datasets were pushed from WFF and ASF between 1 June and 29 June 1997.
Lessons learned include the need for a simulated input dataset for testing
NSCAT and the need for coordinating documentation (in this case, providing the
OCTS instrument specification and test bench measurements).
CNES explained the POLDER Processing Center, covering the ground segment
architecture, chronology of significant events, processing status, and
operation team. The Center programs and checks the Polarization and
Directionality of the Earth's Reflectances (POLDER) instrument, produces and
delivers higher level products from Level 0 data, and archives the mission
data. POLDER's image quality validation phase began on 23 September 1996 and
was completed on 27 May 1997. Level 2 and Level 3 product delivery is set to
begin in July 1998. CNES commented on the excellent operational relationship
between CNES and JAXA, and noted that all problems encountered were resolved.
In regard to the POLDER Instrument Control Programming System, in general,
most files, except housekeeping data telemetry files, were processed with no
problems. No instrument anomalies occurred during TACC monitoring of the
real-time telemetry data, although CNES opened 19 Instrument Anomaly reports
(none of which had mission consequences). The Operation Coordination Letter
(OCL) proved to be a very satisfactory means of communication.
Japan Resources Observation System Organization (JAROS) presented the IMG
Instrument Operation Report (see the handout with the same title for details).
IMG performs global monitoring of greenhouse effect gases with a normal cycle
of 4 days of observation followed by a 10-day pause. IMG suffered a traveling
mirror alignment anomaly in which there was a permanent alignment shift that
degraded all data and a fluctuating alignment angle that degraded data
quality.
This anomaly considerably reduced the volume of valid data obtained.
Nonetheless, and in spite of the shortened operational period due to the ADEOS
loss, valued observation data was obtained.
The Earth Remote Sensing Data Analysis Center (ERSDAC) reported the IMG
data processing status, covering the data volume, product quality, data
processing, version upgrades, product distribution, system reduction plan, and
lessons learned. (Refer to the handout "IMG Data Processing Status" for
details.) By the end of August 1997, 65,956 units of Level 0 data had been
received from JAXA EOC, of which 46,088 units were processed by the IMG Data
and Information System (IMGDIS). (A "unit" of IMG data consists of two
calibration observations and six ground observations.) Most of the data loss
was due to no data being received at the ground stations or poor data
reception. Some data was also lost in ring-shaped regions around the ground
stations (see Figure 1 of the handout). IMG data processing algorithms for
levels 0, 1, and 2 data were upgraded. The Level 1 algorithm was recently
upgraded a second time, and data reprocessing started on 9 March 1998.
Preliminary products were released to PIs and internal users on 16 January
1997; delivery to other researchers began on 27 June 1997. IMGDIS operations
will have to be cut back due to the ADEOS loss. The current plan is to
maintain full functions until the end of August 1998; after that, only the web
function will be maintained (until March 1999). Lessons learned include the
needs to have flexible sensor operations, to immediately change the
observation plan in response to special events such as volcanic eruptions or
large fires, and to quickly deliver Level 0 data and IMG standard products.
The National Institute for Environmental Studies (NIES) presented the
status of the Retroreflector in Space (RIS), describing its purposes, the
flight element, ground segment, operation, results, and unresolved issues.
RIS, a single-element, hollow corner-cube retroreflector, makes long-path
(Earth-satellite-Earth) absorption measurements of atmospheric trace species
and performs laser ranging experiments to accurately determine the ADEOS
orbit. The ground segment consists of a CO2 laser transmitter and receiver for
spectrum measurement and a second harmonic Nd:YAG laser transmitter and
receiver for active tracking and ranging. Active tracking with an accuracy of
30 microradians was achieved, and the absorption spectrum of atmospheric ozone
was measured. The accuracy of orbital predictions based on laser ranging was
ten times better than conventional RARR methods.
ILAS performs spectrographic measurements of ozone-chemistry related
gases, aerosols, PSCs, temperature, and pressure in the high-latitude
stratospheres. The Version 1.00 algorithm has been operating since initial
checkout; Version 2.00, since 4 March; and Version 3, since 5 August.
Version 3.10 is in testing. Raw data acquired at EOC amounted to 6891 events.
(An event is a measurement; there are two events per orbit.) Of these events,
5626 were processed to Level 2 using Version 3.00. ILAS products, UKMO
products, polar vortex edge information, and sun spot image data are being
distributed to the Science Team, Validation Team, and PIs through the ILAS
home page. These will be available to general researchers from May 1998.
Problems encountered include late Level 0 data delivery, data frames
partially missing, and irregularly or incorrectly delivered mission
management information files. The ILAS instrument worked very well and
acquired 6000 data sets (sunrise and sunset events).
The JAXA Office of Research and Development made a presentation on the
Technical Data Acquisition (TEDA) equipment status, covering the objectives,
operations, data processing, measurement data and conclusions. TEDA equipment
measures the space radiation environment, radiation effects on electrical
parts, and contamination and potential charge on spacecraft surfaces using a
high-energy ion telescope (HIT), dose monitor (DOM), single-event upset
monitor (SUM), potential monitor (POM), contamination monitor (COM), and
sensor contamination monitor (COM-S). (Refer to the handout "Operations Status
Report - TEDA mounted on onboard of ADEOS Satellite" for details.) TEDA
successfully acquired data during the nine-month life of ADEOS.
FINAL ACTION ITEM REVIEW - Six action items were identified, all for JAXA.
JAXA will:
1. Clarify the conditions for retransmitting NRT data in the ADEOS-II
program.
2. Continue contingency studies for ADEOS-II.
3. Explain TOMS Level 0' data loss.
4. Provide POLDER Level 0' data statistics.
5. Add a description of the POLDER Level 0 synchronization problem.
6. Continue to study the power suspension problem in the ADEOS-II program.
As the representative of the AO sensor providers, GSFC noted that how will
JAXA deal with the lessons learned? Many ideas are applicable to ADEOS-II.
We were hoping JAXA would review the lessons learned and give us feedback.
In answer to this, JAXA will summarize the lessons learned and give them to
the ADEOS-II office. They will advise us, and we will inform the participants
by OCL.
CLOSING REMARKS
JAXA noted with sorrow that this was the first and last OSR and that
ADEOS was the "Big Bang" for JAXA in this field of activity and thanked
everyone for their support.
Earth Observation Research Center