December 21, 1998
Here is an update of our current understanding of the impact of the voltage converter anomaly on the CERES instrument on TRMM.
- Ground tests are now through 3 months of testing on the potentially more sensitive A/D converter and are
through 2 months of testing on two sets of the 7 components powered by the anomalous converter.
All tests are being conducted at 30V, while the expected range of converter output voltage is 23-29V
to these parts when the voltage converter completely loses regulation. This indicates that it is
unlikely that the higher voltage level will degrade the data or lifetime of the instrument.
- It has been determined from analysis of the August data (initial degradation) and the 5 day test in
- the calibration and radiometric performance of the CERES instrument is not affected by the higher
than normal output voltage of the converter (as expected).
- the voltage output of the converter is expected to continue to degrade with additional operation.
The instrument only monitors this converter output voltage up to 20V, while total deregulation implies
voltages as high as 29V. From data taken to date, we anticipate crossing the 20V level in roughly 10-30
days of further operation. The length of time depends on whether further degradation takes place
linearly in time: which cannot be assured.
- Ground testing has, however, identified a risk to the lifetime of the instrument once the converter has
lost the ability to regulate its output voltage. Once in an unregulated state, if voltage input to the
converter drops to zero, and then later is restored, the inrush of current into a capacitor in the converter
may be sufficient to melt leads: leaving an open circuit, and terminating the life of the CERES instrument.
Power has been removed from the CERES instrument once during the TRMM mission, and while the cause of this
inadvertent shutdown of the CERES instrument by the spacecraft has since been fixed, it remains true that
the default method to "safe" or shutdown the spacecraft is to remove power (voltage) to all of the instruments.
The past occurrence of power loss to the CERES instrument was early in the TRMM mission when the spacecraft
inadvertently switched to this "safe" mode. Since CERES is a very small part of the TRMM spacecraft power
budget, we will request the TRMM project to change the "safe" procedure for future operations such that
they switch CERES to a standown mode with minimum power instead of turning the instrument off. In this case,
there would be no additional risk to the CERES instrument, even with the voltage regulator in an
uncontrolled state. This change would probably require modifying the spacecraft software and the
impact/risk is being assessed by the TRMM project. We expect to know more on this critical issue in January.
- We currently plan to turn on the CERES instrument for 2 days in mid-January to perform an intercalibration
between CERES and the SCARAB radiation budget instrument on the Russian RESOURC satellite. Starting in
mid-November 1998, SCARAB has been collecting 4 days (Mon through Thursday) out of every week of data
(limitations at the Russian downlink station) but hopes to increase this to 100% in the future. Therefore,
we should have some broadband measurements during the peak of LaNina currently expected for Dec/Jan.
La Nina is then predicted to disappear by Summer, 1999. CERES is planning to perform 2-day intercalibrations
of CERES/SCARAB every 1-3 months until the CERES EOS-AM instrument is operating, at which time we will turn
on the TRMM CERES instrument and leave it running indefinitely. The number of the calibrations done depends
on the experience we gain in analyzing the January test, and in the amount of further degradation of the
voltage converter compared to instrument test performed in mid-October.
- We plan to replace the voltage converters on the EOS-AM and PM CERES instruments. The suspect optical
coupler and its mounting method were both changed in the design of this converter several years ago and the
current parts are not susceptible to this potential failure mode.
- The final report of the tiger team is expected late January.
- Given the above scenario, what are some of the CERES science objectives that will not be affected,
and which science objectives will be impacted by the probable loss of data from Sept 1998 through Aug 1998?
- No loss of 1998 El Nino tropical data: captured the El Nino peak in Jan through March 1998 as well
as the decay through July, 1998.
- No loss of stable calibration/continuity for global change long-term climate record: CERES calibration
unaffected by voltage converter degradation and overlap with EOS-AM CERES instruments will establish
calibration continuity to ~ 0.1%.
- No loss of optimal multiple satellite data because of delay in EOS-AM launch to summer,1999. Multiple
satellite CERES data is more accurate than single satellite data, and is more useful for shorter time
scale studies (daily to seasonal scale).
- Reduced accuracy (~ factor of 2) for La Nina tropical data: SCARAB began routine data collection
November 16, 1998 and will cover the rest of the gap through EOS-AM. Accuracy reduction is one satellite
(SCARAB on RESOURC) versus two satellites (TRMM + RESOURC) for diurnal sampling. Other accuracy issue
is whether SCARAB can increase sampling from 4 days out of 7 (ground station limitation).
- Delays by 1 year, the new CERES angular models required to obtain a factor of 3 to 4 improvement in
TOA flux accuracy, and therefore delays the expected improved constraints on surface and atmosphere
- Lost year of data degrades statistical accuracy of final CERES angular models by sqrt(1/3.5) = 15%.
This assumes the VIRS instrument performs nominally through the expected 3.5 year TRMM mission lifetime.
- Lose overlap with two field experiments planned for spring 1998: Nauru99, and INDOEX. Radiation data
for these experiments will be provided by SCARAB.
- Note that the scientific impacts listed assume:
- EOS-AM launch by July 30, 1999, CERES TRMM turn on Sept 1, 1999 to match the turn on of the CERES
instruments on EOS-AM (30-day outgassing period).
- CERES intercalibration with SCARAB is successful using 1.5 day periods every 23 or 46 days from January
through August, 1999.
- TRMM spacecraft can change the spacecraft "safing" procedure to turn the CERES instrument to its reduced
power "safe" mode instead of removing power to the CERES instrument.
Return to CERES +15V Power Converter Anomaly Page