Shelter from the Storm: Protecting the Chandra X-ray Observatory from Radiation

The surprisingly high solar activity over the past year has had a surprisingly large effect on Chandra operations. The following article provides some detail on how the Chandra staff handle periods of high radiation. This article was adapted from a paper at the ADASS XI conference (2001), which will appear in the PASP as ``Shelter from the Storm: Protecting the Chandra X-ray Observatory from Radiation", Robert A. Cameron, David C. Morris, Shanil N. Virani, Scott J. Wolk, William C. Blackwell, Joseph I. Minow and Stephen L. O'Dell -Editor

Chandra's elliptical orbit takes it through the Earth's magnetosphere and radiation belts and into the solar wind beyond the magnetosphere. Thus Chandra is exposed to both a continuous proton flux from the Sun, in the form of the steady solar wind, and from episodic solar flares and large Coronal Mass Ejections (CMEs), and also trapped protons in the magnetosphere and magnetosheath.

The density of damaging 100-200 keV protons in Chandra's environs increases greatly ( $\sim\!\!\times10^5$) in the radiation belts and throughout Chandra's orbit after CME events, compared to the normal proton density during Chandra's science operations, at altitudes above $\sim\!70,000$ km.

After launch, it was discovered that 100-200 keV protons can reflect off Chandra's x-ray optics and reach the ACIS CCD detector. These protons can cause lattice damage in the front-illuminated silicon CCDs, which results in degraded charge transfer efficiency and hence degraded energy resolution for the ACIS detector.

Although significant damage occurred in the early mission, the damage rate has been slowed dramatically by several operational changes. The Science Operations and Flight Operations Teams constantly monitor Chandra's radiation environment. The infamous SCS 107 (Stored Command Sequence 107) has been executed a total 21 times on Chandra due to high radiation, since it was loaded into the on-board computer in December 1999. It has been automatically triggered 13 times by high radiation detected by the EPHIN radiation sensor, and 8 times by ground command by direction of the SOT. Despite popular folklore among SOT and FOT members, not all of these radiation events occurred on a Friday night with associated 3am telecons! More than 1.1 million seconds of science observing time has been lost so far due to these SCS 107 runs. But the operational life of ACIS has been greatly extended, with the projected damage to ACIS over 10 years of mission being equal to the damage from the first 8 orbits.

Operational changes related to radiation protection are described in the following sections.

Radiation Protection System Components

The three principal components of the overall system for protecting Chandra from radiation are:

$\bullet$improvements to the flight software for autonomous protection

$\bullet$implementation of ground radiation monitoring operations

$\bullet$the development of an improved model for predicting solar wind and magnetospheric proton fluxes in the Chandra orbit.

Autonomous Protection

Because the Chandra Operations Control Center is not continuously in communication with Chandra, an enhanced autonomous radiation protection system had to be added to Chandra's capabilities after launch. Flight software changes were implemented in November 1999, as a new Stored Command Sequence (SCS 107) in the On-Board Computer (OBC).

Features:

$\bullet$ Provided an autonomous safing capability for the protection of science instruments (SIs). This does not replace scheduling of SI protection activities through perigee passages, but provides backup protection in case of schedule failures (e.g. if the mission schedule is incorrectly built or stops running).

$\bullet$ Extended the spacecraft safing capabilities to include safing of the SIs in the event of spacecraft failures interrupting the science mission.

$\bullet$ Provided better use of EPHIN (on-board radiation detector) energy ranges, for detecting low energy protons, and adjusted the particle count rate thresholds to prevent false triggers outside perigee.

$\bullet$ Accompanied by improved Flight Operations Team monitoring of SI safety, and improved ground-based SI safing procedures.

$\bullet$ Accompanied by improved review of mission command loads to check correct SI safing command sequences and times for perigee passages.

A universal SI safing response applies to all SI safing triggers. For autonomous safe operation, the mission timeline must be stopped. The same radiation safing thresholds are applied for all SI combinations. There is no autonomous recovery from SI safing.

Events which can trigger the science instrument safing actions:

$\bullet$ Observatory Safe Mode

$\bullet$ Normal Sun Mode transition

$\bullet$ Bright Star Hold

$\bullet$ RADMON high radiation detection

$\bullet$ EPHIN failure

$\bullet$ On-Board Computer standby timeout

Science instrument safing actions performed by SCS 107:

$\bullet$ Disable on-board radiation monitoring

$\bullet$ Terminate science mission activity

$\bullet$ Move Science Instrument Module (SIM) to non-instrument viewing position

$\bullet$ Safe ACIS detector

$\bullet$ Safe HRC detector

$\bullet$ Move SIM to HRC viewing position

$\bullet$ Disable SI safing process (SCS 107)

Mission Planning and Ground Operations

Following early mission damage to ACIS, the Chandra Science Operations Team

http://cxc.harvard.edu/mta/sot.html

(SOT) and Flight Operations Team (FOT) quickly implemented a system of protecting ACIS during each perigee passage, when Chandra is below $\sim\!75,000$ km. The science instrument protection commands used in SCS 107 are included in the mission command timeline for each perigee passage.

For solar flares and CMEs, the SOT also maintains a series of real-time proton flux and fluence monitors and predictors at the Chandra X-ray Center

http://cxc.harvard.edu/mta/RADIATION/ (CXC).

SOT members are alerted through automatic paging software and then assess past and future fluence estimates in conjunction with Chandra's ground contact schedule and science mission timeline. The SOT then provides real-time direction to the FOT if instrument safing by ground command is required.

Source data:

$\bullet$ Advanced Composition Explorer

http://sec.noaa.gov/ace/

(ACE): Electron, Proton and Alpha Monitor (EPAM) P3 channel: 112-187 keV protons.

$\bullet$ Chandra Radiation Model (CRM).

$\bullet$ Air Force and Costello Kp geomagnetic activity index.

$\bullet$ GOES-8 and GOES-10 proton monitors.

Monitors:

$\bullet$ 2 hour ACE EPAM P3 average flux.
This monitor has an alert threshold = $5\times10^4$ cts/cm²-s-sr-MeV.

$\bullet$ Chandra orbital proton fluence. ACE EPAM P3 and CRM fluxes are integrated from perigee to perigee. Fluxes are attenuated by SIM and OTG transmission factors during integration.
Monitor alert thresholds = $1\times10^9, 3\times10^9, 9\times10^9\cdots$ cts/cm²-sr-MeV.

$\bullet$ Air Force and Costello Kp estimates. This monitor uses a seasonally variable threshold matrix for alerts.

Improved Radiation Modeling and Predictions

Chandra Project Science is leading a team to develop and implement an improved model for low energy proton fluxes in the Chandra orbit: the Chandra Radiation Model

http://wwwastro.msfc.nasa.gov/xray/ACIS/fluence/ (CRM). Design Objectives:

$\bullet$ Improve science observing efficiency with low risk of RADMON triggers from trapped radiation.

$\bullet$ Improve real-time predictions of Chandra's proton fluence, for health and safety decision making.

$\bullet$ Refine understanding of the effect of the particle environment on ACIS for improved budgeting and control of lifetime accumulated damage.

Current Chandra mission planning and scheduling uses the NASA AP-8 and AE-8 radiation models. AP-8 provides poor fidelity above geostationary orbit and poor modeling of 100 keV protons. AE-8 is used for scheduling science operations. Mismatches between AE-8 predictions and observed EPHIN rates have resulted in required 10 ks ``padding'' on both the predicted radiation belt entry and exit times.

Characteristics:

$\bullet$ Model 100-200 keV protons, compatible with ACE EPAM P3 data.

$\bullet$ 9-30 ${\rm R}_E$ geocentric altitude coverage.

$\bullet$ Indexed by Geocentric Solar Magnetospheric (GSM) coordinates and Kp.

$\bullet$ Flux predictions identified by region: magnetosphere, magnetosheath, solar wind, for detailed estimation of global radiation environment.

$\bullet$ Probabilistic flux estimates: mean, 50%, 95% fluxes.

$\bullet$ Multiple species and energies: p: 100-200 keV (future: p >1 MeV, He, CNO).

Data source: 5 year dataset (1995-2000) from the Ion Composition Subsystem (ICS) on the Energetic Particles and Ion Composition (EPIC) instrument on the Japan ISIS/NASA Geotail satellite. Ion and electron spectrograms from the Comprehensive Plasma Instrument (CPI) Hot Plasma Analyzer provide the region identification.

Validation: correlation of mission fluence against aggregate ACIS damage shows CTI fluence sensitivity of $5\times10^{-17}$ CTI/(p/cm²-sr-MeV). This is consistent with CTI increase and fluence in early mission unprotected perigees.

- Rob Cameron, on behalf of the Science Operations and Flight Operations Teams.

See also ``Report from the Project Science"