Phase 0 was the on-orbit checkout, which completed on August 4, 2008. The spacecraft subsystems were evaluated and checked first. Subsequently, the instruments were turned on, tested, and calibrated, and the main observation modes were verified. The LAT event data accumulated during this phase is proprietary to the LAT instrument team until the beginning of Phase 2.
Phase 1 began as soon as Fermi was declared operational. During this phase Fermi carries out a one-year validation and sky survey program. The performance of the science instruments was fully characterized and validated, and the data processing pipelines are being refined based on operational experience. The spacecraft then operated in a sky survey mode designed for relatively uniform sky coverage. The survey could be interrupted by Target-of-Opportunity (TOO) observations to follow 5-10 bright transients (typically for a few orbits), or by autonomous repointing for GRB followup. The LAT event data accumulated during this phase is also proprietary to the LAT instrument team until the beginning of Phase 2. The LAT team release spectra and lightcurves for bright transients. A preliminary Bright Source List was released on February 6, 2009. GBM burst data is released as soon as it is processed.
During Phase 2, because of the LAT's wide FOV, survey mode was the baseline operational mode although pointed observations could be proposed through the guest investigator program. In practice, pointed observations have been rare and have largely been discontinued after an anomaly disabling one solar array drive occurred in 2018 (Cycle 11). The expected frequency, criteria and figures-of-merit for the different observation modes will be announced to the scientific community.
The GI program will go through approximately annual cycles. The first GI cycle coincided with Phase 1, and selected Phase 1 guest investigations were not to interfere with the validation of the instruments or the all-sky survey. During Cycle 1 the scientific community had access to LAT event data, but will have access to spectra and lightcurves of selected sources and transients. During subsequent cycles GIs may propose deviations from survey mode, and will have access to all the LAT and GBM science data.
As a member of the general scientific community your primary contact with the Fermi mission is through the Fermi Science Support Center (FSSC) at the Goddard Space Flight Center (GSFC). The FSSC provides the scientific community with data, the analysis software (herein "Science Tools" or sometimes "Fermi Tools"), and documentation. Members of the FSSC manage the Fermi Guest Investigator (GI) program on behalf of NASA HQ. The FSSC also creates the observing timeline based on successful GI observing proposals and the default observing mode.
The data telemetered down from Fermi enters the Fermi ground system through the Mission Operations Center (MOC), staffed by the Flight Operations Team (FOT) and also hosted at GSFC. The MOC 'cleans up' the telemetry, monitors the spacecraft through the housekeeping portion of the telemetry, and transmits the appropriate telemetry to the other ground system elements. The MOC also sends commands to the spacecraft, particularly a weekly load of observing and operational commands.
Each instrument team maintains an Instrument Operations Center. The LAT Instrument Science and Operations Center (LISOC) is located at SLAC in Palo Alto, CA, while the Fermi GBM operates the GBM Instrument Operations Center (GIOC) in Huntsville, AL. The Instrument Operations Centers receive the 'cleaned up' telemetry, monitor their detectors through the housekeeping portion of the telemetry, process the science data, and transmit the resulting science data products to the FSSC. The LAT science data processing is quite extensive, starting with event reconstruction from the 'hits' in different parts of the LAT and ending with a characterization of these events. The Instrument Operations Centers also serve data to members of their teams.
This model of the ground system is more decentralized than usual, with the instrument teams taking a more active role in the routine processing of the science data. Data processing occurs at each ground system element, leading to a straightforward set of data 'levels.' The MOC removes corrupted or duplicate telemetry packets, and ensures that the packets are in time order; the resulting 'cleaned up' telemetry is called Level 0 data. The Instrument Operations Centers process the instrument-dependent data, creating event lists; the data that leaves is called Level 1 data. Finally, processing by scientists using the Science Tools results in Level 2 data. As LAT data is released to the public, the FSSC will create and post count and exposure maps on weekly, monthly, and yearly timescales.
Catalogs and compendia of Level 2 data are considered Level 3 products.
The FOVs of Fermi's two detectors are very large and therefore the concept of an 'observation' of a given source is not very meaningful. Useful LAT data can be taken over approximately 30% of the sky at any time (although the effective area will vary over the FOV), and therefore while data are accumulated for a given source, data are also accumulated for a large number of other sources. As described in the mission overview, survey mode will predominate during at least the early part of the mission. In this mode the LAT will alternate pointing 35 degrees (default) above or below the orbital plane; this observing mode provides uniform sky coverage every 2 orbits (~3 hours) with the equivalent of 30 minutes of on axis exposure. Pointed and modified survey observations will be permitted during Phase 2, but must be justified. Even during these deviations from the standard survey mode, counts accumulated from one source will be interspersed with counts from many other sources.
Thus instead of thinking about the LAT data as consisting of a number of independent observations of discrete sources, the LAT data should be conceptualized as a continuous event list from the beginning of the mission to the end with counts from your source appearing with greater or lesser frequency depending on the LAT's effective area towards the source at a given time.
During the first 10 years of the mission deviations from the standard sky-survey mode - pointed observations or alternative survey strategies to enhance coverage of selected regions of sky - were possible. As a consequence of a transient phenomenon in the sky, the Fermi Project Scientist (or his/her designee) could declare a 'target of opportunity' (TOO) observation resulting in a pointed observation. Such TOO observations could result from a GI notifying the Project Scientist that the criteria for a TOO observation in a successful GI proposal have been met or a request from the community; these requests for a TOO observation, whether or not previously proposed, will be submitted through a website. In Phase 2 the data accumulated during a TOO observation is not proprietary to the requester; in Phases 0 and 1 the LAT event data are not public.
Similarly, if either the LAT or GBM detected a gamma-ray burst sufficiently bright then the spacecraft could autonomously repoint towards the burst location for five hours (the current default).
Data from the Fermi GBM and the LAT are stored in the spacecraft's Solid State Recorder (SSR) as the data are taken; the instruments do not store these data internally. While the system specifications permit 72 hours from the taking of source data until data are available from the FSSC, the actual data latency is typically much less. Although the SSR can hold up to 30 hours of data, the data is downlinked from the spacecraft through a TDRSS satellite 6-7 times per day on the Ku band. Thus in normal operations the data should reside on the spacecraft for only ~3 hours before being downlinked. While the LAT instrument team is permitted up to 24 hours to process the data, the data pipeline is designed to process the data from a given downlink before the data from the next downlink arrives. Similarly, while the FSSC is permitted 24 hours to ingest data and make it available, the FSSC plans to complete processing the data from one data transfer before the next such transfer.
Thus in practice LAT data latency has been ~12 hours.
The GBM team processes their data after every downlink. Burst data is transferred to the FSSC as soon as processing is completed. Daily data products are packaged into one day units, allowing for a shorter latency than routine daily data. Prompt alerts are issued for each recorded gamma-ray burst via NASA's GCN network.
» Forward to Use of the Analysis Threads
» Back to the Overview of the GBM
» Back to the beginning of the Introduction
» Back to the beginning of the Cicerone