IAC-02-Q.3.1.01 2001 MARS ODYSSEY PROJECT REPORT

David A. Spencer, Odyssey Mission Manager Roger G. Gibbs, Odyssey Project Manager Robert A. Mase, Odyssey Navigation Team Lead Jeffrey J. Plaut, Odyssey Deputy Project Scientist Ronald S. Saunders, Odyssey Project Scientist Jet Propulsion Laboratory California Institute of Technology Pasadena, California

Abstract and morphology, and measure the Mars radiation environment from orbit. In The Mars Odyssey orbiter was launched on addition, the orbiter will serve as a data April 7, 2001, and arrived at Mars on relay for future landed assets. Odyssey was October 24, 2001. The orbiter carries designed and developed through a scientific instruments that will determine partnership between the Jet Propulsion surface elemental composition, mineralogy Laboratory in Pasadena, California, and and morphology, and measure the Mars Lockheed Martin Astronautics in Denver, radiation environment from orbit. In Colorado. addition, the orbiter will serve as a data relay for future surface missions. This paper The Odyssey spacecraft was launched from will present an overview of the Odyssey Kennedy Space Center on April 7,200 1, and project, including the key elements of the successfully captured into orbit about Mars spacecraft design, mission design and on October 24, 2001. Following 76 days of navigation, mission operations, and the aerobraking, a series of five propulsive science approach. The project’s risk maneuvers were performed to attain the management process will be described. near-circular science orbit. The science Initial findings of the science team will be mission formally began on February 19, summarized. 2002. The prime science mission lasts for 917 days, and will be completed on August Introduction 24,2004.

The Mars Odyssey Project is the latest in an The primary science objectives are the ongoing series of robotic missions to Mars following: within NASA’s Mars Exploration Program. The Program goals include the global - Globally map the elemental composition of observation of Mars, to enable the Mars surface. understanding of the Mars climatic and - Determine the abundance of Hydrogen in geologic history, including the search for the shallow subsurface. liquid water and the evidence of prior or - Acquire high spatial and spectral resolution extant life. The Odyssey orbiter carries of the surface mineralogy. scientific payloads that will determine - Provide information on the morphology of surface elemental composition, mineralogy the surface.

1 - Assess the Mars radiation environment. criteria. The mission success criteria are - Provide data for evaluation of future shown in Table 1. The strategy for meeting landing sites. the project mission objectives is described in the Mission The detailed The mission success criteria as stated in the implementation plan is given in the Baseline Project Policies Document’ are expressed as Reference Mission4 document. “primary” and “full” mission success

Table 1. Mi: ion Success Criteria Full Mission Success Primary Mission Success Carry out a global survey of Mars from Acquire 25% of the planned mission science data the planned science orbit for one Mars from Mars orbit for &o out of three of the orbiter science instrument complements. year. Archive the acquired science data in the Collect 75% of the planned mission science data Planetary Data System. from each of the orbiter science instrument complements. Provide communications relay for surface elements from the U.S. and other spacefaring nations for two Mars years after achieving the science orbit. Archive the acquired science data in the Planetary Data System within 6 months of acquisition.

SDacecraft Develo~ment Descrm..

The Jet Propulsion Laboratory teamed with The system block diagram is shown in its industrial partner, Lockheed Martin Figure 1. The subsystems which comprise Astronautics, to design, build and test the the Odyssey are: command and data Odyssey orbiter. The design drew heavily handling, electrical power, guidance from the heritage designs of the Mars navigation and control, mechanisms, Climate Orbiter and Stardust spacecraft. propulsion, science payload, flight software and fault protection, structure and harness, telecommunication, and thermal control.

2 I I

GRSIHEND/

I

1553 RS422 - Digital or Analog Discrete Multi-function Bus Motor Commands /Telem - Pyrotechnics 8 Actuators 4 Indicates internal redundanq Temperature sensors, heaters, and EPS 110 not shown for clarity Figure 1. System Block Diagram

and Data Handhng: telemetry and sensor data from the guidance, navigation, and control sensors and provides The command and data handling (C&DH) control of the reaction wheels and thrusters, subsystem receives data files and commands collects sensor data from the science from the ground operators via the instruments, and formats data for telecommunication subsystem, stores and transmission to the ground operators via the executes sequenced commands and telecommunication subsystem. autonomously generated commands, collects

3 MSO1 ORBITER C&DH BLOCK DIAGRAM

I

Communications

Figure 2. Command and Data Handling Block Diagram

There are seven modules in the C&DH. The processor card contains a RAD6000 The uplinWdownlink card provides the processor, 128 megabytes of random access interface to the telecommunication memory and 3 megabytes of non-volatile subsystem, both x-band and UHF. The memory. There is a redundant processor uplinWdownlink card provides command card. decoding on data received from the ground operators and encodes data transmitted. The The inputloutput card collects analog data uplinWdownlink card is redundant. from sensors distributed on the spacecraft and converts this signal to a data word(s); The mass memory card stores data for later provides the control of the Mil-Spec 1553 transmission to ground operators. The data and multi-function data busses; and collects volume capability is 1024 Mbytes. discreet (binary) telemetry. The inputloutput card is redundant. The module interface card arbitrates if there is contention between the two processors, The payload and attitude control interface provides a critical oversight function on the card collects telemetry from and provides processor (the “heartbeat monitor”), and commands to the science instruments and maintains the master spacecraft clock and the guidance, navigation, and control status of the mission phase. sensors and actuators. The payload and attitude control interface card is redundant.

4 Pow= the solar array is capable of producing approximately 540-780 watts of electrical The electrical power subsystem (EPS) power, depending on solar distance and generates, stores, and distributes the incidence angle. When illuminated by the electrical power used on Odyssey. sun the solar array provides the electrical Electrical power is generated by the 7.4 power requirements of the Odyssey and square meter solar array. This solar array is keeps the battery fully charged. comprised of gallium arsenide solar cells bonded to three composite panels. At Mars,

Charge Control Unit (CCU) CCU Card "A" Command & Data Orbiter Handling (C&DH) - 7.4 M2 +/ r

Power Distribution & Drive Unit GN&C

MAD Relays N1H2 Battery Assembly Pwr I- Power MFB Secondary - Control Pwr 22 Cells, 16A-Hr ea.. Telemetryl- Pyro Initiator Unit (PIU) PIU Power Pyro Initiator Card (1) Prop Valve Driver Card (1) Thrusters

Figure 3. Electrical Power Subsystem Block Diagram

A 22 cell, 16 amp-hr nickel-hydrogen condition of the battery as monitored k the battery provides the power requirements voltage, temperature, and pressure of the when the solar array is not generating battery. The charge control unit is sufficient electrical energy, such as when redundant. Mars eclipses the sun in orbit, or the spacecraft must change orientations for a The power distribution and drive unit thruster maneuver. provides electrical power switching and distribution, motor control drive for The charge control unit controls the energy actuators, and regulated bus voltage via high provided to the battery according to the efficiency dc/dc converters. The power

5 ,

distribution and drive unit is redundant. The spacecraft body, solar array normal vector, motor and drive unit provides relays and high gain antenna centerline, and science diodes to ensure redundant capability even instrument boresights either individually or in the presence of electrical shorts in the collectively to track stationary or moving actuation devices. vectors (example: control the THEMIS boresight to track Mars nadir, the solar array The pyrotechnic initiator unit actuates the normal vector to track the sun, and the high release and activation devices (separator gain antenna boresight to track Earth), nuts, thermal wax actuators, and NASA controls the propulsion subsystem to impart standard initiators) and controls the velocity changes or to unload the propulsion engine valves. The pyrotechnic momentum energy in the reaction wheels. initiator unit is redundant. The sun sensors provide pointing data on the position of the sun. The star tracker provides Guidance Navip-.. position and illumination magnitude data of the star field. The inertia measurement unit The guidance navigation and control uses ring laser diode gyros and an subsystem maintains knowledge of the accelerometer to provide spin and orientation of the spacecraft relative to an acceleration rates. These sensors are inertial reference coordinate system, redundant. maintains the knowledge of the positions of the Earth, sun, and Mars, controls the

6 Figure 4. Guidance Navigation and Control Block Diagram

The control algorithms, hosted in the The position of the high gain antenna and command and data handling subsystem, solar array are controlled via the electrical process these sensor inputs and provide power subsystem. control data for the actuators.

The reaction wheels provide three axis roll control of the spacecraft. The thrusters, part of the propulsion subsystem, provide The mechanisms on Odyssey provides the capability to control roll orientation in pitch, ability to restrain through the launch yaw, and roll axis and changes in spacecraft environment and then release the solar array, velocity. The thrusters also provide the high gain antenna, and the Gamma Ray capability to reduce the momentum stored in Spectrometer boom; the two axis gimbals to the reaction wheels. The reaction wheels control the solar array and high gain have functional redundancy via a spare antenna; the positioning of the Gamma Ray wheel oriented in a skewed position thereby Spectrometer to a near and extended providing a control component in all three position. axis. The thruster configuration provides functional redundancy. The solar array is restrained by separation nuts. Upon release of the nuts, springs energy is used to provide the rotational

7 I .

motion of the panels about their hinges. Dampers limit the rate of deployment. At The Gamma Ray Spectrometer boom is full deployment the hinges latch into final restrained by a latch, which is released by position, the stroke of a high force thermal actuator. The thermal actuator uses a low melting The high gain antenna is restrained by point solder to achieve actuating force. separation nuts. After release of the nuts the After release, the boom deployment is antenna boom initial motion is provided by driven by spring force; the deployment spring energy, the subsequent rotation is speed is limited by an eddy current damper, provided by a redundant stepper working through a gearbox and connected motodgearbox within the deployment hinge. via a tether to the end of the boom.

The two axis gimbals control the pointing on the solar array and high gain antenna. These gimbals utilize redundant stepper motors and position encoder electronics.

Figure 5. Gamma Ray Spectrometer Boom Deployed

8 I

The fuel tanks are symmetrically positioned ProDulsion to minimize changes in mass properties as The propulsion subsystem provides changes the fuel and oxidizer are consumed during to the spacecraft velocity on the trajectory to the mission. The fuel and oxidizer tanks are Mars, reduces the differential velocity pressurized prior to launch. This initial between the spacecraft and Mars to allow pressure provides the required fuel flow for the Mars gravitational field to capture the thruster operation during the cruise phase. spacecraft in orbit, reduces ("unloads") the A helium tank, operating through a pressure momentum stored in the reaction wheels, regulator, is isolated from the tanks until and modifies the orbit characteristics during prior to the mars orbit insertion (MOI) the orbital mission. maneuver. The helium tank provides a near constant pressure condition during the MOI The spacecraft uses a dual fuel system, maneuver; following MOI the helium utilizing mono-methyl-hydrazine fuel and pressure source is again isolated from the nitrogen tetroxide oxidizer. The main tanks. The spacecraft will operate for the engine operates in a bi-propellant mode. remainder of the mission in a blow-down The remaining thrusters use the fuel only in mode. a mono-propellant mode.

LEGEND WL-

WN&"

I I I 1 Fud PIA

ARCUK LEROS-lb 154 Ibf Bi-Prop @ins with kat Shield

Figure 6: Propulsion Schematic

The main engine uses both hydrazine and Newtons thrust. The main engine is used nitrogen tetroxide and provides 695 only for the MOI maneuver, where it thrusts

9 for approximately 20 minutes. Following MOI the main engine and the nitrogen tetroxide tanks are isolated by valves for the remainder of the mission. The spacecraft structure provides a lightweight and rigid means of supporting There are four rocket engine modules the elements of the Odyssey and is designed (REM). Each REM is comprised of a 0.9 N and tested to withstand the loads of the and a 22 N thruster. The 0.9 N thrusters launch environment. provide roll control. The 22 N thrusters provide velocity changes and, during the There are two principle structural modules. MOI main engine operation, attitude control. The propulsion module contains the tanks, pressure regulator, filters, pyrotechnic valves, engines, and associated assembles. The equipment module is composed of an equipment deck that supports engineering The software stores the command sequence, components and the radiation science executes the guidance, navigation and experiment and the science deck that control functions, controls the science supports the thermal emission imaging sensors, collects science data and spacecraft system, gamma ray spectrometer, the high telemetry, and provides a data stream to the energy neutron detector, the neutron telecommunication subsystem for spectrometer and the star camera. transmission. The structure is composed of composite The fault protection subsystem utilizes construction techniques with metallic hardware and software sense and status data attachment points. to assess hardware performance, detects loss of communication with ground operators, The harness provides the electrical routing ensures spacecraft health and ability to of all power and data communication lines receive commands from ground operators. on the spacecraft.

10 Solar Array Gamma Sensor Head

Battery

Neutron Spectrometer\

Battery Enable)# Plugs fill

------MARIE PDDU p UHF Antenna $!* HGA Deployment Hinge

I UVL ILIL~R Vressurant Tank Structural Coordinate System

Figure 7. Structure, Launch Configuration

11 (GRS HARNESS) N. SPECT CEB GRS HEND

I (MAIN HARNESS) nl, I C&DH I

WHEEL ELEC. (HGA HARNESS)

(SIA HARNESS) SOLAR ARRAY PIU

~~ II ll (PYROHARNESS) I+ '-1 , , 1 , , 1 REM T-O T-O Figure 8. Harness Block Diagram

antenna, or the wide coverage low gain antenna. Use of the parabolic 1.3 meter The telecommunication subsystem is diameter high gain requires that the antenna comprised of two separate radio systems. actively tracks the Earth; the low gain 4.4 The X-band (4 GHz) radio system receives centimeter wide patch antenna allows commands and data files from ground communication over much greater operators via the Deep Space Network spacecraft orientations. (DSN), transmits data files to the ground operators via the DSN, and provides precise The small deep space transponder (SDST) velocity and positional data via radiometric demodulates the RF signal and provides a techniques. The UHF radio system (400 symbol stream to the command and data MHz) provides a radio relay capability handling subsystem for decoding. The between Odyssey and future assets at Mars, SDST receives an encoded data stream from such as landers, rovers, or balloons. the command and data handling subsystem and provides a modulated, low level RF The X-band system receives signals by signal to the solid state power amplifier either the highly directional high gain (SSPA). The SSPA amplifies this signal to

12 ,

approximately 12 Watts. The amplified or medium gain- antenna for transmission to signal is then routed to either the high gain Earth.

Figure 9. Telecom Block Diagram

The thermal control subsystem maintains the temperature of the spacecraft components within acceptable temperature limits through Following a 7 month Phase A study period, the range of mission phases, which include where initial requirements and capabilities the near-Earth, interplanetary, and Mars were defined, the spacecraft design effort orbit environments. proceeded to Phase B. In Phase B, the system and science requirements were The spacecraft design uses a combination of established, a technical baseline design and passive and active means of thermal control. configuration were defined, orders were Multi-layer blankets limit the loss or placed for long lead time parts, and the 11 absorption of heat energy by means of month Phase B study concluded with a thermal radiation. Electrical heating System Preliminary Design Review. elements are controlled by means of either mechanical switches or by software Phase C/D began October 12, 1998, managed control. The solid state power approximately 3 1 months before launch. amplifier and battery both utilize The System Critical Design review was held thermostatically controlled louvers for beginning on April 16, 1999. The temperature control. Assembly, Test, and Launch Operations

13 began on January 18,2000. The orbiter was the coast of North America and achieve the air shipped to the launch site January 4, desired parking orbit inclination. A thermal 2001. Final orbiter tests, instrument conditioning roll was employed during the integration, spin balancing, fueling, and coast period as the spacecraft flew over integration with the launch vehicle were Europe, followed by the second stage restart. achieved prior to launch. Over the Middle East, the third stage burn injected the upper stagehpacecraft stack Launch and Cruise Phase onto the required escape trajectory, and after spinning up, the spacecraft separated nominally from the upper stage.

Odyssey was launched from Space Launch Complex 17A (SLC-17A) at Cape Performance Canaveral Air Force Station in Florida, on a Delta I1 in the 7925 configuration. The Planetary Protection policies require that, mission had designed a twenty-one day after injection, both the upper stage and the launch period that opened on April 7, 2001 spacecraft must be on a trajectory biased and closed on April 27, 2001. There were away from Mars such that the probability of two instantaneous launch opportunities each impacting Mars is less than 1 in 10,000. day corresponding to a launch azimuth of The size of the aimpoint bias is dependent 65", a long coast trajectory profile, and park on the expected injection dispersions. As orbit inclinations of 52" and 49". The arrival propellant is required to correct for the bias, date was dependent on the launch date, and it is desirable to minimize the aimpoint Mars arrival was designed to occur between correction. October 17 and October 28, 2001. Launch occurred on the first available opportunity, The biased injection targets are expressed in on the morning of April 7,2001. terms of the energy (C3), declination @LA), and right ascension (RLA) of the outgoing A typical Eastern Range flight profile was hyperbolic trajectory asymptote. The actual employed to achieve a 185 km circular injection result can be compared against the parking orbit at the first cutoff of the second target and expected dispersion statistics and stage (SECO-1). Two plane-change presented as a sigma-level miss from the (dogleg) maneuvers were employed to fly up target:

Table 2. Injection Parameters Injection Achieved Delta from Miss vs Parameter Target Expected c3 10.767 0.075 1.0 s DLA -5 1.669' 0.058" 1.0 s RLA 235.169" 0.128" 0.3 s

The injection dispersions can be mapped dispersions are presented in Figure 10, along along the nominal trajectory to the Mars with the achieved. The launch dispersion target plane to illustrate the expected arrival happened to occur in a favorable direction. conditions. The target and expected

14 2001 M~NWpoy Launch Pwfom": Launchd on 07-April-2001.lrtOppodunity I

-I I

Figure 10. Odyssey Launch Performance

Following a successful separation from the cruise attitude and flight software launch vehicle upper stage, the spacecraft configuration. After several hours, the outer configured itself for the journey to Mars. gimbal controlling the High-Gain Antenna, The first activity was to deploy the solar experienced higher than expected arrays to begin recharging the onboard temperatures, so the spacecraft was returned battery. Once the solar array was deployed, from the cruise attitude back to the initial the attitude control subsystem performed a acquisition attitude. This marked the short series of Sun-search slew maneuvers to conclusion of the initial acquisition activities find the sun and to establish attitude as the spacecraft proceeded with the planned reference. Once inertial attitude knowledge cruise operations. is established, with the solar arrays on the Sun, and the Medium-Gain antenna pointed at the Earth, communications with the se Overviely spacecraft could begin. The relatively short interplanetary cruise The first DSN station to contact the phase of the mission lasted for 200 days. spacecraft was DSS-34, the 34-meter Beam Activities during the cruise phase included Wave Guide antenna at the Canberra, initial deployments and checkout of the Australia complex, approximately 50 spacecraft in its cruise configuration, minutes after liftoff. Once initial contact was checkout and calibration of the spacecraft established, the command was sent to and payload subsystems, and navigation transition the telecom system to a two-way activities necessary to determine and correct coherent mode of operation, to begin the flight path to Mars. collection of Doppler and ranging data for Navigation purposes. The spacecraft was Communication with the spacecraft is then commanded out of the planned launch accomplished via the Deep Space Network safe-mode software state to the nominal (DSN) of ground-based radio antennas,

15 distributed around the world. Engineering due to higher than expected temperatures on telemetry, science data, and radiometric the HGA outer gimbal. On April 24, as tracking data are collected during each spacecraft attempted to transition to a tracking pass. One contact or trucking puss modified cruise attitude that would shade the per day with a 34-meter antenna was gimbal, the transition sequence was aborted standard for the cruise phase, with by a safe-mode entry unrelated to the continuous tracking provided around the activity. The spacecraft memory had been critical events such as launch, maneuvers, corrupted by a high-energy particle, which and final approach. resulted in a processor reset. Following the reset, the spacecraft entered safe mode, and One peculiarity of the high negative the flight team quickly recovered to nominal declination trajectory is that for the first two mode, in the desired cruise attitude. For the months of cruise, only the southern remainder of cruise, the spacecraft was hemisphere stations were able to view the configured such that the stowed high-gain spacecraft. During that time, the Canberra antenna pointed towards the Earth, and the tracking stations had very long view periods, solar array pointed generally towards the greater than 16 hours. Late in May, Sun with an offset angle profile. Goldstone came into view, and Madrid could not view the spacecraft until early The Neutron Spectrometer (NS) and High June. A tracking station in Santiago, Chile Energy Neutron Detector (HEND) were was contracted to supplement the DSN for powered on as planned on May 2. On May 4 the first month of cruise. the first planned thruster calibration performed. A checkout of the High-Gain The launch vehicle placed the spacecraft on Antenna receivekransmit capability was a trajectory to Mars that did not require as performed on May 9, and on May 24 the much propellant as planned. So the first communications path was switched from the scheduled trajectory correction maneuver LGA/MGA to the HGA to accommodate the (TCM) at Launch+9 days (April 16) was increasing range of the spacecraft from the delayed to Launch+46 days (May 23). The Earth. project realized a propellant savings due to the delay. The first trajectory correction maneuver executed nominally on May 23, firing the The first science instrument activity was an monopropellant TCM thrusters for 91 opportunity to image the Earth-Moon seconds. A test of the DOR tones in the system with the THEMIS instrument. The Small Deep Space Transponder (SDST) was instrument was powered on and performed a performed on May 24. And the first actual thermal calibration on April 17, ten days ADOR (Delta-Differenced One-way after launch, and successfully imaged the Ranging) measurement was performed on Earth and the Moon in a single frame in both June 4, utilizing both the Goldstone, the visible and infrared spectrums. On April California and Canberra, Australia 23 the MARIE instrument was powered on complexes. A total of 45 ADOR and began cruise science data collection. measurements were collected throughout the remainder of the cruise campaign. Early attempts to transition the spacecraft to the desired cruise orientation, with the High Beginning on June 1, a series of tests of the Gain Antenna pointed at Earth, were aborted UHF relay system were conducted with the

16 Stanford radio telescope. Ultimately three survivakommunications mode to download uplin k/downlin k tests were performed the collected data. A tiger-team was between June 6 and June 21 in addition to established to form a plan of action, and it the initial compatibility test on June 1. A became clear that the quickly approaching calibration of the solar array passive orbit insertion and aerobraking events would restraint was performed on June 20. This test take a higher priority. So the ground analysis involved articulating the solar array into and continued, but commanding to troubleshoot out of the restraint that would be used to the anomaly was postponed until the hold it during the critical orbit insertion and completion of aerobraking. aerobraking phases. The venting of the main engine in Following the development and uplink of a preparation for the large bi-propellant orbit safe-mode patch to protect the GRS insertion burn was performed on August 17. instrument, a low temperature anneal was A test of the 110 kbps downlink data rate performed on June 13. The GRS door was needed for THEMIS science operations was opened on June 25, and the high voltage successfully demonstrated on August 24. ramped up two days later. The THEMIS Also in August, a series of Star Camera instrument performed a visible exposure (SCAM) calibration tests were executed on calibration on June 15 and a star calibration the spacecraft. A number of star camera on June 22 to better characterize the outages had been experienced throughout instrument performance. cruise when the SCAM was not explicitly shaded from the Sun by the HGA. It was The second trajectory maneuver executed at also suspected that stray light from the open 86 days past launch on July 2, and marked GRS door was saturating the SCAM images the transition from early cruise to late cruise. even when the SCAM was shaded by the The mission plan had been designed to HGA. The calibration confirmed that the finish the majority of the checkout and orbit insertion and aerobraking attitudes calibrations prior to this time to provide a would be safe, and that the closure of the quiescent spacecraft in preparation for the GRS anneal door would resolve the issue for coming arrival and orbit insertion events. the remainder of cruise. However, a number of significant activities were executed in the final 114 days prior to In preparation for final approach to Mars, encounter. the spacecraft instruments were turned off, and the spacecraft configured for arrival. A follow-up calibration of the RCS thrusters The GRS door was closed on August 31, was performed on August 8, followed two which was followed by a turn to the late days later by a solar radiation pressure cruise attitude on September 4. This late calibration. These tests were designed to cruise configuration was designed to ensure that the force modeling employed by minimize the buildup of torque due to solar the Navigation Team and the resultant pressure effects on the offset solar array. trajectory perturbations were consistent with And the final MOI checkout occurred on the spacecraft experience. September 6.

On August 13 communications was lost with The third trajectory correction maneuver the MARIE instrument as it hung in a executed on September 17, thirty-seven days transition from science mode to prior to arrival. The HEND and NS were

17

. . . .. I

turned off on September 24. The final The goal of the active calibration effort was trajectory correction maneuver executed on to completely characterize the magnitude October 12, just twelve days prior to arrival. and direction of the thrust vector for each The operational MOI sequence was loaded RCS thruster pair. The calibration was onboard the spacecraft on September 15, designed to fire thruster pairs in sequence to nine days prior to MOI, and the spacecraft spin up, then spin down each reaction wheel. remained in a quiescent state for the The translational velocity change was then remainder of the flight, until two days out measured with the Doppler, and the body when the solar array was stowed in and wheel rates were captured in telemetry. preparation for the MOI burn. This activity This sequence was performed in an Earth- marked the beginning of the MOI sequence pointed attitude, as well as three off-Earth of events. attitudes. The combination provided a viewing profile that enabled the Doppler to er Callbratrons sense the vector components of the velocity change from three nearly orthogonal Reaction Wheel Assemblies (RWAs) attitudes. provide primary attitude control and are desaturated by the RCS thrusters. This The passive calibration was performed three event, known as an angular momentum months prior to encounter to ensure that the desaturation (AMD) is accomplished by thruster behavior had not changed firing the attitude control thrusters to unload significantly. It involved all of the data the momentum. The thrusters fire in pairs to collection, analysis, and interaction between desaturate each spacecraft axis sequentially, the teams that was required for the active but are not coupled. Desaturation events calibration, but did not involve any occurred on a daily basis throughout cruise. spacecraft attitude changes. Because each thruster firing imparted a net change in velocity (AV) to the spacecraft, v Correction the thruster telemetry was recorded and downlinked for flight team evaluation. Although the ideal Orbiter trajectory does Although the net translational AV from each not require any deterministic deep space event was small (less than 10 mdsec) the maneuvers to reach Mars, a schedule of four cumulative trajectory perturbation was quite trajectory correction maneuvers (TCMs) was large, on the order of 10,000 km. So careful established to provide for sufficient control trending and calibration was required to of the arrival conditions’. meet the delivery accuracy requirements. TCM-1 was designed as part of a multi- Two in-flight thruster Cali bration activities maneuver optimization strategy to correct were performed. An active calibration the injection errors, aimpoint bias and other occurred shortly after launch, which trajectory errors while maintaining involved slewing the spacecraft to view the appropriate conditions for planetary RCS thrusting from several different angles. protection. While the TCM-1 aimpoint is Monitoring continued throughout cruise, and selected as part of the maneuver one passive calibration was performed optimization strategy with some which did not involve attitude changes. consideration for planetary protection requirements, the TCM-2 aimpoint was explicitly biased to satisfy overall planetary

18 protection requirements for the cruise phase. maneuvers to ensure a telecom link for TCM-3 and TCM-4 corrected for the TCMs 2 and 3. TCM-4 was a statistical remaining trajectory errors and targeted clean-up maneuver, so could not be directly to the desired encounter conditions constrained a-priori, but a strategy was in preparation for the Mars Orbit Insertion developed to maintain communications. (MOI) burn. The AV required for the cruise phase turned All maneuvers were executed with the four out to be substantially less than planned, 22 N TCM thrusters. In all cases, the AV again due to the positive injection results. direction was constrained to ensure that the The table below presents the planned (99%) medium-gain antenna could maintain and actual AV and fuel usage associated communications with Earth at the burn with each maneuver, as well as the angle attitude. This constraint was incorporated from the antenna boresight to the Earth. into the aimpoint biases for the early

Table 3. TCM Performance

Planned Off-Earth

TCM-1 48ds 3.6m/s 1.2 kg 7" TCM-2 38ds 0.9ds 0.3 kg 14" TCM-3 3.5 m/s 0.45 m/s 0.15 kg 2" TCM-4 0.5 m/s 0.08 m/s 0.04 kg 20" Total 53ds 5ds 1.7 kg The total propellant budget for the cruise phase was dominated by the anticipated maneuvers, but it also included fuel The primary navigation responsibility during allocations for momentum management, as the cruise phase was to accurately determine well as for specific events and contingency. and control the trajectory of the spacecraft to Of the total cruise allocation of 21.9 kg, only deliver it to the desired aimpoint at Mars 3 kg were actually spent. This windfall was encounter6. This was accomplished by recognized shortly after launch, and the tracking the spacecraft radio signal to unspent fuel was termed strategic determine the orbit and designing propulsive propellant, as the project then had the maneuvers to alter the trajectory. Four decision of how best to utilize this maneuvers were scheduled to achieve the propellant to mitigate risk, modify the necessary delivery accuracy, with a fifth science strategy, or potentially extend the maneuver as a contingency in the final hours mission. The allocation of strategic prior to encounter. propellant was an ongoing analysis that continued throughout the cruise and Although standard radiometric orbit aerobraking phases. determination techniques were employed to navigate the spacecraft, the traditional Doppler and ranging measurements were supplemented by a series of interferometric measurements known as delta-differenced

19 ,

one-way ranging (ADOR). This The 30 delivery requirements were to meet measurement is independent of the the targeted altitude to within +25 km, the traditional radiometric measurements, and targeted inclination to within 0.2", and the provides crucial out-of-plane trajectory closest approach time to within f 10 information that is difficult to determine seconds. The table below presents the target from more traditional data types. This and achieved values and demonstrates that technique has been utilized with success in the requirements were met. Figure 11 the past, but the aging hardware and illustrates the target, the constraints and the software system was completely re-built in actual delivery resulting from TCM-4. The preparation for this mission. A total of 45 final 0.08 m/s maneuver was designed to ADOR measurements were planned and correct the residual targeting error left after obtained over a five-month period. TCM-3.

Table 4. Delivery Accuracy

ved Delta Altitude 300.0 km 300.7 km 0.7 km Inclination 93.47" 93.51' 0.04" Periapsis 02:29:58 02:29:58 < 1 sec Time (ET)

,I I. ! *I Id, Inclination I CO Knowledge ', Constraint I , I prior to *o 2~ TCM-4 Desian , I , ._,_II - - - - .----.-----:------_____*_---- - ______. I I t------. TCM-4 Target I 1 Altitude Deltvery 30 Constraint -----t /i, e25 km I !

I L, -475 -400 -3w -8w -460 -a -376 -aa6 B*T (km)

Figure 11. Final Delivery Accuracy

20 at 02:26:57 UTC on October 24 to initiate the Mars Orbit Insertion bum, and the Orbit Insertion phase began with the engine fired for just over twenty minutes. initiation of the Mars Orbit Insertion (MOI) Throughout the burn, the spacecraft sequence, nine days prior to encounter. At maintained a constant angular rate, termed a two days out, the solar array was stowed to pitchover, to increase the efficiency of the configure the spacecraft for the MOI burn. capture maneuver. Ten minutes into the The final TCM-5 maneuver opportunities at burn, the spacecraft passed behind Mars, as 24 hours, and six hours prior to MOI were viewed from the Earth, and communications not needed, and fault-protection was ceased as planned. The bum terminated disabled at 22 hours out. Fifteen minutes when the on-board accelerometers detected prior to the start of the burn, the engine was a 15% thrust decay, indicating that the filled, and shortly thereafter, the fuel and oxidizer had been depleted. Pyros were then oxidizer tanks were pressurized. Seven fired to permanently isolate the pressurant minutes prior to the burn, the spacecraft from the propellant tanks. Ten minutes after slewed to the pre-bum attitude, and at this the end of the burn, the spacecraft came point could only communicate via the MGA back into view of the Earth, and in a carrier-only configuration. The main communications was re-established. engine valves were opened simultaneously

-c 4

i Figure 12. Mars Orbit Insertion

The nominal main engine thrust of 703 N, that the accelerometer did not accurately and Isp of 322.4 sec, along with the 121.3 kg sense the depletion event. of available oxidizer, defined the expected MOI burn time of 1183 seconds, expected The post-maneuver analysis indicated that total propellant expenditure of 266 kg, and the main engine burned for 1219 seconds, AV of 1420 ds.Min and Max timers were expended 146.3 kg of fuel, 121.4 kg of set at 11 15 and 1225 seconds respectively to oxidizer, and produced a total AV of 1433.1 ensure proper bum termination in the event m/S.

21 Based on statistics associated with the main engine performance, expected RCS performance during the bum, and spacecraft mass properties, the capture orbit period was expected to be 19 hours i5 hours. Based on The aerobraking phase began just two days the planned aerobrake profiles and after orbit insertion. Over the next three constraints, the project was able to define a months, a total of 332 passes through the maximum capture orbit period of 22 hours Martian atmosphere slowed the vehicle, from which aerobraking could be safely reducing the orbit period to just under 2 initiated. If the capture orbit period hours6. The aerobralung phase required 24- exceeded this value, then a Period Reduction hour per day operations at both JPL and Maneuver (PRM) would need to be executed LMA and daily support of teams across the prior to the initiation of aerobraking to country. reduce the orbit period to an acceptable level. If needed, the Period Reduction The aerobraking campaign was subdivided Maneuver (PRM) would have executed into three phases. The Walk-Zn phase about 48 hours after MOI in mono- gradually lowered the periapsis altitude from propellant mode on the TCM thrusters. the 300 km capture orbit altitude down to However, the achieved capture orbit period the 110 km altitude desired to initiate main- was determined to be below the mean at phase aerobraking. Main-Phase constituted 18.6 hours and eliminated the need for a the bulk of the aerobraking mission both in period reduction maneuver. terms of time, and number of drag passes. The aerobraking rate was determined by the One other event that was closely monitored heat rate corridor, and the bulk of the period in the orbit insertion phase was potential reduction took place during this phase. As close approaches with Phobos, the inner apoapsis decays, the aerobraking rate is moon of Mars. The elliptical, polar capture eventually constrained by orbit lifetime. orbit would permit close encounters with Lifetime in this context is defined to be the Phobos for particular capture orbit periods. time it takes for apoapsis to decay to 300 Fortunately, the 18.6-hour capture orbit was km. Below this altitude, the orbit will out of phase with Phobos, and no collision degrade and the spacecraft will quickly avoidance maneuvers were necessary. spiral in. The WaZk-Out phase of this mission was defined to occur when the orbit lifetime reached 24 hours.

22 .

Nadir I\Velocity Figure 13. Odyssey Aerobraking Configuration

The dominant mission constraint throughout Variability of the Martian atmosphere aerobraking was heating on the solar panels significantly limited the ability to predict the during the drag pass, which limited the total density that would be observed on any given drag that could be achieved with each pass. orbit. Overall, the observed variability On the other hand, a desire to limit the total exceeded 35% (lo),measured as the ratio of number of drag passes, and a power the heating rates on successive passes. To constraint that limited the aerobraking accommodate this uncertainty, 100% margin duration, required that a minimum level of was generally maintained between the drag be attained in order to finish maximum targeted heating rate and the aerobraking successfully. These constraints flight allowable limit. While this margin were satisfied by instituting a range of protected the vehicle from excessive acceptable heating rates (function of density heating, the uncertainty in the periapsis and relative velocity) that were used to times that resulted from the limited define the targeted aerobraking trajectory prediction capability necessitated frequent profile. updates to the on-board sequences.

Energy balance was another key aerobraking Spacecraft Operations constraint. As aerobraking,progressed, the local true solar time (LTST) of the orbit A number of spacecraft activities were decreased from at an average rate of -2 required on each orbit to prepare the vehicle minutes per day due to the motion of Mars for the atmospheric conditions. Thirty about the Sun. As LTST decreases, solar minutes prior to the periapsis time, the occultation duration increases, reducing the catbed heaters were turned on to warm up power collection time to the arrays. A the RCS thrusters, the solar array was constraint was therefore imposed to articulated to the stowed position, maintain the LTST at the descending communications was stopped as the SSPA equator crossing to be later than 2 PM to was turned off, and the spacecraft would 99% confidence to ensure adequate power to slew to the nadir-point attitude. At about the spacecraft. This constraint effectively five minutes prior to contact with the limited the aerobraking duration. atmosphere, the attitude control mode was set to thruster control with loose deadbands.

23 As the spacecraft flew through the accuracy. 225 seconds of timing margin was atmosphere, the atmospheric torque and utilized to ensure that the spacecraft could attitude control thrusting would naturally transition into the proper configuration prior desaturate the reaction wheels. One out of to and following each drag pass, and this the atmosphere, the attitude deadbands were drove the periapsis timing requirement. The tightened, and control returned to the sequence build schedule was driven by the reaction wheels. The catbeds were turned ability to predict the upcoming periapsis off, and the spacecraft began its slew back to times to within the 225 second requirement. Earth-point attitude. Following the slew, communications was re-established as the Atmospheric variability was the largest SSPA was turned back on and the solar source of uncertainty in predicting the orbit array was articulated out of the restraint and timing. If the observed density for a given back to Sun-point. At this point the drag- drag-pass did not match the predlcted value, pass telemetry playbacks and navigation the amount of energy removed from the tracking commenced. Two additional orbit, and therefore the period reduction redundant telemetry playbacks were then achieved by the pass, would also not match scheduled in case the first was lost for any the prediction. Thus, the time of the next reason. Following the last playback, real- periapsis would be different from the pre- time telemetry was resumed until the next pass prediction. If the periapsis timing error activity . was determined to be greater than 225 seconds, the timing for the next sequence On selected orbits, the periapsis altitude was would need to be adjusted to reflect the new adjusted by performing an apoapsis expected periapsis time. maneuver. These aerobraking maneuvers (ABMs) were designed and built prior to Early in aerobraking, the nominal delta- MOI, and were designed to either raise or period per orbit was more than 20 minutes, lower periapsis and to execute at apoapsis. and the difference between the expected So on ABM orbits, an additional sequence periapsis time and the actual time could was built and uplinked. Twenty-five minutes easily be more than 225 seconds after only prior to the apoapsis, the catbeds were one pass; therefore, a new trajectory predict turned on to warm up the TCM thrusters, the was delivered after every drag-pass to build solar array was articulated to the stowed the sequence for the upcoming pass. Later position, communications was stopped as in aerobraking, as the nominal delta-period the SSPA was turned off, and the spacecraft per orbit decreased, the error in the periapsis would slew to the burn attitude. The burn time prediction also decreased so that, would initiate at apoapsis, followed by a gradually, more than one, and up to six slew back to Earth-point. The solar array orbits could be predicted within the 225 was then deployed back to Sun-point, the second constraint. SSPA turned back on to re-establish communications., and playback of the The Periapsis Timing Estimator (PTE) was recorded ABM telemetry data commenced. an onboard software process that was designed to autonomously adjust the Drag sequences were generated and sequence timing based on the time that the uplinked up to four times daily. The duration drag-pass was sensed by the onboard of each sequence was determined by the accelerometers. This capability was orbit period and the orbit timing prediction successfully used to extend the peripaisis

24 timing prediction capability in the later also pre-planned to be sufficiently large to phases of aerobraking when it was necessary raise periapsis altitude out of the to predict many orbits in the future. atmosphere. The 20 m/s ABXl maneuver executed on Apoapsis 336 on Jan 11,2002, It was at about the 9-hour orbit period when raised the periapsis altitude to 201 km, and Odyssey began to experience unusually low marked a successful conclusion to the atmospheric variability over the North Pole. aerobraking phase. We did not rely on this phenomenon ahead of time, but were able to take advantage of it The remaining maneuvers, although termed in flight. The heat-rate upper corridor limit ABX2-5, were really orbit transition was effectively raised, and the aerobraking maneuvers. ABX2 was planned to execute rate was increased for a time. several days after ABXl at a time when the periapsis point had naturally drifted to the The Navigation timing and sequence build equator. At this time, it was optimal to processes operated flawlessly throughout the perform a small inclination change, and aerobraking phase, and with the help of the raise the periapsis altitude again. The Periapsis Timing Estimator (PTE) the orbit inclination change was designed to set up timing requirement was never exceeded. The the desired Local Mean Solar Time (LMST) daily strategic planning process also drift desired for the science orbit. ABX2 was operated as planned, the flight allowable the largest mono-propellant maneuver of the solar array temperature was never exceeded. mission, and the 56 m/s maneuver was executed at apoapsis 393 on Jan 15, 2002. Following ABX2, the periapsis altitude was 419 km, and the inclination was 93.1'. The transition phase was designed to provide the time required to perform the ABX3 was designed at the same time as propulsive burns needed to achieve the final ABX2, and was scheduled to execute just mapping orbit, deploy the high-gain two days later to lower apoapsis and freeze antenna, and configure the spacecraft and the orbit. The frozen orbit condition requires the science payloads for mapping the periapsis point to be at the South Pole, operations. The spacecraft was maintained but at this time periapsis was still close to in an inertially fixed, Earth-pointed attitude, the Equator. The option to wait until with the solar array on the Sun for the periapsis naturally drifted to the South Pole majority of this phase. Momentum was undesirable, as it would delay the start management was accommodated by angular of the science mission by several weeks, and momentum desaturations that occurred up to the natural eccentricity variation would four times daily. require even more propellant to compensate. The 27 m/s ABX3 maneuver executed The primary navigation task during this successfully on orbit 417 on Jan 17, 2002. phase was to design and execute the five This maneuver established the frozen orbit transition maneuvers, beginning with the by rotating the periapsis point to the South aerobraking termination maneuver, ABXl. Pole, and at the same time reducing the apoapsis altitude to 450 km, and periapsis to The ABXl attitude was selected from a pre- 387 km. designed menu that was used for ail aerobraking maneuvers. The magnitude was

25 Following these large transition maneuvers, trim maneuver. The 400 km, polar near- two orbit trim maneuvers were planned to circular, frozen orbit provides the clean up any residual orbit error. ABX 2 and observational geometry desired by the 3 had executed just as planned, however, science instruments. The orbit period of just small execution errors and orbit propagation less than 2 hours results in roughly 12.5 uncertainty left the need for some small revolutions per Martian day, or sol. clean-up maneuvers. A week of tracking and Successive ground tracks are separated in maneuver design was allocated, and the two longitude at the equator by approximately trim burns ABX4 and ABX5 executed on 28.8" and the entire ground track pattern Jan 28 and 30, 2002. The combined AV of nearly repeats every 2 sols, with a 1' shift to the two maneuvers was 3 m/s. the West.

Once the propulsive maneuvers were The science orbit design was negotiated to completed, the high-gain antenna was balance the observational desires of successfully deployed on February 8, 2002. THEMIS with those of GRS. The MARIE Several spacecraft and science payload investigation is relatively insensitive to the housekeeping and checkout sequences were orbit design. The somewhat conflicting accommodated in the following weeks, and requirements that drive the orbit design are the spacecraft turned to nadir-point on that high-quality THEMIS infrared data are February 18,2002. The THEMIS instrument obtained at local true solar times (LTST) began imaging the planet on February 19, earlier than 5 PM, while GRS cooler 2002, and the Gamma Ray Spectrometer constraints require solar beta angles less began data collection, signaling the start of than -57.5'. The LTST and beta angle the 917-day science mission. In early March, profiles are controlled by the orbit the MARIE troubleshooting activity inclination, which affects the orbit nodal returned the radiation monitor to a fully precession rate, the rate at which the orbit operational state, resulting in a complete plane rotates in inertial space. Figure 15 science payload complement for the science displays the time-history of the science orbit mission. LTST and beta angle for the planned science mission. The figure also includes local mean e OrM solar time (LMST), and Mars to Earth range.

The science orbit was established on January 30, 2002 following the final transition orbit

26 3

2

1-

0 I I I I I I I I I I I I I I I I

-90 I I I I I I I I I I I I I I I I

Ahllll C A

I I 1 I I I I 1 I I I I I I I I I -40 ‘

” I I I I Ill I I I III Ill

0 100 200 300 400 500 600 700 800 900 1000 1100 1200 1300 1400 1500 1600 1700 DAYS FROM START OF SCIENCE MISSION (FEB. 20,2002) Figure 14. Science Orbit Geometry

The frozen orbit condition maintains a maneuver to establish a Sun-synchronous relatively fixed eccentricity and argument of orbit that would be desirable for an extended periapsis for a given semi-major axis. The mission. The second is a planetary periapsis point is “frozen” at the South Pole, protection orbit raise maneuver that would and the orbit altitude at any given latitude is occur at the end of mission. Propellant has constant at all longitudes. One key benefit of also been allocated for any unplanned orbit a frozen orbit for the Odyssey mission is that trim maneuvers that may be required to it keeps the Odyssey orbit away from the compensate for the orbit perturbation from orbit of the Mars Global Surveyor, which is the desaturation events. Propellant for also in a similar frozen orbit. momentum management is required to maintain attitude and normal mapping Only two planned maneuvers remain during operations. Contingency propellant has been the mission. The first is a plane-change allocated to accommodate off-nominal

27 operations, or safe-mode entries. The of propellant remains unallocated. The propellant budget was designed to provide a detailed propellant allocations are shown in 99% probability of accommodating a 1374- Table 5 (in kg, and equivalent AV). day (2 Mars-year) mission. The propellant windfall continues, as a significant amount

Table 5. Science Mission Propellant Allocation

Total Fuel Available: 45.1 kg 3 kg (Mar 1,2001)

Fuel (bd DV 0 Contingency 8.2 kg 40 ds Safe Mode 5.0 kg 20 ds Momentum Mgmt. 11.5 kg 47 m/s Orbit Trim Maneuvers 3.7 kg 18 ds Extended Science 1.9 kg 10 ds PQ Orbit Raise (EOM) 3.4 kg 18 ds

Unallocated Fuel Remaining: 11.4 kg

Science Mission and visual wavelengths to determine surface mineralogy and morphology, provide global The 2001 Mars Odyssey mission contributes gamma ray and neutron observations for a to the NASA Mars Exploration Program full Martian year, and study the Mars goals by a direct search for water in the radiation environment from The near-surface of Mars at present and a search science payload on 2001 Mars Odyssey for evidence of past water in the surface includes a gamma ray spectrometer, a multi- mineralogy and morphology. In particular, spectral thermal imager, and a radiation 2001 Mars Odyssey carries instruments that detector. An overview of these science will observe the Martian surface at infrared instruments is given in Table 6.

28 Table 6. Science Payload Overview

Instrument I Description I Principal Investigator THEMIS I Will determine the mineralogy of the Martian I Philip Christensen, (Thermal Emission surface using multispectral, thermal-infrared Arizona State University Imaging System) images that have 9 spectral bands between 6.5 and 14.5 pm. It will also acquire visible-light images with 18-m pixel resolution in either monochrome or color. GRS Will perform full-planet mapping of elemental William Boynton, (Gamma Ray abundance with an accuracy of 10% or better University of Arizona (GRS Team Spectrometer) and a spatial resolution of about 300 km, by Leader, PI for Gamma Sensor) remote gamma ray spectroscopy, and full-planet mapping of the hydrogen (with depth of water William Feldman, inferred) and COz abundances by remote Los Alamos National Laboratory neutron spectroscopy. (Neutron Spectrometer Scientist)

Igor Mitrofonov (High Energy Neutron Detector, PI) MARIE Will measure the accumulated absorbed dose Gautam Badhwar, (Martian Radiation and dose rate tissue as a hnction of time, Johnson Space Center Environment determine the radiation quality factor, determine (PI, deceased) Experiment) the energy deposition spectrum from 0.1 keV/pm to 1500 keV/pm, and separate the Cary Zeitlin, PI contribution of protons, neutrons, and ME National SpaceBiomedical particles to these quantities. Research Institute Baylor University

Ther- .. Svstm (THEMIS)

THEMIS addresses the Odyssey objectives exploration of Mars. Specific science of globally mapping the elemental objectives of the THEMIS experiment are composition of the surface, acquiring high to: spatial and spectral resolution images of the surface mineralogy, and providing 1 ) Determine the mineralogy and information on the morphology of the petrology of localized deposits Martian surface. THEMIS will characterize associated with hydrothermal or sub- the Martian surface environment by aqueous environments, and to providing high- spat i a1 and identify sample return sites likely to high-spectral-resolution mineralogical and represent these environments. morphological data by means of visible and 2) Provide a direct link to the global infrared imagery. Mineralogical and hyper-spectral mineral mapping from morphological measurements will help to the Mars Global Surveyor (MGS) determine a geologic record of past liquid thermal emission spectrometer (TES) environments. Furthermore, mineralogy and by utilizing the same infrared petrology data will help identify potential spectral region at high (100 in.) Mars landing sites. This experiment will spatial resolution. provide essential information for selecting 3) Study small-scale geologic processes future landing sites aimed at exobiologic and landing site characteristics using

29 morphologic and thermo-physical hyper-spectral mapping. Furthermore, the properties. THEMIS filters were optimized utilizing 4 )Search for pre-dawn thermal knowledge of Martian surface minerals anomalies associated with active determined from TES data, and TES global subsurface hydrothermal systems. maps will allow efficient targeting of areas with known concentrations of key minerals. To accomplish these objectives, THEMIS Also, THEMIS will achieve infrared will determine surface mineralogy using signal-to-noise ratios of 33 to 100 for multispectral thermal-infrared images in 9 surface temperatures (235-265 K) typical spectral bands from 6.5-14.5 pm with 100- for Odyssey’s mapping, -4:30 p.m., orbit. m pixel resolution. THEMIS will also In addition, this orbit is ideally suited to the acquire visible images at 18 dpixel in up to search for pre-dawn temperature anomalies 5 spectral bands for morphology studies and associated with active hydrothermal landing site selection. The THEMIS systems, if they exist. The visible imager thermal-infrared spectral region contains the will have a signal-to-noise ratio of greater fundamental vibrational absorption bands than 100 at 5:OO p.m. local time. The that provide the most diagnostic information THEMIS instrument weighs 10.7 kg, is 28 on mineral composition as all geologic cm wide x 30 cm high x 31 cm long, and materials, including carbonates, consumes an orbital average power of 5.1 hydrothermal silica, sulfates, phosphates, W. hydroxides, silicates, and oxides have strong absorptions in the 6.5-14.5 pm region. Thus, silica and carbonates, which are key ectrometer (m diagnostic minerals in thermal spring deposits, will be readily identified using The GRS instrument suite consists of three thermal-infrared spectra. Remote sensing instruments including a gamma sensor head studies of terrestrial surfaces, together with (GSH), a neutron spectrometer (NS), and a laboratory measurements, have high-energy neutron detector (HEND). demonstrated that 9 spectral bands are These instruments address the 2001 Mars sufficient to detect minerals at abundances Odyssey objectives of globally mapping the of 5-10%. The use of long wavelength elemental composition of the surface, and of infrared data has additional advantages over determining the abundance of hydrogen in shorter-wavelength visible and near-infrared the shallow subsurface. Thus, this data because it can penetrate further through instrument suite plays a lead role in atmospheric dust and surface coatings, and determining the elemental makeup of the the absorption bands are linearly Martian surface proportional to mineral abundance even at very fine grain sizes. When exposed to cosmic rays, chemical elements in the Martian near-subsurface THEMIS was designed as the follow-on to emit gamma rays with distinct energy levels. the Mars Global Surveyor Thermal By measuring gamma rays coming from the Emission Spectrometer (TES), which Martian surface, it is possible to calculate produced a hyper-spectral (286-band) surface elements’ distributions and mineral map of the entire planet. THEMIS abundances. In addition, measuring covers the same wavelength region as the neutrons provides a measurement of TES, eliminating the need for additional hydrogen abundance in the upper meter of

30 subsurface, which in turn provides wide. The NS is 17.3 cm long, 14.4 cm tall inferences about the presence of near- and 31.4 cm wide. The HEND measures surface water. Measuring neutrons also 30.3 cm long, 24.8 cm tall and 24.2 cm provides information about CO, abundances wide. The instrument's central electronics in the upper meter of subsurface. box is 28.1 cm long, 24.3 cm tall and 23.4 cm wide. The experimental objective of the GRS is to determine the elemental composition of Gamma Sensor Head (GSH) Mars' surface by full-planet mapping of elemental abundance with an accuracy of The GSH will detect and count gamma rays 10% or better and a spatial resolution of emitted from the Martian surface. By about 300 km by remote gamma ray associating the energy of gamma rays with spectrometry, and full-planet mapping of the known nuclear transitions and by hydrogen (with depth of water inferred) and determining the number of gamma rays CO, abundances by remote neutron emitted from a given portion of the Martian spectrometry. The instrument is also surface, it is possible to calculate the ratio of sensitive to gamma ray and particle fluxes elemental abundances of the surface and from non-Martian sources and will be able discern their spatial distribution. While the to address problems of astrophysical interest energy represented in these emissions including gamma ray bursts, the determines which elements are present, the extragalactic background, and solar intensity of the spectrum reveals the processes. The GRS principal investigator elements' concentrations. These energies is Dr. William Boynton of the University of will be collected with 600-km resolutions Arizona. over time and used to build up a full-planet map of elemental abundances and their The GRS (as noted above) consists of distributions. The GSH uses a high-purity several components. The GSH is separated germanium detector cooled below 100 K to from the rest of the spacecraft by a 6-m (20- measure gamma ray flux. GSH performance ft) boom, which was extended after Odyssey is a strong function of its temperature, which entered its mapping orbit. This minimizes in turn constrains the spacecraft orbit beta the interference from gamma rays coming angle (angle between orbit normal and from the spacecraft itself. The initial direction to Sun) to insure that the GSH spectrometer activity, lasting about 100 days cooler is shaded from the Sun. The orbit once the spacecraft was in its mapping orbit, beta angle must be less than -57.5" (-56" to provided instrument calibration before the shade the cooler plus pointing uncertainty of boom was deployed. After about four 1.5") in order to acquire useful data. Initial months in the mapping orbit, the boom was instrument calibration will be delayed until deployed and it will remain in this position the beta angle geometry is satisfactory for for the duration of the mission. The NS and GRS data acquisition. A periodic annealing the HEND components of the GRS are of the germanium detector on the GRS may mounted on the main spacecraft structure also be required. Each annealing cycle takes and will operate continuously throughout the approximately 7 days. mission. The entire GRS instrument suite Neutron Spectrometer (NS) weighs 30.2 kg and uses 32 W of power. Along with its cooler, the GSH measures The NS measures neutrons liberated from 46.8 cm long, 53.4 cm tall and 60.4 cm the near-subsurface of Mars by cosmic rays.

31 ,

Since Mars has a thin atmosphere and no electronically. Details of the instrument and global magnetic field, cosmic rays pass the Doppler filter technique for separating unhindered through the atmosphere and thermal and epithermal neutrons have been interact with the surface. Cosmic ray given el~ewhere'~ bombardment of nuclei of subsurface material down to about 3 m produces a large The NS is provided and operated by the Los number of secondary neutrons. These Alamos National Laboratory (LANL). neutrons in turn propagate through the William Feldman at LANL is the NS team subsurface and interact on the way out with leader within the GRS Team. subsurface nuclei. Fast neutrons produced by the cosmic rays may in turn be moderated High Energy Neutron Detector (HEND) by collisions with nuclei before they escape from the subsurface, resulting in neutrons The HEND complements the NS as it with thermal or epithermal energies. The measures higher energy neutrons. HEND flux of secondary neutrons from the surface and NS together will map near-subsurface is the neutron albedo of Mars, which is water and rock terrains. "D consists of a measured by the NS. Goals of the NS are to: set of five particle sensors and their electronics boards. The sensors include Map the distribution of hydrogen three proportional counters and a within 1 m of the surface. scintillation block with two scintillators. Map the seasonal variation of COz Proportional counters and an internal ice and frost that forms on the polar scintillator detect neutrons with different caps during their winter seasons. energies. When all these sensors are on, Map the major compositional HEND measures neutrons over a broad provinces on Mars. energy range from 0.4 eV up to 10.0 MeV. Provide maps of the neutron number HEND also helps calibration of the gamma density and fast-neutron flux at the sensor. surface of Mars for use in converting measured gamma-ray line strengths "D will also provide astrophysical data to elemental abundances. about the nature of gamma-ray bursts and about the physics of solar activity. HEND The NS sensor consists of a cubical block of data from extragalactic gamma ray bursts borated plastic scintillator that is segmented (GRBs) will be used with the data from into four equal volume prisms. In the Ulysses and near-Earth satellites (HE=-2, mapping orbit, one of the prisms faces Wind, etc.). Interplanetary triangulation, a forward into the spacecraft velocity vector, technique involving accurate timing of burst one faces backward, one faces down toward arrival times, allows the sky positions of the Mars, and one faces upward. Neutrons sources of GRBs to be determined with coming directly from Mars will be separated accuracy of several minutes of arc. HEND from those that are reprocessed by the can also observe solar flares from different spacecraft using a combination of velocity points in the Solar system. The filtration (because the spacecraft in orbit simultaneous measurement of gamma rays about Mars travels faster than a thermal and high-energy neutrons from powerful neutron) and self-shielding of one prism by solar flares at Mars combined with those the other three. Fast neutrons are separated from the vicinity of Earth allows a from thermal and epithermal neutrons stereoscopic image of the active region on

32 the Sun. These stereoscopic observations of powerful flares will provide the possibility to build a three-dimensional model of the generation of hard-electromagnetic radiation The MARIE addresses the Odyssey and corpuscular emission in active regions objective of characterizing the Martian near- on the Sun. space radiation environment as related to radiation-induced risk to human explorers. HEND was developed in the Laboratory of As space radiation presents an extreme Space Gamma Ray Spectroscopy at the hazard to crews of interplanetary missions, Space Research Institute (Moscow, Russia). MARIE’S goal is to measure radiation doses HEND is operated by the Russian Aviation that would be experienced by future and Space Agency’s Space Research astronauts and help determine possible Institute in Moscow. Igor Mitrofonov is the effects of Martian radiation on human principal investigator. beings. Hazardous space radiation comes from two sources: energetic particles from the Sun and galactic cosmic rays from beyond our solar system. Both kinds of radiation can trigger cancer and damage the The GRS cooler door was closed on June 1, central nervous system. A spectrometer 2002, and the spacecraft slewed to the safe- inside MARIE will measure the total energy mode (Earth-point) attitude on June 2, in from these radiation sources, both in cruise preparation for GRS boom deployment. A from Earth to Mars and in the Martian orbit. GRS thermal control test and a As the spacecraft orbits Mars, the demonstration of the control modes to be spectrometer sweeps through the sky and used during deployment were conducted on measures the radiation field coming from June 3. The boom deployment release different directions. Specifically, MARIE mechanism was activated on June 4. The goals are to: downlink signal from the spacecraft was lost shortly thereafter, as expected, due to the 1) Characterize specific aspects of the spacecraft motion induced by the boom Martian near-space radiation deployment. Following rate damping and a environment, slew back to Earth-point, high-rate downlink 2 ) Characterize the surface radiation was reestablished, and engineering data for environment as related to the deployment event was downlinked. radiation-induced risk to human Subsequent analysis indicated that the boom exploration, and deployment was nominal, and the inertial 3) Determine and model effects of the measurement unit signature of the boom atmosphere on radiation doses latching in its fully extended position was observed on the surface. observed. Following boom deployment, the GRS cooler door was reopened, and the The principal investigator for the MARIE instrument high voltage was ramped up to experiment is Cary J. Zeitlin of the National begin science data collection. Space Biomedical Research Institute at Baylor University in Houston. The MARIE instrument was provided by NASA’s Human Exploration and Development of Space (HEDS) Program in order to

33 characterize the radiation environment at and map hydrogen abundance in the upper - Mars. The instrument, with a 68-degree 1 m of the ~oil.’~-’~A strong signature of field of view, is designed to continuously hydrogen was detected by all three sensors collect data during Odyssey’s cruise from in the high southern latitudes, south of about Earth to Mars and in Mars orbit. It can store 60” S. latitude. A deficit in the epithermal large amounts of data for downlink neutron flux, which is diagnostic of whenever possible, and will operate hydrogen, was mapped by both “D and throughout the entire science mission. The NS. Figure x is the epithermal neutron map instrument weighs 3.3 kg (7.3 lbs) and uses obtained by NS from 30 days of mapping. 7 W of power. It measures 29.4 cm (1 1.6 in) A deficit in epithermal neutron flux is also long, 23.2 cm (9.1 in) tall and 10.8 cm (4.3 observed in parts of the high northern in) wide. latitudes, but most of the signature is likely masked by the cover of seasonal carbon dioxide frost in the north. The high flux of 1v Science Results thermal neutrons seen by NS in the north is consistent with this interpretation.16 By The Odyssey science mission began on comparing the flux of epithermal and February 19, 2002, with the turn-on of the thermal neutrons, and the H gamma GRS instrument suite and THEMIS. emission line observed by the gamma MARIE began orbital science data sensor, Boynton et al.,15 were able to collection on March 13, 2002. This section estimate the depth and quantity of hydrogen describes some science highlights from the in the high southern latitudes. Model results first several months of the science mission. suggest that the upper meter of the regolith contains a “dry” zone in the upper several 10s of cm, below which is an ice-rich zone GBS with a mass fraction of water ice of 20-50%. Depth to the ice-rich zone decreases with The GRS suite of instruments (“D,NS distance to the south pole. and Gamma Sensor) was used to measure

34 Figure 15: A global view of Mars in in termedi ate-energ y , or epi thermal, neutrons. Soil enriched by hydrogen is indicated by the deep blue colors on the Early THEMIS observations were targeted map, which show a low intensity of at high priority sites, including candidate epithermal neutrons. Progressively smaller landing sites for upcoming Mars lander amounts of hydrogen are shown in the missions. Data were acquired in all camera colors light blue, green, yellow and red. The modes: day and night time infrared and day deep blue areas in the polar regions are time visible. Performance of the camera believed to contain up to 50 percent water system was excellent. Infrared images, both ice in the upper one meter (three feet) of the day and night, showed a remarkable soil. Hydrogen in the far north is hidden at diversity of temperature signatures of this time beneath a layer of carbon dioxide surface materials. Figure y is a mosaic of frost (dry ice). Light blue regions near the several day time single band infrared images equator contain slightly enhanced near- near the hematite-rich candidate landing site surface hydrogen, which is most likely in Terra Meridiani. Temperature variations chemically or physically bound because in day time scenes result from differences in water ice is not stable near the equator. The local slope relative to the sun, thermal view shown here is a map of measurements inertia and albedo of surface materials. In made during the first three months of the Meridiani image, variations in both mapping using the neutron spectrometer thermal inertia and albedo cause strong instrument, part of the gamma ray contrast in temperatures, indicating spectrometer instrument suite. The central differences in the physical properties of the meridian in this projection is zero degrees layered materials currently exposed at the longitude. Topographic features are surface. superimposed on the map for geographic reference.

35 Figure 16: THEMIS infrared imaging exposures taken by the thermal emission shows signs of layering exposed at the imaging system aboard Odyssey during the surface in a region of Mars called Terra first two months of the Odyssey mapping Meridiani. The brightness levels show mission, which began in February 2002. The daytime surface temperatures, which range area shown is about 120 kilometers (75 from about minus 20 degrees to zero degrees miles) across, at approximately 358 degrees Celsius (minus 4 degrees to 32 degrees east (2 degrees west) longitude and 3 Fahrenheit). Many of the temperature degrees north latitude. variations are due to slope effects, with sun- facing slopes warmer than shaded slopes. Images from the THEMIS visible camera However, several rock layers can be seen to were published at the THEMIS website five have distinctly different temperatures, days a week (h ttp://w w w .themi s.as u). indicating that physical properties vary from Figure z is a single band visible image of layer to layer. These differences suggest that another region in Terra Meridiani, showing the environment on this part of Mars varied heavily cratered terrain modified by through time as these layers were formed. channels and gullies. The image is a mosaic combining four

36 37 Figure 17: A THEMIS visible image shows MARIE a region in Terra Meridiani near -12" S, 358" W (2" E). An old, heavily degraded Once nominal operation of MARIE was channel can be seen from the lower restored, the instrument collected radation (southern) portion of the image toward the monitoring data nearly continuously. Figure top. The walls of several craters in this zz shows a sample of MARIE data collected image show vague hints of possible gully during the first several weeks of operations. formation at the bottom of pronounced rock Radiation dose levels were close to those layers, with the suggestion of alcoves above predicted from cosmic ray models. Solar the individual gullies. Image size is 57.8 x activity was high during this time, which is 23.5 km. reflected in the "spikes" of dose rate in the MARIE data.

MARIE Measurements vs HZETRN Model Calculations

- 4 - HZETRN Model Calculations 40 --

3 1 h )r (II 9 30 5E Y

CIa, 2 20 a, 08 10 --

11....1.1.1.1.....1.....11.... 0 I I I I I 0313/02 03/16/02 03/19/02 03/22/02 03/25/02 03/28/02 03/31102 Calendar Date (UTC) Figure 18: Absorbed dose rate as a function calculated from the HZETRN model of time since MARIE was powered up in utilizing the daily solar deceleration March. The gaps in data are related to parameter, phi for the generation of the downloads and housekeeping tasks. The Galactic Cosmic Radiation spectra. solid line is the expected dose rate

38 Risk Manwement risks have significant mission impact and require quick turnaround; on-the-shelf t Trees commands are built and formal contingency procedures are developed for Category A During the development phase, the project risks. Category B risks do not require quick conducted a series of risk reviews to identify turnaround; informal procedures were mission risk areas, and develop approaches written to serve as a reference for these for risk mitigation’. A working group contingencies. Category C risks are not comprised of systems, subsystems, mission critical, and are tracked without operations, and the Assembly, Test and written response plans. Launch Operations (ATLO) team was established to review each mission phase. duc- As part of the risk assessment, a fault tree was developed for each mission phase. The The Odyssey team instituted a program of fault trees were used to identify what testing called “risk reduction testing” or scenarios could lead to a mission failure, and “flight software stress testing”. This test to identify mitigations that prevent or reduce program augmented the project verification the probability of occurrence for that fault. and validation process. Risk reduction It also listed how the mitigations were tested testing began during the development or analyzed. The fault tree is organized by program, and has continued through function; in other words, it looks at what operations. functionally needs to occur to achieve the objectives of that phase. It is complimentary Key characteristics of the risk reduction test to the System Failure Modes, Effects, and program include: Criticality Assessment (FMECA) because it covers failures from functional viewpoint Focus on software performance in (e.g. what has to work) while the System off nominal conditions. FMECA lists how a specific component Core team is independent of project could fail (i.e. provides list of failure modes software development and to feed into fault tree). The fault tree acceptance test team and project contains a verification table listing each fault verification and validation test in the tree, how the project is mitigating that team. fault, and where that mitigation is verified. Multiple combinations of failures, under-performing hardware, and fault protection cases are included in test cases. Mindset of test case definition is The faults identified in the risk assessment “find a way to break it”, initial were grouped into general categories, and conditions allowed to go beyond then rated in terms of likelihood of specification limits. occurrence, impact on the mission, and Review of test case results required recovery time. These ratings performed by the main-line determine the category of the risks; development and V&V team. contingency plans are developed according to the risk category. Category A mission

39 t

6) Results of test cases tracked by a formal action itedfailure reporting system. About one year before launch a “Red Team” 7) Schedule progress tracked and was formed to review the Odyssey project. managed via an earned value The characteristics of a Red Team include: system. 8) Test cases segregated by critical Complete independence from the mission phases. Project and limited independence 9) The suite of risk reduction testing from the Jet Propulsion Laboratory. was run on three test-beds - desktop Membership drawn from JPL, other software simulators, a high fidelity NASA Centers, federally funded spacecraft simulator, and the flight research and development centers, vehicle on an approximate ratio of industry, and Academia. 100:10:1 . Chartered to perform an independent, de tailed, The risk reduction test program differs from comprehensive assessment of the the software acceptance plan or the project entire Odyssey project. verification and validation plan. The latter 0 Red Team report to identify risk primarily verify that the system is designed areas and recommend risk mitigation to specifications and requirements and that actions to improve the prospect for the specifications and requirements are mission success. appropriate and sufficient for the mission. 0 Scope of review included the flight Risk reduction testing is focused on the system (including subsystems), the behavior of the spacecraft as a system when Mission Operations System, and the operating in both nominal and off nominal science instruments. conditions. The Red Team was divided into a lead The risk reduction process proved to be position (non-JPL) with a JPL co-lead. quite valuable to the Odyssey project. Sixteen subteams were formed to penetrate Susceptibilities were identified by the risk specific functions of the Project. The reduction team, which the project then subteams developed findings and risk items alleviated. Risk reduction testing offers the independent of the Red Team leads. Red capability to perform many combinations Team lead and co-lead did not change and permutations of spacecraft operation, findings nor risk ratings submitted by the even in the condition of multiple faults subteams. and/or degraded hardware performance. Ultimately, this resulted in an improved The Red team process was roughly divided characterization of system level into two phases. In phase one, the Red performance, allowing increased robustness Team was presented with design and increased assurance of mission success. information via a technical interchange with the project team. Follow-up actions were completed and the Red Team produced a preliminary findings report. The report categorized identified risks and rated them for mission consequence and likelihood of occurrence.

40 Mars Orbit Insertion and In phase two, the project worked the risk Aerobraking items reported to reduce the likelihood High Gain Antenna Deployment andor the consequences of the identified Transition to Science Mapping risks. The project response varied with the Gamma Ray Spectrometer Boom particular concern. In some cases, the Red Deployment Team was provided with additional information, once the nature of the concern The purpose of a CERR is to demonstrate was understood. Other responses included that all preparation is completed and that the changes to the design, changes to the Project is ready to proceed with low risk. sequence of events, additional qualification The independent review board membership or characterization testing, or further draws from JPL, our industrial partner analysis. Lockheed Martin Astronautics, and Science members. The review board report is The first report identified 117 risk items. As submitted to the project and to the GPMC, the Project and Red Team worked to which authorizes the project to proceed with mitigate these risk items to low risk andor the critical event. low likelihood additional risk items were identified. All Project closures were The CERR is a comprehensive review which submitted to the initiator of the risk item for addresses: concurrence. All risk items, except two, were eventually closed as low residual risk. Mission, spacecraft, and Mission The two exceptions were: 1) the level of Operations performance to date. fidelity of the test bed should be improved, Status of the Incident, surprise, and and 2) the spacecraft should operate in a anomaly (ISA) formal reporting dual processor, hot back-up mode during the system, with particular emphasis on Mars orbit insertion maneuver. These two ISAs which may be relevant to the risk items were the subject of much critical event. discussion. Ultimately the Project, which Status of relevant action items. considered these to be low risk, and the Red Review of development issues, as Team, which rated these as significant required. (medium or high) risk, agreed to disagree. Reporting of personnel staffing and The results were presented to Governing transition plans Program Management Council (GPMC). Status of the operational sequences, The GPMC concurred with the project’s design, liens, test status, approval recommendation to accept the residual risk status. areas identified by the Red Team. Status of the contingency plans and discussion of robustness to faults. .. Event Rea- Reviews Test verificatiodvalidation performed, risk reduction test status. The Project conducted a Critical Event Red Team report, as required. Readiness Review (CERR) prior to all Staffing and training report. critical mission events in the operations Identification of on-call technical phase. The critical events are: experts. Facilities report, includmg status of backup facilities.

41 0 Operational freeze status. managing risk along the way, has been a 0 Work to go plans. complex and arduous undertaking. The 0 Media and public outreach plans, as early science discoveries of the mission are required. a testament to the dedicated work of the distributed Odyssey team. The flight The CERR is typically schedule 4 weeks vehicle is in excellent operating condition, prior to the event. The GPMC reviews the and is poised to provide a rich harvest of results of the CERR typically two weeks science data from Mars for many years to prior to the event. At this time all items come. were 100% closed. Ultimate authority to proceed with the critical events is given by Acknowledgements the GPMC. The work described in this paper was The Critical Events Readiness Review performed at the Jet Propulsion Laboratory, process provided a method to systematically California Institute of Technology, under assess the readiness of the Project and contract with NASA. The authors would provided high value to the project and our like to acknowledge the contributions of the sponsors. spacecraft team at Lockheed Martin Astronautics in Denver, Colorado, and the science investigation teams at the University of Arizona, Arizona State University, NASA The process of developing the 2001 Mars Johnson Space Center, Los Alamos National Odyssey flight vehicle, launching and Laboratory, and the Russian Space Institute. operating the system throughout cruise, aerobraking, and the science mission, and 5Mase, R., Antreasian, P., Bell, J., Martin- Mur, T., Smith, J.C., The Mars Odyssey Navigation Experience, to be presented at the AIMAAS Astrodynamics Specialists ‘Mars Surveyor 2001 Project Policies, JPL Conference, August 54,2002, Monterey, D- 16091, October 2,2000. California.

2Mars Surveyor 2001 Mission Plan, 6Antreasian, P., et al, 2001 Mars Oavssey Revision B, JPL D-1303, August, 2000. Orbit Determination During Interplanetary Cruise, AIAA-2002-453 1, AIMAAS ’Spencer, D., Bell, J., Beutelschies, G., Astrodynamics Specialist Conference, Mase, R., Smith, J.C., 2001 Mars Odyssey Monterey, California, 5-8 Aug 2002. Mission Design, Paper No. AAS 0 1-39 1, AAS/AIAA Astrodynamics Specialists I Smith, J.C., Bell, J., 2001 Mars Odyssey Conference, July 30-August 2,2001, Aerobraking, to be presented at the Quebec City, Quebec, Canada. AIMAAS Astrodynamics Specialists Conference, August 54,2002, Monterey, 4Mars Odyssey 2001 Baseline Reference California. Mission, Rev. B, Doc. No. MSPO1-99-0195, 30 March 2001. ‘Beutelschies, G., “That One ’s Gotta Work ” Mars Odyssey’s use of a Fault Tree Driven

42 Risk Assessment Process, presented at the 14Feldman,W. C., Prettyman, T. H., Tokar, IEEE Aerospace Conference, March 9- 16, R. L., Boynton, W. V., Byrd, R. C., Fuller, 2002, Big Sky, Montana. K. R., Gasnault, O., Longmire, J. L., Olsher, R. H., Storms, S. A., Thomton, G. W., 2001. 'Saunders, R. S., 2000. The Mars Surveyor The Fast Neutron Flux Spectrum Aboard Program - Planned Orbiter and Lander for Mars Odyssey During Cruise, American 2001,3 1st Annual Lunar and Planetary Geophysical Union, Fall Meeting 200 1, Science Conference, March 13-1 7,2000, abstract #P42A-055 0. Houston, Texas, abstract no. 1776. "Boynton, W.V., W. C. Feldman, S. W. "Saunders, R. S., 2001a. 2001 Mars Squyres, T. Prettyman, J. Briickner, L. G. Odyssey Mission Science, American Evans, R. C. Reedy, R. Starr, J. R. Arnold, Geophysical Union, Fall Meeting 200 1, D. M. Drake, P. A. J. Englert, A. E. abstract #P4 1A-08. Meager, Igor Mitrofanov, J. I. Trombka, C. d'Uston, H. WMe, 0. Gasnault, D. K. 11Saunders, R. S., 2001b. Odyssey at Mars - Hamara, D. M. Janes, R. L. Marcialis, S. Cruise and Aerobraking Science Summary, Maurice, I. Mikheeva, G. J. Taylor, R. American Astronomical Society, DPS Tokar, C. Shinohara, 2002. Distribution of Meeting #33, #48.07. Hydrogen in the Near-surface of Mars: Evidence for Sub-surface Ice Deposits 12 Saunders, R. S., and Meyer, M. A., 2001. Science 297: 81-85. 2001 Mars Odyssey: Geologic Questions for Global Geochemical and Mineralogical I6Feldman, W.C., W. V. Boynton, R. L. Mapping, 32nd Annual Lunar and Planetary Tokar, T. H. Prettyman, 0. Gasnault, S. W. Science Conference, March 12-1 6,2001, Squyres, R. C. Elphic, D. J. Lawrence, S. L. Houston, Texas, abstract no. 1945. Lawson, S. Maurice, G. W. McKinney, K. R. Moore, R. C. Reedy, 2002. Global I3Saunders, R. S., Ahlf, P. R., Arvidson, R. Distribution of Neutrons @om Mars: Results E., Badhwar, G., Boynton, W. V., from Mars Odyssey. Science 297: 75-78. Christensen, P. R., Friedman, L. D., Kaplan, D., Malin, M., Meloy, T., Meyer, M., I7Mitrofmov, I., D. Anfimov, A. Kozyrev, Mitrofonov, I. G., Smith, P., Squyres, S. W., M. Litvak, A. Sanin, V. Tret'yakov, A. 1999. Mars 2001 Mission: Science Krylov, V. Shvetsov, W. Boynton, C. Overview, 30th Annual Lunar and Planetary Shinohara, D. Hamara, R. S. Saunders, Science Conference, March 15-29,1999, 2002. Maps of Subsurface Hydrogen from Houston, TX, abstract no. 1769. the High-Energy Neutron Detector, Mars Odyssey Science 297: 78-8 1.

43