The Study of the Anomalous Acceleration of Pioneer 10 and 11
Slava G. Turyshev, John D. Anderson ((Jet Propulsion Laboratory, Caltech)) Michael Martin Nieto ((Los Alamos National Laboratory, U of California))
Journes du GREX 2004 Nice, France, 29 October 2004
THE STUDY OF THE PIONEER ANOMALY THE STUDY OF THE PIONEER ANOMALY
Conclusions & Outline:
The Pioneer 10/11 anomalous acceleration:
aP = (8.74 1.33) 108 cm/s2
A line-of-sight constant acceleration towards the Sun:
We find no mechanism or theory that explains the anomaly Most plausible cause is systematics, yet to be demonstrated
Phys. Rev. D 65 (2002) 082004, gr-qc/0104064
Possible Origin? Conventional Physics [not yet understood]:
Gas leaks, heat reflection, drag force, etc
New Physics [many proposals exist, some interesting] Both are important a win-win situation:
CONVENTIONAL explanation: improvement of spacecraft engineering for precise navigation & attitude control NEW physics: would be truly remarkable

Pioneer 10/11 Mission

Built: TRW (Northrop-Grumman Space Technology) Navigation: Jet Propulsion Laboratory, Caltech Project management: NASA Ames Research Center
Position of Pioneer 10 on 29 October 2004:
Last successful precession maneuver to point the spacecraft to Earth was accomplished on 11 Feb 2000 (distance from the Sun of 75 AU)

Pioneer 10/11 Spacecraft

Pioneer 10/11 were excellent for dynamical astronomy:
Spacecraft design permits precise acceleration estimations, ~108 cm/s2, unlike a Voyager-type 3-axis stabilization
Accurate celestial mechanics experiments - one of the main objectives of the Pioneer extended missions

On-board Power and Heat

Thermal system and on-board power:
Design based on well understood process of on-board nuclear-toelectric energy conversion and heat dissipation within the craft
Pioneer F during checkout tests
The Pioneer F spacecraft during a checkout with the launch vehicle third stage at Cape Kennedy. Pioneer F became Pioneer 10.
Pioneer 10 Launch: 2 March 1972
Pioneer 10/11: Main Missions
Trajectories for Pioneer 10 and 11 during the main mission phase
Trajectories of Pioneers and Voyagers
Ecliptic pole view of Pioneer 10, Pioneer 11, and Voyager trajectories. Digital artwork by T. Esposito. NASA ARC Image # AC97-0036-3.

Detection of the Anomaly

Mid 1979 (search for Planet X with Pioneer 10):
Solar-radiation pressure away from the Sun became < cm/s2 Search for unmodeled accelerations started (~ 20AU)
Early 1980 (Orbit Determination Analysis ODP):
JPL analysis found the biggest systematic error in the accel residuals is a constant bias aP ~ (8 3) 108 cm/s2 directed towards the Sun
An ODP plot of the early unmodeled accelerations of Pioneer 10 and Pioneer 11, from about 1981 to 1989 and 1977 to 1989, respectively

The Observed Anomaly

In the 1995-98, from the JPL-ODP analysis we concluded:
There is an un-modeled acceleration towards the Sun (8.09 0.20) 108 cm/s2 for Pioneer 10 (8.56 0.15) 108 cm/s2 for Pioneer 11 The error is determined with a 5-day BSF with radial accel as a stochastic parameter subject to white Gaussian noise (~500 independent 5-day samples of radial acceleration). NO magnitude variation with distance over a range of 40 to 70 AU
PRL 81(1998) 2858-2861, gr-qc/9808081
The two-way Doppler anomaly to first order in (v/c) behaves as:
Equivalent forms of the Anomaly:
Steady frequency drift: Anomalous acceleration: Clock acceleration:
Modeling the Motion of Pioneer 10/11
Relativistic eq.m. for celestial bodies are correct to (v/c)4:
Relativistic grav. accelerations (EIH) include: Sun, Moon, 9 planets are point masses in isotropic, PPN, N-body metric; Newtonian gravity from large asteroids; terrestrial, lunar figure effects; Earth tides; lunar physical librations
Relativistic models for light propagation are correct to (v/c)2:
Standard Models of Non-Gravitational Forces
Model accounts for many sources of non-grav. forces, including:
Solar radiation and wind pressure; the interplanetary media Attitude-control propulsive maneuvers; gas leakage from the propulsion system DSN antennae contributions to the spacecraft radio tracking data Torques produced by above mentioned forces
Orbit determination procedure, includes:
Models of precession, nutation, sidereal rotation, polar motion, tidal effects, and tectonic plates drift; Model values of the tidal deceleration, non-uniformity of rotation, polar motion, Love numbers, and Chandler wobble are obtained observationally via LLR, SLR and VLBI (from ICRF):

Now [after Pioneer] model can be adjusted to include:
Effects of the recoil force due to emitted radio power Anisotropic thermal radiation of spacecraft
Unknown forces are routinely modeled as stochastic accels:
Exponentially correlated in time, with a variable time constant Stochastic variable was sampled in 0-, 5-,10-day batches
Models Used to Explain the Anomaly
Models and suggestions that failed to explain the anomaly:
Non-gravitational effects: Solar pressure, solar wind, interplanetary medium Precessional attitude control maneuvers and gas leaks Nominal thermal radiation, plutonium half life Some viscous drag force (ULY: solar radiation, maneuvers) Gravity from the Kuiper belt; gravity from the Galaxy Dark Matter distributed in a halo around the solar system Drifting clocks, general relativity, the speed of gravity Hardware problems at the DSN tracking stations Errors in the planetary ephemerides Errors in the values of the EOP, precession, and nutation; Identical design of Pioneer 10/11 spacecraft (GLL, ULY: solar radiation, maneuvers)

Error in JPL's ODP?

Numerous internal checks NASA Grant to The Aerospace Corporation: 1996-1998

The Pioneer Anomaly

The two-way anomaly to first order in (v/c) simply is:

1998.8

CHASMP two-way Doppler residuals (observed Doppler velocity minus model Doppler velocity) for Pioneer 10 vs time. [1 Hz is equal to 65 mm/s range change per second]
Adding one more parameter to the model a constant radial acceleration led to residuals distribution ~ zero Doppler velocity with a systematic variation ~3.0 mm/s. The quality of the fit may be determined by the ratio of residuals to the downlink carrier frequency, 0 2.29 GHz.
Sources of External Systematic Error [PRD, 2002]
Interesting, but not a major source of concern!
Sources of On-board Systematic Error [PRD, 2002]
Pioneer DSN antenna at Goldstone

Pioneer 10/11 spacecraft

A drawing of the Pioneer spacecraft
1987 [97 W] ~32.8% reduction 1998.8 [65 W] 2001
Heat is clearly important source, but:
NOT strong enough to explain the anomaly Exponential decay (or linear decrease) is NOT seen in the anomaly aP
IJMP A 17 (2002) 875-885, gr-qc/0107022
New Focus: the Pioneer 10 Spin History
Most spacecraft show spin-down behavior usually due to structure tiredness, connection loosening, etc.

Pioneer 11 Spin History

Causes for de-spin are different: Pioneer 11 spin increases in between the maneuvers, leaking thruster?

ODP/Sigma residuals

ODP/Sigma Doppler residuals in Hz for the entire Pioneer 10 data span. The two solid vertical lines indicate the boundaries between data Intervals I/II and II/III. Maneuver times are indicated by the vertical dashed lines.

The Pioneer Anomaly: Summary
Our latest result for the Pioneer 10/11 anomalous acceleration:
A line of sight constant acceleration of the s/c toward the Sun: We find no mechanism or theory that explains the anomaly; The most plausible cause is a systematic, yet to be demonstrated.

Behavior of the Anomaly:

We have no real idea how far out the anomaly goes; aP continues out roughly as a constant from ~10 AU; Constancy: temporal and spatial variations less then 3.4%; Amplified (or turned on) for hyperbolic, escape trajectories (?)
Three Different Codes Used:
JPL Orbit Determination Program [DPODP various generations]; Aerospace Corp [CHASPM/POEAS]; GSFC [brewed by Craig Markward in 2003, data from NSSDC].

Next Steps:

Early data processing [work initiated at JPL: fly-byes, entire data set] A European study of the PA recently initiated (ZARM, Bremen)
Meanwhile Pioneer 10 @ Arecibo
Pioneer 10, as seen by 305 m antenna at Arecibo Observatory, Puerto Rico
One data point we need more!

Thank You!

Solutions for Different Data Intervals
Determinations of the anomalous value for aP from Intervals of Pioneer 10 and Pioneer 11 data in units of 108 cm/s2
A Mission to Test the Pioneer Anomaly

Mission Objectives:

To search for any unmodeled small acceleration affecting the spacecraft motion at the level of <0.cm/s2 Determine the physical origin of any anomaly, if found.

Unique Features:

A standard spacecraft bus that allows thermal louvers to be on the sides for symmetric fore/aft thermal rejection. Fore/aft symmetric design with twin antennae (``yo-yo'' concept).
With Off-the-Shelf Technology:
Accuracy a ~ 0.cm/s2 is achievable in about 5 years GIVEN THAT the thrusters are reliable and gas leaks can be eliminated or monitored to a high enough accuracy

New Technology?

FAST ORBIT TRANSFER using solar sails, nuclear propulsion DRAG-FREE systems would help, but are not sufficient DC ACCELEROMTERS are very useful OPTICAL COMM very good, but currently very expensive THRUSTERS: good performance and high repeatability are needed
Minimal investment in new technologies would enable not only to test the Anomaly, but also to uniquely determine its Origin.
Finding Direction of the Pioneer Anomaly
CQG 21 (2004) 1, gr-qc/0308017

1) 2) 3) 4)

Towards the Sun: gravity? Towards the Earth: time? Along the velocity: drag or inertia? On the spin axis: internal systematics?

Directional Modulation of the Anomaly
Clearly different behavior; easy to separate.
Lessons Learned from the Pioneers

Attitude Control:

3D ACCELERATION SENSITIVITY: <0.01 10-8 cm/s2 for each axis Spin-stabilized (preferred) If 3D stabilization use of DC accelerometers and reaction wheels
Navigation & Communication:
3D ACCELERATION SENSITIVITY: <0.01 10-8 cm/s2 for each axis POINTING: control 6 rad; knowledge 3 rad; stability 0.1 rad/s COMM: X and Ka band with significant dual-band tracking DATA TYPES: Doppler, range, DOR, and VLBI

Thermal Design:

ENTIRE SPACECRAFT: heat-balanced & heat-symmetric KNOWLEDGE of all heat sources RTGs, electronics, thrusters, etc ACTIVE CONTROL of all heat dissipation channels within & outward PRECISE KNOWLEDGE of 3D vector of thermal recoil force If spin-stabilized thermal louvers are on the sides of the bus If 3D stabilization harder to balance recoil forces and torques
Investigation emphasized effects previously thought to be insignificant: rejected thermal radiation, gas leaks, radio beam.
Lessons Learned from the Pioneers (2)

On-board Power RTGs:

LOCATION: must provide thermal and inertial balance & stability If spin-stabilized position as farther as practical from the bus If 3D stabilization balance, balance, balance! (see below)

Propulsion System:

Precisely calibrated thrusters, propellant lines & fuel gauges AUTONOMOUS real-time control of their performance
Symmetric Design (Yo-Yo concept):
FORE/AFT SYMMETRIC design with TWO identical Cassegrain antennae transmitting in opposite directions, and ROTATE the craft once in a while (done for Pioneer Earth acquisition maneuver, took ~2.5 hours and 0.5 kg of fuel)

Mission Design:

TRAJECTORY: a hyperbolic solar system escape trajectory >15 AU from the Sun possibly in the plane of ecliptic, co-moving with the solar system's direction within the galaxy FAST TRANSFER ORBIT spacecraft moving with a velocity of 5 AU or more per year, reaching 15 AU in 3 years time or less Heavy class launch vehicle (Delta IV, Proton, Ariane class) Solar sail, or nuclear propulsion at least to 15 AU

Other Possibilities to Study the Anomaly
Experimental possibilities:
The 305-meter antenna of the Arecibo Observatory in Puerto Rico might be able to detect Pioneer's signal for a longer time The existing data for Pioneer 10 [complete to July 2000]
High-rate data from 1978 to Jan 1987: not used in our analysis: Study FLYBYS!
Current or near future missions:
Cassini [RTGs very close]:
Heat recoil force ~40108 cm/s2

GP-B [in orbit]

Acceleration resolution at ~ 1108 m/s2 Earth polar circular ~ 92 min orbit,
LISA Pathfinder [launch 2006]
Acceleration resolution at ~ 11012 cm/s2 Multiple noise cancellation strategies
Technology exists to further test the Pioneer Anomaly
Are There Any Other Possibilities?

Future missions:

JIMO [~2012]:
Nuclear reactor [unlimited power / weight] Focus on technology, very minimal science

Pluto-Kiuper [>2014]:

Cassini spacecraft
The launch data is uncertain at the moment A mission from the Prometheus family?

Solar Probe [>2016]:

a low-mass module may be ejected during solar flyby out of the plane of the ecliptic

GP-B Launch 04-20-2004

Interstellar Probe [>2020]

Todays reality:

The anomaly source is still unknown Analysis of early data (and the entire set) Needs a wider community support Pioneer is a low priority for NASA Designated mission today is hard, but

JIMO spacecraft

Dust in the Kuiper Belt
Possible acceleration caused by dust in the Kuiper belt.
Suggested Explanations: Familiar Physics
A new manifestation of known physics? Interplanetary dust:
i) additional gravitational frequency shift; ii) resistance of s/c antennae as they transverse the dust Contradicts to known properties of the interplanetary medium. Density varies greatly within the KB; not large enough to produce acceleration ~aP
Dark matter hard to understand:
A spherically-symmetric distribution of matter, with produces a constant acceleration inside the distribution. To produce aP even only out to 50 AU would require the total dark matter Ephemeris accuracy allows of DM within orbit of Uranus

Modification of gravity a Yukawa force:
is the new coupling strength relative to Newtonian gravity, and is the new force's range. For instance, = for = 200 AU
Suggested Explanations: MOND & New Physics
MOND (MOdified Newtonian Dynamics):
Gravitational acceleration of a massive body is constant and for. Depending on the value of H, the Hubble constant, Indeed, if H = 82 km/s/Mpc, Variations of MOND to account for the Pioneer anomaly Viking ranging data limit any unmodeled radial acceleration on Earth and Mars to 0.cm/s2 if the anomalous radial acceleration acting on spinning spacecraft is gravitational in origin, it is not universal. for some
Observation aP~cH, stimulated many suggestions:
Gravity of the solar system is not static w.r.t. the cosmic expansion 5-D Kaluza-Klein with a time-varying scale factor for 5-th dimension Effect of a scale-dependent cosmological term in the Grav. action Cosmological models with a time-varying Newtonian G(t)
Suggested Explanations: New Physics
Several scalar-field ideas have also appeared:
Long-range scalar field, with oscillatory decline in aP, d>100 AU Self-interactions of a scalar condensate could be the origins of both Milgrom's inertia modification and also of the Pioneer effect. Flavor oscillations of neutrinos in the Brans-Dicke theory of gravity may produce a QM phase shift of neutrinos A theory of conformal gravity with dynamical mass generation
Phenomenological time models:
Drifting Clocks; Quadratic Time Augmentation; Carrier Frequency Drift; Speed of Gravity Rejected: poor fits / inconsistent solutions among spacecraft
Quadratic in time model (pseudo-acceleration, less likely):
Mimics a line of sight acceleration of s/c, and could be thought of as an expanding space model. Note that aquad affects only the data.
Initial PRL paper was cited ~108 times, including ~ 78 papers with suggested mechanisms to explain the anomaly.
Data Acquisition and Preparation
Data acquisition with JPL's Deep Space Network (DSN):
Goldstone, California; Robledo de Chavela, near Madrid, Spain; Tidbinbilla, near Canberra, Australia. The DSN Frequency and Timing System: At its center is an H-maser that produces a precise and stable reference frequency with Allan deviations of (1.3 1.0)1012, for a 103 sec Doppler integration time (for the S-band) Calculations of the motion of a spacecraft are made on the basis of the range time-delay and/or the Doppler shift in the signals GLL has S-band range data near the Earth. ULY has 2-/3-way S-up/X-down Doppler and range, S-up/S-down: processed S-up/X-down Doppler and range Considerable effort has gone into estimating measurement errors: to provide the data weights necessary to accurately estimate the parameter adjustments and their associated uncertainties To correct for the Earth's tropospheric refraction (affects Doppler observable) the data can be deweighted for low DSN antennae elevation angles to correct for a Doppler bias due to spinning antennae

Radio Doppler and range techniques, the most common for navigation
Data types: Pioneer craft have only 2-and 3-way S-band Doppler
Data preparation and data weighting
Spin calibration of the data:
Data Collection by the DSN
THE PIONEER ANOMALY AND MISSIONS TO TEST IT THE PIONEER ANOMALY AND MISSIONS TO TEST IT
QDP/Sigma 1-day batch residuals
ODP/Sigma 1-day batch-sequential acceleration residuals using the entire Pioneer 10 data set. Maneuver times are indicated by the vertical dashed lines.
Equivalent forms (PRL, 1997): Steady frequency drift: Anomalous acceleration: Clock acceleration: Unknown forces are routinely modeled as stochastic accels:
Basic methods of spacecraft navigation
Relativistic equations of motion; Small non-gravitational forces; Data acquisition and preparation.
The latest (2002) results:
An error budget for the anomaly.

Outline for This Talk:

The Pioneer Anomaly:
Pioneer spacecraft and missions JPL OD process: data and models Initial detection for PA How unique these condition? JPL ODP process Aerospace Corporation Error budget Attempts to explain

Recent Analysis:

Lessons Learned & Next Steps:
Missions of interest A designated mission concept How to get support? NASA? ESA?

This Talk will Cover:

PA history, analysis, lessons learned, and mission to test the PA

A Long Journey Ahead

A group of yellowish stars at the upper right is dominated by the red giant Aldebaran, where Pioneer 10 is heading.

Typical JPL ODP Output

One data point of Pioneer 10 (spacecraft #23).

Typical CHASPM Output

Several data points of Pioneer 10 (spacecraft #23).
Consistency Between ODP and CHASPM

Pioneer 10

ODP: 5-day sample averages of using BSF with a 200-day correl time (dots). Solid lines mean values of aP in three Intervals; dashed lines large BSF computational error bounds. CHASMP: The 200-day accel values using CHASMP solid squares.