Particle Measurements Distribution Particle Measurements, … · 2010. 1. 28. · THEMIS/ESA d...

Particle Measurements DistributionESS 265 Winter Quarter 2010Particle Measurements, Distribution

Functions, Moments and Remote Sensing

Low energy particle instrumentsHigh energy particle instruments

From Counts/s to Distribution Functions: Geometric Factor, MomentsTHEMIS/ESA d THEMIS/SST i f ll d iTHEMIS/ESA and THEMIS/SST: pitfalls and corrections

Decommutation and processing. Data analysis techniques.Application to Remote Sensing

Lecture 04Jan 28, 2010

With significant contributions from:James P. McFadden, Davin Larson,

1

Thomas Moreau, Andrei Runov and Ryan Caron

Types of Low Energy Plasma Instruments

• Langmuir Probes• Retarding potential analyzers• Magnetic spectrographs• Electrostatic analyzers• Energetic neutral atom imagersg g

2

Langmuir Probes, Overviewg ,

• In the ionosphere, where D is ~1-10cm (Ti, Te ~eV, Ne~ 102-106)p D ( )• Scan in voltage, get Ni, Te, Ne• Boom away from spacecraft body, normal to flow• Need to ensure sensor is small relative to ion gyroradius • Missions: AE, DE-2, AE-C, PVOMissions: AE, DE 2, AE C, PVO

3

Langmuir Probes, Issuesg ,• Non exponential fit reveals issues to deal with:

– Cleanliness: Spacecraft conductive to avoid chargingS f t i l l (O b b d t t d ti– Surface materials clean (Oxygen bombardment causes permanent non conductive layer, asymmetry of sensor)

• Avoid by heating up to outgass early in the mission• Avoid by allowing thermal (high energy) electrons to bombard surface

– High altitude missions do not have this problem

W k f i hi l– Work function patchiness low• Order of 100mV is too high (corresponds to 1160K) for E-region temperatures of 300K (0.03eV)• Use vitreous carbon (Weismann & Kintner 1963 on rockets) or highly oriented metals (PVO:

rhenium, molybdenum)– Minimize VxB in high fields

• Venus, Mars OK (low fields) but at Earth, cross field antenna motion causes E along boom– Minimize by tipping boom along B

• Further reading:– Brace, L. H., Langmuir probe measurements in the ionosphere, in Measurement

Techniques in Space Plasmas: Particles Geophys Monogr Ser 102 AGU 1998Techniques in Space Plasmas: Particles, Geophys. Monogr. Ser. 102, AGU, 1998– Krehbiel, J. P., et al., The DE Langmuir probe instrument, Space Sci. Instrumentation,

5, 493, 1981.– Mott-Smith, J. M. and I. Langmuir, The theory of collectors in gaseous discharges,

Phys. Rev., 28, 727, 1926.

4

Retarding Potential Analyzers

• In the ionosphere, mount along ram velocity, measure species densitiesRam speed (7 5km/s) is high or supersonic relative to ion thermal speed or motion

Heelis and Hanson, 1998

– Ram speed (7.5km/s) is high or supersonic relative to ion thermal speed or motion– Spacecraft charging is negative and small relative to motional energy– I-V curve has steps at qVret = ½m(Vsr+Vr)2 – qs ; where: s = sensor potential relative to

plasma, Vsr= ram speed• Ions can be further differentiated with mass spectrograph behind RPAIons can be further differentiated with mass spectrograph behind RPA

– See: Chappell et al., The retarding ion mass spectrometer on DE-1, Space Sci. Instr. 4, 477, 1981

5

RPA/Ion Drift Meters

Heelis and Hanson, 1998,

• In the ionosphere, mounted along ram velocity, measure species velocity– G2 retards lower energy H+, but allows higher energy O+ through– Collimated beam comes through and falls asymmetrically on collectors

G6 l t G3 5 d d t di t ti– G6 suppresses electrons, G3-5 are grounded to remove distortions• Issues: Vt error can be significant when ram direction angle is large• Further reading:

– Heelis and Hanson, Measurements of Thermal Ion Drift Velocity and TemperatureUsing Planar Sensors, in Measurement Techniques in Space Plasmas: Particles,Geophys. Monogr. Ser. 102, AGU, 1998

6

RPAs in tenuous plasmas• In the solar wind, rely on supersonic motion

– G1, G3 are shields (ground); G4 is suppressor (-200V)– G3 is modulator between V1-V2, resulting in dif. Current– North-South RPA measurement results in transverse

speed– Spinning (2.7sec) results in azimuth sectors (11.25o, 45o)– Scanning in velocity/energy results in temperature– Full scan (11V-1.3 kV) initially

• tracking mode after lock on Vram allows data compression

• Further reading:g– Lazarus and Paulanera, A comparison of solar wind

parameters from experiments on the IMP8 and WIND spacecraft, in Measurement Techniques in Space Plasmas: Particles, Geophys. Monogr. Ser. 102, AGU, 19981998

Lazarus and Paulanera, 19987

Magnetic Spectrographs

• For low energy particles (left):– post-acceleration V behind an RPA provides V T and m/q

LIMS Magnetic Spectrograph on CRRES

– post-acceleration Vpa behind an RPA provides V, T and m/q– LIMS: m/q=(Brc)2/(2Vpa), where B is magnetic field, rc magnet curvature

• For higher energy particles (right):– Broom magnet clears electrons– High field bends high energy ions– High field bends high energy ions– Ions that were not bent assumed neutrals (ENAs)

• Further reading:– Reasoner et al., Light ion mass spectrometer for space-plasma investigations: Rev. Sci. Instr.

53(4), p. 441, 1982.( ), p ,

8

Plasma Instruments

Components of a Plasma SensorCo po e ts o a as a Se so

Analyzer selects a subset of particles to be measured.

Detector to amplify an event.Detector to amplify an event.

Analog signal processing to register an event.

Digital electronics to store/compress event data.g

Analyzer

Electrostatic Analyzers

• Electrostatic deflection analyzes velocity distribution– Analyzer constant, K=R1/, where =R2-R1; Outer shell is at 0 Volts, inner shell at potential V.– Electrostatic deflection at entrance aperture can measure incoming ions from different directions if spacecraft non-

spinning– Energy E of the particles of charge q, incident on the MCP is E=-K q V /2

• Further reading:g– Carlson et al., The electron and ion plasma experiment for FAST: Space Sci. Rev. 98, 33, 2001.– McFadden et al., The THEMIS ESA plasma instrument and in-flight calibration, Space Sci. Rev., in press

10

Analyzers – Spherical

Spherical analyzers p yprovide angle focusing.

Analyzers – Spherical Top-Hat

Focal point at ~80o

deflection.

Most compact for largest sensitivity.

Analyzers – Top Hat Response

Top-hat analyzers have been characterized to allow optimizationallow optimization without resorting to simulations.

Time of Flight

• Electrostatic deflection => energy per charge: E/Q Time of flight => energy per mass E/MElectrostatic deflection > energy per charge: E/Q. Time of flight, , > energy per mass E/M– Post-acceleration UACC provides sufficient energy for optimal McP operation and timing electrons at foil– Electrons generated at carbon foil result in energy loss – M/Q=2(E/Q + qUACC)/(d/t)2*

• Further reading:M bi t l 3D l di t ib ti l ith ti f fli ht di i i ti f Cl t FAST d– Moebius et al., 3D plasma distribution analyzer with time-of-flight mass discrimination for Cluster, FAST and Equator-S, in Space Sci. Rev., in Measurement Techniques in Space Plasmas: Particles, Geophys. Monogr. Ser. 102, AGU, 1998

14

Analyzers – Carbon Foils

Plasma Instrument Design: Examples

Measure charged particles from ~1 eV to ~30 keV

Detectors – Dynode Multipliers

To detect a single particle requires aTo detect a single particle requires a detector with gain.

Detectors – Channel Electron Multiplier (CEM)

The channels are curved toThe channels are curved to prevent ion feedback.Tube diameter ~1-2 mm.CEM diameter ~3 cm.

Detectors – Microchannel Plates (MCPs)

Channel diameter ~5-25 um, Plate diameter up to ~10 cm.

~1 kVl tper plate

Detectors – Microchannel Plates

Detectors – Electron Detection Efficiency

Detectors – Ion Detection Efficiency

Neutral Particle Flux measurements

• Variant of RPA to perform neutral flux measurements• Further reading:

– King and Gallimore, Rev Sci Instr. 68(2), 1997

23

Energetic Neutral Atom imagingg g g

First ENA image from ISEE-1Medium Energy Particle AnalyzerRoelof et al., 1987

ENA t S t f CASSINI/MIMI ENA t J it f CASSINI/INCAENA at Saturn from CASSINI/MIMIENA at Earth from IMAGE/HENA

ENA at Jupiter from CASSINI/INCA

24

Energetic Neutral Atom imagingg g g• Ion Neutral Camera (INCA) on Cassini to

Saturn– FOV = 90°×120° OV– Serrated deflector plates at 0 & 66kV

• Deflect 500keV/q particles– Three layer foil at entrance slit suppresses UVy pp– Secondary electrons at entrance steered to McP

• Determine ENA entrance coordinate normal to plane

• Determine START signal to TOF– Two-dimensional imaging MCP provide:

• Particle position at other end of flight path• STOP for TOF

– Time of flight and energy (McP) result in:• Energy and velocity -> mass of ions• Can distinguish between hydrogen and Oxygen

• Further reading:– Mitchell et al., J. Geophys. Res., 109, 2004Mitchell et al., J. Geophys. Res., 109, 2004

25

Energetic Particle Instruments• Solid State Detectors (SSDs) not only detect individual particles they can• Solid State Detectors (SSDs) not only detect individual particles, they can

be used to measure particle energy with good energy resolution.• Typically good for E>20 keV but recent improvements get them to ~2 keV

T o arieties of Silicon Diode Detectors• Two varieties of Silicon Diode Detectors– Implanted Ion (i.e. Canberra PIPS)

• Produced by implanting p-type material into an n-type silicon substrateEas to prod ce pi elated s rfaces• Easy to produce pixelated surfaces

• Very rugged– Surface Barrier

• Chemical process to create diode surface• Chemical process to create diode surface• Easily damaged, sensitive to solvents• Not too common anymore

• Typically both varieties are run fully depleted (electric field extending• Typically both varieties are run fully depleted (electric field extending throughout bulk of material)

• Maximum thickness is ~1000 microns – defines max energy particle that can be stopped within the detector

ESS 261 Energetic Particles26

can be stopped within the detector• Particles can be incident on either side of detector

26

Operation Principle• With the application of a (large• With the application of a (large

enough) reverse bias voltage an electric field is established throughout the entire silicon volume

p n++

--g

(fully depleted detector).• An energetic charged particle will

leave an ionization trail in its wake.

EForward bias

• The electron hole pairs will separate and drift to opposite sides.

• The total charge is proportional to p n++

--

the electronic energy deposited. (3.61 eV per pair for Silicon).

• The signal contains only a few th d l t th i i

E

Reverse biasparticle

thousand electrons thus requiring sensitive electronics.

• The trick is to collect and measure this small signal p n+-

Reverse bias

this small signal. p n+-E

-+

27

Interaction of particles with matter• The way in which energetic particles interact with matter depends

upon their mass and energy.– Photons - have “infinite range”- Their interaction is “all-or-nothing”

They do not slow down but instead “disappear”, typically through one f th i t ti (I I x h i b ti ffi i t )of three interactions (I=I0e-x where is absorption coefficient ):• Photoelectric effect (Low energy: E<~50 keV• Compton Scattering (50 keV ~< E < 1 MeV)• Pair production ( E >2 x 511 keV)Pair production ( E >2 x 511 keV).

• Particles with non-zero mass (Electrons and Ions) will slow down as they pass through matter. They interact with electrons, phonons and nuclei.Electrons and Ions behave differently due to the different mass ratio. The primary interaction of all energetic particles is with the sea of electrons.

– Ions:• Ions interact with a series of distant collisions. Each interaction results in a small

energy loss and very little angular scattering. – They travel in nearly straight lines as e e gy oss a d e y e a gu a sca e g ey a e ea y s a g es asthey slow down. The dispersion is small. (Imagine a fast bowling ball thrown into a sea of slow moving ping pong balls.)

– Electrons:• Electrons can lose a large fraction of their energy and undergo large angle scattering g gy g g g g

with each interaction (Imagine a high speed ping pong ball thrown into the same sea)

28

Photoelectric effectPair production

Compton scatteringCompton scattering

29

(/ often used: mass attenuation coefficient)

30

Interaction of particles with matter:ElectronsElectrons

• When Electron interaction is long-range. Electron-electron energy exchange peaks when incoming particle is close to the target electronexchange peaks when incoming particle is close to the target electron energy. The material electrons have energies that are determined by:

oThe material temperature (phonon-electron interaction couples them) ~kT, ~105m/soThe Fermi energy, i.e., the energy level of free electrons in absolute-zero temperature. This is about 5-10eV, which gives the Fermi speed of 106m/s.

• When an electron hits an atom it can undergo a very large angle deflection, often scattering it back out of the material.

• Bremstrahlung (braking) radiation is produced when electrons undergo extreme accelerations. X-rays are easily generated when energetic electrons strike high Z materials. (a good reason to avoid high Z materials on exposed surfaces)high Z materials on exposed surfaces)

31

Electrons in

Energy lost toEnergy lost toionization(collectable)

Energy lost toBremstrahlungradiation (notradiation (notcollectable)

32

Interaction of particles with matter:Ions

• The stopping power for heavy particles (ions) is given by the Bethe-Bloch formula (1932):

Ions

: where,422

42

Bcm

ezNdxdE A

cmdx e

2 22 CcmZ

2)1(2ln 2

2 ZC

Icm

AZB e

Rate of energy loss is ~ inversely proportional to energy, and proportional to Z (the atomic number) and z2 z the projectile chargeproportional to Z (the atomic number) and z2, z the projectile chargeI = average ionization potential, and C = density and shell corrections.

33

• The range is given by:

Interaction of particles with matter• The range is given by:

dEdER

0 1

– dE/dx is often expressed in units of keVcm2/gr which is dE/dx ti th t i l d it Thi b dl th dE/d i t

dEdx

REstart

times the material density. This bundles the dE/dx curves into groups by normalizing away the material density from the electronic interactions.Th (t i ll i ) i l ft li d t th– The range (typically in cm) is also often normalized to the density and expressed in units of grams/cm2, i.e., the equivalent mass per unit area required to stop the particle.

34

Protons inProtons in

Energy lost to ionization (collectable)Energy lost to ionization (collectable)

Energy lost to phonons (not collectable)

35

Alphas in:Alphas in:

Energy lost to ionization (collectable)Energy lost to ionization (collectable)

Energy lost to phonons (not collectable)

36

Solid State Detector Electronics

Simulated A225Simulated A225response for typical 1MeV electron pulse through a Si detectorSi detector.

The A225 integrates charge, with peak pulse g , p pequal to integrated charge.


A 20ns signal turns into an 8usec pulse!

Front End Counting Electronics

38

Sources of Noise• Capacitance• Capacitance

– Noise results inuncertainty inabsolute valueof energy

• Leakage (dark) current– When dark current is integrated by A225 results in baseline offset– Baseline restorer restores zero level

Leakage current results in error in absolute signal amplitude– Leakage current results in error in absolute signal amplitude

39

Examples of Detector Systems: WIND/3DPPredecessor of THEMIS/SSTPredecessor of THEMIS/SST

Example: WIND/EPACT

Cross section of the EPACT isotope telescope on Wind. The first two detectorsare two-dimensional position sensitive strip detectors (PSD1, PSD2). They are requiredso that path-length corrections may be made for the angle of incidence and for

if iti i d t t thi k T t i d t k ff i lnon-uniformities in detector thickness. Tungsten rings are used to mask off circular areasfor each PSD. There are 6 solid-state detectors increasing in thickness with depth in thestack in order to minimize Landau fluctuations. From von Rosenvinge et al. [1995].

41

Example: THEMIS/SST

Foil Collimator

Attenuator

F il

(for electrons)

Foil

MagnetDetector Stack

Att tAttenuator

Open Collimator(for ions)(for ions)

42

• Distribution functions– Phase space densityp y– Flux– Integral and differential flux– Count rates

• From counts to distribution function– Geometric factor– Dead time corrections– Moments

• Further reading:– Curtis et al., On board analysis techniques for space plasma particle

instruments, Rev. Sci. Instrum., 60, 372, 1989.– ISSI book on: “Calibration of Particle Instruments in Space Physics” West,

Evans and Steiger, Bern, Switzerland, 2007.

43

Ph d it d ti l flPhase space density and particle flux• Distribution function

– f(r,v), units: ))s

cm(1/(cm ))s

cm((cmparticles/ 3333

– Normalized such that:

• Particle flux (a.k.a. flux, or flux density)Definition:

vdfdNvdfN 33

vdfvJdvdvfN*VJ 33

– Definition:

– Jx, Jy, Jz Units:

vdf vJdvdvfN*VJ s)(cmparticles/ 2

– Provides velocity, or 1st moment of distribution:

• Differential (particle) flux– Definition:

/NJV

particles1itJd i

Definition:

– Alternate (more common):

keV s cmp.,.,

keV s cm:units, 22 ei

dE

keVstr s cmparticles i.e., ,

keVstr s cm1 :units ,

dEdΩJd

22

44

Momentum flux• Momentum flux tensor (in plasma rest frame):

3(symmetric). Units:– Provides second moment of distribution or temperature:

• Temperature: df/dfT/NΠ 33

3V/cmor s,nanoPascal evdf v vd vdvvfNT 33

• Temperature:

• Ambient flow must be subtracted: vdf /vd)V-v()V-v(fT/NΠ 33tottot

vdf/vdvvfT/NΠ 33

– Note: When energy range spans two instruments, e.g., ESA and SST, below and above energy E0 (velocity V0) we must sum the measured momentum flux tensor and divide by sum of densities:

E

0

0

0

E

3E

0

3

E

3tottot

E

0

3tottot

SSTESASSTESA

vdfvdf

)vd)V-v()V-v(f vd)V-v()V-v(f()N)/(NΠΠ(T

– Where:

0E0

0

0

0

3E

3

E

3E

0

3

SSTESASSTESAtot

vdfvdf

)vdvf vdvf()N)/(NJJ(V

0E0

vdfvdf

45

Energy flux

EdNQ

Energy Flux: keVkeV 23 )(

Units: scmcm cms 23 )(

3EfdEdNQd

dEQd

keVscmkeV

2

k V

Differential Energy Flux: (A) Units:

ddQd

sKeVstrcmkeV

2Differential Directional Energy Flux: (B)Units:

46

Typical Values [1]

Figure Source: ISSI book (see reference on p. 2)

Typical Values [2]: conversion to counts/s

ESS 261 Low Energy Particles48Figure Source: QUATRO/THEMIS proposal

From Distribution Function to Counts [1]16 )/( f

20 )(

N

16 )/( scmf 2

33),( theNvrf

tn

muT

th2

dxe x2 Nfd 3

TVN ,, 0

and ,, uversusf

Note:

Given

Such that

Can get

In practice, distribution functions of ions and electrons exhibit high energy “tails”, i.e., fluxes at high energies, that are much higher than Maxwellian. These high-energy “tails” are best

,, Eversusf

g e e g es, t at a e uc g e t a a e a ese g e e gy ta s a e bestdescribed by exponential drop-offs, or power-law spectra, which are better reproduced by “kappa” distributions:

which are Maxwellian-like at low energies and exhibit exponential drop-off at high energies.In the limit of k approaching infinity, the kappa distribution approaches a Maxwellian, above except for a normalization constant (2/m)3/2p ( )

49

From Distribution Function to Counts [2]

12 )(

strKeVscm

dEdJd

Diff Flux

ddEfddEfdEfJddfJd

)( ˆ)()()(

323

3

2

dEdEfEm

mdEfmm

ddEf

)(ˆ22

)( ˆ2

2

)()( ˆ

2

22

2

22 2

mm 22

)(2)(ˆ2

22 EfEEfEJd

),,( EE

)()( 22 f

mf

mdEd),,(

TVN ,, 0

Given can get ,, Eversus

dEdJd

0 dEd

50


tk VkeV

dEdQd

2

Diff. Dir. Energy Flux

strkeVscmdEd 2

2

dEdEfEEddEfE

ddEfEddfEdEfQd

m)()()(2

)()()()(

2

22

2

222

223

JdEEfEEQd

)(22

ddE

EEfEEmdEd

)(2

51


Particle Count Rate

Geometric factor for electrostatic Analyzer

GdEd

QdEGdEd

JdsR

**)( 1

Geometric factor for electrostatic AnalyzerG = [Area ∙ Solid Angle ∙ Bin Energy Width/Energy of channel]

cm2 str eV eV

i 2 V/ Vunits: cm2str · eV/eV

Multiplied by units of results in (#/s))( 2 eVstrscmeV

dEdQd

G is a function of energy channel, as well as detector direction.Thus, depends on specifics of instrument.),,( EG

52

From Distribution Function to Counts [4, continued]

For electrostatic analyzers the inherent energy resolution δE is proportional toFor electrostatic analyzers, the inherent energy resolution δE is proportional to

the energy: where Δ is gap, R is radius. This is because

RVqE2 RqVE 2

E 1and , where K is the analyzer constant. So ΔE /E is constant (ΔE =

energy bandwidth) and is part of the geometric factor of the instrument. For

KREE 1

Solid State detectors, the G factor is given for the integrated directional flux.

d ff h h b f d f = detection efficiency. This is the number of counts registered for a given

number of particles that hit the detector. This gets us from # particles/s to #

counts/s. Typical values for electrostatic analyzers is 0.6 – 0.7.53

From Distribution Function to Counts [5]Particle Counts

RunitlessC )(

is the measurement interval (seconds)

RunitlessC )(

Note: fEdEd

JdEscorrectiontimedeadwithdEd

QdGR 2*) (

diff. energy flux diff. flux D. F.

UNITS: )(;);/(# 22 eV

eVstrcmGstrseVcm

eVdEd

QdscountsR

NOTE: count/s are dead time corrected.This dead‐time correction is necessarybefore using the above equation.

“C” is # counts per sample.C = R * AT, where AT is accumulation time.

From Distribution Function to Counts [6]Dead-time correctionsDead time corrections

In addition to geometric factor, detector efficiency and energy bandwidth, we need to apply a dead time correction to the counts measured to get the counts in the plasma.

The dead time results from the processing that needs to take place once an event has been registered, as the DPV is occupied.

Thus. The total dead time is a function of the number of counts.

Total DT = Cm EPT. Live time (LT) is the time the detector was in operation.m ( ) p

LT = AT – DT, when AT is the accumulation interval, or accumulation time.

EPT t i ti f 1 tEPT = event processing time for 1 count.

55

From Distribution Function to Counts [7]Dead-time correctionsDead time corrections

Then the actual count rate is:

DTAT

CLTCR mm

real

EPTRR

CATC

EPTCATCR

measured

measured

ATEPT

m

m

m

mreal

11

/

always < 1i.e., you cannot have in a given Δt more particles than EPT’sthan EPT s.

Thus, dead time correction consists of dividing the measured rate by the quantity:(1 – Rmeasured · EPT)

NOTE: = max count rate (because )EPT

1 ratecountAT

countsofATEPT

ATEPT

max#max11

56


ESS 261 Low Energy Particles57

From Distribution Function to Counts [8, continued]


From Distribution Function to Counts [9]: Summary

ESS 261

From Distribution Function to Counts [9] continued


Moments [1] From: Calibration ofParticle Instruments inSpace Physics, ISSI, P. 6.


Moments [2]


Moments [3]


Moments [4]

64

In practice, on board moment computations:


THEMIS ESA: Requirements and Specs• The ESA instrument measures 3-D electron and ion energy

distribution functions over the Energy range ~5 eV to 30 keV (25 keV for ions).

• Typical energy sweep has 16 or 32 energy samples and logrithmic sweep (dE/E~constant).

• A full 4-pi distribution measurement is produced during each spin.

• Sweep rate of 32/spin gives dense sample of 3 D particle• Sweep rate of 32/spin gives dense sample of 3-D particle distributions. Ion sensor is capable of 64 sweeps/spin to provide adequate sampling of solar wind.

• The ETC board compresses the raw measurements into three selectable “distributions arrays” that come down in separate packets (full, burst and reduced packets) at selectable rates.

• The ETC also computes moments with corrections for detector efficiency and spacecraft potential.

THEMIS ESA: Block Diagram

• Electronics functional design is identical to FAST (with ACTEL upgrades)Three circuit modules plug together for efficient assembly and test• Three circuit modules plug together for efficient assembly and test

• MCP pulse amplifiers are Amptek A121 with programmable gain• All discrete logic, counters, and HV DAC drivers are Actel FPGAs

HV & LV li t d i b ilt t UCB SSL• HV & LV supplies are mature designs built at UCB SSL

67

ESA S/C INTERFACE

Ion and Electron ESAs are packaged together.

180 deg x 6 deg field of viewg g

Sensors have different anode patterns.

Ion ESA 16anodes Electron ESA 8 AnodesIon ESA 16anodes Electron ESA 8 Anodes

68

ESA Data TypesTHEMIS produces four basic ESA data products for each species with resolution that depends upon instrument mode:

Full distributions Low cadence ( 32 or 128 spin resolution) full resolutionFull distributions – Low cadence (~32 or 128 spin resolution), full resolution energy-angle (32E x 88A) distributions, taken continuously over the orbit.

Burst distributions – High cadence (every spin), Full resolution energy-angle (32E x 88A) triggered snapshots (~10 minutes each).

Reduced distributions•High cadence (every spin) lower resolution during Slow Survey (goal: full•High cadence (every spin), lower resolution during Slow Survey (goal: full orbit)

•32E x 1A for both i+&e-•High cadence (every spin), lower resolution during Fast Survey (12hrs per orbit)

•24E x 50A i+ and 32E x 6A e- or 16Ex88A e- depending on probe and orbit

Onboard Moments – High cadence (every spin) density, velocity, pressure, heat flux. 69

ESA Data TypesTypical 88 solid-angle map used for sorting full and burst data

70

Full Distributions, Magnetosphere,Ions and Electrons

Full Distributions, Solar Wind,Ions Only Limited to 16E’s

ESA Ion Data TypesIons and Electrons Ions Only, Limited to 16E s


Magnetosphere Ions Magnetosphere ElectronsESA Reduced Data Types


ESA Ion Angles in Various ModesFast Survey Electrons:

Fast Survey Ions:


On board /collection/processing

• Collection of counts, count rates• Sectoring, Angle maps, Square-root compression

– See modes, energy, sector maps in:• class_materials as etcmap_*.xls

• Burst strategy and compression by averaging in energy/angle

• Compression by moment and pitch angle spectra p y p g pcomputation– See: Curtis et al. Rev. Sci. Instr. And– Abiad, technical memo: thm sys 105a etc req1.7.pdfb ad, tec ca e o t _sys_ 05a_etc_ eq pd

• Corrections (scpot) and pitfalls (cold ions)

74

TransmissionTransmission• Packets, headers, time series compression, lossy

icompression• Encoding (CCSDS, Viterbi, convolutional)

– Consultative Committee for Space Data Systems [a NASA standard]standard]

• On-board data: Ground decommutation and processing– Moments

• L0 (.pkt) and L1 (.cdf) contain same (raw) data in DSL coordinates( p ) ( ) ( )• L2 (.cdf) contain calibrated data in other coordinates (GSE, GSM)

– Particle distributions• L0 packet files

– Raw, sqrt-compressed counts (efficient for loading)Raw, sqrt compressed counts (efficient for loading)– Files organized by APID, but transparent to user

• L1 CDF files– Raw, uncompressed counts– All quantities in one file

• L2 CDF files– Omni-directional spectra (will eventually contain DF’s)– Derived, ground-processed moments in useful coordinates

ESA In-Flight CalibrationESA in-flight calibration and instrument maintenance requires:

1. Getting the timing straightened out - finished1. Getting the timing straightened out finished2. MCP Bias Voltage determination - ongoing3. Sunlight contamination - minimal4 Spacecraft potential correction - in place4. Spacecraft potential correction in place5. Dead-time corrections - in place6. Energy efficiency corrections - in place7 Relative efficiency corrections - in place7. Relative efficiency corrections in place8. Absolute efficiency - in place9. Electron-Ion cross calibration - in place10 Inter-spacecraft cross calibrations - in place10.Inter-spacecraft cross calibrations - in place

76

Sunlight & Photo-electronsNo measureable direct sunlight

contamination. Spacecraft photo-electrons must be

sc_pot=20V

eliminated.Photo-electrons are produced in the

aperture, on the spacecraft and on langmuir probes

Wire boom the spacecraft and on langmuir probes.boom photo-e

s/c photo-e

Axial boom photo-e

p

sc pot=45V

Wire boom

sc_pot=11Vsc_pot=45Vphoto-e

77

Correcting for s/c potentialThm_load_esa_pot.pro

• Software currently assumes s/c potential isassumes s/c potential is 1.15 times the measured average spin plane boom potential (V1234) plus 1

lt t t t ti lvolt contact potential. 1.15 calibration factor determined from density comparisons.comparisons.• Similar to simulations conducted by Cully.• s/c potential corrections are important forare important for electrons and cold ions.

78

Correcting for dead timeTotal dead-time is a combination of “preamplifier dead-time” and “detector dead-time”, which is difficult to calibrate and dependsdifficult to calibrate and depends on MCP bias voltage and preamplifier threshold.

For the above case, high density results in significant electron ESA deadtime. This allowed an estimate of total deadtime (~0.16 us) by eliminating the slope in the Ni/Ne d it ti ( i ht b tt l)density ratio (right, bottom panel).

79

Detector Energy Efficiency 1Default ion efficiencyDefault ion efficiency assumes lab results from Funsten.

For log energy and 2 kV pre-acceleration, the efficiency is shown belowefficiency is shown below.

Energy (eV)

[From ISSI book, p.17]

80

Detector Energy Efficiency 2

Electron efficiency assumes:“Relative electron detectionRelative electron detection efficiency of microchannel plates from 0-3 keV”, R.R. Goruganthu and W. G. Wilson, g ,Rev. Sci. Instrum. Vol. 55, No. 12 Dec 1984.

Efficiency formula: (1-exp(-k*delta/delta max))/(1-exp(-k))

Energy (eV)

(1 exp( k delta/delta_max))/(1 exp( k))

81

Detector Energy Efficiency 3Ion efficiency corrected – work in progress:

Default ion efficiency produces a slopeproduces a slope in the Ni/Ne ratio (right bottom).

Default

Proposed ion efficiency

Energy (eV)

efficiency eliminates this slope. Difference assumed to beProposed assumed to be due to fringing fields at the detector.

Proposed

Energy (eV)Energy (eV)

82

Detector Energy Efficiency 4ff ffElectron efficiency: The proposed electron energy efficiency

curve seems to produce more consistent electron-ion density ratios over a number of orbits. More investigation is needed, but difference appears to be due to secondary e production in ESAdifference appears to be due to secondary e- production in ESA.

Current e- efficiency Proposed e- efficiency

Energy (eV) Energy (eV)83

Relative Anode Calibration


ESA N, V Cross Calibration

Electron-Ion density ratios should agree at the ~5% level.

Ne should be less than Ni since Ni is calculated assuming H+ and efficiency is species dependent, corrected for H+ .

Variations in calculated electron velocity are due to small density fluctuations during a s/c spin.

85

Cross-CalibrationFig. 14 McFadden et al.Ion-electron sensor cross-calibrations use Ni/Ne on same / ( f ) ithis/c (see f,g) within

magnetosheath (12:30-17:00) and match the result to the expexted ratio based on e pe ed a o based oupstream alpha content, ~0.99. Inter-s/c comparisons match electron densities within solar wind (>18:00UT) as shown inwind (>18:00UT) as shown in fig. h. Note: Solar wind ion densities are not used as they are underestimates due to narrowness of SW beam


Cold ions or conics?Differences between electron and ion densities are often due to missed cold ions, but can l b d talso be due to

ions with higher mass.

Conics often appear with a broader energybroader energy, are generally field aligned, and tend to be oxygen.

87

THEMIS/SST Sensor Unit Schematic

Thick DetectorAl/Polyamide/Al FoilFoil Detector

Open Detector

FoilFoilCollimator(electron side)

Sm-Co Magnet

AttenuatorAttenuator

Open Collimator (ion side)

g

88

Each sensor unit is a:THEMIS/SST Sensor Unit Details

• Each sensor unit is a:– Dual-double ended solid state telescope– Each double ended telescope (1/2 sensor) has:– Each double ended telescope (1/2 sensor) has:

• Triplet stack of silicon solid state detectors• Foil (on the side measuring electrons)

– Filters out ions <~350 keV – Leaves electron flux nearly unchanged

• Magnet / Open (on the side measuring ions)g ( g )– Filters out electrons <400 keV – Leaves ion flux nearly unchanged

• Mechanical Pinhole attenuatorMechanical Pinhole attenuator– Reduces count rate during periods of high flux– Reduces radiation damage (caused by low energy ions) during

periods of high flux


p g• Collimators• Preamplifier / shaping electronics

Detector Pixelation• Detectors similar to STEREO/STE

– Produced at LBNL/Craig Tindall PI

Active areaActive area

Guard ring 5 mm

10 mm

Additional Pixels not used for Themis

90

Detector Wiring

4 5 V

-35 V

~200 A Polysilicon +4.5 V

-2.5 V

F Outnp

y

F 225FB

Pixelated side ~1200 A Dead layer

T Out

F Test inpnnp

T 225FB

y

O Out

T Test in

pnO 225F

B

O Test in~200 A Polysilicon + ~200 A Al

Outside Grounded

300 micron thick detectors

91

SST Detector Mechanical Design/Connections

• Typical Electrical Connection Between Detector and Flex-Circuit

Wirebond Loop

(NOT t l t l(NOT to scale – actual loop height < 300 micron)

Kapton Flex-Circuit

Detector (pixelated side)

92

SST Detector Mechanical Assembly

Detectors (4)

BeCu Gasket (3)• DFE Board Subassembly

KaptonHeater

Spring Clamp

PEEK Spacer (4)

Spring Plate (2)Kapton Flex-Circuit (4) AMPTEK Shield

Thermostat

• Detector Board Composition (exploded view)93

SST Detector Mechanical: Real Life

SST Mechanical Design• Bi-Directional Fields-of-View

95

SST Mechanical Design

• Sensor Orientation Relative to Spacecraft Bus

96

SST Mechanical Design• Attenuator Actuation – CLOSED position

Honeywell SwitchHoneywell Switch Honeywell Switch (extended-position)

y(compressed-position)

SMA Actuator (retracted)

SMA Actuator (extended)

( )

97

SST Mechanical Design• Attenuator Actuation – OPEN position

Honeywell SwitchHoneywell Switch Honeywell Switch (compressed-position)

y(extended-position)

SMA Actuator (extended)

SMA Actuator (retracted)

( )

98

• Linear Actuators

– Shaped Memory Alloy (SMA) actuator– Single direction 125 gram pull-force

• Required force < 42 gram => F.S. > 3.0– Operating temp range: -70°C to +75°C

Extended Position

Retracted Position

Relative Size(commercial model shown)

99

SST Accommodation

SSTs

Angelopoulos, 2008

SST Accommodation

Angelopoulos, 2008 101

Particle Ground processing [1]Particle Ground processing [1]• Loading and viewing on-board processed data [1]

There is a single routine for loading on board moments– There is a single routine for loading on-board moments• thm_load_mom, level=1 (loads L1 or L2 data)

– Products introduced (in either case)» 1 thb_p[e,s,t][i,e]m_density x 6» 2 thb_peim_flux (particle flux in #/cm2/s)

3 thb i ft ( t fl i V/ 3)» 3 thb_peim_mftens (momentum flux in eV/ cm3)» 4 thb_peim_eflux (particle flux in #/cm2/s)» 5 thb_peim_velocity (km/s)» 6 thb_peim_ptens (eV/cm3)» 7 thb_peim_ptot (trace of pressure tensor)»» …» 43 thb_pxxm_pot (probe potential subtracted, in Volts) » 44 thb_pxxm_qf » 45 thb_pxxm_shft

– Note: [e,s,t] correspond to ESA, SST, Total; [i,e] to ions, electronsE g thb ptim velocity is the total velocity from the ESA and SST combined– E.g., thb_ptim_velocity is the total velocity from the ESA and SST combined

– There are two routines for introducing ESA distributions• thm_load_esa_pkt (loads L0 data) • thm_load_esa, level=… (loads L1 and L2 data, will become prime in future)

– There is a single routine for introducing SST distributionsThere is a single routine for introducing SST distributions• thm_load_sst (loads SST L1 data)

102

Ground processing [2]Data structure at right is loaded into IDL memory by the below routines.

PROJECT_NAME STRING 'THEMIS'SPACECRAFT STRING 'c'DATA_NAME STRING 'IESA 3D Reduced'APID INT 455UNITS_NAME STRING 'compressed'UNITS_PROCEDURE STRING 'thm_convert_esa_units'VALID BYTE Array[27909]TIME DOUBLE Array[27909]

thm_load_esa_pkt.pro – creates 6 IDL ESA data t t t i i ti

END_TIME DOUBLE Array[27909]DELTA_T DOUBLE Array[27909]INTEG_T DOUBLE Array[27909]DT_ARR FLOAT Array[32, 88, 8]CS_PTR LONG Array[27909]CS_IND INT Array[27909]CONFIG1 BYTE Array[27909]CONFIG2 BYTE Array[27909]

structures containing time, counts, energy & angle arrays, geometric factors, etc.

AN_IND INT Array[27909]EN_IND INT Array[27909]MD_IND INT Array[27909]NENERGY INT Array[6]ENERGY FLOAT Array[32, 6]DENERGY FLOAT Array[32, 6]EFF FLOAT Array[32, 6]NBINS INT Array[8]

thm_load_esa_pot.pro – adds the spacecraft potential to the ESA data structures

THETA FLOAT Array[32, 88, 8]DTHETA FLOAT Array[32, 88, 8]PHI FLOAT Array[32, 88, 8]DPHI FLOAT Array[32, 88, 8]PHI_OFFSET FLOAT Array[27909]DOMEGA FLOAT Array[32,

88, 8]GF FLOAT Array[32, 88, 8]to the ESA data structures.

thm_load_esa_mag.pro adds the magnetic field to

GEOM_FACTOR FLOAT 0.00110000DEAD FLOAT 1.60000e-007MASS FLOAT 0.0104389CHARGE FLOAT 1.00000SC_POT FLOAT Array[27909]B_GSE FLOAT Array[27909, 3]MAGF FLOAT Array[27909, 3]DAT0 BYTE Array[6, 16]– adds the magnetic field to

the ESA data structures.DAT1 BYTE Array[13753, 32]DAT2 BYTE Array[1, 96]DAT3 BYTE Array[1, 192]DAT4 BYTE Array[1, 1152]DAT5 BYTE Array[14150, 1200]

Ground processing [3]Once the data are loaded, use these routines to return an ESA data structure at a single time.

Example IDL commands

get_thc_peir.prop – particle (f – fields)e – ESA (s – SST)i ion (e electron)

t=‘2007-5-5/12:31’dat=get_thc_peir(t)dat=get_thc_peir() dat=get thc peir(/ad)i – ion (e - electron)

r – reduced (f-full,b-burst)

Routines that operate on

dat=get_thc_peir(/ad)dat=get_thc_peir(/re)

Routines that operate on structures

n_3d_new.prov 3d new.pro

print,n_3d_new(dat)print,n_3d_new(dat,energy=[0,20])

_ _ p

Loop routines for time series plotsget_2dt.pro

get_2dt,’n_3d_new’,’thc_peir’get_en_spec,’thc_peir’,name=‘test’

get_en_spec.pro104

Ground processing [4]L di d i i b d d d t [4]• Loading and viewing on-board processed data [4]– Generic routines

• To “get” any distribution function type:» dat=thm_part_dist(‘thb_peif’,gettime(/c))» ;or: ctime,t & dat=thm_part_dist(‘thb_peif’,t)» ;or: ctime,t & dat=get_thb_peif(t)» wset,1 & spec3d,dat» wset,2 & plot3d,dat,units=‘counts’ ; convert units» ; other types: eflux,flux,df,rate ; yp , , ,

• To obtain moments type (e.g. n, v, t):» thm_part_spec_calc,probe='b',moments=['density','flux'],instrument=

['peif','psif']• Unit Conversions:

– To copy/convert to/from units (eflux,flux,df,counts,rate), use function conv_units:

» dat_new=conv_units(dat,‘df’)

105

Ground processing [5]Loading and viewing on board processed data [5]• Loading and viewing on-board processed data [5]– Generic routines

• To “view” distribution function contents type:» dat=thm_part_dist(‘thb_peif’,gettime(/c))» help, dat, /str» help, dat, /str» print, dat.energy, dat.theta, dat.phi ; to view energy/angle bin centers

** Structure <13c28790>, 35 tags, length=140952, ….:PROJECT_NAME STRING 'THEMIS'SPACECRAFT STRING 'b'

ENERGY FLOAT Array[32, 88]DENERGY FLOAT Array[32, 88]EFF DOUBLE Arra [32 88]SPACECRAFT STRING b

DATA_NAME STRING 'IESA 3D Full'APID INT 454UNITS_NAME STRING 'counts'UNITS_PROCEDURE STRING 'thm_convert_esa_units'VALID BYTE 1

EFF DOUBLE Array[32, 88]BINS INT Array[32, 88]NBINS INT 88THETA FLOAT Array[32, 88]DTHETA FLOAT Array[32, 88]PHI FLOAT Array[32 88]

TIME DOUBLE 1.1746494e+009 ; seconds since 1970DELTA_T DOUBLE 3.0899630INTEG_T DOUBLE 0.0030175420DT_ARR FLOAT Array[32, 88]CONFIG1 BYTE 2CONFIG2 BYTE 1

PHI FLOAT Array[32, 88]DPHI FLOAT Array[32, 88]DOMEGA FLOAT Array[32, 88]GF FLOAT Array[32, 88]GEOM_FACTOR FLOAT

0.00153000CONFIG2 BYTE 1AN_IND INT 1EN_IND INT 1MODE INT 2NENERGY INT 32

DEAD FLOAT 1.70000e-007MASS FLOAT 0.0104389CHARGE FLOAT 1.00000SC_POT FLOAT 0.000000MAGF FLOAT Array[3]DATA FLOAT Array[32 88]DATA FLOAT Array[32, 88]

106

Commands to load & plot ESA data:

Ground processing [6]

startdate = '2007-07-31/0:00'sc='b'ndays=1ndays 1thm_inittimespan,startdate,ndaysthm_load_state,probe=scth l d kt bthm_load_esa_pkt,probe=scthm_load_esa_mag,sc=scthm_load_esa_pot,sc=sc

tplot,[‘thb_pe??_en_counts’]

Default plots are in “counts”. Note discontinuity in ion 24 energy modeNote discontinuity in ion 24 energy mode.Calibrated data in energy flux do not have this discontinuity.

107

Ground processing [7]ESA Plot T pes and Units

Once the data are loaded, there are several plotting routines.

ESA Plot Types and Units

dat=get_thb_peif()

spec3d,datp ,

spec3d,dat,units=‘df’

Each angle bin is a differentEach angle bin is a different color curve.

For 3-D angle maps use:

plot3d,dat

108

Ground processing [8]Loading and viewing on board processed data• Loading and viewing on-board processed data– ….– Generic routines

• To introduce s/c potential and magnetic field type:» ctime t & dat=thm part dist(‘thb peif’ t)» ctime,t & dat=thm_part_dist( thb_peif ,t)» get_data,’thb_pxxm_pot’,data=thb_pxxm_pot_str» it=where((thb_pxxm_pot_str.x gt t(0)-3.) and (thb_pxxm_pot_str.x lt t(0)+3.))» dat.sc_pot=median(thb_pxxm_pot_str.y(it))» get_data,’thb_fgs_dsl’,data=thb_fgs_dsl_str» jt=where((thb fgs dsl str x gt t(0)-1 5) and (thb fgs dsl str x lt t(0)+1 5))» jt=where((thb_fgs_dsl_str.x gt t(0)-1.5) and (thb_fgs_dsl_str.x lt t(0)+1.5))» dat.magf(*)=thb_fgs_dsl_str.y(jt,*)

• If sc_pot not available, use a guess (e.g., sc_pot=15V)» dat.sc_pot=10. ; Volts

• Compute the density, temperature for that instant» print 'ion density 1/cc ' n 3d(dat)» print,'ion density 1/cc = ',n_3d(dat)» print,'ion temperature eV = ',t_3d(dat)

• Another way is to use generic tool to introduce scpot and mag:» thm_part_spec_calc,probe='b',scpot_suffix=‘_pxxm_pot’,mag_suffix=‘_fgs_dsl’,» moments=['density',‘velocity‘,’t3’,’magt3’],instrument=['peif','psif']

109

Ground processing [9]Plot3d acting on data structure “dat” produces the below plot.Example IDL command: plot3d,dat,/zero

MagneticMagnetic field indicated by plus andby plus and diamond

110

Ground processing [10]IDL routines for time series plotsplots

get en specget_en_spec

get_2dt

111

Ground processing [11]• Loading and viewing on-board processed datag g p

– Viewing Distribution Function Cuts• Use (crib thm_crib_esa_slice2d provided in themis software distribution):• ; choose which spacecraft (a-e): sc = ‘b’ • ; choose type of data - f-full, r-reduced, b-burst: typ = 'b'• thm_esa_slice2d,sc,typ,current_time,timeinterval,thebdata='th'+sc+'_fgs_ds_ _ _ _ _

l',species=species,range=range,rotation=rotation,angle=angle,filetype=filetype,outputfile=outputfile;,nosmooth=1

• Note, rotations:– ; 'BV': x = V_para and the bulk velocity in the x-y plane. (DEFAULT)– ; 'BE': x = V_para and the VxB in the x-y plane.– ; 'xy': x = V x and y = V y; xy : x V_x and y V_y.– ; 'xz': x = V_x and y = V_z.– ; 'yz': x = V_y and y = V_z.– ; 'perp': x-y plane is perp. to B, x is velocity projection on plane.– ; 'perp_xy': x-y plane is perp. to B, x is x-axis projection on plane.– ; 'perp_xz': x-z plane is perp. to B, y is z-axis projection on plane.– ; 'perp yz': x-y plane is perp to B x is y-axis projection on plane; perp_yz : x y plane is perp. to B, x is y axis projection on plane.

• Loading and viewing on-board processed dataGround processing [12]g g p

– Viewing Pitch-Angle vs Energy• Use (crib crib_get_en_pa_spec.pro)Credits: Written by Arjun Raj, 1999; last modified by: Tai Phan on August 31,

2007, bundled by Christine Gabrielse, Aug. 22, 2008IDL d d d t; IDL codes needed to run:

; get_en_pa_spec_themis.pro; dat_avg.pro; angl.pro; dotp.pro; fixangdata.pro; get_counts.pro


Data Analysis ToolsData Analysis Tools• Higher level products and visualization

– Particle spectrograms in various coordinates.– Code also in class materials : idl/thm crib spectrograms proCode also in class materials : idl/thm_crib_spectrograms.pro

• DSL coordinates– Energy, theta/phi angle spectrograms– ;DSL coordinates– ; energy spectrogram– thm part getspec, probe=['b'], trange=['07-03-23/11:10','07-03-23/11:30'], $t _pa t_getspec, p obe [ b ], t a ge [ 0 03 3/ 0 , 0 03 3/ 30 ], $– data_type=['psif'],/energy, $– phi=[-135,-45], theta=[-45,45], erange=[25000,500000],suff='_dusk' – thm_part_getspec, probe=['b'], trange=['07-03-23/11:10','07-03-23/11:30'], $– data_type=['psif'],/energy, $– phi=[45,135], theta=[-45,45], erange=[25000,500000],suff='_dawn' – ; phi spectrogram– thm_part_getspec, probe=['b'], trange=['07-03-23/11:10','07-03-23/11:30'], $– data_type=['peir'],angle='phi', $– phi=[0,360], theta=[-90,90], erange=[1.5e4,2.5e4]– ; theta spectrogram

thm part getspec probe=['b'] trange=['07 03 23/11:10' '07 03 23/11:30'] $– thm_part_getspec, probe=[ b ], trange=[ 07-03-23/11:10 , 07-03-23/11:30 ], $– data_type=['peir'],angle='theta', $– phi=[0,360], theta=[-90,90], erange=[1.5e4,2.5e4]– tplot,'thb_fgs_gsm thb_psif_en_eflux_dusk thb_peir_an_eflux_*'– tlimit,['07-03-23/11:12','07-03-23/11:22']

114

Data Analysis Tools [2]• DSL coordinatesDSL coordinates

– Energy, theta/phi angle spectrograms– ;DSL coordinates (results)

115

Data Analysis ToolsData Analysis Tools• Higher level products and visualization

– Particle spectrograms in various coordinates• FAC coordinates (field aligned)• FAC coordinates (field aligned)

– Energy, pitch angle (pa) / gyro(velocity)phase angle spectrograms– ; Energy spectrogram– thm_part_getspec, probe=['b'], trange=['07-03-23/11:10','07-03-23/11:30'],$ – data_type=['psif'], /energy, $y gy– pitch=[0,45], suff='_para', $ – erange=[5000,25000],regrid=[32,16]– ; Gyro(velocity)phase spectrogram– thm_part_getspec, probe=['b'], trange=['07-03-23/11:10','07-03-23/11:30'],$ – data type=['psif'] angle='gyro' $– data_type=[ psif ], angle= gyro , $– pitch=[45,135], other_dim='ygsm', suff='_perp', $ – erange=[100000,150000],regrid=[32,16]– ; Pitch angle spectrogram– thm_part_getspec, probe=['b'], trange=['07-03-23/11:10','07-03-23/11:30'],$

$– data_type=['peer'], angle='pa', $– erange=[15000,25000],regrid=[32,16]– tplot,'thb_fgs_gsm thb_psif_en_eflux_para thb_psif_an_eflux_gyro_perp

thb_peer_an_eflux_pa'– tlimit,['07-03-23/11:12','07-03-23/11:22']tlimit,[ 07 03 23/11:12 , 07 03 23/11:22 ]

116

Data Analysis Tools• FAC coordinates (field aligned)FAC coordinates (field aligned)

– Energy, pitch angle (pa) / gyro(velocity)phase angle spectrograms (results)

117

Data Analysis Tools• Higher level products and visualization

– Particle spectrograms in various coordinates• FAC coordinates (field aligned) (Look in: thm_fac_matrix_make)

– other_dimension:» ; 'Xgse', (DEFAULT) translates from gse or gsm into FAC» ; Definition(works on GSE, or GSM): X Axis = on plane defined by Xgse - Z» ; Second coordinate definition: Y = Z x X gse» ; Second coordinate definition: Y = Z x X_gse» ; Third coordinate, X completes orthogonal RHS» ; 'Rgeo',translate from geo into FAC using radial position vector» ; Rgeo is radial position vector, positive radialy outwards.» ; Second coordinate definition: Y = Z x Rgeo (westward)» ; Third coordinate, X completes orthogonal RHS XYZ.» ; 'mRgeo' opposite to above» ; mRgeo , opposite to above » ; mRgeo is radial position vector, positive radially inwards.» ; 'Phigeo', translate into FAC using azimuthal position vector» ; Phigeo is the azimuthal geo position vector, positive Eastward» ; First coordinate definition: X = Phigeo x Z (positive outwards)» ; Second coordinate, Y ~ Phigeo (eastward) completes orthogonal RHS XYZ» ; 'mPhigeo' opposite to above» ; 'mPhigeo', opposite to above» ; Second coordinate, Y ~ mPhigeo (Westward) completes orthogonal RHS XYZ» ; 'Phism', translate into FAC using azimuthal Solar Magnetospheric vector.» ; Phism is "phi" vector of satellite position in SM coordinates.» ; Y Axis = on plane defined by Phism-Z, normal to Z» ; Second coordinate definition: X = Phism x Z; Third completes orthogonal RHS;

'mPhism' opposite to abo e; 'mPhism', opposite to above» ; mPhism is minus "phi" vector of satellite position in SM coordinates.» ; 'Ygsm', translate into FAC using cartesian Ygsm position as other dimension.» ; Y Axis on plane defined by Ygsm and Z» ; First coordinate definition: X = Ygsm x Z» ; Third completes orthogonal RHS XYZ

118

Example of Msheath/Mpause

ESS 261 Low Energy Particles119First limitation: s/c charging prevents cold ions from reaching sensor (2204-2207UT)Second limitation: s/c potential below lowest e- energy; cold electrons missed (2211-2214UT)

Example of Msheath/Mpause/Plumes


SST Products• Products: Full Reduced (Burst is same as full)• Products: Full, Reduced (Burst is same as full)

– Full: 16E x 64A– Reduced: 16E x 6A , orReduced: 16E x 6A , or

16E x 1A (omni)• Modes: Slow Survey, Fast Survey, Particle

Burst• Slow Survey:

– Full distributions (ions and electrons) at 5min resolution( )– Reduced, omnidirectional distributions: every spin

• Fast Survey:– Ions: Full distributions every spin– Electrons: Reduced distributions (16E x 6A) every spin

• Burst:– Ions: same as above– Electrons: Full distributions every spin

121

Data Analysis Tools• Pitfalls• Pitfalls

– Sun contamination• ; Sun contamination is masked on board but often fails

; Use keyword: mask_remove to removed masked bins and interpolate across y _ psectors

• ; Sun contamination is lefted unmasked recently (and most of the time) on board ; There is code to recognize the faulty bins (saturated) and remove them altogether.; This is called : method_sunpulse_clean='spin_fit' , or ‘median’ and tells the; programs to remove data beyond 2sigma away from spin-phase fit/median.

• ;Sun contamination/saturation also affects other channels due to electronic noise.;The code can remove the typical noise value and provide the remaining good; signal (assuming no saturation). The keyword is: enoise_bins and the; procedure is documented in: thm_crib_sst_contamination.pro

122

Sun contamination (thm_crib_sst_contamination.pro)– ;PROCEDURE: thm_crib_sst_contamination

;Purpose: 1 Demonstrate the basic procedure for removal of sun contamination– ;Purpose: 1. Demonstrate the basic procedure for removal of sun contamination,– ; electronic noise, and masking.– ; 2.. Demonstrate removal of suncontamination via various methods. – ; 3. Demonstrate the correction of inadvertant masking in SST data– ; 4. Demonstrate scaling data for loss of solid angle in SST measurements.– ; 5. Demonstrate substraction of electronic noise by selecting bins in a specific region– ; 6. Show how to use these techniques for both angular spectrograms,energy spectrgrams, and

moments.– ;SEE ALSO:– ; thm_sst_remove_sunpulse.pro(this routine has the majority of the documentation)– ; thm_part_moments.pro, thm_part_moments2.pro, thm_part_getspec.pro– ; thm_part_dist.pro, thm_sst_psif.pro, thm_sst_psef.pro,thm_sst_erange_bin_val.pro– ; thm_crib_part_getspec.pro

Sun contamination (sst_remove_sunpulse.pro)– ; Routine to perform a variety of calibrations on full distribution sst data These can; Routine to perform a variety of calibrations on full distribution sst data. These can

remove sun contamination and on-board masking. They can also scale the data to account for the loss of solid angle from the inability of the sst to measure directly along the probe geometric Z axis and the inability to measure directly along the probe geometric xy plane (ie X=0 Y=0 Z = n or X=n Y=m Z=0 are SST 'blind spots')


geometric xy plane.(ie X 0,Y 0,Z n or X n,Y m,Z 0, are SST blind spots ) THM_REMOVE_SUNPULSE routine should not generally be called directly. Keywords to it will be passed down from higher level routines such as, thm_part_moments, thm_part_moments2, thm_part_dist,thm_part_getspec, thm_sst_psif, and thm_sst_psef

Data Analysis Tools• Pitfalls• Pitfalls

– Sun contamination– Read crib sheets:

thm crib sst contamination pro andthm_crib_sst_contamination.pro, anddocumented procedure: thm_sst_remove_sunpulse.pro

» ; » edit3dbins,thm_sst_psif(probe=sc, gettime(/c)), bins2mask» ; ON: Button1; OFF: Button2; QUIT: Button3» ; ON: Button1; OFF: Button2; QUIT: Button3» print,bins2mask» thm_part_getspec, probe=probe, trange=[sdate, edate], $» theta=[-45,0], phi=[0,360], $» data type=['psif'] start angle=0 $» data_type=[ psif ], start_angle=0,$»

angle='phi',method_sunpulse_clean='median',tplotsuffix='_ex2_t1',$» enoise_bins =

bins,enoise bgnd time=times,mask remove=.99bins,enoise_bgnd_time times,mask_remove .99» tplot

124

Ground processing (particles only)• Pitfalls• Pitfalls

– Sun contamination: Bin selection » ;

125

Density CorrectionInterpolate densities• Interpolate densities

• Add• date='2008-03-01'• startdate = '2008-03-01/00:00'• timespan,startdate, 4.0, /hour• Trange=['08-03-01/00:00','08-03-01/04:00']

T ['08 03 01/01 40' '08 03 01/02 40']• Tzoom=['08-03-01/01:40','08-03-01/02:40']

• ;... select exact time interval to calculate join ESA/SST moments• tbeg = time_double(date+'/00:00')• tend = time_double(date+'/04:00')• ;select a probe• sc='b'• thm_load_state,probe=sc,coord='gsm',/get_support

h l d fi l l 1 b• thm_load_fit, level=1, probe=sc,datatype=['efs', 'fgs'],/verbose

• thm_cotrans,strjoin('th'+sc+'_fgs'),out_suf='_gsm', in_c='dsl', out_c='gsm'

• ;• ; SST now• thm_load_sst,probe=sc,lev=1• thm_part_moments, probe = sc, instr= ['ps?f'], $

$

Ni

• moments = ['density', 'velocity', 't3'], $• mag_suffix='_peir_magt3', $• scpot_suffix='_peir_sc_pot';,/median• ; work in gsm• thm_cotrans,'th'+sc+'_ps?f_velocity',

in_coord='dsl',out_coord='gsm',out_suffix='_gsm'• ;• ; ESA now

Ne

• thm_load_esa,probe=sc• ; Interpolate densities• tinterpol_mxn,'th'+sc+'_peer_density',

'th'+sc+'_peir_density',/overwrite,/nan_extrapolate• tinterpol_mxn,'th'+sc+'_ps?f_density',

'th'+sc+'_peir_density',/overwrite,/nan_extrapolate• …• ; ...total ion density• totNi = sst_i_n.y + esa_i_n.y

126

Velocity CorrectionI t l t d iti• Interpolate densities

• Add flux• ;• ;

l• ; ...sst Flux• sstFi = sst_i_v.y*0.• sstFi[*,0] = sst_i_n.y*sst_i_v.y[*,0]• sstFi[*,1] = sst_i_n.y*sst_i_v.y[*,1]• sstFi[*,2] = sst_i_n.y*sst_i_v.y[*,2]

• ; ...esa Flux; ...esa Flux• esaFi = esa_i_v.y*0.• esaFi[*,0] = esa_i_n.y*esa_i_v.y[*,0]• esaFi[*,1] = esa_i_n.y*esa_i_v.y[*,1]• esaFi[*,2] = esa_i_n.y*esa_i_v.y[*,2]

• ; ...total ion density• totNi = sst_i_n.y + esa_i_n.y

• store_data, 'th'+sc+'_Ni',$data=x:esa_i_n.x, y:totNi

• options, 'th'+sc+'_Ni', 'ytitle', $'Ni !C!C1/cm!U3'

• ylim 'th'+sc+' Ni' 0 01 1 1• ylim, th +sc+ _Ni , 0.01, 1., 1

• ; ...total ion velocity (GSM)• totVi = esa_i_v.y*0.• totVi[*,0] = (sstFi[*,0]+esaFi[*,0])/totNi• totVi[*,1] = (sstFi[*,1]+esaFi[*,1])/totNi• totVi[*,2] = (sstFi[*,2]+esaFi[*,2])/totNi

127

Pressure CorrectionRemove SST noise• Remove SST noise

• Interpolate pressures• Then add• ;• ; SST now• ; SST now• thm_load_sst,probe=sc,lev=1• thm_part_moments, probe = sc, instr= ['ps?f'], $• moments = ['density', 'velocity', 't3'], $• mag_suffix='_peir_magt3', $• scpot_suffix='_peir_sc_pot';,/median• ; …interpolate• ; … add• ; ...pressure• ; ...SST: perpendicular temperature only• sst_Tperp = .5*(sst_i_t3.y[*,0]+sst_i_t3.y[*,1])• sst_i_p_nPa = 0.16*.001*sst_i_n.y * sst_Tperp • ; perp. pressure in nPa• store_data, 'th'+sc+'_psif_p_perp_nPa', $• data=x:sst_i_n.x, y:sst_i_p_nPa• options, 'th'+sc+' psif p perp nPa', $p _p _p_p p_• 'ytitle', 'sst Pi !C!CnPa'

• ; ...ESA: scalar temperature• esa_Ti = total(esa_i_T.y,2)/3.• store_data,'Ti_th'+sc+'_peir', $• data=x:esa_i_n.x, y:esa_Ti• ; ...ESA ion pressure:• esa i p nPa = 0.16 *.001 * esa i n.y*esa Ti_ _p_ _ _ y _• ; scalar pressure in nPa• store_data, 'th'+sc+'_peir_p_nPa', $• data=x:esa_i_n.x, y:esa_i_p_nPa• options, 'th'+sc+'_peir_p_nPa', $• 'ytitle', 'esa Pi !C!CnPa'

• ; ...Total ion pressure• totPi = sst i p nPa + esa i p nPa_ _p_ _ _p_• store_data, 'th'+sc+'_i_p_nPa', $• data=x:esa_i_n.x, y:totPi• options, 'th'+sc+'_i_p_nPa', 'ytitle', 'Pi !C!CnPa'

128

Finite gyroradius techniquesI G di l d t t h i• Ion Gyroradius large compared to magnetospheric boundaries– Can be used to remotely sense speed

To Tail

and thickness of boundaries– Assumption is that boundary is sharp

and flux has step function across• Application at the magnetopause

THEMIS• Application at the magnetopause• Application at the magnetotail

– Can also be applied to waves ifparticle gradient is sufficiently high

• Application on ULF waves atinner magnetosphere

To Earth

Method exploits finite iongyroradius to remotely sense

To EarthTo Sun

gyroradius to remotely senseapproaching ion boundary andmeasure boundary speed (V )129

At the magnetotail325kmi,thermal-tail (4keV,20nT)= ~325km

i,super-thermal (50keV,20nT)= ~2200km

Plasma Sheet Thickness ~ 1-3 REBoundary Layer Thickness 500 2000kmBoundary Layer Thickness ~500-2000kmCurrent layer Thickness ~ 500-2000km

Waves Across Boundary: ~1000-10,000kmAlong Boundary: Normal : 1 10Along Boundary: ~Normal : 1-10

RE

For magnetotail particles, the current layer andFor magnetotail particles, the current layer and plasma sheet boundary layer are sharp compared to the superthermal ion gyroradius and the magnetic field is the same direction in the plasma sheet and outside (the lobe). Thisthe plasma sheet and outside (the lobe). This means we can use the measured field to determine gyrocenters both at the outer plasma sheet and the lobe, on either side of the hot magnetotail boundary.magnetotail boundary.

130

Sid Vi ( l ti ) 52oSide View (elevations)

SpinSST:El ti

25o

52

SpinAxis

Elevationdirection(DSL) 25o

To Sun( DSL) -25o

-52o

ESA:Elevation 11 25o

33.75o

Elevationdirection(DSL)

11.25o

131

Top View (sectors)p ( )For ESA and SST (0=Sun)

Spin axisp

11.25o

33 75

To Sun (0o)

Spin motiondirection ( DSL)

33.75o

Normal to Sun +90oNormal to Sun, +90o

132

TH-BTH-B

(a)

(b)

(a)

(b)

B fielda im th You care to time this!

(c)(c)

Particle motion direction

azimuth(solid white)

You care to time this!(+/- 90o to Bfield azimuth)

(d)

(e)

(d)

(e)

Particle motion directionCoordinate: ( DSL)Energy: 125-175keV

( )( )

Note: direction dependson spin axis. -B field

i thazimuth(dashed white)

133

Multiple spacecraft, energies, elevations

A

B

D….

E

Elev: 25deg E=30-50keVElev: 25deg, E=80-120keV

134

Vi_const 310km/sec/keV fci_cons 0.0152Hz/nT B 30nTTi 40keV rho_ion 683kmTi 100keV rho_ion 1081km Ti 150keV rho_ion 1323km Ti 300keV rho_ion 1872km

SC E (keV) detectord (deg) r timeB 40 SPW -128.0 683.4 11:19:29B 40 SPE -52.0 683.4 11:19:39B 40 SEW -155.0 683.4 11:19:18B 40 SEE -25.0 683.4 11:19:42B 40 NPW 128 0 683 4 11:19:29

Note:NEE= North-Equatorial, EastNPW=North-Equatorial, WestAngles measured from East directionB 40 NPW 128.0 683.4 11:19:29

B 40 NPE 52.0 683.4 11:19:38B 40 NEW 155.0 683.4 11:19:24B 40 NEE 25.0 683.4 11:19:43B 100 SPW -128.0 1080.5 11:19:17B 100 SPE -52.0 1080.5 11:19:42B 100 SEW -155.0 1080.5 11:19:20

Angles measured from East direction-25deg elevation, 90deg East = SEE+52deg elevation, 90deg East = NPE… Spin axis

B 100 SEE -25.0 1080.5 11:19:45B 100 NPW 128.0 1080.5 11:19:20B 100 NPE 52.0 1080.5 11:19:45B 100 NEW 155.0 1080.5 11:19:23B 100 NEE 25.0 1080.5 11:19:48B 150 SPW -128.0 1323.4 11:19:10B 150 SPE 52 0 1323 4 11:19:44

pNPW

NEWNPE

NEEB 150 SPE -52.0 1323.4 11:19:44B 150 SEW -155.0 1323.4 11:19:14B 150 SEE -25.0 1323.4 11:19:51B 150 NPW 128.0 1323.4 11:19:23B 150 NPE 52.0 1323.4 11:19:45B 150 NEW 155.0 1323.4 11:19:13B 150 NEE 25.0 1323.4 11:19:48

B

SEW SEEB 300 SPW -128.0 1871.5 11:19:10B 300 SPE -52.0 1871.5 11:19:44B 300 SEW -155.0 1871.5 11:19:14B 300 SEE -25.0 1871.5 11:19:51B 300 NPW 128.0 1871.5 11:19:23B 300 NPE 52.0 1871.5 11:19:45B 300 NEW 155 0 1871 5 11 19 13

SEW

SPW

SEE

SPEB 300 NEW 155.0 1871.5 11:19:13B 300 NEE 25.0 1871.5 11:19:48 Boundary

135

Spin axis

BNPW

NEW

NPE

NEEV: NEE Part. direction

NEW NEE

YSC

SEW

SPW

SEE

SPEZd

SPW SPE

C ld/ l

n

Hot/dense plasma

Cold/tenuous plasmaY

GCNEE

Y

n

Y

Show: d=*sin(-)Note: d negative if moving towards spacecraft

BoundaryNote: d negative if moving towards spacecraft

136

P d• Procedure– For a given , determine variance of data for all – Find minimum in variance, this determines (boundary direction)– Speed distance as function of time determines boundary speed

– intro_ascii,'remote_sense_A.txt',delta,rho,hh,mm,ss,nskip=13,format="(25x,f6.1,f8.1,3(1x,i2))"– ;– angle=fltarr(73)– chisqrd=fltarr(73)– for ijk=0,72 do begin

il fl t(ijk*5)– epsilon=float(ijk*5)– get_d_vs_dt,epsilon,hh,mm,ss,rho,delta,dist,times– yfit=dist & yfit(*)=0.– chi2=dist & chi2(*)=0.– coeffs=svdfit(times,dist,2,yfit=yfit,chisq=chi2)– angle(ijk)=epsilon– chisqrd(ijk)=chi2chisqrd(ijk) chi2– endfor– ipos=indgen(30)+43– chisqrd_min=min(chisqrd(ipos),imin)– plot,angle,chisqrd– print,angle(ipos(imin)),chisqrd(ipos(imin))– ;– stop

137

ZDP d

Y

D

BA

• Procedure– Note two minima (identical solutions)

• One for approaching boundary at V>0• One for receding boundary at V<0

– Convention that d<0 if boundary

1000

km V ~ 70km/sAConvention that d<0 if boundary

moves towards spacecraftallows us to pick one of the two(positive slope of d versus time)

= 280o= 280o

Var

ianc

e,

2V

aria

nce,

2

V

Boundary orientation,

V

Boundary orientation,

138

Probe: TH-BAngle to Y east=280degm

)

Probe: TH-BAngle to Y east=280degm

) Angle to Y_east 280degD0 = -2224 kmV0 = 69.9 km/stcross= 11:19:31.81

dista

nce

(km Angle to Y_east 280deg

D0 = -2224 kmV0 = 69.9 km/stcross= 11:19:31.81

dista

nce

(km

Bou

ndar

y d

Bou

ndar

y d

Time since 11:19:00Time since 11:19:00

tcross V [km/s] [deg]

D 11:19:27.6 75 270

B 11:19:31.8 70 280

A 11:19:38.4 80 275

Table 1. Results of remote sensing analysis on the inner probes

Timing of the arrivals of the other signatures at the inner three spacecraft

139

At the magnetopause200kmi,sheath (0.5keV,10nT)= ~200km

i,m-sphere (10keV,10nT)= ~1000km

Magnetopause Thickness ~ 6000kmCurrent layer Thickness 500kmCurrent layer Thickness ~ 500km

FTE scale, Normal 2 Boundary: ~6000kmAlong Boundary: Normal : 1 3 RAlong Boundary: ~Normal : 1-3 RE

For leaking magnetospheric particles, the currentcurrentlayer is sharp compared to the ion gyroradius andthe magnetic field is the same direction in the sheath and the magnetopause outside thesheath and the magnetopause outside the current layer. This means we can use the measured field outside themagnetopause to determine gyrocenters both at the magnetopause and the magnetosheath onthe magnetopause and the magnetosheath on either side of the hot magnetopause boundary.

140

YgseYgse

Magnetopause encounter on July 12, 2007

C

DTH-B AE

Ygse

C

DTH-B AE

Ygse

(a)(b)(a)(b)

XgseXgse

(c)(c)

XgseXgse

(d)(d)

(e)

(f)

(e)

(f)

Magnetic field angle is 60deg below spin plane and +120deg in azimuth i.e., anti-Sunward and roughly tangent to the magnetopause. The particle velocities, centered at 52deg above the

(g)(h)(g)(h)

spin plane, have roughly 90o pitch angles, with gyro-centers that were on the Earthward side of the spacecraft. The energy spectra of the NP particles show clearly the arrival of the FTE ahead of its magnetic signature, remotely sensing its arri al d e to the finite g roradi s (h)(h)sensing its arrival due to the finite gyroradius effect of the energetic particles. T=55s, i,100keV, 28nT) =1150km, V=40km/s

At the near-Earth magnetosphere

142

At the near-Earth magnetosphereRemote sensing of wavesgin ESA data, at the mostappropriate coordinateSystem, I.e, field alignedcoordinates

timespan,'7 11 07/10',2,/hours & sc='a'

coordinates. gyro=0o => Earthward particles

thm_load_state,probe=sc,/get_supp

thm_load_fit,probe=sc,data='fgs',coord='gsm',suff='_gsm'

thm_load_mom,probe=sc ; L2: onboard processed moms

thm_load_esa,probe=sc ; L2: gmoms, omni spectra

tplot 'tha fgs gsm tha pxxm pot tha pe?m densitytplot,'tha_fgs_gsm tha_pxxm_pot tha_pe?m_density tha_pe?r_en_eflux'

;

trange=['07-11-07/11:00','07-11-07/11:30']

thm_part_getspec, probe=['a'], trange=trange, angle='gyro', $$

pitch=[45,135], other_dim='mPhism', $

; /normalize, $

data_type=['peir'], regrid=[32,16]

tplot,'tha peir an eflux gyro tha fgs gsm tha pxxm pottplot, tha_peir_an_eflux_gyro tha_fgs_gsm tha_pxxm_pot tha_pe?m_density tha_pe?r_en_eflux'

Particle Measurements Distribution Particle Measurements, … · 2010. 1. 28. · THEMIS/ESA d...

Documents

Transcript of Particle Measurements Distribution Particle Measurements, … · 2010. 1. 28. · THEMIS/ESA d...