Variability Timescales inthe M87 Jet

D. E. Harris (SAO)C. C. Cheung (GSFC)

L. Stawarz (KIPAC)

INTRODUCTION, METHODS, & PROBLEMSINTRODUCTION, METHODS, & PROBLEMS THE TEV CONNECTION: SITE OF EXCESSTHE TEV CONNECTION: SITE OF EXCESS SIGNATURES OF ESIGNATURES OF E22 LOSSES LOSSES IMPULSIVE BRIGHTENING IN HST-1IMPULSIVE BRIGHTENING IN HST-1

Chandra RG workshop 2008 July

BACKGROUND

• Nucleus: varies between 0.2and 1.5 keV/s

• HST-1 at 0.86” (60pcprojected) is the site of giantflare: more than factor of 50!

• Knot D is affected by HST-1• Knot A shows ACIS

contamination

Chandra RG workshop 2008 July

BACKGROUND: RADIO

• From VLBA observations at 1.5 GHz, we findsuperluminal proper motions in HST-1 (~4c).

• For no acceleration, component was ‘emitted’(from upstream end) ~ 2002 (prior to peak ofmajor flare).

Chandra RG workshop 2008 July

Methods: Intensities for light curves:-pileup mitigation-

• Filter evt1 for GTI, but not grade (recover migration)• Multiply each event by its energy; sum from 0.2 - 17 keV• For the nucleus, subtract 5% of HST-1

Chandra RG workshop 2008 July

Problem: Pileup can produce a secondarypeak in PSF from release of trapped charge

at readout time

Chandra RG workshop 2008 July

Problem: Cross-talk from HST-1• The PA of the secondary

peak rotates during theobserving season.

• Before mid-March themajor effect is on thenucleus; after that date itaffects mainly knot D.

• For this reason, the bestdata for the nucleus isrestricted to times whenHST-1 < 4 keV/s

Chandra RG workshop 2008 July

Methods: First Derivative of LC• To study timescales, we

calculate the slope betweenadjacent (black) and everyother (red) observation.

• The fractional change per yearfor increasing intensities isfpy(+) = [(I2/I1) -1] / ΔT,and for decreasing intensitiesfpy(-) = [1 - (I1/I2)] / ΔT

• Thus a value of fpy=±1corresponds to a change by afactor of 2 in a year.

Chandra RG workshop 2008 July

The TeV connectionThe HESS group reported a

higher gamma ray flux in2005. They argued thatbecause of rapidvariability, the likely originwas the nucleus (close toSMBH) in spite of thesimilarity of their γ-raylight curve to the X-raylight curve of HST-1.

Chandra RG workshop 2008 July

Comparing variability timescales betweenthe nucleus and HST-1

• Both the 2005 HESS ‘event’and the 2008 Feb (MAGIC &VERITAS) TeV flaringdisplayed variability over a fewdays.

• The Chandra data indicate thatalthough the nucleus neverexperienced a giant flare likethat in HST-1, its ‘flickering” ischaracterized by larger valuesof fractional change per year.

Chandra RG workshop 2008 July

The First TeV Event2005 Apr (peak of HST-1

flare): Although there ispossible contamination ofthe nucleus by HST-1di/dt(nucleus) reached+50 (doubling time of < 7days). [During a shortcampaign of weeklyChandra observations]

Chandra RG workshop 2008 July

The Second TeV Event

2008 Feb: At the time ofthe TeV flaringobserved by MAGICand VERITAS thenucleus reached itshighest observed leveland di/dt = 11.8±0.5.[Normal monitoring;every 6 weeks]

Chandra RG workshop 2008 July

The TeV Connection: Summary

In favor of the nucleus:

• Easier to get source size ~ fewlight days

• Faster flickering in X-rays• Coincidence in 2008 Feb TeV

flaring with X-ray level and di/dt

In favor of HST-1:

• The observed (radio) size of HST-1 is of order 0.7 light years andthe inferred size from the X-rayvariability is of order 0.2 lightyears (∗δ).

• Coincidence of peaks in lightcurve (2005).

• With an SSC model, we expect toget IC scattering at TeVfrequencies.

• There is much less of a problemgetting TeV photons out of HST-1than out of the nucleus.

Chandra RG workshop 2008 July

Signatures of E2 Losses for HST-1

• If all frequencies decaytogether, expansionlosses dominate or thebeaming factor ischanging.

• If E2 losses (synchrotron& IC) are important, weexpect longer decaytimes for lowerfrequencies.

Chandra RG workshop 2008 July

The 2005 Decay• During the decay of the

giant flare, there is ashoulder in the UV lightcurve.

• fpy = -3.5±0.2 for X-raysfpy = -0.85 ± 1.16 for UV

• If the X-ray value were toarise from synchrotronlosses, B ≈ 1 mG for δ=4,the same field strengthfound by equipartitioncalculations.

Chandra RG workshop 2008 July

Comparative di/dt values: X-ray, UV, &Radio

• Granted the large error bars &poor sampling in the radio, themost negative values of fpyare -5 for X-rays; -2 for UV,and -1 for radio.

• We suspect that whileexpansion losses play asignificant role, E2 lossescontribute to the frequencydependence.

• These data are inconsistentwith a simple expansion for asimple power law.

Chandra RG workshop 2008 July

Discovery of impulsive brighteningin HST-1

Chandra RG workshop 2008 July

The River Analogy

• A jet is like a smoothly flowing river; observableemission arises only where there is white water.

• Even though HST-1 increased its emission bymore than a factor of 50, the power sodissipated is less than one percent of the totalpower believed to be flowing down the jet.

• A large rock may have been dropped into theriver; or perhaps there was a release of waterfrom an upstream dam………

Chandra RG workshop 2008 July

A Wager

Although I have not been able to identify acriterion for deciding the outcome, I wagerthe classical bottle of wine that:

The oscillations in brightening and fading ofHST-1 arise from a local instability ornaturally occurring characteristicfrequency rather than from a change in thepower flow of the entire jet.

[this is a personal wager, and does not involve my co-authors!]

FIN