TheUnexpected2012DraconidMeteorStorm - arXiv · arXiv:1311.1733v1 [astro-ph.EP] 7 Nov 2013 Mon....

arX

iv1

311

1733

v1 [

astr

o-ph

EP]

7 N

ov 2

013

Mon Not R Astron Soc 000 1ndash15 (2013) Printed 27 April 2018 (MN LATEX style file v22)

The Unexpected 2012 Draconid Meteor Storm

Quanzhi Ye1⋆ Paul A Wiegert1 Peter G Brown12

Margaret D Campbell-Brown1 and Robert J Weryk11Department of Physics and Astronomy The University of Western Ontario London Ontario N6A 3K7 Canada2Centre for Planetary Science amp Exploration The University of Western Ontario London Ontario N6A 5B8 Canada

Accepted 1970 January 1 Received 1970 January 1 in original form 1970 January 1

ABSTRACT

An unexpected intense outburst of the Draconid meteor shower was detected by theCanadian Meteor Orbit Radar (CMOR) on October 8 2012 The peak flux occurredat sim1640 UT on October 8 with a maximum of 24 plusmn 03 hrminus1

middot kmminus2 (appropriateto meteoroid mass larger than 10minus7 kg) equivalent to a ZHRmax asymp 9000plusmn 1000 using5-minute intervals using a mass distribution index of s = 188plusmn 001 as determinedfrom the amplitude distribution of underdense Draconid echoes This makes the out-burst among the strongest Draconid returns since 1946 and the highest flux showersince the 1966 Leonid meteor storm assuming a constant power-law distribution holdsfrom radar to visual meteoroid sizes The weighted mean geocentric radiant in the timeinterval of 15ndash19h UT Oct 8 2012 was αg = 2624 plusmn 01 δg = 557 plusmn 01 (epochJ20000) Visual observers also reported increased activity around the peak time butwith a much lower rate (ZHRsim 200) suggesting that the magnitude-cumulative num-ber relationship is not a simple power-law Ablation modeling of the observed meteorsas a population does not yield a unique solution for the grain size and distributionof Draconid meteoroids but is consistent with a typical Draconid meteoroid of mtotal

between 10minus6 to 10minus4 kg being composed of 10 ndash 100 grains Dynamical simulationsindicate that the outburst was caused by dust particles released during the 1966 per-ihelion passage of the parent comet 21PGiacobini-Zinner although there are dis-crepancies between the modelled and observed timing of the encounter presumablycaused by approaches of the comet to Jupiter during 1966ndash1972 Based on the resultsof our dynamical simulation we predict possible increased activity of the Draconidmeteor shower in 2018 2019 2021 and 2025

Key words meteors meteoroids comets individual 21PGiacobini-Zinner

1 INTRODUCTION

The October Draconid meteor shower (DRA or IAUMDC009 usually referred to as ldquoDraconidsrdquo some older literaturemay also refer to it as the ldquoGiacobinidsrdquo) is an annual meteorshower produced by comet 21PGiacobini-Zinner and activein early October It was first observed in the 1920s (Davidson1915 Denning 1926) and produced two spectacular meteorstorms in 1933 and 1946 (cf Jenniskens 1995) Aside fromthe two storms and a few outbursts it has been quiet inmost years with hourly rates no more than a few meteorsper hour

A review of the historic observations and studies ofthe Draconids may be found in our earlier work (Ye et al2013) in this introductory section we mainly address thestrong returns in 2011ndash2012 The outburst of the 2011 Dra-

⋆ E-mail qye22uwoca

conids was predicted by a number of researchers (Maslov2011 Vaubaillon et al 2011 to name a few) and analy-ses made by various observing teams indicate that the pre-dictions were quite accurate in both timing and meteorrates (such as Kero et al 2012 Koten et al 2012 Sato et al2012 Vaubaillon et al 2012 and many others)

In contrast no significant Draconid outburst was pre-dicted for 2012 Maslov (2011) suggested encounters withthe material ejected by Comet Giacobini-Zinner in 1959 and1966 (in the form of the 1959- and 1966-trails) at 15ndash17h UTon Oct 8 2012 but he noted that ldquovisually-detectable activ-ity is unlikelyrdquo predicting the maximum Zenith Hourly Rate(ZHR) to be 05 and 02 respectively for each encounter Theprediction by Maslov is summarized in Table 1

In fact an intense outburst of the Draconid meteorshower was subsequently detected by the Canadian MeteorOrbit Radar (CMOR) on Oct 8 2012 (Brown amp Ye 2012)As seen by CMOR the meteor rate started increasing at

ccopy 2013 RAS

2 Q-Z Ye et al

sim 15h UT quickly reached a maximum around 1640 UTand returned to the background rate around 19h UT Pre-liminary analysis indicated that the peak ZHR was wellabove the storm threshold (1 000 meteors per hour) Un-fortunately this was mostly a radar event in addition tothe poor timing of the peak (which only favored Siberiaand Central Asia in terms of darkness) visual and videoobservers were largely caught unprepared The quick lookanalysis carried out by the International Meteor Organiza-tion (IMO) revealed a peak of ZHR=324 plusmn 66 centered at1651 UT based on observations from only 4 observers1 Thecamera set up by the Petnica Meteor Group in Serbia alsodetected some activity at 17ndash18h UT2

Due to the lack of observations by other techniques theradar observations are essential for the study of this out-burst We will mostly follow the methodology used in ourearlier study of the 2011 event (Ye et al 2013) to analyze the2012 event as recorded by CMOR The goals of the presentpaper are to estimate the basic observational characteristicsof the outburst such as mean radiant mass distributionflux variation and to apply the Campbell-Brown amp Koschny(2004) meteoroid ablation model to try to constrain thephysical characteristics of the meteoroids We will also inves-tigate the cause of this outburst through numerical modelingof the stream

2 INSTRUMENTATION AND DATA

REDUCTION

The instrumental and data reduction details of this workis similar to that given in Ye et al (2013) as appliedto CMOR observations of the 2011 Draconids Tech-nical details of the entire system may also be foundin Hocking et al (2001) Jones et al (2005) Brown et al(2008) and Weryk amp Brown (2012) Here we only summarizea few key concepts

CMOR is a six-site interferometric radar array locatednear London Ontario Canada It operates at 1745 2985and 3815 MHz but only the 2985 MHz data is used inthis study The 2985 MHz component transmits with 12kW peak power with a pulse repetition frequency of 532 HzThe radar ldquoseesrdquo a meteor by detecting the reflection of theoutgoing beam from an ionized meteor trail therefore radarmeteors are detected 90 away from their radiants (such areflection condition is called specular) As such when the ra-diant is close to the zenith it becomes difficult for the radarto detect echoes from a particular stream as the echoes areat low elevation and large ranges when the radiant is close tothe zenith Detected echoes are processed in an automaticmanner to eliminate questionable detections correlate thesame events observed at different sites and calculate meteortrajectories Usually raw streamed data is not permanentlykept but we intentionally saved the raw data from the 2012Draconid outburst for further analysis

Meteor echoes seen by radar may be classified as ei-ther underdense or overdense echoes based on their ap-parent amplitude-time characteristics The deciding factors

1 httpwwwimonetlivedraconids2012 2 httpwwwmeteorirswordpressi-ove-godine-pojacana-aktivnost-drakonida

underdense

Am

plitu

de (r

aw u

nit)

overdense

Time (raw unit)

twisted

Figure 1 Examples of meteor echoes as seen by CMOR Top un-derdense trail middle overdense trail bottom overdense wind-twisted trail

are many (see Ye et al 2013 for a more theoretical introduc-tion) but for a given shower overdense echoes are typicallyassociated with larger meteoroids than underdense echoesThe underdense echoes are always characterized by a rapidrise and an exponential decay in amplitude (Figure 1a) mak-ing them relatively easy to identify automatically Ideallythe overdense echoes are characterized by a rapid rise aplateau and a steady decay in amplitude (Figure 1b) but inreality trails distorted by upper atmospheric winds and ex-hibiting multiple reflection points are often seen (Figure 1c)making them difficult to characterize with automatic algo-rithms

Noting the advantages and weakness of the automaticroutines we take two approaches for the initial data reduc-tion

(i) For the first approach we try to separate Draconidechoes from other echoes in the automatically processeddata This is done in several steps First we select all me-teor radiants measured with time-of-flight (tof) velocitieswithin 10 of the apparent common radiant (ie an accep-tance radius of 10 around the apparent center of the clus-ter of meteor radiants) and 20 of the annual geocentricvelocity vg = 204 km middot sminus1 (Jenniskens 2006) The veloc-ity restriction is intentionally broad reflecting the spreaddue to measurement error and the high decelerations ofDraconid meteoroids a consequence of their fragility (egJacchia et al 1950 Campbell-Brown et al 2006) All these

ccopy 2013 RAS MNRAS 000 1ndash15

The Unexpected 2012 Draconid Meteor Storm 3

Table 1 The Draconid 2012 outburst prediction by Maslov (2011)

Trail Time αg δg vg ZHR(Oct 8 UT) (J2000) (km middot sminus1)

1959 1622 1654 2627 +558 210 051966 1537 - - 02

echoes are manually inspected to remove data with prob-lematic time picks (ie an improperly automatically chosentime of occurrence of the specular point) determined by theautomatic algorithms and to ensure the trajectory solutionis correct We do not revise the time picks as we need tokeep the procedure repeatable by automatic algorithms inorder to estimate the uncertainty The uncertainties are es-timated with a Monto Carlo synthetic echo simulator de-veloped by Weryk amp Brown (2012) allowing the derivationof a weighted mean common radiant We then explore theideal radiant size for this outburst event by varying the ac-ceptance radiant interval to search for the acceptance radiusat which the number of radiants reaches an asymptote ap-proaching the background level which in this case is sim 5

(Figure 2) therefore we reduce the acceptance radius from10 to 5 We identify 576 Draconid echoes in this mannerand refer to this dataset as the complete dataset (Table 2)

(ii) For the second approach we try to separate all over-dense Draconid echoes As noted before overdense echoesare sometimes difficult to identify automatically thereforewe separate these echoes by manually inspecting the rawdata from the main site between 15ndash19h UT on Oct 82012 Trajectories of these meteor echoes are manually de-termined when remote site observations are available to de-termine whether they are Draconids In all some 240 over-dense Draconids are identified this way and will be referredas the overdense dataset

Since the velocities in the complete dataset are deter-mined using the tof method which depends on trail geom-etry and is uncertain in part due to height averaging wealso measure the speed of the meteoroids using the Fres-nel phase-time method and Fresnel amplitude-time method(simplified as ldquopre-t0rdquo and ldquoFresnelrdquo method hereafter seeCeplecha et al 1998 sect461 and sect462 for an overview) Theadvantage of these two methods is that they only depend onthe observation from the main site which does not introducethe geometry or height averaging effects of the tof tech-nique In total we gathered 360 echoes with pre-t0 speedsand 25 echoes with Fresnel speeds where 13 echoes haveboth pre-t0 and Fresnel speeds Numerically the two speedsare equivalent as the deceleration of the meteoroid is mini-mal during its passage over the Fresnel zones (amounting tosim 1 minus 2 km of trail length Figure 3) we therefore assumethe pre-t0 or Fresnel speed is the meteoroidrsquos speed at itsspecular height For the 13 echoes where both pre-t0 andFresnel velocities are measured preference is given to theone with smaller uncertainty

The raw number of echoes as a function of time on Oct8 2012 in each dataset is given in Figure 4

2 4 6 8 10

0

10

20

30

40

50

60

70

Rad

iant

den

sity

per

incr

emen

t are

a (d

eg-2

)

Radiant size

Oct 8 Average of Oct 7 and 9

Figure 2 Radiant densities centered at λminusλ⊙ = 53 β = +79

in sun-centered ecliptic coordinates as a function of radiant sizeon the outburst date Oct 8 2012 Shown for comparison is theaverage of the same ecliptic sun-centered radiant location on Oct7 and Oct 9 2012 when no significant Draconid activity waspresent It can be seen that the radiant size on Oct 8 and thebackground rate meets at sim 5

3 OBSERVATIONAL RESULTS

31 Radiant and Orbit

The weighted mean radiant of all meteors in the thecomplete dataset is αg = 2623 plusmn 01 δg = +557 plusmn 01

(J2000 epoch) (Figure 5) The weighted mean radiant ofmeteors detected within 15ndash19h UT is αg = 2624 plusmn 01δg = +557 plusmn 01 (J2000 epoch)

The derivation of true orbit requires additional effort indata reduction as the fragility of Draconid meteoroids intro-duces an underestimation of the deceleration correction asnoted by Campbell-Brown et al (2006) and Ye et al (2013)that leads to a reduction in estimating the true out of at-mosphere speed of the meteoroids To solve this problemwe make use of the echoes detected at higher heights asthey do not penetrate very deep into the atmosphere andtherefore suffer less deceleration We also need to select theechoes with pre-t0 or Fresnel velocity measured to minimizethe error caused by deceleration undercorrection We onlyconsider the echoes detected within 15ndash19h UT (ie close tothe peak time) to minimize the possible contamination ofbackground Draconid meteors not associated with the out-burst We then choose 98 km as the cut-off height as it isthe highest possible cut-off with a statistically significantsample (only two echoes are detected at gt 99 km) Tak-ing all these constraints into account a total of 14 echoesare selected from the complete dataset We note that a sig-


4 Q-Z Ye et al

Table 2 Characteristics of the 2012 Draconid radiant as observed by CMOR The mean radiants are calculated using error weightingsaccording to the Monte Carlo error simulations by the synthetic echo simulator (see sect21 of this paper as well as Weryk amp Brown 2012Ye et al 2013)

Time period αg δg Nradiants

(J2000)

Oct 8 (all day) 26232 plusmn 010 +5567 plusmn 005 576Oct 8 (15ndash19h UT) 26235 plusmn 010 +5567 plusmn 005 545

20 21 22 23 24 2520

21

22

23

24

25

Fres

nel v

eloc

ity (k

m s

-1)

Pre-t0 velocity (km s-1)

Fitting slope=1008+-0009

Figure 3 The Fresnel velocities versus pre-t0 velocities for the13 high altitude echoes having both techniques for velocities mea-sureable

1500 1600 1700 1800 19000

10

20

30

40

50

60

Raw

num

ber o

f ech

oes

Time Oct 8 (UT)

Complete Overdense

Figure 4 The raw number of Draconid echoes detected byCMOR in the complete and the overdense dataset binned in10-minute intervals The background level is normally ≪ 1 Dra-conid per 10-minute bin

nificant fraction (514) of these have local zenith angle ηabove 50 which suggests that the undercorrection effect isagain significant due to the longer atmospheric path of themeteoroid entering the atmosphere at such a shallow angle

270o 265o 260o 255o

50o

55o

60o

Geo

cent

ric D

ecl

(J20

00)

Geocentric RA (J2000)

Median error

Figure 5 Distribution of individual Draconid radiant of thecomplete dataset

Hence we remove these five echoes and create an additionalsubset and present both results together (Table 3)

We note that the mean orbit derived from the nine-meteor sample is consistent with the orbit of 21PGiacobini-Zinner at its 1959 and 1966 apparitions Additionally the vgderived from the nine-meteor sample agrees with that de-rived from the multi-station video observations of the 2011Draconid outburst (vg = 209 plusmn 10 km middot sminus1 as noted byJenniskens et al 2011) within uncertainty This agreementis reassuring as the video observations have more completecoverage of the trail length of each meteoroid compared toradar allowing the deceleration correction for individual me-teoroids to be established with higher confidence along withthe pre-atmosphere velocity (vg) and orbit We believe thatthe orbit derived from the nine-meteor sample represents thebest mean orbit for the 2012 Draconid outburst as observedby CMOR

32 Mass Index

The mass index can be determined using either underdenseechoes or overdense echoes For underdense echoes the cu-mulative number of echoes with amplitude greater than Afollows the relation N prop Aminuss+1 (McIntosh 1968) while for

diffusion-limited overdense echoes N prop τminus 3

4(sminus1) (McIntosh

1968) where τ is the echo durationWe first use the underdense echoes detected between

15ndash19h UT in the complete dataset to determine s Echoeswithin a range interval between 110ndash130 km from the main



Landscape table to go here

Table 3


6 Q-Z Ye et al

1000 10000

10

100

Cum

ulat

ive

echo

num

ber

Amplitude (raw unit)

s=188 001

Figure 6 Cumulative echo numbers as a function of amplitudefor Draconid echoes with ranges between 110ndash130km The massindex is determined to be 188plusmn 001 by fitting the linear portionof the curve as shown

site are selected to avoid contamination from overdense-transition echoes (see Blaauw et al 2011 for discussion)We determine the mass index to be 188 plusmn 001 by fittingthe linear portion of the data in Figure 6 The uncertaintygiven here is the fitting uncertainty only and could be sev-eral times smaller than the real one due to the small samplesize

We then use the overdense dataset as a check on themass index value determined using the amplitude distribu-tion alone (Figure 7) The trail of electrons can fall to anundetectable density either through diffusion or chemical re-combination The latter is more likely for very dense trailsproduced by larger meteoroids at low altitudes The possibleturnover between diffusion- and chemistry-limited regime orthe ldquocharacteristic timerdquo can be seen at about 25 s in Fig-ure 7 which agrees with the value derived for 2011 (27sYe et al 2013) Using the few data points in the possiblediffusion-limited regime the mass index can be estimatedto be around 17ndash18 However we should note that this re-sult is doubtful due to the presence of a sudden steep risein cumulative number at τ ltsim 2 s The possibility of un-derdense contamination can be ruled out by examining thetheoretical underdense region in the height-duration distri-bution (Figure 8) A possible explanation of this behavioris the lack of long overdense echoes due to the radiant ge-ometry at the time of the outburst as well as abundance ofsmaller meteoroids of this event preventing us from gettingenough statistics for longer duration echoes Alternativelythe lack of a power law fit at the high mass end of the mete-oroid distribution (ie overdense echoes) may indicate thatthe size distribution within the stream does not follow apower-law in contrast to the behavior seen in 2011 where aclear fit existed to the smallest overdense echoes durationsIn this case the upper upturn below τ ltsim 2 indicates anoverabundance of smaller Draconids in the outburst

1 10

10

100

Cum

ulat

ive

num

ber

Echo duration (s)

tc~25 s

diffusion-limited regime ()

chemistry-limited regime ()

Figure 7 Cumulative echo numbers as a function of echo du-ration for overdense Draconid echoes The possible characteristictime can be seen near tc = 25 s

Hei

ght (

km)

Echo duration (s)

overdense

underdense

Figure 8 The height range of the selected overdense echoes inthe overdense dataset The shaded area marks the underdenseregion defined by McKinley (1961) If this population were mainlyunderdense a clear duration vs height trend would be presentedin the graph

33 Flux

The flux can be calculated from the number of echoes de-tected per unit time divided by the effective collecting areaof the radar for the Draconid radiant The effective collectingarea of CMOR is calculated following the scheme describedin Brown amp Jones (1995) Simply put the collecting areafor a given radiant is the integration of the magnitude ofthe gain over the ldquoreflecting striprdquo of the radar wave (a 90

great circle perpendicular to the radiant direction) whichtakes into account both the radiant geometry and mass in-dex The ZHR (ie the number of meteors that an averageobserver would see in one hour given that the sky is clearand dark and the radiant is at the zenith) can then be cal-culated by using its relationship to the actual meteoroid flux



(Koschack amp Rendtel 1990) as computed by the radar andmass index using the fact that the limiting meteor radiomagnitude of CMOR is sim 85

As shown in Figure 9 the main peak based on 5-minaveraged CMOR data occurred at 1638 UT (solar lon-gitude λ⊙ = 195622) Oct 8 with maximum flux of24plusmn 03 hrminus1 middot kmminus2 (appropriate to meteoroid mass largerthan 10minus7 kg) equaling ZHRmax = 9000 plusmn 1000 The IMOdata shows several ldquopeak-letsrdquo from 1625ndash1655 UT withZHRmax around 500ndash600 The existence of these ldquopeak-letsrdquois likely a statistical oddity due to the limited number of ob-servers (only one observer for each of these ldquopeak-letsrdquo) Forexample the time range around 1640 UT (which containedthree observers reporting 10 Draconids total observed) cor-responds to a ZHR of 185plusmn 55

We note that these ZHRs assumed a single power-law fitfrom fainter radar meteors to equivalent radio magnitude of+65 where the ZHR is defined As noted earlier an apparentdeviation is notable between larger visual meteoroids andsmaller radar meteoroids from a pure power-law hence theeffective ZHR quoted from CMOR observations are likelyupper limits

34 Ablation Modeling

To model the structure of the Draconid meteoroidswe used the meteoroid ablation model developed byCampbell-Brown amp Koschny (2004) The velocities and elec-tron line densities as measured by CMOR are used as con-straints to fit the model The velocities used here are ap-propriate to the echo specular height (ie pre-t0 or Fresnelvelocities) the electron line densities are computed from theobserved amplitude value using the formula appropriate toeither underdense echoes (when q lt 24times1014 mminus1) or over-dense echoes (when q gt 24times 1014 mminus1) The methodologyis essentially the same as for the 2011 outburst (Ye et al2013) except that we do not model individual events for de-celeration as no such events are found for the 2012 outburstWe use the parameters suggested by two previous stud-ies of the Draconids Borovicka et al (2007) and Ye et al(2013) while other parameters are left fixed as suggestedin Campbell-Brown amp Koschny (2004) The parameters aresummarized in Table 4

We first tried to find the optimal combination of grainmass and grain number by fixing the total mass at 10minus5 kgand varying these two variables We choose 10minus5 kg here asthis value is a reasonable compromise between the detectionlimit (sim 10minus7 kg) and the underdenseoverdense transitionlevel (sim 10minus4 kg appropriate to meteors with q asymp 1014 mminus1

or peak visual brightness of +4 magnitude) We used thetransition level as our upper limit as most of the echoesin the complete dataset are underdense echoes Finally welook for a modeling fit that minimizes the deviation betweenthe model and trend of the data points with respect to theheight However from Figure 10 we cannot identify a uniquefit the range of optimal fit seems to lie somewhere betweenngrain = 10 and ngrain = 10 000

We then plot the modeling fit for mtotal = 10minus6 to10minus4 kg for ngrain = 10 to 10 000 to further examine thegoodness of each parameter set (Figure 11) Again we donot see an obviously unique solution but it seems that thengrain = 10 and ngrain = 100 scenarios produce a better

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

75

80

85

90

95

100

105

1012 1013 1014 101575

80

85

90

95

100

105Velocity (km s-1)

Hei

ght (

km)

ngrain=1 ngrain=10 ngrain=100 ngrain=1 000 ngrain=10 000 ngrain=100 000

Electron line density (m-1)

Hei

ght (

km)

Figure 10 Sensitivity test for different number of grains ngrainat a fixed total mass m = 10minus5 kg

qualitative fit to both velocities and electron line densitiesThis generally agrees with our previous finding for the 2011outburst which find the optimal ngrain = 100 it also agreeswith the photographic result for the 2005 outburst to someextent (Borovicka et al 2007) but we note that the fits arenot well constrained from our observations therefore theycan only be used to broadly establish the fact that the 2012Draconid meteoroids were not radically different in physicalmakeup from those detected during the 2011 outburst

4 DYNAMIC MODELING

To better understand the 2012 Draconids outburst nu-merical simulations were performed of the parent comet21PGiacobini-Zinner The simulations included a SolarSystem of eight planets whose initial conditions were derivedfrom the JPL DE405 ephemeris (Standish 1998) with theEarth-Moon represented by a single particle at the Earth-Moon barycenter The system of planets parent and mete-oroids were integrated with the RADAU method (Everhart1985) with a time step of seven days used in all cases

The comet orbital elements used in these simula-tions are derived from Marsden amp Williams (2008) whereorbital elements are provided for each appearance of21PGiacobini-Zinner from its discovery in 1900 through2005 This extensive data set is important as the parent


8 Q-Z Ye et al

100

1000

10000 IMO visual

1954o 1955o 1956o 1957o 1958o

100

1000

10000 ZHR

Zeni

th H

ourly

Rat

e(ZHR

hr-1

)

CMOR

Flux (hr -1 km-2)

12 13 14 15 16 17 18 19 20Time Oct 8 (UT)

001

01

1

1954o 1955o 1956o 1957o 1958o0

100

200

300

Solar Longitude (J20000)

Col

lect

ing

area

(km

2 )R

adiant elevation0o

20o

40o

60o

80o

Figure 9 The ZHRflux profile of the 2012 Draconid outburst as observed by CMOR (binned in 5-minute intervals error bars representPoisson errors only) comparing to the visual profile reported by the visual observers to the International Meteor Organization (IMO)The radar flux is appropriate for meteoroids with m gt 10minus7 kg the visual flux is appropriate for meteors with V lt 2minus 3 mag

Table 4 Input parameter for the ablation model as used in Ye et al (2013)

Parameter Value Unit

Deceleration corrected apparent velocity vinfin 2327 km middot sminus1

Zenith angle η 55 degBulk density ρbulk 300 kg middotmminus3

Grain density ρgrain 3 000 kg middotmminus3

Heat of ablation qheat 3times 106 J middot kgminus1

Thermal conductivity κ 02 J middotmminus1 middot sminus1 middotKminus1

being a Jupiter-family comet is extensively perturbed andhas time-varying non-gravitational parameters

An initial survey of the meteoroid complex used the cat-alog orbit for 21PGiacobini-Zinnerrsquos 2005 perihelion pas-sage as a starting point Simulations of meteoroids releasedduring each of 21PGiacobini-Zinnerrsquos perihelion passagesback to 1862 were examined for clues as to the origin ofthe material that produced the 2012 outburst These simu-lations revealed that the 1966 apparition was the source ofthe 2012 event and as a result the orbital elements given for

the 1966 perihelion passage were used in subsequent simu-lations These orbital elements are listed in Table 5

No non-gravitational forces due to outgassing (egMarsden et al 1973) were applied to the comet in ourfinal simulation results Some test simulations that didinclude non-gravitational forces using the values for21PGiacobini-Zinner from Yeomans (1986) as listed inMarsden amp Williams (2008) were run but these did not pro-duce any noticeable differences in the results

The simulations were run with the two-stage refinement



14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

1012 1013 1014 1015

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

1012 1013 1014 1015

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

1012 1013 1014 1015

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

1012 1013 1014 1015

Hei

ght (

km)

mtotal=10-6 kg

mtotal=10-5 kg

mtotal=10-4 kg

ngrain=10

Hei

ght (

km)

ngrain=102

Hei

ght (

km)

ngrain=103

Hei

ght (

km)

Velocity (km s-1)

ngrain=104


Figure 11 Sensitivity test for different number of grains ngrain but for varying mgrain (ranges from 10minus10 to 10minus5 kg) The corre-sponding mtotal range is from 10minus6 to 10minus4 kg


10 Q-Z Ye et al

Name Perihelion q (AU) e a (AU) ω Ω i

21P 2005 Jul 27605 1037914 0705691 3526613 1725429 1954301 318109

21P 1966 Mar 282844 0933501 0729394 3449669 1729153 1966674 309382

sim 09968 plusmn 00003 0725 plusmn 0003 362 plusmn 004 17275 plusmn 009 19565 plusmn 004 3179 plusmn 002

obs 09952 plusmn 00003 0710plusmn 0032 322plusmn 061 17226 plusmn 035 195620 plusmn 0001 3145 plusmn 069

Table 5 A list of the orbital elements used for comet 21PGiacobini-Zinner for dynamic modeling of the Draconid meteoroid streamformation From Marsden amp Williams (2008) CMOR observations are extracted from the nine-meteor solution in Table 3

procedure described in Wiegert et al (2013) Briefly in thefirst stage the comet orbit is integrated backwards to thedesired starting point in this case back approximately 150years to 1862 The comet is then integrated forward againreleasing meteoroids at each perihelion passage as it does soAll meteoroids which pass sufficiently close to the Earthrsquosorbit during the simulation are collected this is our list ofldquobulls-eyesrdquo Bulls-eyes are those that have a minimum or-bital intersection distance (MOID) of no more than 002 AUand that are at their MOID with the Earth no more thanplusmn7 days from when our planet is there

In the second stage the integration of the comet for-ward in time is repeated However in this case instead ofreleasing particles randomly as determined by a cometaryejection model particles are released only near initial con-ditions in which a bulls-eye was produced in the same timestep in the first simulation These second generation parti-cles are released at the same position (that of the nucleustaken to be a point particle) but are given a random changein each velocity component of up to plusmn10 of that of theoriginal bulls-eye These second generation meteoroids in-evitably contain far more particles that reach our planet Ofthis sample those which pass closest to the Earth in spaceand time will be considered to constitute the simulation out-burst

In our initial simulations at each perihelion passage anumberM = 104minus105 particles is released in each of the foursize ranges with radii r from 10minus5minus10minus4 m 10minus4minus10minus3 m10minus3 minus 10minus2 m and 10minus2 minus 10minus1 m the whole range span-ning from 10 microm to 10 cm in diameter They are chosenso that the distribution of particle radii is flat when binnedlogarithmically by size A power-law size distribution withmass index s can be recovered after the simulation is com-plete by giving each particle a weight proportional to rminus3s+3

(Wiegert et al 2009) if desired However we do not applysuch a weight to the results reported here because the out-burst is found to consist of such a narrow range of particles(radii almost exclusively from 100 microm to 1 mm) that theapplication of such a weighting is likely inappropriate

Post-Newtonian general relativistic corrections and ra-diative (ie Poynting-Robertson) effects are also includedThe ratio of solar radiation pressure to gravitational force βis related to the particle radius r (in microm) through β = 19rfollowing (Weidenschilling amp Jackson 1993) though our ex-pression assumes a particle mass density ρ = 300 kg middotmminus3

for Draconid meteoroids as was reported by Borovicka et al(2007)

The comet is considered active (that is simulated me-teoroids are released) when at a heliocentric distance of3 AU or less during the first simulation stage the simu-lated parent releases particles at a uniform rate during this

part of its orbit While active particles are released withvelocities from the prescription of either Crifo amp Rodionov(1997) or the revised Whipple model of Jones (1995) Weuse a nucleus radius for 21PGiacobini-Zinner of 1000 m(Tancredi et al 2000) but in the absence of other detailswe assume a Bond albedo for the nucleus of 005 a nucleusdensity of 300 kg middotmminus3 and an active fraction of the cometrsquossurface of 20 where needed in the above ejection modelsThe Brown amp Jones (1995) model was found to reproducethe duration of the outburst as observed by CMOR slightlybetter and the results reported here use the Jones model

A supplementary integration of the comet orbitbackwards for a thousand years allowed a determina-tion of the Lyapunov exponent using the algorithm ofMikkola amp Innanen (1999) The e-folding timescale is ap-proximately 30 years not unexpected for a Jupiter-familycomet with a node near that giant planet Our primary re-sult that the outburst originated from the 1966 perihelionpassage of the parent comet is thus two e-folding times intothe past and thus not unduely affected by chaotic effectsThe short Lyapunov time is a result of the frequent closeencounters that 21P suffers with Jupiter The one of mostrelevance here is the only close encounter occuring in the1966-2012 time frame a close approach to within less than16 Hill radii in 1969 This encounter strongly affects boththe parent and the meteoroids released during the 1966 per-ihelion passage

5 COMPARISON OF SIMULATIONS WITH

OBSERVATIONS

The simulation output for the year 2012 is presented in Fig-ure 12 which shows the nodal intersection points of the sim-ulated meteoroids overplotted on the Earthrsquos orbit The dotsshown are those test particles within 002 AU of Earthrsquos or-bit and reaching that point within plusmn7 days of our planetbeing at the same solar longitude These are colour-codedby the perihelion of the parent that released them The blackpoints represent those which are closest to intersecting theEarth within ∆r = plusmn0002 AU and ∆t = plusmn1 day thesewill be taken to constitute the simulated outburst

Only one perihelion passage has a significant popula-tion of particles that meet the outburst criteria 1966 Thelocation of the peak corresponds to the time that CMORdetected the outburst though the nodal footprint is slightlyoff the Earthrsquos orbit this will be discussed further below



0950 0960 0970 0980025

00

260

027

00

280

x (AU)

y (A

U)

2012minus10minus08 0 UT




radar peak

1966

181718521933194019591966othersclosest

Figure 12 The nodal footprint of the Draconid meteor stream for 2012 from simulation The most populous perihelion passages areindicated by specific colors all others are in grey Black is reserved for those particles which meet the most stringent encounter criteriaand are deemed to comprise the simulated shower Only the 1966 perihelion passage has a substantial number of such particles Thetiming of the observed CMOR radar peak is indicated by a dotted line See text for more details

51 Why was there no prediction of an outburst

The largest difference between the simulations and observa-tions is the fact that the simulated meteors approach but donot quite intersect the Earthrsquos orbit This likely contributedto the absence of theoretical predictions of the outburstMaslov3 did predict activity for the 2012 Draconids at this

3 httpferajnarodruRadiantsPredictions1901-2100engDraconids1901-2100predenghtml

time in particular in the radio meteor range which he at-tributed to the 1959 and 1966 trails which our simulationsconfirm

In order to investigate the question of why the simu-lated nodal footprint of the 1966 streamlet does not inter-sect the Earthrsquos orbit we performed additional simulationswith higher ejection velocities to determine the conditionsrequired for these meteoroids to reach the Earth Additionalejection velocities around 50ndash100 m middot sminus1 are needed to movethe simulated 1966 stream to Earth orbit crossing Parti-


12 Q-Z Ye et al

cles in our model are ejected with a range of velocities Vej

which depends on their size and heliocentric distance Parti-cles from 10minus5 to 10minus4 m have Vej = 210plusmn 80 ms (averageplusmn one standard deviation) 10minus4 to 10minus3 m Vej = 66 plusmn 25ms 10minus3 to 10minus2 m Vej = 21plusmn 8 ms and 10minus2 to 10minus1 m= 66plusmn25 ms The three largest size ranges can be broughtto Earth intersection by increasing the ejection velocity to50 to 100 ms while 30 ms proves just insufficient Thusthe discrepancy in the nodal position (which amounts toroughly 0001 AU or about half the distance to the Moon)could be accounted for by relatively high ejection velocities(as Maslov also noted) However we note that the close en-counter with Jupiter suffered by 21PGiacobini-Zinner andits daughter particles in 1969 as well as uncertainties inthe comet orbit itself (ie its non-gravitational parameterschanged markedly in the 1966ndash1972 time frame Yeomans(1986)) make it difficult to pinpoint a single cause We con-jecture that inevitable errors in the Jupiter family cometrsquosorbit coupled with its frequent approaches to Jupiter pro-duce uncertainties in the location of the streamrsquos node largeenough to confound attempts to model the shower accu-rately

Nonetheless we can conclude with some confidence thatthe outburst observed by CMOR was produced by the 1966trailet This is because a comparison of the properties ofthese particles are consistent to a high degree with the Dra-conid outburst observed by CMOR This includes the timingof the outburst its radiant velocity and the overall orbitalelements which we present below

52 Timing and radiants

Figure 13 shows the timing of the shower The radar andsimulation results are offset slightly with the simulationsshowing activity peaking about 003 of solar longitude(about 45 minutes) earlier than was observed though theprecise width and peak of the simulated shower is somewhatsensitive to our choice of ∆r and ∆t

The radiants are shown in Figure 14 The locationsof the CMOR radiant is indicated by the error bars Theellipse indicates one standard deviation of the simulatedradiant The observed and simulated radiants thus haveαg and δg which are consistent to within one sigma Thegeocentric velocity (far from Earth) of simulated meteors2093plusmn 003 km middot sminus1 These agree with the radar-measuredvelocities within their respective uncertainties

53 The ldquoabsentrdquo storm for visual observers

The size distribution of the outburst in the simulation is con-centrated in a narrow range of meteoroid sizes from about01 to 07 mm corresponding to a typical β asymp 5times10minus3 Thissize selection arises essentially from the timing constraintParticles of different sizes tend to have different arrival timesat Earth even if released from the parent at the same timeand with the same ejection velocity because a particlersquos or-bital period is a function of β The need to arrive at thecorrect time to produce the outburst thus will restrict β toa narrow range This size distribution is qualitatively con-sistent with CMOR observations Because of the strong sizeselection for this outburst the concept of a power-law dis-tribution of sizes must be used with great care in this case

Solar longitude (deg J2000)

Cou

nts

1954 1955 1956 1957 1958 1959

050

010

0015

00

Figure 13 The timing of the Draconid outburst of 2012 Theblack histogram shows the simulation the grey band indicatesthe times where CMOR saw significant shower activity and thewhite line indicates the radar peak

2626 2624 2622 2620 2618

552

554

556

558

560

562

Right Ascension (deg J2000)

Dec

linat

ion

(deg

J20

00)

Figure 14 The simulated radiant The grey points indicate thoseparticles passing within plusmn002 AU and plusmn7 days of Earth Theblack points indicate those meeting the most stringent criteria(∆r = 0002 AU ∆t = plusmn1 day) and are deemed most likely tohave comprised the material producing the outburst The greenerror bars indicated the CMOR radiant the red ellipse is the onestandard deviation position of the simulated radiant



5eminus05 5eminus04 5eminus03 5eminus02

510

2050

100

200

500

β

ejec

tion

velo

city

(m

s)

Figure 15 The phase space of β and ejection speed sampled bythe simulation (grey) and those that are part of the simulatedshower (black) demonstrating the reason for dynamical filteringof the outburst to a narrow range of meteoroid sizes

The orbital elements for the core of the simulated out-burst is listed in Table 5 Given the match in the timingand width of the outburst as well as consistency betweenthe radar derived sizes and radiants and those of the simu-lations we conclude that the 2012 Draconids outburst con-sisted primarily of particles released during the 1966 perihe-lion passage of comet 21PGiacobini-Zinner

The meteoroids which reach the Earth from the 1966perihelion passage show no concentration of release times inthe pre- or post-perihelion orbital arc The β and ejectionvelocities in the simulation are shown in Figure 15 only anarrow segment of the particles ejected from the comet reachthe Earth during the 2012 outburst

The extensive compilation of historical comet observa-tions of Kronk (1999) was examined to see if the 1966 per-ihelion passage of 21PGiacobini-Zinner displayed any be-haviour that might explain the outburst The apparition wasa faint one and poorly observed but there are no indicationsof an outburst or other unusual dust production during its1966 apparition

54 Future outlook

We extended our simulations to determine whether any fur-ther outbursts of the Draconids are likely between now and2025 This process is fraught with difficulties since we havealready concluded that our knowledge of the parentrsquos orbit isinsufficient to model even the observed 2012 outburst in fulldetail However the desire to better understand this peculiarshower leads us to the attempt nonetheless

Our earlier discussion of the 2012 outburst revealed in-sufficient accuracy either in our knowledge of the cometrsquos dy-namical parameters (ie orbital elements non-gravitational

parameters) or in our ability to simulate the cometrsquos evo-lution To attempt to get around these restrictions we runsimulations based on each of the sets of osculating orbital el-ements listed in Marsden amp Williams (2008) for all fourteenappearances of 21PGiacobini-Zinner from its discovery in1900 through 2005 inclusive Each is integrated backwardsfor 100 years and then forwards again producing dust ateach perihelion passage until the year 2025 The initial sim-ulations and refinement stages are performed in all of thesecases as described earlier In the absence of observation un-certainty and the chaos produced by close encounters withJupiter these simulations would all produce the same re-sults Most ndash but not all ndash simulations produced similar pre-dictions However observational uncertainty and chaos arepresent and thus each of the fourteen orbits is only an ap-proximation to the true orbit at some instant in time andwe find that shower predictions are very sensitive to suchsmall variations in the cometrsquos orbit The results reportedhere for future activity represent a somewhat subjective se-lection of those years where a majority of the simulationsproduce appreciable activity

To better understand the situation we first comparedthe simulations with known outbursts of the Draconids(2005 2011 2012) in the 2001-2012 range Here we wereheartened to able to reproduce the occurrence and absenceof outbursts as observed with one caveat that the simulatednodal footprint of the 2005 outburst was displaced slightlyoutside the Earthrsquos orbit similarly to that of 2012 (The 2011outburst footprint did not require adjustment)

Given these results we provide here a list of years whenthe simulations show the possibility of increased Draconidsactivity where many (though some times not all) the nodalfoot prints are on or near the Earthrsquos orbit These years are2018 2019 2021 and 2025 The strongest activity is expectedfrom the year 2018 but its nodal footprint is offset slightlyfrom the Earth a precise activity level is hard to calculatebut an outburst similar to 2012 is possible The other showeryears are not offset from the Earthrsquos orbit but show lowerrates and smaller particle sizes (Table 6)

6 CONCLUSION

We have reported the analysis of the unexpected intense out-burst of the Draconid meteor shower detected by CMOR onOctober 8 2012 The peak occurred at sim1640 UT (λ⊙ =195622) with a ZHRmax asymp 9000 plusmn 1000 The weightedmean radiant around the peak time (15ndash19h UT) was atαg = 2624 plusmn 01 δg = 557 plusmn 01 (J2000) The massdistribution index was determined to be s = 188 plusmn 001lower than the value found in 2011 indicating that the out-burst was dominated by faint meteors compared to 2011We then used the meteoroid ablation model developed byCampbell-Brown amp Koschny (2004) to model the structureof the Draconid meteoroids which does not reveal uniquesolutions but suggests that the number of grains in eachmeteoroid falls within the range of 10ndash100

Visual observations also showed increased activityaround the peak time but with a much lower activity thanindicated by radar (ZHRsim 200) We intrpret this to indicatea strong size selection in the outburst centred near radarmeteoroid sizes a result consistent with simulations which


14 Q-Z Ye et al

Table 6 Predicted future Draconid activity

Year Trail Start Peak End Max r (mm) Min r (mm) Offset

2018 1953 1952 1954 1956 100 04 Yes2019 1979 1940 1942 1945 2 lt 01 No2021 1979 1935 1938 1945 025 lt 01 No2025 1933 1940 1950 1953 10 03 No

predict a similar narrow range in small particle sizes Theconcept of a single power-law distribution in meteoroid sizesmay not be valid for this event spanning radar to visual sizesindicating that ZHR and flux estimates are similarly uncer-tain

Dynamical simulations of this event show with someconfidence that this outburst was originated from particlesreleased during the 1966 perihelion passage of the parentbody 21PGiacobini-Zinner although there are some un-certainties of the exact timing of the encounter Furthersimulations showed that possible increased activity of theDraconid meteor shower may occur in 2018 2019 2021 and2025

ACKNOWLEDGMENTS

We thank Dr David Asher for his very thorough com-ments and Dr Rainer Arlt for his helps on the IMO visualdata We also thank Zbigniew Krzeminski Jason Gill andDaniel Wong for helping with CMOR analysis and oper-ations Funding support from the NASA Meteoroid Envi-ronment Office (cooperative agreement NNX11AB76A) forCMOR operations is gratefully acknowledged QY thanks21PGiacobini-Zinner for giving him a totally unexpectedsurprise which makes his study and research more cheerful

Additionally we thank the following visual observers fortheir reports which is essential as a confirmation of this out-burst Boris Badmaev Igor Bukva Jose Vicente Diaz Mar-tinez Svetlana Dondokova Francisco Jose Sevilla LobatoLjubica Grasic Ilija Ivanovic Javor Kac Julia MangadaevaOlga Masheeva Aleksandar Matic Jovana Milic GalinaMuhutdinova Francisco Ocana Gonzalez Rolanda OndarVladimir Osorov Galina Otto Sasha Prokofyev MaciejReszelski Alejandro Sanchez De Miguel Nikolay Solomin-skiy Jakub Koukal Michel Vandeputte Salvador AguirreJurgen Rendtel Terrence Ross Branislav Savic AlexandrMaidik Koen Miskotte Snezana Todorovic Sogto Tsuren-dashiev Kristina Veljkovic Allen Zhong Milena Zivoticand Bimba Zurbaev

REFERENCES

Blaauw R C Campbell-Brown M D Weryk R J 2011MNRAS 412 2033

Borovicka J Spurny P Koten P 2007 AampA 473 661Brown P Jones J 1995 Earth Moon and Planets 68 223Brown P Weryk R J Wong D K Jones J 2008 Icarus195 317

Brown P Ye Q 2012 Central Bureau Electronic Tele-grams 3249 1

Campbell-Brown M Vaubaillon J Brown P Weryk R JArlt R 2006 AampA 451 339

Campbell-Brown M D Koschny D 2004 AampA 418 751

Ceplecha Z Borovicka J Elford W G RevelleD O Hawkes R L Porubcan V Simek M 1998Space Sci Rev 84 327

Crifo J F Rodionov A V 1997 Icarus 127 319

Davidson R M 1915 JBAA 25 292

Denning W F 1926 MNRAS 87 104

Everhart E 1985 in Carusi A Valsecchi G B eds Dy-namics of Comets Their Origin and Evolution An efficientintegrator that uses Gauss-Radau spacings Kluwer Dor-drecht pp 185ndash202

Hocking W K Fuller B Vandepeer B 2001 Journal ofAtmospheric and Solar-Terrestrial Physics 63 155

Jacchia L G Kopal Z Millman P M 1950 ApJ 111104

Jenniskens P 1995 AampA 295 206

Jenniskens P 2006 Meteor Showers and their ParentComets

Jenniskens P Barentsen G Yrjola I 2011 Central Bu-reau Electronic Telegrams 2862 1

Jones J 1995 MNRAS 275 773

Jones J Brown P Ellis K J Webster A RCampbell-Brown M Krzemenski Z Weryk R J 2005Planet Space Sci 53 413

Kero J Fujiwara Y Abo M Szasz C Nakamura T2012 MNRAS 424 1799

Koschack R Rendtel J 1990 WGN Journal of the Inter-national Meteor Organization 18 44

Koten P Vaubaillon J Toth J Zenden J McAuliffe JKoschny D Pautet D 2012 LPI Contributions 16676225

Kronk G W 1999 Cometography Vol 1 Cambridge Uni-versity Press Cambridge

Marsden B G Sekanina Z Yeomans D K 1973 78 211

Marsden B G Williams G V 2008 Catalogue ofCometary Orbits 17th edn IAU Central Bureau for As-tronomical Telegrams ndash Minor Planet Center CambridgeMassachusetts

Maslov M 2011 WGN Journal of the International Me-teor Organization 39 64

McIntosh B A 1968 in Kresak L Millman P M edsPhysics and Dynamics of Meteors Vol 33 of IAU Sympo-sium Meteor Mass Distribution from Radar Observationsp 343

McKinley D W R 1961 Meteor science and engineering

Mikkola S Innanen K 1999 74 59Sato M Watanabe J Ohkawa T 2012 LPI Contribu-tions 1667 6319

Standish E M 1998 Technical report Planetary and Lu-nar Ephemerides DE405LE405 NASA Jet Propulsion



LaboratoryTancredi G Fernandez J A Rickman H Licandro J2000 AampAS 146 73

Vaubaillon J Koten P Bouley S Rudawska R MaquetL Colas F Toth J Zender J McAuliffe J PautetD Koschny D Jenniskens P Leroy A Lecacheux JAntierHe K 2012 LPI Contributions 1667 6280

Vaubaillon J Watanabe J Sato M Horii S Koten P2011 WGN Journal of the International Meteor Organi-zation 39 59

Weidenschilling S J Jackson A A 1993 Icarus 104 244Weryk R J Brown P G 2012 Planet Space Sci 62 132Wiegert P Vaubaillon J Campbell-Brown M 2009Icarus 201 295

Wiegert P A Brown P G Weryk R J Wong D K2013 AJ 145 70

Ye Q Brown P G Campbell-Brown M D Weryk R J2013 MNRAS 436 675

Yeomans D K 1986 Quart J Roy Astron Soc 27 116


1 Introduction

2 Instrumentation and Data Reduction

3 Observational Results


32 Mass Index

33 Flux


4 Dynamic Modeling

5 Comparison of simulations with observations



53 The ``absent storm for visual observers

54 Future outlook

6 Conclusion

2 Q-Z Ye et al

sim 15h UT quickly reached a maximum around 1640 UTand returned to the background rate around 19h UT Pre-liminary analysis indicated that the peak ZHR was wellabove the storm threshold (1 000 meteors per hour) Un-fortunately this was mostly a radar event in addition tothe poor timing of the peak (which only favored Siberiaand Central Asia in terms of darkness) visual and videoobservers were largely caught unprepared The quick lookanalysis carried out by the International Meteor Organiza-tion (IMO) revealed a peak of ZHR=324 plusmn 66 centered at1651 UT based on observations from only 4 observers1 Thecamera set up by the Petnica Meteor Group in Serbia alsodetected some activity at 17ndash18h UT2

Due to the lack of observations by other techniques theradar observations are essential for the study of this out-burst We will mostly follow the methodology used in ourearlier study of the 2011 event (Ye et al 2013) to analyze the2012 event as recorded by CMOR The goals of the presentpaper are to estimate the basic observational characteristicsof the outburst such as mean radiant mass distributionflux variation and to apply the Campbell-Brown amp Koschny(2004) meteoroid ablation model to try to constrain thephysical characteristics of the meteoroids We will also inves-tigate the cause of this outburst through numerical modelingof the stream

2 INSTRUMENTATION AND DATA

REDUCTION

The instrumental and data reduction details of this workis similar to that given in Ye et al (2013) as appliedto CMOR observations of the 2011 Draconids Tech-nical details of the entire system may also be foundin Hocking et al (2001) Jones et al (2005) Brown et al(2008) and Weryk amp Brown (2012) Here we only summarizea few key concepts

CMOR is a six-site interferometric radar array locatednear London Ontario Canada It operates at 1745 2985and 3815 MHz but only the 2985 MHz data is used inthis study The 2985 MHz component transmits with 12kW peak power with a pulse repetition frequency of 532 HzThe radar ldquoseesrdquo a meteor by detecting the reflection of theoutgoing beam from an ionized meteor trail therefore radarmeteors are detected 90 away from their radiants (such areflection condition is called specular) As such when the ra-diant is close to the zenith it becomes difficult for the radarto detect echoes from a particular stream as the echoes areat low elevation and large ranges when the radiant is close tothe zenith Detected echoes are processed in an automaticmanner to eliminate questionable detections correlate thesame events observed at different sites and calculate meteortrajectories Usually raw streamed data is not permanentlykept but we intentionally saved the raw data from the 2012Draconid outburst for further analysis

Meteor echoes seen by radar may be classified as ei-ther underdense or overdense echoes based on their ap-parent amplitude-time characteristics The deciding factors

1 httpwwwimonetlivedraconids2012 2 httpwwwmeteorirswordpressi-ove-godine-pojacana-aktivnost-drakonida

underdense

Am

plitu

de (r

aw u

nit)

overdense

Time (raw unit)

twisted

Figure 1 Examples of meteor echoes as seen by CMOR Top un-derdense trail middle overdense trail bottom overdense wind-twisted trail

are many (see Ye et al 2013 for a more theoretical introduc-tion) but for a given shower overdense echoes are typicallyassociated with larger meteoroids than underdense echoesThe underdense echoes are always characterized by a rapidrise and an exponential decay in amplitude (Figure 1a) mak-ing them relatively easy to identify automatically Ideallythe overdense echoes are characterized by a rapid rise aplateau and a steady decay in amplitude (Figure 1b) but inreality trails distorted by upper atmospheric winds and ex-hibiting multiple reflection points are often seen (Figure 1c)making them difficult to characterize with automatic algo-rithms

Noting the advantages and weakness of the automaticroutines we take two approaches for the initial data reduc-tion

(i) For the first approach we try to separate Draconidechoes from other echoes in the automatically processeddata This is done in several steps First we select all me-teor radiants measured with time-of-flight (tof) velocitieswithin 10 of the apparent common radiant (ie an accep-tance radius of 10 around the apparent center of the clus-ter of meteor radiants) and 20 of the annual geocentricvelocity vg = 204 km middot sminus1 (Jenniskens 2006) The veloc-ity restriction is intentionally broad reflecting the spreaddue to measurement error and the high decelerations ofDraconid meteoroids a consequence of their fragility (egJacchia et al 1950 Campbell-Brown et al 2006) All these





1959 1622 1654 2627 +558 210 051966 1537 - - 02






2 4 6 8 10

0

10

20

30

40

50

60

70

Rad

iant

den

sity

per

incr

emen

t are

a (d

eg-2

)

Radiant size










4 Q-Z Ye et al



(J2000)


20 21 22 23 24 2520

21

22

23

24

25

Fres

nel v

eloc

ity (k

m s

-1)




1500 1600 1700 1800 19000

10

20

30

40

50

60

Raw

num

ber o

f ech

oes

Time Oct 8 (UT)

Complete Overdense



270o 265o 260o 255o

50o

55o

60o

Geo

cent

ric D

ecl

(J20

00)


Median error




32 Mass Index









Table 3


6 Q-Z Ye et al

1000 10000

10

100

Cum

ulat

ive

echo

num

ber


s=188 001




1 10

10

100

Cum

ulat

ive

num

ber

Echo duration (s)

tc~25 s




Hei

ght (

km)

Echo duration (s)

overdense

underdense


33 Flux













14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

75

80

85

90

95

100

105

1012 1013 1014 101575

80

85

90

95

100


Hei

ght (

km)



Hei

ght (

km)



4 DYNAMIC MODELING




8 Q-Z Ye et al

100

1000

10000 IMO visual

1954o 1955o 1956o 1957o 1958o

100

1000

10000 ZHR

Zeni

th H

ourly

Rat

e(ZHR

hr-1

)

CMOR

Flux (hr -1 km-2)

12 13 14 15 16 17 18 19 20Time Oct 8 (UT)

001

01

1

1954o 1955o 1956o 1957o 1958o0

100

200

300


Col

lect

ing

area

(km

2 )R

adiant elevation0o

20o

40o

60o

80o
















14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

1012 1013 1014 1015

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

1012 1013 1014 1015

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

1012 1013 1014 1015

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

1012 1013 1014 1015

Hei

ght (

km)

mtotal=10-6 kg

mtotal=10-5 kg

mtotal=10-4 kg

ngrain=10

Hei

ght (

km)

ngrain=102

Hei

ght (

km)

ngrain=103

Hei

ght (

km)

Velocity (km s-1)

ngrain=104




10 Q-Z Ye et al


21P 2005 Jul 27605 1037914 0705691 3526613 1725429 1954301 318109

21P 1966 Mar 282844 0933501 0729394 3449669 1729153 1966674 309382














OBSERVATIONS





0950 0960 0970 0980025

00

260

027

00

280

x (AU)

y (A

U)





radar peak

1966

181718521933194019591966othersclosest








12 Q-Z Ye et al










Cou

nts

1954 1955 1956 1957 1958 1959

050

010

0015

00


2626 2624 2622 2620 2618

552

554

556

558

560

562


Dec

linat

ion

(deg

J20

00)





510

2050

100

200

500

β

ejec

tion

velo

city

(m

s)





54 Future outlook






6 CONCLUSION




14 Q-Z Ye et al



2018 1953 1952 1954 1956 100 04 Yes2019 1979 1940 1942 1945 2 lt 01 No2021 1979 1935 1938 1945 025 lt 01 No2025 1933 1940 1950 1953 10 03 No



ACKNOWLEDGMENTS



REFERENCES







































1 Introduction




32 Mass Index

33 Flux


4 Dynamic Modeling





54 Future outlook

6 Conclusion




1959 1622 1654 2627 +558 210 051966 1537 - - 02






2 4 6 8 10

0

10

20

30

40

50

60

70

Rad

iant

den

sity

per

incr

emen

t are

a (d

eg-2

)

Radiant size










4 Q-Z Ye et al



(J2000)


20 21 22 23 24 2520

21

22

23

24

25

Fres

nel v

eloc

ity (k

m s

-1)




1500 1600 1700 1800 19000

10

20

30

40

50

60

Raw

num

ber o

f ech

oes

Time Oct 8 (UT)

Complete Overdense



270o 265o 260o 255o

50o

55o

60o

Geo

cent

ric D

ecl

(J20

00)


Median error




32 Mass Index









Table 3


6 Q-Z Ye et al

1000 10000

10

100

Cum

ulat

ive

echo

num

ber


s=188 001




1 10

10

100

Cum

ulat

ive

num

ber

Echo duration (s)

tc~25 s




Hei

ght (

km)

Echo duration (s)

overdense

underdense


33 Flux













14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

75

80

85

90

95

100

105

1012 1013 1014 101575

80

85

90

95

100


Hei

ght (

km)



Hei

ght (

km)



4 DYNAMIC MODELING




8 Q-Z Ye et al

100

1000

10000 IMO visual

1954o 1955o 1956o 1957o 1958o

100

1000

10000 ZHR

Zeni

th H

ourly

Rat

e(ZHR

hr-1

)

CMOR

Flux (hr -1 km-2)

12 13 14 15 16 17 18 19 20Time Oct 8 (UT)

001

01

1

1954o 1955o 1956o 1957o 1958o0

100

200

300


Col

lect

ing

area

(km

2 )R

adiant elevation0o

20o

40o

60o

80o
















14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

1012 1013 1014 1015

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

1012 1013 1014 1015

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

1012 1013 1014 1015

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

1012 1013 1014 1015

Hei

ght (

km)

mtotal=10-6 kg

mtotal=10-5 kg

mtotal=10-4 kg

ngrain=10

Hei

ght (

km)

ngrain=102

Hei

ght (

km)

ngrain=103

Hei

ght (

km)

Velocity (km s-1)

ngrain=104




10 Q-Z Ye et al


21P 2005 Jul 27605 1037914 0705691 3526613 1725429 1954301 318109

21P 1966 Mar 282844 0933501 0729394 3449669 1729153 1966674 309382














OBSERVATIONS





0950 0960 0970 0980025

00

260

027

00

280

x (AU)

y (A

U)





radar peak

1966

181718521933194019591966othersclosest








12 Q-Z Ye et al










Cou

nts

1954 1955 1956 1957 1958 1959

050

010

0015

00


2626 2624 2622 2620 2618

552

554

556

558

560

562


Dec

linat

ion

(deg

J20

00)





510

2050

100

200

500

β

ejec

tion

velo

city

(m

s)





54 Future outlook






6 CONCLUSION




14 Q-Z Ye et al



2018 1953 1952 1954 1956 100 04 Yes2019 1979 1940 1942 1945 2 lt 01 No2021 1979 1935 1938 1945 025 lt 01 No2025 1933 1940 1950 1953 10 03 No



ACKNOWLEDGMENTS



REFERENCES







































1 Introduction




32 Mass Index

33 Flux


4 Dynamic Modeling





54 Future outlook

6 Conclusion

4 Q-Z Ye et al



(J2000)


20 21 22 23 24 2520

21

22

23

24

25

Fres

nel v

eloc

ity (k

m s

-1)




1500 1600 1700 1800 19000

10

20

30

40

50

60

Raw

num

ber o

f ech

oes

Time Oct 8 (UT)

Complete Overdense



270o 265o 260o 255o

50o

55o

60o

Geo

cent

ric D

ecl

(J20

00)


Median error




32 Mass Index









Table 3


6 Q-Z Ye et al

1000 10000

10

100

Cum

ulat

ive

echo

num

ber


s=188 001




1 10

10

100

Cum

ulat

ive

num

ber

Echo duration (s)

tc~25 s




Hei

ght (

km)

Echo duration (s)

overdense

underdense


33 Flux













14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

75

80

85

90

95

100

105

1012 1013 1014 101575

80

85

90

95

100


Hei

ght (

km)



Hei

ght (

km)



4 DYNAMIC MODELING




8 Q-Z Ye et al

100

1000

10000 IMO visual

1954o 1955o 1956o 1957o 1958o

100

1000

10000 ZHR

Zeni

th H

ourly

Rat

e(ZHR

hr-1

)

CMOR

Flux (hr -1 km-2)

12 13 14 15 16 17 18 19 20Time Oct 8 (UT)

001

01

1

1954o 1955o 1956o 1957o 1958o0

100

200

300


Col

lect

ing

area

(km

2 )R

adiant elevation0o

20o

40o

60o

80o
















14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

1012 1013 1014 1015

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

1012 1013 1014 1015

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

1012 1013 1014 1015

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

1012 1013 1014 1015

Hei

ght (

km)

mtotal=10-6 kg

mtotal=10-5 kg

mtotal=10-4 kg

ngrain=10

Hei

ght (

km)

ngrain=102

Hei

ght (

km)

ngrain=103

Hei

ght (

km)

Velocity (km s-1)

ngrain=104




10 Q-Z Ye et al


21P 2005 Jul 27605 1037914 0705691 3526613 1725429 1954301 318109

21P 1966 Mar 282844 0933501 0729394 3449669 1729153 1966674 309382














OBSERVATIONS





0950 0960 0970 0980025

00

260

027

00

280

x (AU)

y (A

U)





radar peak

1966

181718521933194019591966othersclosest








12 Q-Z Ye et al










Cou

nts

1954 1955 1956 1957 1958 1959

050

010

0015

00


2626 2624 2622 2620 2618

552

554

556

558

560

562


Dec

linat

ion

(deg

J20

00)





510

2050

100

200

500

β

ejec

tion

velo

city

(m

s)





54 Future outlook






6 CONCLUSION




14 Q-Z Ye et al



2018 1953 1952 1954 1956 100 04 Yes2019 1979 1940 1942 1945 2 lt 01 No2021 1979 1935 1938 1945 025 lt 01 No2025 1933 1940 1950 1953 10 03 No



ACKNOWLEDGMENTS



REFERENCES







































1 Introduction




32 Mass Index

33 Flux


4 Dynamic Modeling





54 Future outlook

6 Conclusion



Table 3


6 Q-Z Ye et al

1000 10000

10

100

Cum

ulat

ive

echo

num

ber


s=188 001




1 10

10

100

Cum

ulat

ive

num

ber

Echo duration (s)

tc~25 s




Hei

ght (

km)

Echo duration (s)

overdense

underdense


33 Flux













14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

75

80

85

90

95

100

105

1012 1013 1014 101575

80

85

90

95

100


Hei

ght (

km)



Hei

ght (

km)



4 DYNAMIC MODELING




8 Q-Z Ye et al

100

1000

10000 IMO visual

1954o 1955o 1956o 1957o 1958o

100

1000

10000 ZHR

Zeni

th H

ourly

Rat

e(ZHR

hr-1

)

CMOR

Flux (hr -1 km-2)

12 13 14 15 16 17 18 19 20Time Oct 8 (UT)

001

01

1

1954o 1955o 1956o 1957o 1958o0

100

200

300


Col

lect

ing

area

(km

2 )R

adiant elevation0o

20o

40o

60o

80o
















14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

1012 1013 1014 1015

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

1012 1013 1014 1015

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

1012 1013 1014 1015

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

1012 1013 1014 1015

Hei

ght (

km)

mtotal=10-6 kg

mtotal=10-5 kg

mtotal=10-4 kg

ngrain=10

Hei

ght (

km)

ngrain=102

Hei

ght (

km)

ngrain=103

Hei

ght (

km)

Velocity (km s-1)

ngrain=104




10 Q-Z Ye et al


21P 2005 Jul 27605 1037914 0705691 3526613 1725429 1954301 318109

21P 1966 Mar 282844 0933501 0729394 3449669 1729153 1966674 309382














OBSERVATIONS





0950 0960 0970 0980025

00

260

027

00

280

x (AU)

y (A

U)





radar peak

1966

181718521933194019591966othersclosest








12 Q-Z Ye et al










Cou

nts

1954 1955 1956 1957 1958 1959

050

010

0015

00


2626 2624 2622 2620 2618

552

554

556

558

560

562


Dec

linat

ion

(deg

J20

00)





510

2050

100

200

500

β

ejec

tion

velo

city

(m

s)





54 Future outlook






6 CONCLUSION




14 Q-Z Ye et al



2018 1953 1952 1954 1956 100 04 Yes2019 1979 1940 1942 1945 2 lt 01 No2021 1979 1935 1938 1945 025 lt 01 No2025 1933 1940 1950 1953 10 03 No



ACKNOWLEDGMENTS



REFERENCES







































1 Introduction




32 Mass Index

33 Flux


4 Dynamic Modeling





54 Future outlook

6 Conclusion

6 Q-Z Ye et al

1000 10000

10

100

Cum

ulat

ive

echo

num

ber


s=188 001




1 10

10

100

Cum

ulat

ive

num

ber

Echo duration (s)

tc~25 s




Hei

ght (

km)

Echo duration (s)

overdense

underdense


33 Flux













14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

75

80

85

90

95

100

105

1012 1013 1014 101575

80

85

90

95

100


Hei

ght (

km)



Hei

ght (

km)



4 DYNAMIC MODELING




8 Q-Z Ye et al

100

1000

10000 IMO visual

1954o 1955o 1956o 1957o 1958o

100

1000

10000 ZHR

Zeni

th H

ourly

Rat

e(ZHR

hr-1

)

CMOR

Flux (hr -1 km-2)

12 13 14 15 16 17 18 19 20Time Oct 8 (UT)

001

01

1

1954o 1955o 1956o 1957o 1958o0

100

200

300


Col

lect

ing

area

(km

2 )R

adiant elevation0o

20o

40o

60o

80o
















14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

1012 1013 1014 1015

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

1012 1013 1014 1015

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

1012 1013 1014 1015

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

1012 1013 1014 1015

Hei

ght (

km)

mtotal=10-6 kg

mtotal=10-5 kg

mtotal=10-4 kg

ngrain=10

Hei

ght (

km)

ngrain=102

Hei

ght (

km)

ngrain=103

Hei

ght (

km)

Velocity (km s-1)

ngrain=104




10 Q-Z Ye et al


21P 2005 Jul 27605 1037914 0705691 3526613 1725429 1954301 318109

21P 1966 Mar 282844 0933501 0729394 3449669 1729153 1966674 309382














OBSERVATIONS





0950 0960 0970 0980025

00

260

027

00

280

x (AU)

y (A

U)





radar peak

1966

181718521933194019591966othersclosest








12 Q-Z Ye et al










Cou

nts

1954 1955 1956 1957 1958 1959

050

010

0015

00


2626 2624 2622 2620 2618

552

554

556

558

560

562


Dec

linat

ion

(deg

J20

00)





510

2050

100

200

500

β

ejec

tion

velo

city

(m

s)





54 Future outlook






6 CONCLUSION




14 Q-Z Ye et al



2018 1953 1952 1954 1956 100 04 Yes2019 1979 1940 1942 1945 2 lt 01 No2021 1979 1935 1938 1945 025 lt 01 No2025 1933 1940 1950 1953 10 03 No



ACKNOWLEDGMENTS



REFERENCES







































1 Introduction




32 Mass Index

33 Flux


4 Dynamic Modeling





54 Future outlook

6 Conclusion










14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

75

80

85

90

95

100

105

1012 1013 1014 101575

80

85

90

95

100


Hei

ght (

km)



Hei

ght (

km)



4 DYNAMIC MODELING




8 Q-Z Ye et al

100

1000

10000 IMO visual

1954o 1955o 1956o 1957o 1958o

100

1000

10000 ZHR

Zeni

th H

ourly

Rat

e(ZHR

hr-1

)

CMOR

Flux (hr -1 km-2)

12 13 14 15 16 17 18 19 20Time Oct 8 (UT)

001

01

1

1954o 1955o 1956o 1957o 1958o0

100

200

300


Col

lect

ing

area

(km

2 )R

adiant elevation0o

20o

40o

60o

80o
















14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

1012 1013 1014 1015

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

1012 1013 1014 1015

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

1012 1013 1014 1015

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

1012 1013 1014 1015

Hei

ght (

km)

mtotal=10-6 kg

mtotal=10-5 kg

mtotal=10-4 kg

ngrain=10

Hei

ght (

km)

ngrain=102

Hei

ght (

km)

ngrain=103

Hei

ght (

km)

Velocity (km s-1)

ngrain=104




10 Q-Z Ye et al


21P 2005 Jul 27605 1037914 0705691 3526613 1725429 1954301 318109

21P 1966 Mar 282844 0933501 0729394 3449669 1729153 1966674 309382














OBSERVATIONS





0950 0960 0970 0980025

00

260

027

00

280

x (AU)

y (A

U)





radar peak

1966

181718521933194019591966othersclosest








12 Q-Z Ye et al










Cou

nts

1954 1955 1956 1957 1958 1959

050

010

0015

00


2626 2624 2622 2620 2618

552

554

556

558

560

562


Dec

linat

ion

(deg

J20

00)





510

2050

100

200

500

β

ejec

tion

velo

city

(m

s)





54 Future outlook






6 CONCLUSION




14 Q-Z Ye et al



2018 1953 1952 1954 1956 100 04 Yes2019 1979 1940 1942 1945 2 lt 01 No2021 1979 1935 1938 1945 025 lt 01 No2025 1933 1940 1950 1953 10 03 No



ACKNOWLEDGMENTS



REFERENCES







































1 Introduction




32 Mass Index

33 Flux


4 Dynamic Modeling





54 Future outlook

6 Conclusion

8 Q-Z Ye et al

100

1000

10000 IMO visual

1954o 1955o 1956o 1957o 1958o

100

1000

10000 ZHR

Zeni

th H

ourly

Rat

e(ZHR

hr-1

)

CMOR

Flux (hr -1 km-2)

12 13 14 15 16 17 18 19 20Time Oct 8 (UT)

001

01

1

1954o 1955o 1956o 1957o 1958o0

100

200

300


Col

lect

ing

area

(km

2 )R

adiant elevation0o

20o

40o

60o

80o
















14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

1012 1013 1014 1015

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

1012 1013 1014 1015

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

1012 1013 1014 1015

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

1012 1013 1014 1015

Hei

ght (

km)

mtotal=10-6 kg

mtotal=10-5 kg

mtotal=10-4 kg

ngrain=10

Hei

ght (

km)

ngrain=102

Hei

ght (

km)

ngrain=103

Hei

ght (

km)

Velocity (km s-1)

ngrain=104




10 Q-Z Ye et al


21P 2005 Jul 27605 1037914 0705691 3526613 1725429 1954301 318109

21P 1966 Mar 282844 0933501 0729394 3449669 1729153 1966674 309382














OBSERVATIONS





0950 0960 0970 0980025

00

260

027

00

280

x (AU)

y (A

U)





radar peak

1966

181718521933194019591966othersclosest








12 Q-Z Ye et al










Cou

nts

1954 1955 1956 1957 1958 1959

050

010

0015

00


2626 2624 2622 2620 2618

552

554

556

558

560

562


Dec

linat

ion

(deg

J20

00)





510

2050

100

200

500

β

ejec

tion

velo

city

(m

s)





54 Future outlook






6 CONCLUSION




14 Q-Z Ye et al



2018 1953 1952 1954 1956 100 04 Yes2019 1979 1940 1942 1945 2 lt 01 No2021 1979 1935 1938 1945 025 lt 01 No2025 1933 1940 1950 1953 10 03 No



ACKNOWLEDGMENTS



REFERENCES







































1 Introduction




32 Mass Index

33 Flux


4 Dynamic Modeling





54 Future outlook

6 Conclusion


14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

1012 1013 1014 1015

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

1012 1013 1014 1015

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

1012 1013 1014 1015

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

14 16 18 20 22 24 26 28 30 3275

80

85

90

95

100

105

1012 1013 1014 1015

Hei

ght (

km)

mtotal=10-6 kg

mtotal=10-5 kg

mtotal=10-4 kg

ngrain=10

Hei

ght (

km)

ngrain=102

Hei

ght (

km)

ngrain=103

Hei

ght (

km)

Velocity (km s-1)

ngrain=104




10 Q-Z Ye et al


21P 2005 Jul 27605 1037914 0705691 3526613 1725429 1954301 318109

21P 1966 Mar 282844 0933501 0729394 3449669 1729153 1966674 309382














OBSERVATIONS





0950 0960 0970 0980025

00

260

027

00

280

x (AU)

y (A

U)





radar peak

1966

181718521933194019591966othersclosest








12 Q-Z Ye et al










Cou

nts

1954 1955 1956 1957 1958 1959

050

010

0015

00


2626 2624 2622 2620 2618

552

554

556

558

560

562


Dec

linat

ion

(deg

J20

00)





510

2050

100

200

500

β

ejec

tion

velo

city

(m

s)





54 Future outlook






6 CONCLUSION




14 Q-Z Ye et al



2018 1953 1952 1954 1956 100 04 Yes2019 1979 1940 1942 1945 2 lt 01 No2021 1979 1935 1938 1945 025 lt 01 No2025 1933 1940 1950 1953 10 03 No



ACKNOWLEDGMENTS



REFERENCES







































1 Introduction




32 Mass Index

33 Flux


4 Dynamic Modeling





54 Future outlook

6 Conclusion

10 Q-Z Ye et al


21P 2005 Jul 27605 1037914 0705691 3526613 1725429 1954301 318109

21P 1966 Mar 282844 0933501 0729394 3449669 1729153 1966674 309382














OBSERVATIONS





0950 0960 0970 0980025

00

260

027

00

280

x (AU)

y (A

U)





radar peak

1966

181718521933194019591966othersclosest








12 Q-Z Ye et al










Cou

nts

1954 1955 1956 1957 1958 1959

050

010

0015

00


2626 2624 2622 2620 2618

552

554

556

558

560

562


Dec

linat

ion

(deg

J20

00)





510

2050

100

200

500

β

ejec

tion

velo

city

(m

s)





54 Future outlook






6 CONCLUSION




14 Q-Z Ye et al



2018 1953 1952 1954 1956 100 04 Yes2019 1979 1940 1942 1945 2 lt 01 No2021 1979 1935 1938 1945 025 lt 01 No2025 1933 1940 1950 1953 10 03 No



ACKNOWLEDGMENTS



REFERENCES







































1 Introduction




32 Mass Index

33 Flux


4 Dynamic Modeling





54 Future outlook

6 Conclusion


0950 0960 0970 0980025

00

260

027

00

280

x (AU)

y (A

U)





radar peak

1966

181718521933194019591966othersclosest








12 Q-Z Ye et al










Cou

nts

1954 1955 1956 1957 1958 1959

050

010

0015

00


2626 2624 2622 2620 2618

552

554

556

558

560

562


Dec

linat

ion

(deg

J20

00)





510

2050

100

200

500

β

ejec

tion

velo

city

(m

s)





54 Future outlook






6 CONCLUSION




14 Q-Z Ye et al



2018 1953 1952 1954 1956 100 04 Yes2019 1979 1940 1942 1945 2 lt 01 No2021 1979 1935 1938 1945 025 lt 01 No2025 1933 1940 1950 1953 10 03 No



ACKNOWLEDGMENTS



REFERENCES







































1 Introduction




32 Mass Index

33 Flux


4 Dynamic Modeling





54 Future outlook

6 Conclusion

12 Q-Z Ye et al










Cou

nts

1954 1955 1956 1957 1958 1959

050

010

0015

00


2626 2624 2622 2620 2618

552

554

556

558

560

562


Dec

linat

ion

(deg

J20

00)





510

2050

100

200

500

β

ejec

tion

velo

city

(m

s)





54 Future outlook






6 CONCLUSION




14 Q-Z Ye et al



2018 1953 1952 1954 1956 100 04 Yes2019 1979 1940 1942 1945 2 lt 01 No2021 1979 1935 1938 1945 025 lt 01 No2025 1933 1940 1950 1953 10 03 No



ACKNOWLEDGMENTS



REFERENCES







































1 Introduction




32 Mass Index

33 Flux


4 Dynamic Modeling





54 Future outlook

6 Conclusion



510

2050

100

200

500

β

ejec

tion

velo

city

(m

s)





54 Future outlook






6 CONCLUSION




14 Q-Z Ye et al



2018 1953 1952 1954 1956 100 04 Yes2019 1979 1940 1942 1945 2 lt 01 No2021 1979 1935 1938 1945 025 lt 01 No2025 1933 1940 1950 1953 10 03 No



ACKNOWLEDGMENTS



REFERENCES







































1 Introduction




32 Mass Index

33 Flux


4 Dynamic Modeling





54 Future outlook

6 Conclusion

14 Q-Z Ye et al



2018 1953 1952 1954 1956 100 04 Yes2019 1979 1940 1942 1945 2 lt 01 No2021 1979 1935 1938 1945 025 lt 01 No2025 1933 1940 1950 1953 10 03 No



ACKNOWLEDGMENTS



REFERENCES







































1 Introduction




32 Mass Index

33 Flux


4 Dynamic Modeling





54 Future outlook

6 Conclusion










1 Introduction




32 Mass Index

33 Flux


4 Dynamic Modeling





54 Future outlook

6 Conclusion

TheUnexpected2012DraconidMeteorStorm - arXiv · arXiv:1311.1733v1 [astro-ph.EP] 7 Nov 2013 Mon....

Documents

Transcript of TheUnexpected2012DraconidMeteorStorm - arXiv · arXiv:1311.1733v1 [astro-ph.EP] 7 Nov 2013 Mon....