Activity of the Eta-Aquariid and Orionid meteor showers
A. Egal, P. G. Brown, J. Rendtel, M. Campbell-Brown, P. Wiegert
AAstronomy & Astrophysics manuscript no. Egal2020b © ESO 2020June 16, 2020
Activity of the Eta-Aquariid and Orionid meteor showers
A. Egal , , (cid:63) , P. G. Brown , , J. Rendtel , M. Campbell-Brown , , and P. Wiegert , Department of Physics and Astronomy, The University of Western Ontario, London, Ontario N6A 3K7, Canada Institute for Earth and Space Exploration (IESX), The University of Western Ontario, London, Ontario N6A 3K7, Canada IMCCE, Observatoire de Paris, PSL Research University, CNRS, Sorbonne Universités, UPMC Univ. Paris 06, Univ. Lille, France Leibniz-Institut f. Astrophysik Potsdam, An der Sternwarte 16, 14482 Potsdam, Germany, and International Meteor Organization,Eschenweg 16, 14476 Potsdam, GermanyReceived XYZ; accepted XYZ
ABSTRACT
Aims.
We present a multi-instrumental, multidecadal analysis of the activity of the Eta-Aquariid and Orionid meteor showers for thepurpose of constraining models of 1P / Halley’s meteoroid streams.
Methods.
The interannual variability of the showers’ peak activity and period of duration is investigated through the compilation ofpublished visual and radar observations prior to 1985 and more recent measurements reported in the International Meteor Organization(IMO) Visual Meteor DataBase, by the IMO Video Meteor Network and by the Canadian Meteor Orbit Radar (CMOR). Thesetechniques probe the range of meteoroid masses from submilligrams to grams. The η -Aquariids and Orionids activity duration, shape,maximum zenithal hourly rates (ZHR) values, and the solar longitude of annual peaks since 1985 are analyzed. When available,annual activity profiles recorded by each detection network were measured and are compared. Results.
Observations from the three detection methods show generally good agreement in the showers’ shape, activity levels, andannual intensity variations. Both showers display several activity peaks of variable location and strength with time. The η -Aquariidsare usually two to three times stronger than the Orionids, but the two showers display occasional outbursts with peaks two to fourtimes their usual activity level. CMOR observations since 2002 seem to support the existence of an ∼
12 year cycle in Orionids activityvariations; however, additional and longer term radar and optical observations of the shower are required to confirm such periodicity.
Key words. meteors, meteoroids – comets: individual: 1P / Halley
1. Introduction
Comet 1P / Halley is known to produce two annual meteor show-ers, the η -Aquariids in early May and the Orionids in late Oc-tober. These two related showers are often collectively termedthe Halleyids. Both showers exhibit a total duration of about 30°in solar longitude ( L (cid:12) ) and a complex fine structure character-ized by several subpeaks of variable location and intensity. Theaverage activity levels of the Orionids and the η -Aquariids aresimilar. The activity of the showers is frequently characterizedby the zenithal hourly rate (ZHR), which represents the numberof meteors an observer would observe per hour under standardreference conditions.The maximum activity of the Orionids usually varies be-tween 15 and 30 meteors per hour near 209° of solar longitude( L (cid:12) , ecliptic J2000), while the η -Aquariids maximum rates aretwo to three times higher, near 45.5° solar longitude. Both show-ers are frequently cited as being rich in small (masses ≈ − kg)particles (Jenniskens 2006; Campbell-Brown and Brown 2015;Schult et al. 2018).Despite moderate average annual activity, the two showersare known to produce occasional outbursts. Recent examples in-clude the Orionids apparitions of 2006 and 2007, when the peakZHR exceeded 60 (Rendtel 2007; Arlt et al. 2008), and the 2013 η -Aquariids outburst that reached an activity level of 135 mete-ors per hour (Cooper 2013). Mean motion resonances of Halley’smeteoroids with Jupiter were identified as responsible for these (cid:63) e-mail: [email protected] periods of enhanced activity (Rendtel 2007; Sato and Watanabe2007, 2014; Sekhar and Asher 2014).The Orionids and η -Aquariids are of particular interest be-cause of their connection to comet 1P / Halley. Meteor observa-tions can be used to constrain the nature of the meteoroid trailsthat are ejected by the comet (meteoroids size, density variationsalong the stream, etc.) and characterize its past activity. Con-straining meteoroid stream models using meteor observationscould therefore provide insights into 1P / Halley’s past orbitalevolution (Sekhar and Asher 2014; Kinsman and Asher 2017),which is mostly unknown prior to 1000 BCE.The two showers are also of concern from a spacecraft safetyperspective. The Orionids and especially the η -Aquariids areamong the more significant impact hazards out of all the ma-jor showers throughout the year. This is a consequence of sev-eral characteristics of the showers including their long duration,comparatively high flux, the occurrence of occasional outburstsand their high velocity ( ∼
66 km / s). In spite of their significance,few predictions of future Halleyid activity are currently availablein the literature. Moreover, no comprehensive numerical modelsof the streams have been published which utilize the full suiteof modern observational data, in part because no compilation ofsuch data have become available. This is a major goal of the cur-rent work.Most of Halley’s meteoroid stream models were developedin the 1980s, when the expected return of the comet in 1986 re-newed the interest in the study of these showers. Several charac-teristics of the Halleyids (similar duration and average intensity,complex structure) were successfully explained by the long-term Article number, page 1 of 25 a r X i v : . [ a s t r o - ph . E P ] J un & A proofs: manuscript no. Egal2020b evolution of meteoroids under the influence of planetary pertur-bations (mainly induced by Jupiter, see McIntosh and Hajduk1983 and McIntosh and Jones 1988).However, almost no modeling e ff ort had been made to ex-plain the annual activity variations of the Orionids and η -Aquariids noticed by Hajduk (1970) and Hajduk (1973), untilthe observation of the recent Orionids outbursts. By investigatingthe role of resonances highlighted in previous works, numericalsimulations of Sekhar and Asher (2014) successfully reproducedthe apparition dates of recent and ancient Orionids outbursts, andpredicted a future outburst of the shower in 2070. Unfortunately,no similar analysis of the η -Aquariids has been conducted bythose authors, in part because of the small number of reliableobservations of this shower compared to the Orionids (Sekharand Asher 2014).In this work, we investigate the long-term activity of the Ori-onids and η -Aquariids as measured by visual, video and radarobservations. Published and original observations of both show-ers are compiled and analyzed to examine the structure, the du-ration, and the annual variation of their activity profiles. Follow-ing the work of Campbell-Brown and Brown (2015) which fo-cused solely on the η -Aquariids up to 2014, we provide recentmeasurements of the η -Aquariids and a complete set of Orionidsrecorded by the Canadian Meteor Orbit Radar (CMOR) between2002 and 2019, which constitutes the longest consistent obser-vational set of the Halleyids to date. Results from CMOR arecompared to the visual observations contained in the IMO Vi-sual Meteor Data Base (VMDB) and measurements of the VideoMeteor Network (VMN).The specific goal of this paper is to measure characteristics ofthe Halleyids including their long-term average ZHR-equivalentprofile, activity shapes of annual apparitions, maximum activitylevels each year, and the location of the peak using an integratedmulti-instrument analysis of the showers. In particular we wishto compare results across the di ff erent detection systems to betteridentify biases.The structure of this paper is as follows: sections 2, 3 and4 review the discovery circumstances, observational history andgeneral characteristics of comet 1P / Halley and previously pub-lished analyses of the Orionids and the η -Aquariids meteorshowers. Sections 5 and 6 present the methodology for selec-tion, processing and analysis of the showers’ visual, video. andradar observations since 2002. The conclusion of our analysis ispresented in Section 7. To aid future modeling e ff orts, individ-ual activity profiles per year including specific numerical mea-surements for peak activity and location found in this work areavailable in Appendices A, B, C and D.
2. History of 1P/Halley and early observations of theHalleyids / Halley is a famous comet, evolving in a retrograde orbit witha period of about 76 years. The comet was named after Ed-mond Halley, who recognized in 1705 the comets of August-September 1682, October 1607 and August 1531 as being thesame object. Halley estimated the orbital period of the comet andanticipated a return in 1758. His eponymous comet was recov-ered in December 1758, a few years after Halley’s death, becom-ing the first comet whose return close to the Sun was successfullypredicted (Hughes 1987b).Since Edmond Halley’s discovery, several ancient obser-vations of comets have been linked to older apparitions of 1P / Halley (Yeomans and Kiang 1981; Hughes 1987b). The firstidentified observations of 1P / Halley date back to 240 BCE inChinese records (Kiang 1972) and 164 BCE in Babylonianrecords (Stephenson et al. 1985). With some adjustments to thecomet’s eccentricity around AD 837, Yeomans and Kiang (1981)determined a reliable set of orbits for 1P / Halley until 1404 BCE.Because of a close encounter with Earth in 1404 BCE and thelack of older observational constraints on the comet’s motion,the orbital elements of 1P / Halley have not been precisely deter-mined prior to this epoch (Yeomans and Kiang 1981).In March 1986, an international fleet of spacecraft called the"Halley Armada" approached the comet to examine its nucleus.The Armada was composed of five main probes, Giotto (Euro-pean Space Agency), Vega 1 and Vega 2 (Soviet Union) and Sui-sei and Sakigake (Japan), supported by additional measurementsof the international ISEE-3 (ICE) spacecraft and NASA’s Pio-neer 7 and Pioneer 12 probes. On the ground, observations of thecomet and its associated meteoroid streams were collected andarchived by the International Halley Watch (IHW) organization,an international agency created to coordinate comet Halley’s ob-servations. Reviews of the extensive research conducted duringthe comet 1986 apparition can be found for example in Whip-ple (1987) and Edberg et al. (1988). The next apparition of thecomet is expected in 2061. / Halley’s meteoroid streams are responsible for two observedmeteor showers on Earth, the Orionids (cf. Section 3) and the η -Aquariids (cf. Section 4). The η -Aquariids occur at the de-scending node of the comet, while the Orionids are connectedto 1P / Halley’s ascending node.The evolution of 1P / Halley’s nodes as a function of time ispresented in Figure 1. The motion of the comet was integratedfrom 1404 BCE to AD 2050 using a RADAU15 (Everhart 1985)integrator with an external time step of 1 day. Orbital solutionsof Yeomans and Kiang (1981) were used as initial conditionsfor the comet apparitions before 1910, while the JPL J863 / . Results arepresented in the barycentric frame (ecliptic J2000).As shown in Figure 1, the orbital precession of 1P / Halleyleads to considerable change in the location of its nodes since1404 BCE. In 1404 BCE, the comet’s descending node was lo-cated well outside the Earth’s orbit, while the ascending nodewas much closer to Earth’s orbit. Currently, the comet crossesthe ecliptic plane at an ascending node far outside the Earth’s or-bit (1.8 AU), and descends below the ecliptic at about 0.85 AU.The current high ascending nodal distance of 1P / Halley fromEarth’s orbit explains why the link between the Orionids showerand 1P / Halley was di ffi cult to establish (cf. Section 3).Explaining the similar activity levels of the η -Aquariids andOrionids (Lovell 1954; Hajduk 1970, 1973) with such di ff erentnodal distances has challenged researchers for many decades.If the showers were produced by a stream centered on the cur-rent orbit of the comet, the η -Aquariids would be considerablystronger than the Orionids and their durations significantly dif-ferent.Perhaps the most successful model of the Halleyids is thatof the theoretical shell model of McIntosh and Hajduk (1983). Itwas later confirmed by McIntosh and Jones (1988)’s numericalsimulations and o ff ers an explanation for many characteristics ofthe showers. The model is based on the idea that particles ejected https: // ssd.jpl.nasa.gov / sbdb.cgiArticle number, page 2 of 25. Egal et al.: Activity of the Eta-Aquariid and Orionid meteor showers Fig. 1.
Ecliptic location of 1P / Halley’s ascending and descending nodeas a function of time, in the barycentric frame (ecliptic J2000). TheEarth’s orbit is represented in blue. from the comet evolve at di ff erent rates, with some remainingon orbits where the comet was a long time ago and others pre-cessing more rapidly than the comet (and eventually reaching itsfuture positions). Each meteoroid stream would therefore evolveinto a ribbon-like structure of uniform thickness, producing twometeor showers of similar duration and intensity at the Earth.The superposition of several ribbon-shaped streams, separatedthrough small perturbation relative to the comet orbit, couldbe responsible for the observed filamentary structure of the η -Aquariids and Orionids showers (cf. Sections 3 and 4).The time evolution of the Minimum Orbit Intersection Dis-tance (MOID) between Earth and the comet around each node isprovided in Figure 2. The Earth currently approaches the cometorbit at a minimum distance of 0.154 AU at the time of the Orion-ids and 0.065 AU at the time of the η -Aquariids. 1P / Halley’s de-scending node reached its closest distance to Earth’s orbit aroundAD 500, and at about 800 BCE for its ascending node. The prox-imity of the comet around AD 500 might explain the existenceof strong η -Aquariids outbursts reported in ancient Chinese ob-servations. The long-term (millennium timescale) variation of the strengthof the showers is recorded in ancient visual records, thoughthe biases and incompleteness of these sources imply that somestrong returns may easily have been missed. The analysis of an-cient records of meteor showers reveals that the Orionids andthe η -Aquariids were active a millennium or more ago (Ahn2003, 2004). Orionids outbursts were identified in Chinese andJapanese observations in AD 585, 930, 1436, 1439, 1465, and1623 (Imoto and Hasegawa 1958). Possible strong displays inAD 288 and 1651 are also mentioned in Zhuang (1977).The first Korean record of an η -Aquariid outburst could beas old as 687 BCE (Ahn 2004). Chinese observations allowedthe identification of several η -Aquariids outbursts in 74 BCE, Fig. 2.
Time variations of the Minimum Orbit Intersection Distance(MOID) of comet 1P / Halley and Earth close to the comet’s ascendingand descending node. and AD 401, 443, 466, 530, 839, 905, 927, 934 (Zhuang 1977;Imoto and Hasegawa 1958). The comparison of the numericalintegration of the comet Halley meteoroid stream with Mayahieroglyphic inscriptions seems to indicate that this civilizationalso kept track of observed η -Aquariids outbursts (Kinsman andAsher 2017). In that work, the existence of a strong outburst inAD 461, observed but not classified as an η -Aquariid in Zhuang(1977) or Imoto and Hasegawa (1958), was also discussed.The Halleyids intensity and year-to-year variations cannotbe rigorously estimated from these ancient observations. Indeed,missing records for specific years and missing information aboutthe observing conditions limits the interpretation of the showersannual activity. However, the records are able to highlight the ex-istence of years with particularly strong activity (e.g., "hundredsof meteors scattered in all directions" in AD 585, or "hundreds oflarge and small meteors" in AD 1439, see Imoto and Hasegawa1958), with maximum dates falling a few days earlier than thecurrent shower peaks (Zhuang 1977).
3. Orionids
Despite earlier observations of Orionids meteors, the discoveryof the meteor shower is independently attributed to E. C. Her-rick and Quetelet in 1839 (Lindblad and Porubcan 1999; Kronk2014). The first precise Orionid radiant was determined by A.Herschel in 1864 and 1865 (Denning 1899; Herschel 1865),and shortly after Falb (1868) proposed a connection between1P / Halley, the η -Aquariid, and the Orionid meteor showers. Thesimilarity of the Orionids with the η -Aquariids was noticed againby Olivier in 1911.However, despite the already established connection be-tween the η -Aquariids and 1P / Halley, the link between the Ori-onids and the comet was not immediately accepted because ofthe large orbital distance of 1P / Halley at the time of the shower.The relation between the Orionids and its parent comet wasfinally accepted after decades of controversy (Obrubov 1993;Zhuang 1977; Rendtel 2008).The complexity of the Orionids activity structure was noticedat the beginning of the 20th century. Considerable but variablerates have been reported in early analyses of visual observations
Article number, page 3 of 25 & A proofs: manuscript no. Egal2020b (e.g., Prentice 1931, 1933, 1936). In 1918, Denning emphasizedthat the Orionids radiant appeared to be stationary during the en-tire duration of the shower. The stationary radiant hypothesis wasat the center of a great controversy for many years, until proofof a drift was presented by (among others) Olivier (e.g., Olivier1923, 1925) and Prentice (e.g., Prentice 1939). The existence ofmultiple subradiants within the stream, investigated by severalobservers around the same period, was not confirmed by furtherobservations (Lindblad and Porubcan 1999).The first photographic Orionid was captured in 1922 atthe Harvard Observatory (King 1923 in Lindblad and Porub-can 1999), but only few double station observations of theshower were obtained during the program (Lindblad and Porub-can 1999). In the following decades, the shower was the subjectof several visual (Stohl and Porubcan 1981), telescopic (Zno-jil 1968; Porubˇcan 1973), and radar studies (e.g., Hajduk 1982;Jones 1983; Cevolani and Hajduk 1985).Most of the Orionid data gathered over the period 1900-1967was collected and analyzed by Hajduk (1970). This work com-piled and processed visual and photographic observations of theshower, as well as an extensive set of radar data from Canadaand Czechoslovakia. Despite the heterogeneity of the observa-tional methods and instruments used, Hajduk (1970) managedto estimate the peak meteor rate and solar longitude of the max-imum activity over this time period. The derived visual meteorrates typically ranged between 10 and 30 meteors per hour, withincreased rates during some individual returns.The shower showed several maxima between L (cid:12) = L (cid:12) = Between 1944 and 1950, visual observations carried out at theSkalnaté Pleso Observatory provided the first reliable activityprofile of the shower. From this set of data, Stohl and Porubcan(1981) and Porubcan and Zvolankova (1984) identified a mainpeak of activity at L (cid:12) = L (cid:12) = L (cid:12) = L (cid:12) = L (cid:12) = L (cid:12) = L (cid:12) = L (cid:12) = η -Aquariids.Radar observations carried out by the CMOR radar between2002 and 2008 highlighted an activity lasting from October 11 toNovember 9 (from 198° to 227°, cf. Brown et al. 2010), slightlylonger than the duration reported by Porubcan et al. (1991). Themain maximum was identified around 208° of solar longitude.The maximum Orionid rate recorded by the Middle AtmosphereALOMAR Radar System (MAARSY) between 2013 and 2015was identified as being around L (cid:12) = η -Aquariidswas investigated by Dubietis (2003). Processing the standardizedobservations of the IMO Visual Meteor Data Base (VMDB), theauthor derived the population index and an average peak ZHRfor every apparition of the Orionids between 1984 and 2001. Noreproducible trend in the population index variations was iden-tified. As in previous works, the peak ZHR was found to varybetween 10 to 35 meteors per hour. Based on the annual varia-tion of the ZHR within these bounds, a 12 year periodicity wasproposed (cf. Section 6.3.2).A similar analysis was conducted by Rendtel (2008), whoprocessed visual observations dating back to 1944 along with thedata contained in the VMDB since the 1980s. As expected, thepeak ZHR and the population index of the shower were found tovary over time. An average maximum ZHR of the shower wasestimated to be about 20 to 25 between L (cid:12) = L (cid:12) = . The complex structure ofthe stream is not always discernible in each individual profile,when the low number of observations restricts the time resolu-tion of the activity profiles (Dubietis 2003). The variability inthe number of peaks, in their intensity, and solar longitude fromyear to year is nonetheless clearly noticeable.However, the quality and quantity of visual data in any givenyear are heavily influenced by the lunar phase around the Orionidmaximum. These type of data are further complicated by the of-ten poor weather at the end of October in the Northern Hemi-sphere, where the bulk of Orionid observations are performed.As a result, interannual variability of the shower is very di ffi cultto conclusively prove based on visual observations alone. The Orionids are known to have produced strong outbursts overthe past century, reaching two to four times the usual intensity ofthe shower (Lovell 1954; Hajduk 1970). The analysis of visualand radio measurements of the shower revealed increased me-teor rates around 1934-1936, 1946-1948, 1966-1968, and poten- https: // / members / imo_live_showerArticle number, page 4 of 25. Egal et al.: Activity of the Eta-Aquariid and Orionid meteor showers tially around 1927 as well (Hajduk 1970, Figure 1). A small Ori-onid outburst (ZHR >
30) was reported by European observersin 1993 (Miskotte 1993; Rendtel and Betlem 1993; Jenniskens1995). The low population index at the time of the peak (r = ff erent meteoroid population from the shower background(Rendtel 2007). The outburst was accompanied by an excep-tional number of fireballs recorded by the European Fireball Net-work (Spurný and Shrbený 2008). The following year, a maxi-mum ZHR of about 80 around L (cid:12) = The results of the International Halley Watch campaign demon-strated that, as was the case for the 1910 apparition, the passageof 1P / Halley through perihelion in 1986 yielded no rate enhance-ment of the η -Aquariids or the Orionids (Spalding 1987; Porub-can et al. 1991). In addition, the 1993 Orionids outburst occurredwhen the comet was far from its perihelion position (Jenniskens2006). The existence of Halleyids outbursts is therefore not cor-related with recent perihelion passages of the comet, a conclu-sion that is easy to understand when we consider the significantorbital distance between 1P / Halley and the Earth.The influence of Jupiter on the stream evolution, particu-larly on the spatial density and the meteoroid size distribution,was first investigated by Hajduk (1970) and McIntosh and Haj-duk (1983). After the observation of a strong outburst in 2006,the hypothesis of enhanced Orionid activity caused by mete-oroids trapped in resonant orbits was investigated by several au-thors. Among the possible Mean Motion Resonances (MMR)with Jupiter (Emel’Yanenko 2001), the 1:6 MMR was identifiedby Rendtel (2007) as the probable cause of the 2006 outburstand suspected enhanced Orionid activity between 1933 and 1938(Rendtel 2008). Consistent with that result, Sato and Watanabe(2007) determined that the 2006 outburst was caused by 1:6 res-onant meteoroids ejected between 911 BCE and 1266 BCE. Incontrast, Spurný and Shrbený (2008) suggested the source of theobserved fireballs during this apparition was most likely the 1:5MMR with Jupiter.The resonant behavior of the Orionids was investigated bySekhar and Asher (2014). By performing numerical simulations,the authors highlighted the influence of the 1:6 and 2:13 MMRwith Jupiter on recent and ancient Orionids outbursts. They iden-tified the 2:13 MMR as being responsible for the observed 1993outburst and of a possible older outburst in 1916.Results of their simulations indicated several meteor out-bursts due to particles trapped in the 1:6 resonance between AD1436 and 1440, when Orionid outbursts were reported in ancientChinese records (see Section 2.3). Following a similar approach,Kinsman and Asher (2020)’s numerical simulations pointed to-ward a strong Orionid outburst in AD 585 caused by the center of the 1:6 resonance. In addition, Sekhar and Asher (2014)’s modelalso identified this resonance as causing the observed enhancedactivity of 2006, 2007, 2008 and 2009 (Kero et al. 2011). Theauthors predicted that a future Orionid outburst in 2070 could beproduced by particles currently trapped in the 2:13 MMR withJupiter.
It is typically assumed that the mass distribution of meteor show-ers follows a power law, that is,d N ∝ M − s d M with dN the number of meteoroids of masses between M and M + d M . The exponent s is called the di ff erential mass indexand characterizes the proportion of big and small particles in ashower. Having s values strictly greater than 2 imply a streammass concentrated in small particles, when values strictly lowerthan 2 indicate the opposite. The mass index is related to thepopulation index r , the ratio of the number of meteors of magni-tude M + s (cid:39) + . r (Jenniskens 2006).Several estimates of the mass indices of the η -Aquariids andOrionids are available in the literature. Mass indices deducedfrom visual and radar observations between 1953 and 1980 werecompiled by Hughes (1987a). Estimates ranged from 1.4 to 2 forthe η -Aquariids (mass ranges of [10 − , − ]g and [10 − , − ]grespectively) and 1.85 and 2.51 for the Orionids (masses in[10 − , − , ffi cult (Hughes 1987a), the data hints that the Orionidshave a slightly higher s than the η -Aquariids.In-situ mass distribution indices measured by the Giotto,Vega-1, and Vega-2 spacecrafts when approaching 1P / Halley’snucleus are summarized in Hughes (1987a). Mass distributionindices of the dust changed with the distance between the space-crafts and the nucleus in an uncertain manner. Vega-1 and Vega-2measurements led to s estimates between 1.54 and 1.92 (massesbetween 10 − and 10 − g) with the SP-2 dust detector and be-tween 1.84 and 2.53 (masses between 10 − and 10 − g) with theDUCMA instrument. Giotto / DIDSY results spanned from 1.49to 2.03 for masses between 10 − and 10 − g. Most of the in-situ s estimates are lower than 2 (Hughes 1987a).Confronting the meteoroids mass index measured close tothe comet’s nucleus and deduced from meteor observations isdi ffi cult for several reasons. The observed mass ranges barelyoverlap, preventing the comparison of the s measured for me-teoroids of similar mass. In addition, the variability of in-situ s estimates in function of the distance to the comet prevents a clearestimate of the dust mass index close to the nucleus’ surface. Inconsequence, we can only conclude that mass index estimatesof 1P / Halley’s meteoroids of di ff erent masses and observed atdi ff erent locations before 1986 range between 1.4 and 2.5. Subsequent meteor studies showed better agreement in the massindex measured for specific years, and highlighted the variabilityof s for both showers. From visual data, Dubietis (2003) mea-sured mass indices range from 1.83 to 2.11 for the Orionids be-tween 1984 and 2001, with an average of 1.87. A particularly Article number, page 5 of 25 & A proofs: manuscript no. Egal2020b low value of 1.83 ( r = .
25) was estimated during the 1993 ap-parition of the shower.Similarly, Rendtel (2008) measured s values close to thepeak of maximum activity varying between 1.46 and 1.96 be-tween 1979 and 2006. The lowest mass index was found duringthe 2006 outburst, with an average of 1.69 and a peak value of1.46 (Rendtel 2008) or less (Trigo-Rodríguez et al. 2007). Asin Dubietis (2003), the mass index was found to present smallvariations around an average value of s = .
87 ( r = . η -Aquariids and Orionids s (Blaauw et al. 2011). Mass indices var-ied between 1.93 and 1.65, reaching their minimum value aroundthe peak of maximum activity. The comparison of s around thepeak time between 2007 and 2009 also showed a variability ofthe mass index from year to year (varying from 1.65 to 1.77).These results are consistent with Cevolani and Gabucci (1996),who measured s > . s = .
95, appropriate to meteoroid masses between10 − and 10 − kg (Schult et al. 2018).
4. Eta-Aquariids
The η -Aquariids were the first meteor shower to be linked tocomet 1P / Halley. In 1876, A. S. Herschel calculated the theo-retical radiant of meteors associated with several comets, andestimated that 1P / Halley could be responsible for a shower inearly May (Kronk 2014). In 1910, the correlation between the η -Aquariids and Halley was established by Olivier (Olivier 1912).The η -Aquariids is the third strongest annual meteor showerobservable at Earth, and one of the most active showers observ-able from the Southern Hemisphere. Visual observations of the η -Aquariids are strongly favored for Southern Hemisphere ob-servers compared to those in the north, where a higher proportionof meteor observers are located. The radiant elongation ( ≤ η -Aquariids records has lim-ited the modeling of this shower (Sekhar and Asher 2014). For-tunately, observational constraints for η -Aquariids modeling areprovided by radar measurements of the shower. In 1947, the η -Aquariids became one of the first streams to be detected usingspecular backscattering radio techniques at the Jodrell Bank Ex-perimental Station (Clegg et al. 1947). Subsequently, multiplespecular radar observations of the shower were conducted be-tween the 1950s and 1990s (e.g., Hajduk and Cevolani 1981;Hajduk and Buhagiar 1982; Hajduk and Vana 1985; Chebotarevet al. 1988). The shower was also the first one to be clearly iden-tified in head echo measurements, in particular by the interfer-ometric 49.92 MHz high-power large aperture radar at the Ji-camarca Radio Observatory (Chau and Galindo 2008). In addi-tion, the η -Aquariids is the strongest stream detected in spec-ular backscatter by the Advanced Meteor Orbit Radar (Galli-gan 2000), which has a limiting sensitivity near +
13. These ob- servations support the idea that the η -Aquariids are particularlyrich in small meteoroids (Jenniskens 2006; Campbell-Brown andBrown 2015), with masses below 10 − kg (Schult et al. 2018). Visual observations of the η -Aquariids during the 20th centuryoriginate from a limited number of sources. As with the Ori-onids, Porubcan et al. (1991) processed observations gatheredduring the International Halley Watch campaign to derive an av-erage activity profile of the shower between 1984 and 1987. Themain peak of activity was identified as a sharp double maximumat solar longitudes of 45.5° and 46.5°, with an average peak ZHRof 50. The total period of activity exceeded one month, with aFWHM of 7 to 8 days. The existence of a small dip just after themaximum was also identified, with the presence of possible sec-ondary maxima of activity. The profile is presented in AppendixA for reference.An analysis of visual observations of the η -Aquariids fromthe Southern Hemisphere (South Africa) between 1986 and1995, and in 1997 and 1998 is presented in Cooper (1996, 1997,1998). The author estimated an average peak ZHR of 60-70 be-tween L (cid:12) = L (cid:12) = L (cid:12) = L (cid:12) =
48° wasalso noticed for the 1997 η -Aquariids by Rendtel (1997). En-hanced activity (ZHR higher than 100) was reported by Cooper(1996) for the 1993 and 1995 apparitions.The comparison of the activity profiles computed by north-ern and southern observers highlight the importance of the loca-tion in the visual observation of the shower. On some occasions,both profiles show good agreement (like in 1993 or in 1997, cf.Cooper 1997; Rendtel 1997). For the other apparitions, observersfrom the Southern Hemisphere were able to provide profiles ofhigher temporal resolution and longer duration, permitting, forexample, the identification of enhanced activity in 1995 that wasmissed by northern observers. Cooper’s activity profiles are pre-sented in Appendix A. In the Appendix, Cooper’s original ac-tivity profiles (computed with a population index r of 2.3 and alimiting magnitude LM = r of 2.46 (see Section 5.2.1) with the relation: ZHR r (cid:39) ZHR r ∗ (cid:32) r r (cid:33) (6 . − LM ) . (1)Following their analysis of the Orionids, Hajduk (1973),Hajduk and Cevolani (1981), Hajduk and Buhagiar (1982)and Hajduk and Vana (1985) analyzed the variations of the η -Aquariids activity in visual and radar data. Again, the maxi-mum activity was detected as a sharp double maximum around L (cid:12) =
45° and L (cid:12) = η -Aquariids displayed considerablevariations in density along the orbit and a possible drift of themain peak’s solar longitude was observed (from 44.7° in 1971to 45° in 1975 and 47° in 1978).The long-term evolution of the η -Aquariids was also investi-gated by Dubietis (2003). As with the Orionids, the author com-puted the shower population index r and average peak ZHR fromvisual observations in the IMO VMDB between 1989 and 2001.The population index of the η -Aquariids displayed a minimumin 1992-1994, when a particularly low r value was also recordedfor the Orionids in 1993 (Rendtel and Betlem 1993). If the η -Aquariids annual ZHR seemed to present some periodic trends, Article number, page 6 of 25. Egal et al.: Activity of the Eta-Aquariid and Orionid meteor showers the reliability of these variations is reduced by the smaller num-ber of available observations. No regular periodicity in the η -Aquariids rates was clearly identified by Hajduk (1973) over theperiod 1900-1967.A detailed analysis of the shower observed by the CanadianMeteor Orbit Radar (CMOR) between 2002 and 2014 is pre-sented in Campbell-Brown and Brown (2015). The variability ofthe activity profiles is an additional indication of the existence offine structure within the stream, as already noted by Blaauw et al.(2011). The main peak was generally localized around L (cid:12) = L (cid:12) = L (cid:12) =
54° was observed in 2002 and 2006. Theanalysis also revealed the existence of two strong outbursts, in2004 and 2013. In 2004, the maximum activity occurred closeto the full moon and no visual observations can confirm or con-tradict the existence of an outburst. To our knowledge, CMORis the only source of η -Aquariids measurements during the 2004apparition. The 2013 outburst, predicted by Sato and Watanabe(2014), was successfully recorded by radio, visual, and video de-tection networks (e.g., Molau et al. 2013; Cooper 2013; Steyaert2014; Campbell-Brown and Brown 2015). Dubietis (2003) estimated a mass index of 1.78 to 1.94 from vi-sual observations of the shower between 1989 and 2001. Thelowest value of 1.78 ( r = .
18) was reached in 1992. Like theOrionids, an average mass index of 1.87 was estimated for the η -Aquariids (Dubietis 2003). However, much lower values of s were measured close to the maximum activity of the shower (forexample a s of 1.72 in 1997, cf. Rendtel 1997).The η -Aquariids recorded by the CMOR radar in 2008 dis-play a mass index varying between 2 (at the beginning and endof the activity) to 1.85 around L (cid:12) =
5. Halleyids observations between 2002 and 2019
In the following sections, we analyze the long-term activity ofthe η -Aquariids and the Orionids, focusing on the period of cov-erage of the CMOR radar (since 2002). CMOR measurementsare of particular importance between 2003 and 2010, when onlya few observations of the Orionids (and even fewer for the η -Aquariids) were published in the literature. When available,CMOR results are compared with visual observations containedin the IMO VMDB and measurements of the VMN network. Inthis section, details about the available observations and our dataprocessing are provided. The analysis of the resulting activityprofiles is presented in Section 6. Activity profiles of the η -Aquariids are available in the IMO Vi-sual Meteor Data Base (VMDB) back to 1989. Because of thedi ffi cult observing conditions from the Northern Hemisphere,several apparitions of the shower were missed or only partiallyrecorded. Observations are in particular missing in 1991, and also in 1993, 1996, 2004 and 2015 when the main activity oc-curred close to the full Moon. Because Northern observers tendto focus their attention on the estimated peak date, the total η -Aquariids duration is hardly retrievable on the sole basis of theseprofiles. Exceptions are years 2012, 2013, 2018 and 2019 whencomplete activity profiles are available on the website.Activity profiles of the Orionids are available in the VMDBgoing back to 1985, and continuously since 1989. When allowedby the Moon phase, visual observations cover the full period ofactivity of the shower (for example in 2006, 2008, 2010, 2012or 2019). These visual profiles are particularly valuable whencomparing the results of di ff erent detection networks. The IMO Video Meteor Network (hereafter called VMN) iscomprised of about 130 cameras dedicated to meteor observa-tions. The network coverage extends mainly over Europe, withadditional cameras located in the United States and Australia.The cameras are capable of recording meteors down to a limitingmagnitude of about 3.0 ± ± . The meteor detection is automatically performedby the MetRec software , which also provides flux estimates.Monthly reports of the VMN are regularly published in WGN,the Journal of the IMO.A complementary web interface allows users to visualize andanalyze the flux profile of the showers recorded by the networksince 2011 . Several filters (choice of time bin, the cameras used,radiant location, limiting magnitude, single or multiyear analy-sis, etc.) can be applied when computing the flux profile. A re-cent modification of the software o ff ers the opportunity to mod-ify the population index of a shower, which is now applied inthe flux density calculation (and not only in the transformationof the flux density into a ZHR). The CMOR radar has been providing consistent single-stationand orbital, multifrequency observations of the Halleyids since2002. The equipment consists of three independent radar sys-tems running at frequencies of 17.45 MHz, 29.85 MHz, and38.15 MHz. A detailed description of the instrument is presentedin Brown et al. (2008, 2010). In this work, the flux computationof the η -Aquariids and Orionids was performed using the 29.85MHz and 38.15 MHz data, as 17.45 MHz su ff ers significant ter-restrial interference, particularly in the early years of operation.The data processing was performed as described in Campbell-Brown and Brown (2015).It is important to note that the hardware, experimental setup(pulse repetition frequency, receiver bandwidth, pulse shape, andduration) as well as the software and detection algorithms usedby the 38 MHz CMOR system were completely unchanged since2002. The 29 MHz system underwent a transmitter power up-grade in the summer of 2009, but the receiver, antennas, andsoftware detection algorithms plus experimental setup remainedunchanged compared to the pre-2009 period.The main change in the systems over time due to hardwareaging is the transmit power output which is directly measuredand recorded for each system every 30 minutes or less and in-cluded as a correction in flux calculations. As a result, we expectthat shower profiles over this time frame can be compared and http: // / imc13 / meteoroids2013_poster.pdf https: // meteorflux.org / Article number, page 7 of 25 & A proofs: manuscript no. Egal2020b di ff erences (particularly those showing up in both systems) con-fidently associated with real flux variability. The comparison of visual, video, and radar data is challenging.Each observation method su ff ers its own biases, related to theobserving conditions (atmospheric conditions, radiant elevation,etc.), the meteoroids’ characteristics (mass, size, deceleration),and instrumental constraints. In addition, the three systems con-sidered here are not equally sensitive to the same mass range,which could prevent a reliable comparison between the systems.Indeed, di ff erences in the shower activity between systems couldsimply reflect di ff erent data processing assumptions, or the pres-ence of di ff erent meteoroid size distributions.As a result, in this study, we focus on a comparison of theglobal characteristics of the Halleyids activity profile (shape,maximum ZHR, and approximate location of the maximum ac-tivity). When possible, activity profiles are determined usinga consistent population index and a consistent time resolution.Though the recomputation of the activity profiles with consistentparameters may reduce our sensitivity to some activity variations(e.g., rebinning with a longer time interval could obscure short-term variations), it is necessary for reliable comparison betweenthe di ff erent data sets. Using radar measurements as our baseline, we assume for therest of this work a constant mass index of 1.9 ( r = .
46) forthe η -Aquariids (Campbell-Brown and Brown 2015) and 1.95( r = .
59) for the Orionids (Schult et al. 2018). These valueswere used for all flux computations and ZHR estimates derivedfrom VMN and CMOR data; no additional correction was ap-plied to these observations, except an additional normalizing fac-tor described in Section 5.2.3.In the VMDB, the choice of the population index and thetemporal resolution of the profile is not directly o ff ered to theuser. As a first approximation, we could attempt to rescale thevisual profiles to our selected population index values. However,such a transformation requires knowledge of the limiting magni-tude ( LM ) of the observation, which is not accessible and variessubstantially with moonlight conditions. In addition, ZHR esti-mates in the VMDB frequently result from an average of sev-eral interval counts, each one processed with di ff erent correctivefactors. Applying a uniform correction (assuming for example aconstant LM value) to the VMDB profiles, recorded in very dif-ferent observing conditions, is therefore not realistic. However,population indices usually applied by the IMO to compute theactivity profiles are r = . s = .
87) for the η -Aquariids and r = . s = .
92) for the Orionids, which are close to our se-lected values. Di ff erences induced by use of these mass indicesas opposed to our values, are lower than the usual uncertainty onthe ZHR computation. A constant bin of 1° in solar longitude was considered for theactivity profile computations of the VMN and CMOR data. Theselection of a rather large time bin was made to ease the com-parison of our results with previous published works (Koschackand Roggemans 1991; Dubietis 2003). Because of the limitedvisibility of the η -Aquariids radiant from CMOR’s latitude of +
43, the time resolution for CMOR profiles for this shower arenecessarily limited to half a day anyway.Much as is the case for the shower population index, the res-olution of the profiles available in the VMDB cannot be selecteddirectly by the user. A variable time bin (related to the quantityand frequency of the reported observations) is usually applied todi ff erent portions of the profile to increase the reliability of theZHR rates presented. In this work, no smoothing or interpolationof the available visual profile as a function of solar longitude wasapplied. The individual activity profiles were plotted against thevideo and the radar data without further correction. As a result,the analysis of the main peak time and intensity requires a cau-tion as is described in later sections. When we started comparing the original activity profiles derivedfor each system, an initial divergence of the ZHR-equivalent lev-els recorded was immediately evident. The main peak activitymeasured from CMOR (29 MHz) was systematically 1 to 1.5times higher than the VMN measurements, and 1 to 2.5 timeshigher than the VMDB records. These di ff erences were observedfor both showers and each year considered. They are thereforelikely issues of the limiting sensitivity being systematically un-der or over estimated, calibration di ff erences or observationalbiases, rather than caused by di ff erences in the observed mete-oroids population with mass range which would produce morerandom scatter.Since this work focuses on the relative long-term variabil-ity of the Orionids and η -Aquariids meteor showers, we decidedto scale the activity levels measured by each system to obtainsimilar ZHR rates over a long period of time ( ∼ η -Aquariids. These dif-ferent correction factors for the two showers might be due tothe reduced number of visual observations available for the η -Aquariids. CMOR 29 MHz profiles were multiplied by 0.74 forthe Orionids and 0.45 for the η -Aquariids.As mentioned in Campbell-Brown and Brown (2015), theflux calculated from the 38 MHz data is about two to three timeslower than the flux determined from the 29 MHz system (proba-bly because of uncertainties in the mass index correction factorsand / or initial radius bias). As a consequence, the Orionids pro-files computed from the 38 MHz data needed to be increased bya factor of 1.63, while no modification was required for the η -Aquariids profiles. The normalizing factors for each system aresummarized in Table 1.Each curve and ZHR estimates presented in Section 6 andAppendices B & C were scaled by the factors of Table 1, andare noted ZHR v in the text. Therefore, results and figures of thefollowing sections should not be interpreted to represent abso-lute ZHR estimates of the Halleyids meteor showers. Scaling thevideo and radar profiles to visual observations results from anarbitrary choice to make the longer time base of visual data thestandard, and not from a conviction that visual records bettermatch the real activity of the showers. The purpose of this workis instead to provide consistent measurements of the activity pro- Article number, page 8 of 25. Egal et al.: Activity of the Eta-Aquariid and Orionid meteor showers
VMDB VMN CMOR 29 MHz CMOR 38 MHz η -Aquariids 1 0.6 0.45 1Orionids 1 0.8 0.74 1.63 Table 1.
Normalization factors applied to the η -Aquariids and Orionids activity measured by the VMDB, VMN, and CMOR networks. Each ZHRestimate presented in Section 6 results from the multiplication of the original ZHR measured by each system with the coe ffi cients listed above. files (duration, shape) and year-to-year variations in activity be-tween di ff erent detection networks.
6. Results and analysis
CMOR activity profiles of the η -Aquariids and Orionids between2002 and 2019 are presented in Appendix B, Figures B.1 andB.2. Results of the 29 MHz (blue) and 38 MHz (green) sys-tems are plotted along with the VMDB profiles (in black) andthe VMN observations (beginning in 2011, in red). η -Aquariids The η -Aquariids are generally active between L (cid:12) =
35° and L (cid:12) = L (cid:12) =
40° and L (cid:12) = v estimates usually reach a maximum value of 60to 80 meteors per hour, except for two years of enhanced activity:the 2004 outburst recorded by CMOR and already reported byCampbell-Brown and Brown (2015), and the 2013 outburst pre-dicted by Sato and Watanabe (2014) and well covered by visual,video, and radar observations. The location of the main peakvaries between L (cid:12) =
44° and 47°, with the existence of severalpeaks of lower intensity after the main maximum. Most of thevisual profiles peak around 45° to 45.75° (2001, 2005, secondmaxima in 2006, 2007, 2009-2014 and 2017), with some occur-ring earlier than this time (2003, first maxima in 2006, 2015 and2016) and a few later (2002, 2008, 2018 and 2019). The chang-ing activity profiles of the shower from return to return suggeststhat there is structure in the stream, as such large changes in peaktimes are not common among the major showers (cf. Rendtel andArlt 2008).Good agreement in the duration and peak time in a given yearis found between the di ff erent techniques. VMN measurementsare comparable to the VMDB, at least when the visual obser-vations provide a complete activity profile with good resolution(like from 2017 to 2019). In some years, the di ff erence betweenthe two networks is larger (e.g., in 2011 or 2014). CMOR pro-files typically follow the shape defined by optical measurements,but present more variations in the main peak location and sug-gest the existence of subpeaks.For example, the 2012, 2013, 2017 and 2019 apparitions arevery consistent between the networks, but the radar main peaklocation diverges from the optical data in 2005, 2015, 2016 and2018. Profiles obtained using the 38 MHz and 29 MHz data alsodi ff er slightly from one another, but there is no evidence thatone frequency better reproduces the optical observations for ev-ery apparition of the shower. Sudden gaps in the radar profilesindicate periods for which the records are missing (because ofinstrumental issues or lack of reliable measurements).Unfortunately, radar observations were not available aroundthe estimated peak time in 2002, 2003, 2006 and potentially2010. In addition, the 2009 radar profile largely diverges fromthe VMDB observations around the peak time, in this instancebecause of a large scale equipment change to the 29 MHz CMOR system which occurred in this time frame. As a result, only the38 MHz data is available to characterize this specific apparition. In our data, the Orionids display noticeable activity between L (cid:12) = L (cid:12) = v rates vary between 20and 40 meteors per hour up to 2003 and after 2012, and be-tween 40 and 80 around the 2006 and 2007 resonant return years.Years of enhanced activity tend to present a sharper and less scat-tered profile than apparitions of moderate intensity. As noticedby many observers (cf. Section 3), the Orionids present a broadmaximum between 206° and 211°, with several subpeaks of vari-able intensity and location. The main peak of activity is thereforedi ffi cult to assess for several Orionids apparitions. On some oc-casions, the highest ZHR v rates do not coincide with the centerof the broad maximum (e.g., in 2018, 2019) or the presence ofmultiple subpeaks prevents the identification of the main peak(e.g., in 2009 or 2017). In 2017 for example, the visual profiledisplays two maxima of equivalent strength, clearly separated bya dip already observed in the past (see Section 3).For the Orionids, the observations by di ff erent detectionmethods are less consistent than for the η -Aquariids. This is inpart a consequence of the lower number statistics for the Orion-ids compared to the η -Aquariids. From figure B.2, it is clear thatthe shower does not show a stable activity profile from returnto return. The overall duration and location of the broad max-imum of activity are similar between the systems, but the finestructure of the profiles di ff ers. Clearly, the lower activity levelof this shower increases the apparent discrepancies between ourdi ff erent data sets due to small number statistics.Relatively good agreement is found between the VMN andVMDB profiles in 2014, 2017, and 2019, but not for earlier ap-paritions (e.g., 2011, 2012). Observations from the 38 MHz sys-tem appears to be particularly sensitive to the low-level activityof the shower, leading to very broad profiles. CMOR activityprofiles are similar to visual observations in 2005, 2007, 2008,and 2010. A lack of observations around the peak in 2006 doesnot allow us to clearly define the main peak location and inten-sity, but high ZHR v estimates ( >
70) were recorded at L (cid:12) = L (cid:12) = ff er for otherapparitions of the shower, especially in 2016 and 2018. In 2018,an early peak of activity is noticeable in the VMDB data (ZHR v of 54 ± L (cid:12) = Figure 3 presents the average activity profiles of the η -Aquariidsand Orionids recorded by CMOR over the periods 2002-2007,2008-2013, and 2014-2019. The average profiles display sig-nificant variations as a function of the period considered. EarlyCMOR measurements match published estimates of η -Aquariids Article number, page 9 of 25 & A proofs: manuscript no. Egal2020b
Fig. 3.
Individual and average activity profiles (ZHR v ) of the η -Aquariids (top) and Orionids (bottom) recorded by the radar CMOR between 2002and 2007 (left), 2008 and 2013 (center), and 2014 and 2019 (right). activity (see Section 4). The average η -Aquariids profiles ofboth frequencies from 2002-2007 show an initial maximum at L (cid:12) = L (cid:12) = L (cid:12) = L (cid:12) = L (cid:12) = L (cid:12) = L (cid:12) = η -Aquariids and Orionids are presentedin Figure 4 and 5. The average profile computed from the VMNdata (in red) is also shown for comparison. The average VMDBprofile is shown for the Orionids but because of the limited η -Aquariids data available on the VMDB website, Figure 4 insteadincludes the average IMO observations from 1988 to 2007 deter-mined by Rendtel and Arlt (2008).Averaging the showers’ activity over long time periodssmoothes the year-to-year profiles presented in Appendices Aand B. This is equivalent to applying a low-pass filter to the timeseries, removing most of the information about the fine struc-ture in the stream. The average η -Aquariids profiles measuredby the di ff erent systems show remarkable consistency. The mainpeaks are located at similar solar longitudes and the profiles dis-play similar rising and falling slopes of activity. The consistency Fig. 4.
Average activity (ZHR v ) of the Eta-Aquariid meteor shower, asmeasured with the CMOR 29 MHz system (blue) and the 38 MHz sys-tem (green) between 2002 and 2019, the VMN (red) between 2011 and2019, and visual observations from the IMO from 1988 to 2007 (Rend-tel and Arlt 2008, in black). between measurements of di ff erent systems supports the ideathat there is no significant size sorting of the particles within Article number, page 10 of 25. Egal et al.: Activity of the Eta-Aquariid and Orionid meteor showers
Fig. 5.
Average activity (ZHR v ) of the Orionid meteor shower, as mea-sured with the CMOR 29 MHz system (blue) and the 38 MHz sys-tem (green) between 2002 and 2019, the VMN (red) between 2011 and2019, and visual observations from the VMDB between 2002 and 2019(black). Di ff erences in the 38 MHz profile are increased by the lowernumber of available observations. More information is given in the text. the η -Aquariids, in agreement with earlier studies by Jenniskens(2006) or Campbell-Brown and Brown (2015).The average Orionids profiles are also broadly similar be-tween the detection systems. The CMOR and VMDB de-termined average profiles display a double peaked maximumaround L (cid:12) = L (cid:12) = L (cid:12) = L (cid:12) = ff ers noticeably from these others with lower activity mea-sured between 203° to 206° in solar longitude. Low ZHR v ratesbetween L (cid:12) = L (cid:12) = v gap at L (cid:12) = ff erence.It could be due to an increase in the shower mass index in thisinterval, which would attenuate detection numbers through botha larger initial radius e ff ect and a smaller number of echoes de-tectable by 38 MHz. It may also be a function of the lower sensi-tivity of 38 MHz to the Orionids compared to the sporadic back-ground in general, an e ff ect noticeable in the larger activity scat-ter between L (cid:12) = L (cid:12) = v rates decrease more rapidly invisual and radar data than in the video records. However, it is un-clear if these di ff erences are related to the stream characteristicsor to biases in the shower observations and the data processing. Following Jenniskens (1994), the shape of a meteor showerZHR profile can be approximated by a double-exponential curve,which can be expressed as:
ZHR = ZHR m ∗ B ( L (cid:12) − L (cid:12) m ) B = + B p if L (cid:12) ≤ L (cid:12) m B = − B m if L (cid:12) > L (cid:12) m , (2)where the maximum ZHR ( ZHR m ) at solar longitude L (cid:12) m and theslope coe ffi cients B p and B m are fit to the observed profile.With the presence of a significant plateau in the Orionids ac-tivity profile, we found it preferable to replace L (cid:12) m with two solarlongitudes L (cid:12) m , and L (cid:12) m , delimiting the plateau location. Withthis adaptation, the coe ffi cients B p and B m characterize theslope of the ascending ( L (cid:12) ≤ L (cid:12) m , ) and descending ( L (cid:12) > L (cid:12) m , )branch of the activity profile. The plateau region is modeled bya linear function ZHR = α x + β , where α and β are determinedfrom the estimates of { L (cid:12) m , , ZHR ( L (cid:12) m , ) } and { L (cid:12) m , , ZHR ( L (cid:12) m , ) } to ensure the continuity of the modeled ZHR profile.A standard least-squares fit of each average activity of Fig-ures 4 and 5 was performed to determine the slope parameters B p , B m , B p and B m . Results with the associated formal uncer-tainties of the fit are summarized in Table 2.As mentioned in Jenniskens (2006) and Campbell-Brownand Brown (2015), the η -Aquariids rise of activity is more sud-den than the post-maximum decrease in intensity. Depending onthe detection system, we find the slope coe ffi cient of the ascend-ing branch to vary between 0.104 and 0.180, and for the descend-ing branch to change between 0.065 to 0.108. These estimatesare consistent with Campbell-Brown and Brown (2015) results(cf. [1] in Table 2).As expected, the Orionids display a much more symmetricactivity profile, especially in the radar measurements. Our esti-mates of the ascending branch slope for each system vary be-tween 0.073 and 0.349, for a descending coe ffi cient of 0.054 to0.233. These value span Jenniskens (1994)’s result derived fromvisual observations carried out in the Northern hemisphere, andare a bit lower than the values deduced from Southern Hemi-sphere data (see [2] & [3] in Table 2). Following Dubietis (2003) and Rendtel (2008), we investigate inthis section the long-term evolution of the Halleyids main peak.The variation of the showers’ strength and structure (peak lo-cation) since 1985 is analyzed. We remind the reader that theresults presented in this section reflect the general behavior ofthe showers’ activity (ZHR v ), and not absolute measurements ofthe apparitions’ ZHRs. Assuming a constant population indexwith time influences the general shape of the profiles computed,as well as the main peak intensity and location. In addition, theselection of a large solar longitude binning ( L (cid:12) = v ) rates andcan modify the location of the maximum intensity for each ap-parition of the showers. Figures 6 and 7 illustrate the evolution of the magnitude of the η -Aquariids and Orionids main peak of activity per year. Maxi-mum (ZHR v ) rates measured by the VMN (red), CMOR 29 MHz Article number, page 11 of 25 & A proofs: manuscript no. Egal2020b η -AquariidsSystem B p B m B p B m ZHR m ZHR m L (cid:12) m , (°) L (cid:12) m , (°)29 MHz 0.111 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± B p B m B p B m ZHR m ZHR m L (cid:12) m , (°) L (cid:12) m , (°)29 MHz 0.078 ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± Table 2.
Fit coe ffi cients of the rising and falling slopes of η -Aquariids and Orionids average profiles presented in Figures 4 and 5. Each estimate isfollowed by the fit formal uncertainty. The parameters B p and B m are adaptations of Equation 2’s B p and B m coe ffi cients for profiles containinga dip or a plateau. The modeled maximum ZHR ( ZHR m , ZHR m ) and the solar longitude endpoints of the plateau ( L (cid:12) m , , L (cid:12) m , ) are provided forinformation. [1]: η -Aquariids coe ffi cients provided by Campbell-Brown and Brown (2015). [2,3]: Orionids coe ffi cients determined by Jenniskens(1994) from Northern [2] and Southern [3] Hemisphere observations. Fig. 6.
Annual variation of the η -Aquariids main peak (ZHR v ) between 1985 and 2019. The maximum rates recorded by CMOR 29 MHz (blue)and 38 MHz (green) are plotted along with the results of the VMN system (in red). Gray, black, and empty symbols refer to the maximum ZHR v deduced from visual observations. The new analysis of the IMO VMDB observations since 2001 is characterized by circles of di ff erent colorsdepending on the reliability of the peak observations (filled: reliable estimate, empty: uncertain value, empty and dashed: very uncertain estimate).The gray filled curve represents the approximate ZHR v maximum and uncertainty from the directly available online VMDB profiles (with norigorous processing of the data). Additional measurements (especially before 2001) refer to previous publications. Horizontal error bars imply anaverage estimate of the maximum (ZHR v ) over the indicated period. The bottom symbols represent the phase of the moon around the estimatedmain peak date. More information is given in the text. (blue), and CMOR 38 MHz (green) are plotted along with the re-sult of visual observations (black, gray, and empty symbols). Though the visual activity profiles provided by the IMOVMDB user interface could be used without additional process-ing in sections 6.1 and 6.2, the estimate of the main peak inten- Article number, page 12 of 25. Egal et al.: Activity of the Eta-Aquariid and Orionid meteor showers
Fig. 7.
Annual variation of the Orionids main peak (ZHR v ) between 1985 and 2019. Same notations are used as in Figure 6. sity for each apparition of the shower requires a more cautiousanalysis of the data. One of us (JR) re-processed the observationsavailable in the IMO VMDB with our preferred constant studypopulation indices of 2.46 for the η -Aquariids and 2.59 for theOrionids.The time binning of several parts of the profile was modi-fied to increase the reliability of the main peak location, makingsure that the related ZHR estimate is not based on too few con-tributing intervals. In this re-analysis, the probable main peakintensity, location, and profile FWHM for every apparition ofthe η -Aquariids and Orionids since 2001 was determined.The results are plotted as circles in Figures 6 and 7. The col-ors of the symbols reflect the reliability of the derived maximumZHR v (filled: reliable estimate, empty: uncertain value, emptyand dashed: very uncertain estimate), depending on the structureof the profile, and the number and quality of the available ob-servations. Since these measurements are not directly retrievablefrom the VMDB website, numerical estimates are provided inAppendix D.For comparison, the interval of maximum intensity deducedfrom the VMDB profiles of Figures B.1 and B.2 is representedby the gray filled curve in Figures 6 and 7. The identificationof the main peak was performed using a combination of the fullactivity profile and using observations conducted close to the es-timated peak time ("Peak" tab in the live ZHR website). Themaximum peak ZHR v was retrieved without any smoothing, fit-ting or extrapolation of the original data. When no observationsare available around the supposed maximum of activity, no ap-proximated ZHR v was computed; this is the reason for the largegaps for some years in the gray curve.Additional measurements of the showers’ activity (espe-cially before 2001) published by several authors (Spalding 1987;Koseki 1988; Porubcan et al. 1991; Koschack and Roggemans1991; Rendtel and Betlem 1993; Cooper 1996, 1997; Rendtel 1997; Cooper 1998; Dubietis 2003; Rendtel 2008; Arlt et al.2008; Cooper 2013) were also added in Figures 6 and 7 (filledor empty triangles, squares, etc.) for comparison.From Figure 6, we notice that the annual activity variationsof the η -Aquariids determined by the di ff erent detection net-works is remarkably similar. Results from the 29 MHz systemare in better agreement with optical measurements than for 38MHz (especially in 2008). However, data obtained with both fre-quencies show a similar evolution with time, following the gen-eral trend found in the optical data. Since 1997, annual ZHR v maximum rates vary between an average of 65-70 meteors perhour, with the notable exception of two outbursts in 2004 and2013 (ZHR v ≥ v rates tend to have been observed in the past.This may be due to a real enhancement of the shower aver-age level activity or simply reflect changes in coverage betweenNorthern and Southern observations. No clear periodicity in out-burst years is apparent from this figure. A potential 12 year pe-riodicity of the shower minimum of activity could be imaginedif years 1990-1991, 2002-2003, and 2014-2015 corresponded tothe lowest η -Aquariids rates measured since 1985. However, wefind no conclusive evidence of such periodicity from the obser-vations.The Orionids annual peak activity evolution per year alsoshows relatively good agreement between the di ff erent tech-niques (cf. Figure 7). Before 1990, published peak ZHR ratesare close to 20 meteors per hour on average. A small systematicincrease in the peak of the average activity level is noticeableover the period 1990 to 2001, reaching rates of about 40 mete-ors per hour in 1993 and 1997. A potential minimum of activityis observed in 2002, but is based on uncertain visual observa-tions under poor lunar conditions and a very broad profile fromthe CMOR 38 MHz system. All techniques agree in an enhance- Article number, page 13 of 25 & A proofs: manuscript no. Egal2020b ment of Orionids activity between 2002 and 2013, with the high-est rates of more than 70 meteors per hour reached for the 2006and 2007 resonant years. The existence of an additional increaseof activity in 2016 is observed in the VMN and CMOR data, butvisual observations of the VMDB neither support or contradictthis conclusion.The greatest di ff erence between the four data sets relates tothe most recent apparition of the shower (2019). In this year,CMOR 29 MHz rates were systematically about twice that foundfrom the optical techniques. The variations in the 38 MHz dataare less reliable than the 29 MHz profile for the Orionids becauseof the lower sensitivity of that system and its correspondinglypoorer noise statistics.The existence of a periodicity to Orionid activity, caused byJovian perturbations of the meteoroid stream, has been proposedby several authors (e.g., McIntosh and Hajduk 1983; Dubietis2003; Rendtel 2008; Sekhar and Asher 2014). Visual and radarmeasurements suggested a periodic return of low meteor rates,raising the question of a 12 year variation of the Orionid mini-mum of activity (McIntosh and Hajduk 1983). No periodic vari-ation of the apparition of enhanced meteor rates has been clearlyidentified for the shower. However, increased meteor activityaround 1984-1985 and 1993-1998 led Dubietis (2003) to pro-pose that a similar periodicity is linked to years of maximumOrionid activity.The variance in the magnitude of peak activity between thethree techniques for recent apparitions of the shower precludeany strong statement about the 12 year periodicity in the Orion-ids maximum meteor rates that seems to be supported by Figure7. In particular, when examining years of low peak intensity, wefind a clear minimum in 1990, also reported by Koschack andRoggemans (1991), and another potential minimum in activityin 2002. A 12 year periodicity would lead to a subsequent mini-mum in 2014, which is clearly observed in the VMN and VMDBdata sets, and to a lesser extent in CMOR measurements. The ex-istence of a periodicity in Orionid peak activity is therefore notrejected by our analysis. However, because of the uncertainty ofthe 2002 minimum, no strong conclusion can be drawn - addi-tional observations of several Orionid returns covering anotherfull 12 year cycle are required. Trends in the location of the Orionids and η -Aquariids activitypeaks were investigated by Hajduk (1970) and Hajduk (1973).From his analysis of visual and radar observations, Hajduk(1970) identified a clear variability of the Orionids peak loca-tion for each return of the shower, including a shift by as muchas 5° in solar longitude in some years. When long observationalseries were obtained with the radars (early visual observationswere considered to be doubtful), the author noticed a gradualdisplacement of the activity peaks during successive apparitions.If a similar trend was suspected for the η -Aquariids (cf. Section4), no evidence of it was found in the observed changes in theshower’s main peak of activity (Hajduk 1973).Since the time of those earlier works, the location of theOrionids and η -Aquariids main peak of activity has been re-examined by several authors (e.g., McIntosh and Hajduk 1983;Hajduk 1980; Hajduk et al. 1984; Cevolani and Hajduk 1985,1987; Koschack and Roggemans 1991). Covering periods of 5 to10 years, displacements in solar longitude of 0.3 to 0.75 degreesper year and 0.42 to 0.92 degrees per year have been claimed forthe η -Aquariids and Orionids respectively (McIntosh and Haj-duk 1983). As noted earlier, the identification of the showers’ main peakof activity is highly dependent on the time resolution of the ac-tivity profiles, the selection of the population index (that mayvary with time), and the level of continuity in the monitoring ofthe showers’ activity. As a consequence of these factors, in thiswork we choose to examine the evolution of maximum activ-ity regions (sometimes comprising several peaks) instead of themain peak location as these are likely to produce more robustresults.To aid in the visualization of the location of the solar longi-tude L (cid:12) , max of maximum activity over time, the individual activ-ity profiles of Appendix B were converted into the intensity mapsas shown in Figures 8, 9, and C.1. In these figures, the ZHR v ofeach apparition of the shower (rows) as a function of the solarlongitude (columns) is colored as a function of the shower inten-sity (yellow: maximum meteor rates recorded, black: low meteorrates or no measurements available). To examine the variabilityin the activity peak locations, the profiles were normalized bythe maximum ZHR v recorded for each year of the shower (com-puted in Section 6.3.1 in these figures, but the reader is remindedof the peak ZHR values by a black curve adjacent to the map).Since CMOR records provide the longest consistent obser-vational set for the Halleyids in our study, we consider the mea-surements of the 29 MHz system as a reference for the rest of thissection. Despite the ability of the 38 MHz data to reproduce theabsolute intensity variations of the showers (cf. Section 6.3.1),the 29 MHz system was selected because of the lower statisticalnoise in the Orionids profile.Figure 8 highlights the evolution in L (cid:12) of the peak activ-ity regions recorded by CMOR for the Orionids (left panel)and η -Aquariids (right panel). The maximum activity of the η -Aquariids shows a moderate variation between L (cid:12) =
44° and L (cid:12) = L (cid:12) = − . The resultant fre-quency fit corresponds to a period of 11.88 years, very close toJupiter’s orbital period of 11.86 years.Figure 9 compares the resulting sinusoidal curve to the Ori-onids intensity maps as measured by CMOR (left) and theVMDB (right). For clarity, the activity maps of Figure 9 werenormalized by the highest ZHR measured during the entire pe-riod of observation for the technique considered.In the VMDB map, global periods of enhanced activity (e.g.,around 1998 and 2007) match with times that the sine modelis at lowest solar longitudes, while higher modeled solar longi-tudes are close to the reported minima of activity. On the otherhand, the correlation between the yearly maximum activity lo-cation and the sinusoidal fit is less clear in these observations.However, because of the low signal-to-noise ratio for the Ori-onids and the presence of multiple gaps in the VMDB profiles,no robust conclusions can be drawn from the intensity map ofFigure 9.To investigate the veracity of a periodic displacement inshower activity in more detail, we present in Figure 9 (bottompanel) a comparison of our sinusoidal model (gray curve) with Article number, page 14 of 25. Egal et al.: Activity of the Eta-Aquariid and Orionid meteor showers
Fig. 8.
Intensity maps of the normalized Orionids (left) and η -Aquariids (right) activity as measured by the CMOR 29 MHz system since 2002. Foreach apparition of the shower, the activity profiles of Figures B.2 and B.1 are normalized by the maximum ZHR v value determined in the previoussection. This is represented by the black curve to the right of each map. Timing of maximum activity are highlighted by light colors (yellow toorange), while low meteor rates (or the absence of measurements) are represented by dark colors (purple to black). reported past maxima locations of the shower (symbols). The es-timated maximum peak location and FWHM of the profile deter-mined from the rigorous analysis of the Orionids visual data inSection 6.3.1 are represented by circles (empty or filled) and ver-tical lines in the figure. Additional published observations since1985 were added to the visual data set and summarized in TableD.2.Between 1960 and 1970, our model is compared to the solarlongitude of the maximum activity recorded by radar observa-tions as reported by Hajduk (1970). Results of the simultaneousmeasurements of the Budrio and Ondrejov radars, discussed inHajduk et al. (1984) and Cevolani and Hajduk (1985), are pre-sented as inverted triangles over the period 1976-1982. Comple-mentary visual observations, listed by Koschack and Roggemans(1991), were added to Figure 9 when a single maximum of ac-tivity (or double-peak maximum) could be identified from theobservations.From this figure, it is clear that the sinusoidal fit does not re-produce all the peak locations reported for the Orionids since1960. However, the modeled curve falls within the estimatedFWHM ranges from visual observations (except for the 1993 and2014 apparitions). This could be interpreted as potentially indi-cating the location of secondary maxima not always capturedin the existing data. The ∼
12 year periodicity determined fromCMOR measurements reproduces very well Hajduk (1970) radarobservations between 1961 and 1967, but deviates more fromthe estimates of Cevolani and Hajduk (1985) and Koschack andRoggemans (1991) made between 1975 and 1983. A secondaryfit solution, leading to a period of around 11.1 years, o ff ers acorrect match of CMOR and the 1975-1983 radar observations,but does not reproduce the 1961-1967 results. Without accurateestimates of the shower FWHM and measurement uncertainties before 1985, the agreement between the sinusoidal model andthese older radar observations cannot be clearly established.From earlier studies and recent visual observations of theshower, there is no evidence of a ∼
12 year periodicity of the Ori-onids main peak location (right hand side of the bottom panelof Figure 9). However, the variations presented in Figure 9 donot exclude the existence of such periodicity either, especiallywhen we consider the di ffi culty of identifying the main peak ofactivity. The period of the modeled fit together with the fact thatsuch a trend is most noticeable in the longest and most consis-tent set of observations so far (CMOR), lends support to the ex-istence of periodic oscillations over many years in the timing ofthe Orionids maximum activity. We therefore suggest that ourmeasurements point toward a cyclical variation of the locationof the Orionids most active period, but emphasize that this trendneeds to be confirmed by future radar and optical observations.
7. Conclusions
In this study, we have measured long-term trends in the activ-ity of the η -Aquariids and Orionids meteor showers as observedwith visual, video, and radar techniques. Results from the IMOVMDB and CMOR databases were compared for each annualapparition of the showers since 2002, along with VMN observa-tions since 2011. Despite the di ff erent biases inherent to each de-tection method, observations from the three data sets show goodagreement in the general shape, activity level, and annual inten-sity variations of both showers. This consistency among systemssensitive to di ff erent size meteoroids suggests that there is no sig-nificant size sorting of the particles within the meteoroid stream.The analysis of the Halleyids activity since 2002 is generallyconsistent with previous published observations of the showers. Article number, page 15 of 25 & A proofs: manuscript no. Egal2020b
Fig. 9.
Top: Intensity maps of the Orionids as measured by CMOR 29 MHz (left) and from visual observations (right). Here individual profileswere normalized by the maximum ZHR v ever recorded by the network, and not by the maximum ZHR v of a particular apparition. The modeledsinusoidal curve, best fit of Figure 8, is presented in cyan. Bottom plot: comparison of the same modeled solar longitude variation (gray) asshown in the top plot, but now compared to the visual observations of Table D.2 (empty and filled circles) and including radar observations ofHajduk (1970) (triangles, since 1960), Hajduk et al. (1984) and Cevolani and Hajduk (1985) (inverse trangles), and complementary observationssummarized in Koschack and Roggemans (1991) (squares, before 1985). The main characteristics of the η -Aquariids and Orionids as de-rived from our analysis are:1. The η -Aquariids are generally active between L (cid:12) =
35° and L (cid:12) = L (cid:12) = L (cid:12) = v rates vary around an average of 65to 70 meteors per hour for the η -Aquariids, and between 20to 40 meteors per hour for the Orionids.5. Several outbursts, caused by meteoroids trapped in resonantorbits with Jupiter, were observed for the η -Aquariids (in2004 and 2013) and the Orionids (in 2006-2007). ZHR max-imum rates then reached two to four times the usual activitylevel of the showers.6. The average profile of the η -Aquariids is asymmetric, with arise of activity more sudden than the decreasing activity. The profile of the Orionids is more symmetric around the broadmaximum of activity. The general shape of the average activ-ity profiles can di ff er as a function of the period considered.No clear periodicity in the annual activity level or the mainpeak location of the η -Aquariids can be inferred from our anal-ysis. Consistent radar observations of the Orionids since 2002support the existence of a periodic displacement in the loca-tion of the solar longitude of the shower’s peak. A period of ∼ ∆ L (cid:12) of about0.53°.yr − was estimated from CMOR 29 MHz data. The ex-istence of such periodicity cannot be established from existingvisual observations, and needs to be confirmed by future radarand optical observations of the shower. Acknowledgements.
This work was supported in part by NASA Meteoroid En-vironment O ffi ce under cooperative agreement 80NSSC18M0046 and contract80MSFC18C0011, by the Natural Sciences and Engineering Research Councilof Canada (Grants no. RGPIN-2016-04433 & RGPIN-2018-05659), and by theCanada Research Chairs Program. We are thankful to the referee for his carefulreview that helped improve the manuscript. Article number, page 16 of 25. Egal et al.: Activity of the Eta-Aquariid and Orionid meteor showers
References
Ahn, S.H., 2003. Meteors and showers a millennium ago. MNRAS 343, 1095–1100. arXiv:astro-ph/0304543 .Ahn, S.H., 2004. A Catalogue of Meteor Showers and Storms in Korean,Japanese, Chinese Histories. Journal of Astronomy and Space Sciences 21,529–552.Arlt, R., Rendtel, J., Bader, P., 2008. The 2007 Orionids from visual observa-tions. WGN, Journal of the International Meteor Organization 36, 55–60.Babadzhanov, P.B., Chebotarev, R.P., Hajduk, A., 1977. Simultaneous radarmeteor observations at Ondrejov and Dushanbe. Bulletin of the AstronomicalInstitutes of Czechoslovakia 28, 286–288.Babadzhanov, P.B., Chebotarev, R.P., Hajduk, A., 1979. Long-base observa-tions of the Orionid meteor shower. Bulletin of the Astronomical Institutesof Czechoslovakia 30, 225–227.Blaauw, R.C., Campbell-Brown, M.D., Weryk, R.J., 2011. A meteoroid streamsurvey using the Canadian Meteor Orbit Radar - III. Mass distribution indicesof six major meteor showers. MNRAS 414, 3322–3329.Brown, P., Weryk, R.J., Wong, D.K., Jones, J., 2008. The Canadian Meteor OrbitRadar Meteor Stream Catalogue. Earth Moon and Planets 102, 209–219.Brown, P., Wong, D.K., Weryk, R.J., Wiegert, P., 2010. A meteoroid streamsurvey using the Canadian Meteor Orbit Radar. II: Identification of minorshowers using a 3D wavelet transform. Icarus 207, 66–81.Campbell-Brown, M., Brown, P.G., 2015. A 13-year radar study of the η -Aquariid meteor shower. MNRAS 446, 3669–3675.Cevolani, G., Gabucci, M.F., 1996. Mass distribution of meteoroids obtained bya meteor forward-scatter (MFS) radar method. Nuovo Cimento C GeophysicsSpace Physics C 19C, 271–283.Cevolani, G., Hajduk, A., 1985. Radar observations of Orionids at Budrioand Ondrejov in 1981 and 1982. Bulletin of the Astronomical Institutes ofCzechoslovakia 36, 278–281.Cevolani, G., Hajduk, A., 1987. Activity of the meteoric complex of comet Hal-ley. Publications of the Astronomical Institute of the Czechoslovak Academyof Sciences 2, 179–182.Chau, J.L., Galindo, F., 2008. First definitive observations of meteor showerparticles using a high-power large-aperture radar. Icarus 194, 23–29.Chebotarev, R.P., Isamutdinov, S.O., Hajduk, A., 1988. Radar Observations ofEta Aquarids in 1981-1986 at Dushanbe and Ondˇrejov. Bulletin of the Astro-nomical Institutes of Czechoslovakia 39, 82.Clegg, J.A., Hughes, V.A., Lovell, A.C.B., 1947. The Daylight Meteor Streamsof 1947 May-August. MNRAS 107, 369.Cooper, T., 1996. A Decade of Visual eta-Aquarid Meteor Obsevations. WGN,Journal of the International Meteor Organization 24, 157–161.Cooper, T.P., 1997. The 1997 eta Aquarid Meteor Shower. Monthly Notes of theAstronomical Society of South Africa 56, 56.Cooper, T.P., 1998. Observations of eta Aquarids. Monthly Notes of the Astro-nomical Society of South Africa 57, 62.Cooper, T.P., 2013. The 2013 Eta Aquariids from Bredell. Monthly Notes of theAstronomical Society of South Africa 72, 183–189.Denning, W.F., 1899. General Catalogue of the Radiant Points of MeteoricShowers and of Fireballs and Shooting Stars observed at more than one Sta-tion. MmRAS 53, 203.Dubietis, A., 2003. Long-term activity of meteor showers from Comet 1P / Halley.WGN, Journal of the International Meteor Organization 31, 43–48.Edberg, S.J., Newburn, R.L., Rahe, J., 1988. The international halley watch,in: Grewing, M., Praderie, F., Reinhard, R. (Eds.), Exploration of Halley’sComet, Springer Berlin Heidelberg, Berlin, Heidelberg. pp. 981–984.Emel’Yanenko, V.V., 2001. Resonance structure of meteoroid streams, in: Warm-bein, B. (Ed.), Meteoroids 2001 Conference, pp. 43–45.Everhart, E., 1985. An e ffi cient integrator that uses Gauss-Radau spacings, in:Carusi, A., Valsecchi, G.B. (Eds.), Dynamics of Comets: Their Origin andEvolution, Kluwer, Dordrecht. pp. 185–202.Falb, R., 1868. Der Comet Halley und seine Meteoriten. Von Herrn R. Falb.Astronomische Nachrichten 72, 361.Galligan, D.P., 2000. Structural Analysis of Radar Meteoroid Data. Doctoral.Canterbury.Hajduk, A., 1970. Structure of the meteor stream associated with comet Halley.Bulletin of the Astronomical Institutes of Czechoslovakia 21, 37.Hajduk, A., 1973. The Structure of the Eta Aquarid Meteor Stream. Bulletin ofthe Astronomical Institutes of Czechoslovakia 24, 9.Hajduk, A., 1980. The core of the meteor stream associated with Comet Halley,in: Halliday, I., McIntosh, B.A. (Eds.), Solid Particles in the Solar System,pp. 149–152.Hajduk, A., 1982. The Total Mass and Structure of the Meteor Stream Associatedwith Comet Halley. Sun and Planetary System, Fricke, W. and Teleki, G. 96,335.Hajduk, A., Buhagiar, M., 1982. Southern and Northern Hemisphere Obser-vations of the Eta Aquarid Meteor Shower in 1969-1978. Bulletin of theAstronomical Institutes of Czechoslovakia 33, 262. Hajduk, A., Cevolani, G., 1981. Simultaneous radar meteor observations at On-drejov and Budrio. Bulletin of the Astronomical Institutes of Czechoslovakia32, 304–310.Hajduk, A., Cevolani, G., Formiggini, C., Babadzhanov, P.B., Chebotarev, R.P.,1984. Orionid meteor shower activity in 1979 from long-base radar observa-tions. Bulletin of the Astronomical Institutes of Czechoslovakia 35, 1–5.Hajduk, A., Hajdukova, M., Cevolani, G., Formiggini, C., 1987. Activity ofOrionids in 1983-1985 from simultaneous radar observations. Bulletin of theAstronomical Institutes of Czechoslovakia 38, 129–131.Hajduk, A., Vana, J., 1985. Eta-Aquarids 1969-1977 from Ondrejov MeteorRadar Records. Contributions of the Astronomical Observatory SkalnatePleso 13, 61.Herschel, A.S., 1865. Radiant Points of Shooting-Stars. MNRAS 26, 51.Hughes, D.W., 1987a. P / Halley dust characteristics - A comparison betweenOrionid and Eta Aquarid meteor observations and those from the flyby space-craft. A&A 187, 879–888.Hughes, D.W., 1987b. The History of Halley’s Comet. Philosophical Transac-tions of the Royal Society of London Series A 323, 349–366.Imoto, S., Hasegawa, I., 1958. Historical Records of Meteor Showers in China,Korea, and Japan. Smithsonian Contributions to Astrophysics 2, 131.Jenniskens, P., 1994. Meteor stream activity I. The annual streams. A&A 287,990–1013.Jenniskens, P., 1995. Meteor stream activity. II. Meteor outbursts. A&A 295,206–235.Jenniskens, P., 2006. Meteor Showers and their Parent Comets.Jones, J., 1983. Radar observations of the Orionid meteor shower. MNRAS 204,765–776.Kennedy, J., Eberhart, R., 1995. Particle swarm optimization. Proceedings ofIEEE International Conference on Neural Networks , 1942–1948.Kero, J., Szasz, C., Nakamura, T., Meisel, D.D., Ueda, M., Fujiwara, Y., Tera-sawa, T., Miyamoto, H., Nishimura, K., 2011. First results from the 2009-2010 MU radar head echo observation programme for sporadic and showermeteors: the Orionids 2009. MNRAS 416, 2550–2559.Kiang, T., 1972. The past orbit of Halley’s Comet. MmRAS 76, 27.Kinsman, J.H., Asher, D.J., 2017. Evidence of Eta Aquariid outburstsrecorded in the classic Maya hieroglyphic script using orbital integrations.Planet. Space Sci. 144, 112–125. arXiv:1707.08246 .Kinsman, J.H., Asher, D.J., 2020. Orbital dynamics of highly probable but rareOrionid outbursts possibly observed by the ancient Maya. MNRAS 493, 551–558. arXiv:2002.00106 .Koschack, R., Roggemans, P., 1991. The 1990 Orionids. WGN, Journal of theInternational Meteor Organization 19, 115–130.Koseki, M., 1988. Japanese η -Aquarids observations in 1987. WGN, Journal ofthe International Meteor Organization 16, 94–96.Kronk, G.W., 2014. Meteor Showers.Lindblad, B.A., Porubcan, V., 1999. Orionid Meteor Stream. Contributions ofthe Astronomical Observatory Skalnate Pleso 29, 77.Lovell, A.C.B., 1954. Meteor astronomy.McIntosh, B.A., Hajduk, A., 1983. Comet Halley meteor stream - A new model.MNRAS 205, 931–943.McIntosh, B.A., Jones, J., 1988. The Halley comet meteor stream - Numericalmodelling of its dynamic evolution. MNRAS 235, 673–693.Miskotte, K., 1993. High Orionid activity on October 18, 1993. WGN, Journalof the International Meteor Organization 21, 292.Molau, S., Kac, J., Berko, E., Crivello, S., Stomeo, E., Igaz, A., Barentsen, G.,Goncalves, R., 2013. Results of the IMO Video Meteor Network - May 2013.WGN, Journal of the International Meteor Organization 41, 133–138.Obrubov, Y.V., 1993. Long-period evolution and new meteor showers of cometP / Halley, in: Stohl, J., Williams, I.P. (Eds.), Meteoroids and their Parent Bod-ies, p. 69.Olivier, C.P., 1912. The eta Aquarid meteors. AJ 27, 129–130.Olivier, C.P., 1923. The radiant of the Orionids. The Observatory 46, 188–189.Olivier, C.P., 1925. Meteors.Porubcan, V., Hajduk, A., McIntosh, B.A., 1991. Visual meteor results fromthe International Halley Watch. Bulletin of the Astronomical Institutes ofCzechoslovakia 42, 199–204.Porubcan, V., Zvolankova, J., 1984. Activity and Flux of the Orionid MeteorShower from Visual Observations. Contributions of the Astronomical Obser-vatory Skalnate Pleso 12, 279.Porubˇcan, V., 1973. The Telescopic Radiant Areas of the Perseids and the Ori-onids. Bulletin of the Astronomical Institutes of Czechoslovakia 24, 1.Prentice, J.P.M., 1931. Reports of the Sections: Meteor Section. Journal of theBritish Astronomical Association 41, 327–328.Prentice, J.P.M., 1933. The radiants of the Orionid meteor shower. Journal ofthe British Astronomical Association 43, 376–382.Prentice, J.P.M., 1936. The radiants of the Orionid meteor shower. Journal ofthe British Astronomical Association 46, 329–336.Prentice, J.P.M., 1939. Papers Communicated to the Association. Journal of theBritish Astronomical Association 49, 297–299.
Article number, page 17 of 25 & A proofs: manuscript no. Egal2020b
Rendtel, J., 1997. The eta-Aquarid Meteor Shower in 1997. WGN, Journal ofthe International Meteor Organization 25, 153–156.Rendtel, J., 2007. Three days of enhanced Orionid activity in 2006 - Meteoroidsfrom a resonance region? WGN, Journal of the International Meteor Organi-zation 35, 41–45.Rendtel, J., 2008. The Orionid Meteor Shower Observed Over 70 Years. EarthMoon and Planets 102, 103–110.Rendtel, J., Arlt, R., 2008. Handbook for meteor observers.Rendtel, J., Betlem, H., 1993. Orionid meteor activity on October 18, 1993.WGN, Journal of the International Meteor Organization 21, 264–268.Sato, M., Watanabe, J., 2014. Forecast of enhanced activity of eta-Aquariids in2013, in: Jopek, T.J., Rietmeijer, F.J.M., Watanabe, J., Williams, I.P. (Eds.),Meteoroids 2013, pp. 213–216.Sato, M., Watanabe, J.I., 2007. Origin of the 2006 Orionid Outburst. PASJ 59,L21–L24.Schult, C., Brown, P., Pokorný, P., Stober, G., Chau, J.L., 2018. A meteoroidstream survey using meteor head echo observations from the Middle Atmo-sphere ALOMAR Radar System (MAARSY). Icarus 309, 177–186.Sekhar, A., Asher, D.J., 2014. Resonant behavior of comet Halley andthe Orionid stream. Meteoritics and Planetary Science 49, 52–62. arXiv:1303.2928 .Spalding, G.H., 1987. The Activity of the Orionid Meteor Stream in 1985. Jour-nal of the British Astronomical Association 98, 26–34.Spurný, P., Shrbený, L., 2008. Exceptional Fireball Activity of Orionids in 2006.Earth Moon and Planets 102, 141–150.Stephenson, F.R., Yau, K.K.C., Hunger, H., 1985. Records of Halley’s comet onBabylonian tablets. Nature 314, 587–592.Steyaert, C., 2014. The global radio Eta-Aquariids 2013, in: Gyssens, M., Rogge-mans, P., Zoladek, P. (Eds.), Proceedings of the International Meteor Confer-ence, Poznan, Poland, 22-25 August 2013, pp. 73–77.Stohl, J., Porubcan, V., 1981. Orionid Meteor Shower - Activity and MagnitudeDescription. Contributions of the Astronomical Observatory Skalnate Pleso10, 39.Svoreˇn, J., Kaˇnuchová, Z., 2017. Orionids and Eta Aquariids in the IAU MDCdatabase. Planet. Space Sci. 143, 138–141.Trigo-Rodríguez, J.M., Madiedo, J.M., Llorca, J., Gural, P.S., Pujols, P., Tezel,T., 2007. The 2006 Orionid outburst imaged by all-sky CCD cameras fromSpain: meteoroid spatial fluxes and orbital elements. MNRAS 380, 126–132.Whipple, F.L., 1987. The Cometary Nucleus - Current Concepts. A&A 187,852.Yeomans, D.K., Kiang, T., 1981. The long-term motion of comet Halley. MN-RAS 197, 633–646.Zhuang, T.s., 1977. Ancient Chinese records of meteor showers. Chinese As-tronomy 1, 197–220.Znojil, V., 1968. Frequency occurrence of small particles in meteor showers. II.Orionids, (cid:15)
Geminids. Bulletin of the Astronomical Institutes of Czechoslo-vakia 19, 306.
Article number, page 18 of 25. Egal et al.: Activity of the Eta-Aquariid and Orionid meteor showers
Appendix A: Example of Orionids and η -Aquariids activity profiles between 1985 and 2001 Fig. A.1.
Available activity profiles of the Orionids between 1985 and 2001. Original ZHR were rescaled to a common population index of 2.59.Article number, page 19 of 25 & A proofs: manuscript no. Egal2020b
Fig. A.2.
Available activity profiles of the η -Aquariids between 1985 and 2001. Original ZHR were rescaled to a common population index of2.46. When possible, observations from the Northern and Southern Hemisphere are combined (cf. Cooper 1997).Article number, page 20 of 25. Egal et al.: Activity of the Eta-Aquariid and Orionid meteor showers Appendix B: 2002-2019 activity profiles
Fig. B.1.
Activity profiles of the η -Aquariids between 2002 and 2019. Article number, page 21 of 25 & A proofs: manuscript no. Egal2020b
Fig. B.2.
Activity profiles of the Orionids between 2002 and 2019.Article number, page 22 of 25. Egal et al.: Activity of the Eta-Aquariid and Orionid meteor showers
Appendix C: Additional intensity maps of the η -Aquariids and Orionids between 1985 and 2019 Fig. C.1.
VMDB (top), CMOR 38 MHz (middle), and VMN (bottom) intensity maps of the Orionids (left) and η -Aquariids (right). For eachapparition of the shower presented in Appendix A and B, the activity profile is normalized by the corresponding maximum ZHR determined inSection 6.3.1 (reminded by the black curve at the right of the map). Article number, page 23 of 25 & A proofs: manuscript no. Egal2020b
Appendix D: Maximum visual rates of the η -Aquariids and Orionids meteor showers Year L (cid:12) main peak ZHR main peak FWHM r at maximum Source(°, J2000)1984-1987 45.28 ∼
51 - 2.5 [1] F ±
24 - 2.3 [2] F ± ±
36 - 2.41 ± ± F ± F ± F ±
34 – 2.3 [2] F ∼
44 110 – 2.3 [2]1994 43.5 80 – 2.3 [2]1995 ∼
44 104 ±
15 – 2.3 [2] F ∼ ∼ ± ± ± ± ±
13 – 2.46 [7]2002 46 54 ± L (cid:12) ± L (cid:12) ± L (cid:12) ±
12 43.3°- 45.6° L (cid:12) ±
12 43.3°- 45.6° L (cid:12) ±
23 44.5°- 46.4° L (cid:12) ± L (cid:12) ± L (cid:12) ±
15 – 2.46 [7]2011 44.7-45.9 63 ± L (cid:12) ±
20 43.6°- 47.5° L (cid:12) ±
16 – – [8]2013 45.75 130 ±
22 44.0°- 44.6° L (cid:12) ±
14 44.0°- 46.3° L (cid:12) ±
120 – 2.46 [7]2016 44.1 3 ±
17 41.5°- 46.8° L (cid:12) ± L (cid:12) ±
13 44.0°- 47.3° L (cid:12) ± L (cid:12) Table D.1.
Compilation of main peak times and ZHRs of the η -Aquariids meteor shower between 1985 and 2019, deduced from visual obser-vations. When available, the population index r at the peak and the Full Half Width Maximum (FWHM) are also provided. Values from workslabeled "F" were extracted from the published ZHR profiles and not specified explicitly in the source text. Estimates determined in this work aredetailed in Section 6.3.1. References: [1] Porubcan et al. (1991), [2] Cooper (1996), [3] Koseki (1988), [4] Cooper (1997), [5] Rendtel (1997), [6]Cooper (1998), [7] this work and [8] Cooper (2013).Article number, page 24 of 25. Egal et al.: Activity of the Eta-Aquariid and Orionid meteor showers Year SL main peak ZHR main peak duration / FWHM r at maximum Source(°, J2000)1982, 1984-1987 207.30 ∼
18 – 2.5 [1] F ± ± ± ± ± ± L (cid:12) ± ± L (cid:12) ± ±
16 – 2.59 [5]2004 209.4 40 ± L (cid:12) ±
24 – 2.59 [5]2006 208.2 72 ± L (cid:12) ± L (cid:12) ± ± L (cid:12) ±
10 206.1°- 209.6° L (cid:12) ± L (cid:12) ±
10 206.0°- 210.6° L (cid:12) ± L (cid:12) ± L (cid:12) ±
10 – 2.59 [5]2014 208.4 21 ± L (cid:12) ± L (cid:12) ± ± L (cid:12) ± L (cid:12) ± ± L (cid:12) Table D.2.
Compilation of main peak times and ZHRs of the Orionids meteor shower between 1982 and 2019, deduced from visual observations.When available, the population index rr