The duration of this meteor shower covers the period of October 6-10. Maximum currently occurs on October 9/10 (λ=159.40°), from an average radiant of α=262°, δ=+54°. The maximum rate typically reaches 1-2 per hour, but outbursts of hundreds or thousands per hour occurred several times during the 20th century.
This represents the view from mid-northern latitudes at about 9:00 p.m. local time around October 8. The graphic does not represent the view at the time of maximum, but is simply meant to help prospective observers to find the radiant location. The red line across the bottom of the image represents the horizon. (Image produced by the Author using SkyChart III 3.5 and Adobe Photoshop 5.5.)
The Draconids are not a good meteor display for observers in the Southern Hemisphere. Although it is actually above the horizon for a short time, this happens in morning twilight at about 5:00 a.m. local time. The highest altitude in dark skies occurs just before morning twilight begins, at which time the radiant is about 8° below the horizon.
The discovery of this meteor shower resulted from predictions by several astronomers that the periodic comet 21P/Giacobini-Zinner might produce a radiant in early October.
The first to make such a prediction was M. Davidson, who, in 1915, examined the periodic comets observed since 1892 to isolate any that might be capable of producing meteor showers. One of the comets which met the established criteria was 21P/Giacobini-Zinner. Davidson found that the comet's orbit would be fairly close to Earth on October 10, 1915, and predicted that, if the debris from this comet had spread outward by about two million miles, a shower might be active from α=267°, δ=+50°. This was about two years after the parent comet had passed perihelion. During the first half of October, W. F. Denning recorded "a number of meteors" from α=267°, δ=+49°.
Davidson revised his prediction in 1920 (primarily due to a discovered error in his earlier prediction), saying that the distance between the orbits of the comet and Earth amounted to 5 1/2 million miles. He subsequently suggested that maximum would most likely occur on October 9 from α=251.5°, δ=+55.9°. Later that same year, Denning became the first observer to make a definite observation of this new shower, as he observed 5 meteors from α=268°, δ=+53° during October 6-9. These meteors were described as slow. Giacobini-Zinner had passed perihelion during the Spring of 1920.
The 1926 Return
The comet was next expected at perihelion at the end of 1926, and predictions for a meteor shower in October of that year were made by both Davidson and A. C. D. Crommelin. Both men gave October 10 as the expected day of maximum, and they gave similar radiants of α=261°, δ=+53.5° and α=265°, δ=+54°, respectively. The orbits of the comet and Earth were found to intersect in 1926, though the expected shower would occur 70 days before the comet passed that point of its orbit. Around October 10.4, observers in England were made aware that some unusual activity was present.
What gained the attention of many observers on October 9, 1926, was the appearance of a fireball. The event was noted by hundreds in the British Isles and 35 reports allowed the radiant to be determined as α=262°, δ=+55°. The meteor moved slowly and lit up the sky. A persistent train lasted about 30 minutes "during which time it underwent curious changes of form and exhibited drift amongst the stars."
Several other radiants were determined in 1926, which acted to confirm the predictions of Davidson and Crommelin. J. P. M. Prentice (Stowmarket, England) observed for 3 hours centered on October 9.9, and detected 16 meteors from α=263°, δ=+54°. He described the meteors as slow and estimated the radiant diameter as 6 deg. He claimed that the hourly rate may have been near 17 had his observations been continuous. Observations made by A. King (Ashby) and Denning (Bristol) gave a radiant of α=255°, δ=+56°.
The years following 1926 were closely monitored by a few observers trying to catch another glimpse of the Draconid shower. During 1927-1932, Prentice observed extensively around October 9-10, but no activity was detected.
The 1933 Return
Comet 21P/Giacobini-Zinner passed perihelion on July 15, 1933, and, at the time of the predicted maximum of October 9, Earth crossed the area of the comet's descending node just 80 days after the comet. Astronomers were not prepared for what was in store for them, but, as evening twilight fell over Europe, observers noted the beginnings of something unusual. Within just a couple of hours the number of Draconids skyrocketed, and, at 20:00 UT, one of the best displays of the 20th century was in progress. Some of the more impressive statistics follow:
- From Ireland, W. Ellison (Armagh Observatory) reported that meteors fell as frequently as snowflakes and he gave the radiant as α=26°, δ=+55°. W. Milligan (near Omagh) saw thousands, including 100 during one 5-second interval. The radiant was α=264.5°, δ=+54.5°.
- At Birchircara, Malta, R. Forbes-Bentley observed over 22,500 meteors in just a few hours and estimated the peak rate hit 480 per minute at 20:15 UT. The meteors were described as mostly faint with only 5% reaching first magnitude. The radiant was estimated as α=262.5°, δ=+55°.
- In Russia, N. S. Sytinskaja (Leningrad) collected numerous observations between 18:00 and 22:00. At maximum, rates reached 100 per minute at Leningrad, 300 per minute at Pulkovo and 200 per minute at Odessa.
- From Spain, P. M. Ryves (Zaragoza) estimated a maximum rate of 100 per minute and an average radiant of α=266°, δ=+53.5°. He said the meteors were generally faint with the "great majority of the meteors being 3rd to 5th mag."
As can be gathered from these reports (and many others not included here), the shower's maximum rate reached 100 per minute, or about 6000 per hour, around 20:15 UT on October 9. The meteors were slow, generally faint and were usually yellow.
Following the 1933 appearance, the Draconids again fell to nonexistence. The comet's next perihelion date was February 17, 1940, and there were numerous predictions of a possible strong return in October 1939. However, Earth crossed the comet's orbit 136 days ahead of the comet which meant bad news for meteor enthusiasts, as no storm---or even a small shower---appeared.
The 1946 Return
During 1940 to 1945, activity continued to be absent, but astronomers were already making predictions for the very favorable 1946 return. In that year, the comet was expected at perihelion on September 18, so that Earth would cross the comet's orbit just 15 days after the comet!
The Draconids were best placed for observers in the Western Hemisphere during October 9/10, with excellent meteor counts not only coming from all across the United States, but Canada and even Venezuela. European observers did detect the Draconids, but the radiant was very low over the horizon and, though spectacular, it was at least one-fourth the strength of that seen in America. Observations were also made in Czechoslovakia for slightly more than one hour prior to morning twilight. Some of the more interesting observations are as follows:
- At the University of Oklahoma Observatory, B. S. Whitney led a team of 10 observers during the period of 1:23 and 4:34 UT. Taking counts every 10 minutes, they determined that maximum came around 3:50 UT (October 10) when estimated hourly rates were near 3000.
- From Southern California, numerous counts were tabulated at Griffith Observatory. From that observatory, R. Michaelis and K. Bouvier observed rates of 55 per minute between 3:45 and 4:02 UT. E. L. Forsyth (Fallbrook) detected 63 per minute at 3:50 and G. W. Bunton (Sunland) observed 180 per minute around this same time.
- In England, Prentice evaluated the British observations and noted a maximum rate of 965 per hour. The radiant was then very low over the horizon and it was determined that the ZHR was actually 2250.
- At Skalnate Pleso Observatory (Czechoslovakia) observers fought clouds, moonlight and morning twilight to observe the Draconids. Maximum occurred at 3:53 UT with a corrected rate of 6800 per hour. Two secondary maxima occurred at 3:23 and 3:40. Analysis revealed the half-time of the shower to have been 0.65 hours.
What may have been the highlight of this event was the appearance of a large blue-white fireball over Southern California at 3:38 UT. Forsyth said it left a yellow train which lasted over three minutes. As the train drifted and became diffuse, it took on the shape of a horseshoe.
The 1946 event marked an important first for meteor astronomy---the detection of a meteor shower by radar. In the United States alone, 21 radar systems were operated at frequencies of 100, 600, 1200, 3000 and 10000 Mc/sec. From these instruments only the radar operating at 100 Mc/sec detected meteor echoes. The majority of all meteor activity occurred between 3:00 and 4:30 UT on October 10. Other radar equipment operating in London and the Soviet Union operated at frequencies between 3.5 Mc/sec and 212 Mc/sec and confirmed that maximum occurred between 3:00 and 4:00. Most interesting was a record obtained by J. A. Pierce, who used a 3.5 Mc/sec pulsed ionospheric sounder and found that meteors were so numerous that a temporary ionosphere was formed at a height of 90 km. The meteoric ionosphere lasted three hours and was confirmed elsewhere.
Following 1946, both visual and radio-echo techniques were utilized in searches for this shower during 1947-1951. Visual observers detected no meteors possibly associated with the Draconids, while radio-echo observations at Jodrell Bank detected "no activity during the Giacobinid epoch in excess of the background sporadic rate (that is, not greater than 4 or 5 per hour)."
The 1952 Return
A possible shower was predicted for October 9, 1952. Various calculations revealed Earth would cross the comet's orbit 193 days ahead of the comet. In addition, the closest distance between the orbits of Earth and 21P/Giacobini-Zinner was very similar to that of 1933, or about 0.0057 AU; however, on this occasion, the comet's orbit would actually pass inside of Earth's orbit. Visual observations by British observers revealed only the barest hint of activity shortly after sunset on October 9/10, but, just a few hours earlier, daylight observations had been made using the radio-echo apparatus at Jodrell Bank in England.
The Jodrell Bank team first noted the Draconid rate rising above that of the sporadic background at 14:20 UT on October 9. Meteor echoes were counted during 10-minute intervals: 3 were noted at 15:00, there were 6 at 15:10, 10 at 15:20, 11 at 15:30 and 17 appeared at 15:40. The highest rates occurred at 15:50, when a 10-minute rate of 29 was reported---indicating an hourly rate of 174. The following decline in activity was very rapid, and one-half hour after maximum the 10-minute rate had declined to only 3. The last definite sign of activity occurred at 16:40, when the rate was 2. The Jodrell Bank observers concluded that maximum occurred at a solar longitude of 196.25 deg from RA=262 deg, DEC=+54 deg.
The 1959 return of 21P/Giacobini-Zinner was very favorable and, since Earth arrived at the comet's orbit just 21.6 days before the comet passed through the region some believed a meteor storm would occur. However, the comet's perihelion distance had been pulled closer to the sun so that the closest distance between the orbits of Earth and the comet was 0.058 AU. Subsequently, no shower was observed. The comet's 1966 return also failed to produce meteors due to unfavorable geometric conditions.
21P/Giacobini-Zinner passed only 0.58 AU from Jupiter during 1969, which acted to increase its perihelion distance to 0.99 AU---Draconid showers were again possible. Searches for activity began during October 1971, with Earth crossing the comet's orbit 309 days before the comet. No notable activity was observed during October 7 to 10, by members of the American Meteor Society, as hourly rates remained around one.
The 1972 Return
The 1972 Draconids were looked for with much anticipation. Not only was Earth going to cross the comet's orbit 58.5 days after the comet, but the two orbits were separated by only 0.00074 AU! Unfortunately, despite these promising statistics, the shower was quite a disappointment. Observers in the United States obtained the highest visual rates when 10-15 per hour were detected on October 8/9. Maximum had been predicted for 17:00 UT on October 8, which made Japan the best location for observations. Unfortunately, the Japanese observers were met with cloudy skies. Despite this hindrance, the Hiraiso Branch of the Radio Research Laboratory operated a 27.1 MHz radar. A peak of 84 returns in 10 minutes was noticed at 16:10 UT on October 8, followed by a secondary peak of 69 returns in 10 minutes at 21:00 UT. Further predictions for Draconid showers in 1978-1979 and 1985-1986, were not met with noticeable displays.