The recently published Yemeni observing report about SN 1006 from al-Yamani clearly gives AD 1006 Apr $17 \pm 2$ (mid-Rajab 396h) as first observation date. Since this is about 1.5 weeks earlier than the otherwise earliest reports (Apr 28 or 30) as discussed so far, we were motivated to investigate an early sighting in more depth. We searched for additional evidences from other areas like East Asia and Europe. We found that the date given by al-Yamani is fully consistent with other evidence, including: (a) SN 1006 "rose several times half an hour after sunset" (al-Yamani), which is correct for the location of Sana in Yemen for the time around Apr 17, but it would not be correct for late Apr or early May; (b) the date (3rd year, 3rd lunar month, 28th day wuzi, Ichidai Yoki) for an observation of a guest star in Japan is inconsistent (there is no day wuzi in that lunar month), but may be dated to Apr 16 by reading wuwu date rather than a wuzi date; (c) there is observational evidence that SN 1006 was observed in East Asia early or mid April; for the second half of April, a bad weather (early monsoon) period is not unlikely -- there is a lack of night reports; (d) the observer in St. Gallen reported to have seen SN 1006 "for three months", which must have ended at the very latest on AD 1006 Jul 10, given his northern location, so that his observations probably started in April. We conclude that the correctly reported details give quite high confidence in the fully self-consistent report of al-Yamani, so that the early discovery date should be considered seriously.

We present here an Arabic report about supernova 1006 (SN 1006) written by the famous Persian scholar Ibn Sina (Lat. Avicenna, AD 980-1037), which was not discussed in astronomical literature before. The short observational report about a new star is part of Ibn Sina's book called al-Shifa', a work about philosophy including physics, astronomy, and meteorology. We present the Arabic text and our English translation. After a detailed discussion of the dating of the observation, we show that the text specifies that the transient celestial object was stationary and/or tail-less ("a star among the stars"), that it "remained for close to three months getting fainter and fainter until it disappeared", that it "threw out sparks", i.e. it was scintillating and very bright, and that the color changed with time. The information content is consistent with the other Arabic and non-Arabic reports about SN 1006. Hence, it is quite clear that Ibn Sina refers to SN 1006 in his report, given as an example for transient celestial objects in a discussion of Aristotle's "Meteorology". Given the wording and the description, e.g. for the color evolution, this report is independent from other reports known so far.

Given that the cause for the strong increase in 14C in AD 774/5 in Japanese and German trees is still a matter of debate (e.g. short Gamma-Ray Burst or solar super-flare), we have searched in Arabic chronicles for reports about unusual transient celestial events. In the History of al-Tabari we found two (almost identical) reports about such an event. The group around caliph al-Mansur observed a transient event while on the way from Baghdad to Mecca on AD 775 Aug 29 - Sep 1 (Julian calendar). A celestial object (kawkab) was seen to fall or set (inqadda), and its trace (atharuhu) was seen for at least tens of minutes (up to 70-90 min) during morning twilight. The reports use the Arabic words kawkab and athar(uhu), which were also used in the known Arabic reports about supernovae SN 1006 and 1054, so that one might consider an interpretation as a nova-like event. The kawkab (celestial object) was observed only during the morning twilight at a brightness of probably between about -3 and 0 mag. Such a brightness and time-scale would be expected for optical kilonovae (at 3 to 9 kpc) in the context of short Gamma-Ray Bursts. There are no similar reports from eastern Asia for this time. However, the short reports are fully consistent with a bolide: The word kawkab can be used for meteor, the verb inqadda normally means "falling down", the word atharuhu can mean "its trace". We therefore prefer the interpretation as bolide. We discuss in detail how to convert the Muslim calendar date to a date in the Julian calendar using first the calculated Islamic calendar and then considering the time when the crescent new moon could be visible at the given location.