Sunday, May 12, 2019

Sunday rant about a Netflix eclipse

In the latest Netflix mystery show 'The Society' - "constructed around the premise of what happens to 200 teenagers who think they’re going on a school field trip, and are instead deposited in a carbon copy version of their small town, except with no adults and no younger children around" as summarized here - an unexpected annular solar eclipse takes place at the beginning of episode two.

To anyone with basic knowledge of astronomy - which includes none of those teenagers, sadly, or the showrunner, obviously - this would immediately raise a number of questions and potentially go a long way to answer what's the main question of the show at this early stage: are they still in the original New England town after all (which is isolated by dense forest on all sides now, though), are they in a copy of it somewhere else on this planet, are they in some parallel Universe (as happens on Netflix all the time) or are they, like, dead?

The dates of solar eclipses can be predicted since several millennia and their ground tracks since at least 300 years: nothing would be easier than to check whether on that date there was an annular eclipse in New England (the last one there was exactly 25 years ago, BTW, on 10 May 1994). Or elsewhere on Earth. Or nowhere at all - which would hint at either a parallel or fake reality, Truman Show-style.

Alas, the only question the solar eclipse raises on The Society is whether some of the teenagers had 'caused' it with an esoteric ritual they had been performing at the time. Hey, they had been going to school, remember? Wonder what they had learned there about the ways the Universe works ... basic 'sky literacy' was obviously not included. And for the showrunner ... he apparently also lives in a pre-scientific world where eclipses happen at will just for dramatical effect.

(Other obvious experiments not performed by the teenagers at this point ten days into the mystery: getting a shortwave radio receiver which at night would catch radio broadcasts from other countries and continents even, in case other humans still existed. Or using a GPS receiver - built into many cell phones as well as car navigation systems - to read out their coordinates ... or detect the disappearance of all GPS satellites. Now that would scare me ...)

UPDATE: Near the end of the episode someone does finally find out that the eclipse was not supposed to have happened (referring to the next total one in the U.S. coming in 2024 only, though the one shown had been annular) - and he boldly concludes that they are no longer on Earth and perhaps even in another solar system or parallel universe. That eclipses also happen elsewhere on Earth - who cares ...

UPDATE 2 / SPOILER: In Episode 7 (!) the one nerd of the group finally reveals the eclipse anomaly to all, half a year later - and also reports (as apparently nobody else had noted) that there are no artificial satellites in the sky and, most importantly, that while the star pattern is fine for the northern hemisphere, Betelgeuze has shifted a bit. Parallel word hypothesis confirmed, apparently.

Friday, July 21, 2017

The #mu69occ campaign: Occam's razor wins again ...

On July 17 the third and final attempt to observe a stellar occultation by the next target of the New Horizons mission in the Kuiper Belt, 2014 MU69, succeeded with several telescopes seeing the star blink out briefly - in contrast to a similarly effortful attempt on June 3. In spite of placing two 'fences' of portable telescopes perpendicular to the expected shadow track, none of them saw anything. When this was finally acknowledged over a month later, the respective press release from NASA and the JHU APL was full of wild speculations about how MU69's shadow could have slipped somehow through the telescope fences and tried to spin the campaign's negative outcome as a success. Of course it wasn't: you can only claim success when at least one telescope has seen the target star blink out.

Furthermore occulation observers know from decades of experience - especially from the past century - that by far the most likely explanation for a total failure is that the shadow track had not been where it had been calculated: this could be due to either bad astrometry of the star or a bad orbit of the occulter. This, however, wasn't even mentioned as a possibility in the official word - and when I asked scientists involved in the campaign about it, I got either no answer at all or a snarky one at best. Apparently the use of unpublished Gaia astrometry of the star and HST astrometry of MU69 to determine its orbit had made it all but impossible in their minds that the June 3 shadow track could have shifted beyond all telescope fences.

Alas, this is exactly what had happened, New Horizons' Alan Stern has acknowledged now: "we [...] didn't put telescopes in the right place because back then we didn't have the MU69 orbit prediction well enough in hand. Subsequent HST June-July data helped with that." So there, the - by far - most likely explanation for the June 3 failure was the right one indeed and all the speculations were wrong! Fortunately the negative observations from back then (and whatever SOFIA saw or didn't on July 10; that's still kept secret) can now be put in context as the offset of the telescopes from the true shadow track can now be calculated and secondary bodies or debris or even a ring can be excluded at that relative location from 2014 MU69. A happy ending for the campaign and for New Horizons - and a lesson for press release writers trying to make a failed observation look good ...

Sunday, June 11, 2017

German artist triggers solar eclipse detective work

A brief article on an art exhibition in Zurich (a press release here and here) has just sent me on an unusual eclipse chase. The photo artist Thomas Ruff has a weakness for space topics, and one of the objects now depicts a total solar eclipse! Ruff's modus operandi here is to take old press photos and copy the caption from the back side onto the front side as well - which so happens to be readable only in fragments on the image of a solar corona shown in the article (which often hides behind a paywall anyway), though.

In particular the year of the eclipse is not evident - but one can read that the image was taken from a KC-135 airplane over the Atlantic near South America. According to this article (page 5) and this one (page 6) the KC-135 astronomy program covered 6 eclipses from 1965 to 1980 - and the only eclipse fitting geographically [ADDENDUM: no; see below] would be the one in 1973. Which, of course, is the same that was also observed from a Concorde by other astronomers: the KC-135 observers actually watched the latter pass way above them at an amazing speed.

I may add that I actually once met Ruff in a professional context ... in 1991 at the Kunstverein in Bonn, Germany, where he had an exhibition before becoming a celebrity showing the "Sterne" series. Together with a friend we created an astronomy-didactical show in which we tried to convey the depth in these huge prints of ESO sky images that contained stars, galaxies but also occasional satellite tracks. I don't recall much but one detail: I had prepared a square meter of paper with a square millimeter marked on it to demonstrate what a million - being the difference in area - meant. We couldn't come up with a means to demonstrate a billion, though ...

ADDENDUM: Airborne eclipse expert Glenn Schneider has pointed out in personal messages that the 1966 TSE is a much better fit when a KC-135 also observed - several text fragments that can be deciphered match actual information on that flight. "The name of the place is not very legible," says Schneider, but it sure looks like Rio Grande which "is on the coast (not too far away) and is very close to 240 miles from greatest eclipse" - where you logically would fly with a research aircraft. And from what one can read the plane was "240 miles southeast of" said town.

Schneider also points out that while the 1973 TSE was visible low from parts of South America, that's a long way "to where the Concord flew across Africa! You are right that in 1973 the guys in the KC-135 watched the Concorde streak by overhead at 55,000 ft - but they were over Africa, not South America or the nearby coastal waters. Finally I will add, the corona in that picture, though not a very good picture, just does not look like the 1973 corona. I did not see the 1966 eclipse as it pre-dated my eclipse chasing started in 1970, so I cannot comment on that directly. But I did see 1973, and that just doesn't look the same..."

Saturday, May 27, 2017

Another impact flash on Jupiter - and I helped confirm it!

This afternoon Marc Delcroix, Damian Peach and Ricardo Hueso Alonso spread the news that Sauveur Pedranghelu in Corsica had probably recorded another impact flash on Jupiter yesterday - but there was no confirmation from another site, essential to rule out some flash (e.g. from an artificial satellite) close to Earth. I distributed the news and call for observations in various other Facebook groups and also on the Facebook page of Abenteuer Astronomie, a German astronomy magazine I'm with.

Within hours in the comments to the latter posting I learned that Andre Fleckstein had indeed filmed the same flash! The time was 19:25:18 UTC on 26 May 2017, i.e. smack in the middle of the time interval given by Pedranghelu, and the duration was about one second. The quality of the video was described as poor, due to bad seeing, but one could see the flash well while it was running. Fleckstein also provided a quick screenshot, upside down w.r.t. Pedranghelu's image - and the flash is clearly there, as confirmed by Peach and Alonso. The Fleckstein video will now be processed further, and scientific analysis will follow.

ADDENDA: a stacked version of Fleckstein's data, now in the same orientation as Pedranghelu's image, and a further processed version together with an explanation - and yet another confirmation of the impact, as a video clip and also nicely stacked the same way as the other two videos! Also a first report on and some more analysis of this triple success - and a depiction as space art, by a well-known Jupiter observer. Meanwhile an image sequence starting minutes after the impact or this hi-res image a bit later aren't showing any traces, typical for such events.

Saturday, February 11, 2017

Stop. That. Hype.

And so it happened again: when last night's penumnbral lunar eclipse coincided with the perigee of a faint comet (i.e. two less-than-stellar events for the public at large) all hell broke loose in the "science" and general media around the globe. The claim went that three "dazzling" (for everything in astronomy "dazzles", right? Except black holes perhaps. Which suck) sky phenomena were visible at the same time: a "snow moon", a lunar eclipse and a "mysterious green" comet. Oh yeah?
  • That "snow moon" term which suddenly invaded even German 'news' stories is of dubious provenance and in any case related to U.S. East coast folklore at best. Adopting it blindly around the globe makes no sense culturally, let alone geographically (Africans and Australians might agree). And why on Earth does each and every Moon need some fancy nickname nowadays?

  • Talking about three sky events makes no sense either as the lunar eclipse is an effect of said full moon. And regarding its relevance confusion abounded: some mixed it up with total eclipses, others claimed nothing would be seen at all - while in fact from past experience one could predict a quite distinct darkening so close to the umbra. Which was the case indeed, obvious - though not dramatic - even under bad conditions.

  • The worst mistake, however, was throwing the poor comet 45P/HMP into the mix, which has faded (in absolute brightness) since perihelion and lost most of its tail by now. Close to Earth it grew into a fuzzy blob half a degree in diameter but only of 7th magnitude: completely drowned out by the bright sky the full or nearly full moon causes. To advertise it in connection with a a full moon / non-total lunar eclipse was sheer madness.
Now you might say, so what, any astronomy in the media is good astronomy - but that is dead wrong. As one can read today in social media in reports by amateur astronomers talking to lay people the disappointment runs deep: "the media" or worse "the astronomers" predicted something exciting, and it didn't occur. In these times of - deserved or undeserved - widespread growing distrust in journalism even such small sins should be avoided. So please think first before promising things the sky won't or can't deliver - it'll hurt us all in the end ...

Thursday, December 1, 2016

Emperor Nero's "planetarium" - reality or a myth?

A report on Nero exhibitions in Germany this year made me aware of the claim that the octogonal dining room - coenatio principalis - of his palace Domus Aurea in Rome had either one rotating dome or several nested ones in order to display celestial motions 1),2),3). Or - since there is no archaeological evidence for such mechanisms - that there was a fixed dome with moving lanterns instead 4).

The only contemporary source seems to be one line in a text by Gaius Suetonius Tranquillus - and Ref. 5 says that scholars "have long debated whether this planetarium-like aspect of the room was a marvel of Roman engineering or simply a figment of Suetonius' often whimsical imagination." Is there a consensus by now about what clever astronomy display system was or wasn't installed in the Domus Aurea? And what about an even better 'planetarium' in Domitian's Domus Augustana Ref. 1 mentions?

1) Dewdney, Acquainted with the night
2) von Stuckrad, Das Ringen um die Astrologie
3) Merola, Rome's Domus Aurea
4) Goesl, Modern Projection Planetariums as Media of Iterative Reinvention
5) VROMA, Photographs of Domus Aurea

Monday, November 7, 2016

Why you should embrace the 14 November "supermoon" - and why not

On November 14 a slightly unusual sky event awaits us: a perigee full moon (i.e. a near-coincidence of full moon and the closest point of the Moon in its elliptical orbit around the Earth) which is the closest between 1948 and 2034 - though beating those full moons only by a few hundred kilometers i.e. a fraction of one percent of the distance. Here's what that means and what it doesn't. In a nutshell:
  • The full moon of November 14 is the largest in the sky of the year, and it is possible to notice with the unaided eye that its angular diameter and especially area are larger that at other times, by up to 14 and 30 percent, respectively, relative to full moons at apogee i.e. when farthest from the Earth.

  • Typically 3 or 4 subsequent full moons each year occur pretty close to perigee (as this diagram clarifies) and thus look indistinguishable to the eye; all of them are colloquially known as 'supermoons' these days, a decidedly non-astronomical term reluctantly picked up in astronomy outreach in recent years.

  • While the November 2016 full moon holds a proximity record for several decades in both directions, it is totally indistinguishable for the eye from the perigee moons of any other year (and there are many 'close calls' much nearer than the above-mentioned record years: for example next year already).
So: enjoy the show but be aware that something looking exactly the same happens several times each year. Now for the hard numbers (disclaimer: I'm taking the 1948/2034 claim reported in many places for granted and haven't checked it year by year; also distances and times listed and calculated in different places deviate by tiny amounts) ...
  • On 2016 November 14 full moon occurs at 13:54 UTC, when the distance between the centers of Moon and Earth is 356,520 km.

  • On 1948 January 26 full moon occured at 7:12 UTC, when said distance was 356,490 km: in 2016 it thus stays only 0.008% farther away.
  • On 2034 November 25 full moon occurs at 22:34 UTC, when the distance will be 356,446 km: in 2016 it stays 0.02% farther away, still not a difference 'to write home about.'

  • In 2015 at the closest full moon (which coincided with a total eclipse) it stayed 0.1% farther away than in 2016 and looked exactly the same when out of eclipse.
  • In 2017 at the closest full moon the distance will be 356,605 km or only 0.02% farther away than in 2016 (i.e. by same factor by which 2034 will be closer).

  • In October 2016 the full moon was only 0.55% farther away than it will be on 14 November 2016.
  • In December 2016 the full moon will be 0.82% farther away than on 14 November 2016; even that won't be evident in any way: thus three supermoons in a row.
In contrast in April 2016 the full moon was 13.9% more distant than it will be in November when it will thus have a 29.7% larger disk area and brightness. Claims vary how evident this is to casual but attentive observers and all too often "experts" just write it off - but those who actually look, including this blogger, can find it quite evident. So don't let the Moon orbit ellipticity visibility deniers distract you from giving it a try!

Finally some more math, inspired by a message received after posting a draft of this analysis: remember that all numbers above refer to the distance between the centers of Earth and Moon – while most observers (minus the residents of ISS and Tiangong-2) reside on the surface of the former which rotates quite rapidly, namely once per day. We are some 6370 km from the center and typically at an angle to the line connecting the centers of both. Which means a lot.

Take 14 November 2016: For the center of the Earth the distance to the Moon changes little during the day, beginning at 356,472 km at 0:00 UTC, reaching the minimum (the excitement is all about) at about 11:30 UTC and rising again to 356,788 km at 24:00 UTC – the distance shrinks by ~230 km and rises again by ~280 km during that (UTC) day.

But now go to Hawaii, well placed for actually seeing the Moon at perigee: At 0:00 UTC it is 361,133 km away, at local(!) perigee at 10:00 UTC 350,175 km and at 24:00 UTC 361,995 km – the distance first shrinks by 11,000 km, then increases again by 12,000 km! This diurnal effect just dwarfs the tiny differences between the various “supermoons” over the years and centuries where we are talking of a few dozen kilometers.

Oh, and at which latitude you sit also makes a difference. For example in Tahiti at 17°S the 14 November Moon culminates (58° high) at 9:35 UTC and is 351,043 km away. In the same longitude it culminates in the zenith at 12°N and 350,141 km distant – but at 50°S it culminates only at 26° and 353,669 km distant, i.e. 3500 km farther away at the same time. Once again a difference two orders of magnitude more than the differences between the various supermoons …

So to conclude: forget the “largest full moon in decades” meme – it’s mathematically correct but dwarfed in magnitude by effects of your place on the planet and the time on perihelion day. But embrace the fact that a few full moons each(!) year are significantly closer than the others. They are not exactly “super” but perhaps a bit ‘superior’ to the others and can be a bit more impressive than other full moons. That’s all, folks …

Main sources: basic information on ‘supermoons’ (date in 1948 off), distances of many full moons and many calculations performed on JPL’s HORIZONS. A selection of stories on the 2016 supermoons (often in denial of the visibility of the perigee effect and sometimes with funny astronomical misunderstandings): here, here, here, here, here, here, here, here, here, here, here, here, here, here, here and here. Stories - by this blogger - on the actual visibility of the perigee effect: here, here, here, here (bottom) and here.

Sunday, August 7, 2016

The Perseids of 2016 ... beyond the hype

It's one of those years again when we are getting not only the usual bubbly stories about a summer sky full of "shooting stars" (with often highly misleading illustrations in the form of composite images with numerous meteors in one frame or completely made-up artist's views sometimes resembling the end of the world) from mainstream media and certain astronomy website alike but this year, we are told, the Perseids will be super-awesome, much stronger than usual, the best in a decade, unlike any before. Wow ...

Reality check: page 12 of the 2016 Meteor Shower Calendar of the International Meteor Organization. The key sentence there: "Results from Mikhail Maslov and Esko Lyytinen indicate that we will cross a part of the stream which was shifted closer to the Earth’s orbit by Jupiter in 2016. As a consequence, the background ZHR may reach a level of 150–160." Which would be 1/5 to 1/4 more meteors at the peak than an typical year with a ZHR (zenith hourly rate) of 120 or so - this is not a dramatic increase, on a par with some recent years and well below e.g. the 1993 Perseids show which reached a maximum ZHR over 400.

Maslov's current calculation can be found in more detail here: He sees the - somewhat - higher than usual peak at 12:40 UTC on August 12 which for Europe means that the nights Aug. 11/12 and 12/13 should be comparable. Taking into account the radiant altitude and the bright Moon - which sets only after midnight, meteor party planners beware! - one can hope for actual maxmimum hourly rates under otherwise excellent conditions in the 70s, i.e. on average one meteor per minute: see the first and third diagram at the bottom of this Dutch website, blue = what you would see.

There is, however, a minority view based on a NASA model mentioned in this presentation from 2015 which sees a somewhat higher maximum ZHR (around 200) half a day earlier (around 0:30 UTC on August 12): should that happen European observers would be in a sweet spot an see twice as many meteors per hour in the wee hours of August 12 (middle diagram on the Dutch page). It is this vague possibility that much of the (extra) hype this year is based upon, but be warned that NASA's model has had a worse track record that what goes into the IMO Calendar.

So far I've seen only this one article strongly arguing against the PER 2016 hype and pointing out (some of) its problems, though it still uses exaggerated ZHR numbers and doesn't discuss the competing models. Some further information pages and articles of widely varying quality and in several languages about the 2016 Perseids can be found here, here, here, here, here, here, here, here, here, here, here, here, here, here, here, here and here - but from the preceding paragraphs you now know what to believe and where to be skeptical. And what really happens can be followed here, with a few hours delay while visual reports from experienced observers are being ingested.

Monday, June 20, 2016

When the full moon meets the summer solstice

It's one of these days when you would want to grab your 'fellow' journalists, shake them and shout: look at the numbers, st*pid! Countless web stories tell us that today (20 June 2016) full moon and summer solstice fall on the same day and that this is super-rare and thus newsworthy. Alas a simple check of the actual times tells you that full moon was at 11:03 UTC and solstice will be at 22:34 UTC - which is already on 21 June in all of Asia and most of Europe. So much for the coinciding dates which are true mostly for the Americas but simply do not exist for the majority of humanity.

So one has to ignore absolute dates and just go after the time difference to figure out - if one so desires - whether the 11 1/2 hours time difference today is a rarity. A table full moon times and a solstice & equinox calculator allow for a quick check: In 1910 there was a 12 1/2 hour difference (June 22/20 vs. 7 UTC), in 1929 a 6 hour difference (June 22/4 vs. 21/22 UTC), in 1948 a 40 minute difference (June 21/12 UTC), in 1967 a 2 1/2 hour difference (June 22/5 vs. 3 UTC), in 1986 an 11 hour difference (June 22/4 vs. 21/16 UTC) and in 1997 a 13 hour difference (June 20/19 vs. 21/8 UTC). Oh, and there was 2005 with a 22 hour difference (June 22/4 vs. 21/6 UTC): In Chicago e.g. solstice was at 1:46 a.m. CDT and full moon at 11:15 p.m. CDT - on the same day, 21 June (though in subsequent nights).

So this year's half-day difference isn't so rare at all: We had comparably close pairs of full moons and equinoxes in 1997 and 1986 and much closer pairs in 1967 and especially 1948. It is particularly obnoxious that the 1997 case - a mere one Metonic cycle ago - is flatly ignored in the "reporting" today. The reason, though, is obvious and casts a sharp light on how media mechanisms work: since full moon was 5 hours before midnight UTC while solstice was 8 hours after midnight UTC the pair appeared on two different dates also in most of the U.S. and so wasn't "important" (and the scanty 2005 case was overlooked, too). In contrast to the current 'sensation'. Sigh ...

Friday, June 17, 2016

Professional astronomer lets the crowd fund her science

A few hours ago a bold astronomical crowd-funding campaign succeeded when over USD 100,000 were pledged to support an intense photometry campaign for a single star. This raises public involvement in professional science to a new level but it raises some questions, too. Normally a professional astronomer who believes to have found a promising observing project applies for telescope time at a professional observatory which costs nothing - but getting enough time (or time at all) is not guaranteed. Alternatively a dedicated instrument could be built for which funding would be obtained via a grant from a funding organization if the latter deems it a worthy project - or, as it is possible these days, observing time could be bought from a private telescope network. Again the usual approach would be to try to get a grant to obtain the needed cash.

But in this case the astronomer didn't expect to be lucky either way - and instead asked the public at large to fund the telescope time buy. That this worked out so well in the end was due to the enormous hype that had been building (or built deliberately) around the star in question, which is the famous KIC 8462852, of course, with its erratic dips in brightness discovered by the Kepler satellite (and citizen scientists looking at its lightcurves). Its behavior is not fully explained, but some comet debris clouds are the likely culprit - and yet this star has been firmly associated the potential 'alien megastructures' in the public mind. Without these wild speculations - not exactly supported by the scientists in question but not actively discouraged either and rehashed in the media again and again - and also an added layer of drama about historical data and a long-term brightness trend or lack thereof the crowd-funding would have hardly raised a dime.

So there, the pay-per-view telescope network will soon monitor KIC 8462852 with high cadence and enough photometric precision to catch further dimmings (some of which were so strong that no Kepler would have been needed to detect them) - if any occur in the bought time interval, of course. In the best of all worlds, the dimmings (for which amateurs with their own telescopes are on the look-out as well) will return in time and display some property not seen in the Kepler data which will lead to a viable explanation. Equally likely is that nothing happens, the money is gone and a null result remains which wouldn't constrain modelmaking much. KIC 8462852 as 'star of mystery' for the public at large is a unique case in the history of astronomy so far: whether such a let's-all-fund-my-science-pet-project approach could - and should - be applied to other astronomical problems is anything but clear. The outcome and aftermath of the observing run will certainly shape opinions eventually: both amongst astronomers and the public asked to pay.

Monday, June 13, 2016

Involuntary data journalism ...

This doesn't happen - fortunately! - too often: while preparing a "simple" popular article about a paper making headlines already some real scientific detective work ensued that led to fundamental new insights and a revision of some conclusions of said paper ... The topic were the new light pollution atlas and the associated claims - pimped further by the media - that most of this planet was so bright by now that no sensible sky observations were possible anymore, esp. for Europeans and Americans.

Obviously nonsense, so what went wrong? In a first step the unfortunate mix of three different measures for the night sky brightness in the paper - absolute full, absolute artificial increment and relative increment - had to be cleared up which was trivial compared to mastering the formulae to convert between the three different absolute methods in use. That done the paper's key contents could be condensed into this master table which was then - crucially - amended with my own SQM measurements in two Dark Sky Places in Germany and on Rhodes in the past two years.

Since I had been present during these measurements I knew what actual sky appearance they meant - and that finally connected the numbers in my table and in the paper's main table and graphics with the real sky. It turned out that the paper's authors had been way too demanding in what a non-light polluted sky had to be like (and they had also been a bit too conservative re. the visibility of the Milky Way). Using my own experience and their - calibrated better than ever - numbers the article could finally be written after several hours of quite exciting "data journalism" and practical math. You're welcome!