Rainbows above the Buckingham Palace and Windsor Castle at the Moment of the Queen's Death

LilWabbit

Senior Member
From the same Physics Stack Exchange discussion cited earlier on the first page of this thread:

Article:
There are many obvious reasons why rainbows are not very common or, more precisely, are not very commonly observed by any particular individual: it has to be rain, it has to be sun, it has to be this or that part of the day, East or West, dark background, etc. and that individual has to be in the right spot at the right time and actually look.


This is pretty much basics and identical to what @Ann K recaps here:

If you'll permit a childish simplification, here's a quick and dirty sketch made from the weather map of the USA yesterday. The essential requirements are rain, the end of the cloud cover, and the sun low enough to come underneath. It wasn't a completely solid rain cloud, but fairly close.

But the author continues with droplet size and droplet density as additional criteria:

Article:
If all those other things line up, will we always see a rainbow or does it have to be a particular type of rain or particular type of water droplets? Here are couple of additional factors affecting the probability/visibility of a rainbow:

Droplet size and shape. For the brightest rainbow, droplets have to be 1-2mm. Droplets smaller than 0.05mm will create a white rainbow, etc. More details could be found here.

Droplets density. For the rainbow to work, a photon has to hit one droplet and fly right back to your eye, i.e., it has to be scattered once. If the density of droplets is high, like in clouds, a photon will likely hit multiple droplets and end up moving at a random angle, breaking the 42 degrees requirement.


But ... this one particular band stretched from Tallahassee, Florida to Washington DC at the time I looked at the weather map, roughly a distance greater than Land's End to John O'Groats, and rainbows would have been in sight of many, many millions of people, had they looked out the window.

That's an opinion based on an assumption of alignment of some but not all the rainbow criteria mentioned in the foregoing.

And once again: Even if there were a huge contiguous geographic zone optimal for viewing a rainbow by millions of people at around Tallahassee, it doesn't really address in any helpful way the probability of a rainbow appearing at Buckingham Palace .
 
Last edited:

Mendel

Senior Member.
I'm not very woke but if I were, what's the gender-neutral formulation of strawman?

A straw person? :)
a straw effigy is what you set up when you find it difficult to burn the real thing


While we fully agree that all scientific claims require sound objective evidence to back them up, Carl Sagan's popular slogan, whilst poignantly poetic, is not an established scientific standard.
Both parts of this statement are false.

Scientific claims are merely required to be falsifiable (by 'sound objective evidence') to be "scientific".

For a scientific theory be convincing, I'd say it is necessary for it to either explain common evidence better than other theories, or to explain evidence other theories can't.
Article:
The Sagan standard, according to Tressoldi (2011), "is at the heart of the scientific method, and a model for critical thinking, rational thought and skepticism everywhere".[2]
Consider yourself contradicted.

The rainbow, of course, already has an accepted simple explanation.
 
Last edited:

Mendel

Senior Member.
The best explanations are able to predict measurement outcomes.
Yes.

Can you predict rainbows from the deaths of a particular set of people?

or predict deaths from rainbows?

if you can't, your explanation is useless
 
Last edited:

LilWabbit

Senior Member
Scientific claims are merely required to be falsifiable (by 'sound objective evidence') to be scientific.

Correct. Almost as if you were quoting me from our earlier exchanges. No contradiction.

For a scientific theory be convincing, I'd say it is necessary for it to either explain common evidence better than other theories, or to explain evidence other theories can't.

Yes, and usually the standard for that is it's ability to predict observations or measurement outcomes better than its rivals. Nothing to do with extraordinariness.

The Sagan standard, according to Tressoldi (2011), "is at the heart of the scientific method, and a model for critical thinking, rational thought and skepticism everywhere"

That's a quote from some psychologist
 

LilWabbit

Senior Member
Yes.

Can you predict rainbows from the deaths of a particular det of people?

or predict deaths from rainbows?

if you can't, your explanation is useless

By contraposition of a chance occurrence predicting poorly the said co-occurrence that standard would be logically met.

But thus far we have insufficient data to establish whether chance occurrence is a poor predictor. It may not be.
 

deirdre

Senior Member.
"you ascribed gender to an entity about which we know nothing,
no. im saying lil wabbit was not necessarily speaking of an entity. if you dont know if it is a he, she or it (ie the universe), you say it.
at no point did I capitalise "Him" - that's a fabrication purely of your own.
i have no choice but to capitalize Him.

Stop doing this to yourself.
stop being a hypocrit.
 

deirdre

Senior Member.
Define "relatively common". Respect to what?

it's your quote. you tell me.
https://www.metoffice.gov.uk/weather/learn-about/weather/optical-effects/rainbows/double-rainbows

A double rainbow is a wonderful sight where you get two spectacular natural displays for the price of one.

Surprisingly, this phenomenon is actually relatively common, especially at times when the sun is low in the sky such as in the early morning or late afternoon.
Content from External Source

and rainbows would have been in sight of many, many millions of people, had they looked out the window.
i dont think we can say that from a weather map. but mostly his quote was double rainbows.
 

Mendel

Senior Member.
Correct. Almost as if you were quoting me from our earlier exchanges. No contradiction.
Well, you wrote that "all scientific claims require sound objective evidence to back them up", and that's an entirely different criterium not rooted in reality.

Yes, and usually the standard for that is it's ability to predict observations or measurement outcomes better than its rivals. Nothing to do with extraordinariness.
"No" to both.
Obviously it happens, but I feel it's more often that a revolutionary theory is not just, say, predicting something more accurately, but predicting something entirely different than the current theories — something extraordinary, one might say.

That's a quote from some psychologist
Better than no quote at all, wouldn't you say? ;)

Tressoldi specialises in ESP, he probably encounters extraordinary scientific claims more often than most scientists.
 
Last edited:

deirdre

Senior Member.
could you give an example of something that is intelligent (wabbit's claim), yet not an entity?
i did already in my comment.
just like i told Phil, before he pitched a fit, that wabbit was at fault for earlier mentioning "divine" in his wording.

(and yet people constantly accuse me of not comprehending what i read. sigh)
 

LilWabbit

Senior Member
Well, you wrote that "all scientific claims require sound objective evidence to back them up", and that's an entirely different criterium not rooted in reality.

Nah, just a more roundabout way of saying the same.

"No" to both.
Obviously it happens, but I feel it's more often that a revolutionary theory is not just, say, predicting something more accurately, but predicting something entirely different than the current theories — something extraordinary, one might say.

That's all part and parcel of the word "better" I used, rather than "accuracy" which was your own "burning effigy". :p

Plus what you' feel' what a revolutionary theory is or isn't, isn't a particularly convincing counter-argument anyhoo.

Better than no quote at all, wouldn't you say? ;)

What you "feel" is just as convincing to me as to what Tressoldi says about what science is.

Tressoldi specialises in ESP, he probably encounters extraordinary scientific claims more often than most scientists.

Extraordinary is subjective, and hence has nothing to do with the method of natural science. Sagan's statement is a skepticist ideological tenet rather than a basis of all science.

Street hawkers in Beijing grilling grubs in skewers is extraordinary to most. It still happens. That the earth is round was an extraordinary claim. Nowadays claiming the opposite is. It doesn't matter which one is 'more' extraordinary. What matters is which one is consistent with all (not just some) relevant observations (adequacy principle) and generates better predictions (which implies new predictions amongst other things) -- 'the prediction of the theory'.
 

deirdre

Senior Member.
what, "the universe"?
I asked you a simple question. I'd like to accuse you of depriving me of a simple answer, which means we get this unnecessary exchange detracting from the point you were making
well if you knew i said the universe then why unnecessarily ask me for an example? jeez. if you disagree with my example, then just say why.
 

Mendel

Senior Member.
well if you knew i said the universe then why unnecessarily ask me for an example? jeez. if you disagree with my example, then just say why.
I didn't recognize you meant for the universe to be that example because I don't consider the universe to be intelligent.

That you thought of it as such was my best guess after you gave me the clue that there was an example in that post; I did not know that when I first read it.

(Note to self: @deirdre thinks the universe is smart)
 

FatPhil

Senior Member.
I dont deny anything, forgive me for quoting you appropriately.

You used different capitalisations each time you quoted me. One of them must have been inappropriate. (Clue: it was the one where you changed the semantics of my quote. Don't do that, m'kay.)
 

deirdre

Senior Member.
You used different capitalisations each time you quoted me. One of them must have been inappropriate. (Clue: it was the one where you changed the semantics of my quote. Don't do that, m'kay.)
you're exaggerating. i used air quotes around MY "Him" one time. i grant that i should have known to write [airquotes], but you are making a huge fuss over nothing. you know what your 'he' signifies. I'm tired or your pointless derailment.
 

jplaza

Member
Some more maps indicating cloud cover over England and Southern England at 1730 hrs UTC (1830 hrs London time) on 8 September 2022. Windsor is roughly at Slough. Difficult to come by sharper maps indicating microclimates and small clouds.

Zoom Earth Weather Map.JPG

Sat24 Weather Map Large.JPG

It's impossible to draw conclusions on the probability of seeing rainbows based on these maps, but they do indicate a lot of choppy cloud cover over Southern England.

I would also like to revisit the matter of probability distributions and why the Normal/Gaussian Distribution also came to mind earlier (visually) with its bell-curve, whilst inapplicable to the specific variable of rainbow appearances and me confusing terms.

If the historical frequency of the 2 somewhat random variables deaths of monarchs and rainbows appearing at the seats of their reign roughly follows Discrete Uniform Distribution overtime and/or Poisson Distribution;

Does the probability distribution of days when the two variables fail to coincide on the same day follow roughly the Gaussian/Normal Distribution with a low standard deviation or just another Uniform Distribution? (Ignoring for a moment the non-trivial fact that the 8 September coincidence occurred not only within the same day but within minutes of the official announcement).

In other words, most of the values in the resulting bell-curve (unless it's a not a bell-curve but rather a uniform distribution with a pyramid-like graph), indicating the probability of coincidence within a fixed time-period, say a week or a month, tend to be close to the mean (a low standard deviation) of 0 likelihood of coincidence while the likelihood of coincidence increases as we move either backwards (left) or forwards (right) along the time coordinate (x) by centuries or millennia from a particular documented coincidence.

I guess Binomial Distribution could also present the same data in a different manner. Open to corrections.
Why don't you keep it simple?
You have a distribution when you have multiple outcomes for an "event".

Take a normal card deck. Pick a random card and try to guess its color (diamons, hearts, spades or clubs). You have a chance of 1/4 (25%) of guessing right. That's the probability for a single event. You don't have a "probability distribution" for guessing one single card. Either you do, or not.

Pick 10 cards and try to guess their color: you may guess correctly 0,1,2,3,4,5,6,7,8,9 or 10. Multiple outcomes. And for each outcome you can calculate its probability. That's the "probabiltiy distribution": probability of guessing 0,1,2,3,4,5,6,7,8,9 or 10 cards.


For the rainbows, first, you need to define probabilities in a meaningful way, so there is a way to calculate it.

What does "the probability of seeing a rainbow is 30%" mean? It means in average if you randomly pick 100 events, you would see the rainbow 30 times. But what's "the event" here? You could say it's "a day". Pick any random day and check whether there was a rainbow or not. Pick another day, check,... repeat until you have enough events as to calculate a probabilty (Out of N days, in x of them there was a rainbow, so x/N is the probability of seeing a rainbow in a random day. There's no distribution of anything here. You either see it or not for a single day).

Or you can define you event as "a rainy day". And do the same, picking now random rainy days, and check in how many you have a rainbow. (Note that the definition of this probability is different to the previous example.)

Or you can add more conditions: "A rainbow seen from the seat of the reign in the UK". So pick N random days and check in how many a rainbows were seen from the seat of the reign of UK.

More conditions: "What is the probability of seeing a rainbow from the seat of the reign in UK in any day of September"... so pick N days of september of any year, etc....

Now, if you choose conditions that are totally unrelated to the appearance or not a rainbow, the probability won't change: What's the probability of seeing a rainbow the days I am wearing a red t-shirt? And the days I wear a yellow t-shirt? Should be the same.

But if you choose conditions that are related, then it will change: Whats the probability of seeing a rainbow any day in September? And any day in December?

So what you want to check if the death of a king/queen is related to the appearance of a rainbow. Just calculate "the probability of seeing a rainbow any day" and compare to "the probability of seeing a rainbow the day a king/queen dies".

Or add any conditions you think is relevant:
- "Probability of seeing a rainbow from Buckingham any day of september?" vs "probability of seeing a rainbow from Buckingham any day of september that a king/queen died"

Or maybe you want to go "all-in":
- "Probability of seeing a double rainbow from Windsor and Buckingham on Sep 8th at 18:30" vs "probability of seeing a double rainbow from Windsor and Buckingham on Sep 8th at 18:30 concurrently with the announcement of the death of the Queen". So you would need to take "N random Sep 8th days from any year when a double rainbow was visible from Windsor and Buckingham at 18:30", and compare to "N random Sep 8th days from any year when a double rainbow was visible from Windsor and Buckingham at 18:30 at the same time the death of a queen was being announced ".

If the death (or announcement of the death) is related to the appearance of a double rainbow, those probabilities (either large, small, or minuscule) should be different.

Problem is the more conditions you add, you may end up with a single event from which you can't calculate any probability at all.
 

jplaza

Member
it's your quote. you tell me.
In that context, "relatively common" respect to seeing a single rainbow. Seeing a "double rainbow" may not be as common a seeing a single one, but they are not extraordinary to see. I have seen quite a few. I can't say whether for each double rainbow I have seen 2, 3, 4, or 5 single ones, but the idea is that they are frequent, even if not as much as single ones.

Now back to you:
How a double rainbow seen from Buckingham is extraordinary, or not "relatively common"? respect to what?
 

deirdre

Senior Member.
but they are not extraordinary to see
the rainbow was never extraordinary. the timing of the rainbow and location was extraordinary.

respect to what?
the chances of seeing a double rainbow at Buckingham Palace with a concurrent rainbow at Windsor (i'll leave out the "when a commemoration flag was lowered"..even though technically that should be included.)
 

Ann K

Senior Member.
Problem is the more conditions you add, you may end up with a single event from which you can't calculate any probability at all.
And "the death of a monarch" coupled with a weather report for the day is so severely limited in examples that any "statistics" that someone claims are highly suspect. He (Lilwabbit) keeps talking about the probability of a rainbow, when it's the "monarch dies" value that makes this whole thing an idle speculation.
 

jplaza

Member
the chances of seeing a double rainbow at Buckingham Palace with a concurrent rainbow at Windsor
And some other places totally irrelevant that we don't know about because the journalists cherry picked the ones that looked better in the headlines.
 

deirdre

Senior Member.
And some other places totally irrelevant that we don't know about because the journalists cherry picked the ones that looked better in the headlines.
no. you said "respect to what?". respect to landmarks where mourners had gathered.

:) it's fine if you don't think the event was extraordinary.
 

Earlscourt

New Member
in my opinion its meaningless. Rainbows are basically an every other day occurrence across the UK, believe me. lol

On the subject of rainbows, I actually witnessed a tiny rainbow form, from the ground up, (actually viewing both ends) I believe not many people have ever seen that happen, I have images of it. It grew and eventually the other end, spread over the top of a nearby hill. It was pretty astounding to watch! Before you ask, there was no gold on either end....sadly
 

LilWabbit

Senior Member
And "the death of a monarch" coupled with a weather report for the day is so severely limited in examples that any "statistics" that someone claims are highly suspect.

I haven't claimed any statistic. Please stop distorting my words, thank you. I'm trying to explore a scientifically credible way to establish a statistical probability, which is not easy given the limited amount of data and the complexity of the factors involved. I have merely said that with the limited data rainbows appearing at Buckingham palace seems somewhat rare. But it does happen. Not just on the 8th.

He (Lilwabbit) keeps talking about the probability of a rainbow, when it's the "monarch dies" value that makes this whole thing an idle speculation.

It would be idle if it weren't for its concurrence with the two rainbows appearing at the same monarch's chief residences. That you don't want to cannot see even the seeming extraordinariness of such a concurrence in no way implies the crowds gathered were unreasonable for being amazed, or the news networks covering the rainbows were just looking for cheap scoops while deliberately ignoring other facts that would have reduced their seeming extraordinariness.
 

Ann K

Senior Member.
i haven't claimed any statistic. Please stop distorting my words, thank you. I'm trying to explore a scientifically credible way to establish a statistical probability
if you want "statistal probability", you're going to need some statistics. If you want it to be meaningful, you're going to need more than a single event.
 

Ann K

Senior Member.
On the subject of rainbows, I actually witnessed a tiny rainbow form, from the ground up, (actually viewing both ends)
I had a ninth-floor dorm room in college, and once saw a double rainbow that was nearly a complete circle from that vantage point.
 

LilWabbit

Senior Member
So you would need to take "N random Sep 8th days from any year when a double rainbow was visible from Windsor and Buckingham at 18:30", and compare to "N random Sep 8th days from any year when a double rainbow was visible from Windsor and Buckingham at 18:30 at the same time the death of a queen was being announced ".

These would indeed represent more relevant parameters for the probability calculation. The gist of their logic is not new to this discussion. Similar probability calculations to what you're suggesting were already roughly presented in post #16 on the first page of this thread:

"On how many of those rainy September days are rainbows reported on the annual average and at what frequency around London City and Windsor?"

Or with some further elaboration (excluding the misleading and unnecessary term "reported") in post #40 on the first page:

"If rainbows appear above the Buckingham Palace on every other September late afternoon on an average year, then we can comfortably dismiss any hypothesis of an intelligent timing by 'a higher power'."

In other words, if from N random 8th of September days from any year in half of them a double rainbow appears at Buckingham Palace, then we'll have roughly the same result of rainbows appearing on every other September day (15 days) in most years. Provided of course that the average historical weather for the 1st of September and the 30th of September in London isn't significantly different (mostly grey and rainy :p). It's probably easier to find data for every September afternoon of the last ten years than to go through all the 8 Septembers of the past two centuries in hope of finding out whether a double rainbow was visible on any of those afternoons at the Palace.

1. On Mean Historical Frequencies

@jplaza wrote:

If the death (or announcement of the death) is related to the appearance of a double rainbow, those probabilities (either large, small, or minuscule) should be different.

Or, in other words, for the timing of the Queen's death (and the announcement that followed) to be likely related to the simultaneous appearance of rainbows at Buckingham Palace and Windsor Castle, the probability of their appearance should not fall within the mean historical frequency of their appearance at these locations on 8 September late afternoon.

If the mean frequency is roughly every other year, then nothing extraordinary is happening. If it's roughly every ten years with the last one occurring last year, then it's a tad more extraordinary. If the mean frequency is 10,000+ years for a double rainbow to appear at 1830 hrs at the approximate location of today's Buckingham Palace with a concurrent rainbow appearing at the location of the Windsor Castle, with the last joint appearance occurring 451 years ago, then their appearance would represent a significant deviation from a discrete uniform distribution -- in other words, a very likely non-random occurrence. Or, by contraposition, a very unlikely but possible random occurrence.

Even if by 2022 it was becoming statistically increasingly likely for a double rainbow to 're-appear' at Buckingham Palace after a 100 year mean interval since the last one appeared, say, on 8 September 1867, it would still seem extraordinary that it would coincide the passing of its chief resident. The coincidence with the death of a British monarch is highly unlikely due to the low and random frequency of also the latter variable (i.e. death of British monarchs). This is why some manner of discussion on the probability distribution of both variables seems relevant. As well as searching for evidence as to how frequently rainbows have actually been visible at these locations during September.

2. Historical Data vis-à-vis Mathematical Model

To establish a statistically significant deviation or lack thereof we must have reliable data on multiple rainbow appearances (yes, more than just this single event is required) from a very long period of time from that location, which is obviously virtually impossible to come by. Hence, such a distribution could alternatively be modelled based on a careful analysis of September weather patterns and conditions at that location. Which is also difficult due to the many factors involved. Flippant subjective claims "it's very common", "I've seen a lot this year where I live", to hastily justify a personal bias against uncomfortable hypotheses just won't do if we wish to even feign scientific commitment. A proper debunk requires more even if your subjective intuitions about their commonality were correct.

3. Aunty Pat Revisited

Yes, the probability of the 8 September rainbows is roughly the same as rainbows appearing at any particular location in the City of London. In other words, a rainbow appearing at 1815 hrs on 5 September at the Westfield Stratford City shopping mall in East London has roughly the same probability as the 8 September Buckingham Palace rainbow. Rainbows have likely appeared at Westfield Mall before. If 1815 hrs on 5 September 2022 represents a significant deviation from the usual historical September distribution of rainbow appearances at the location of the mall, then the rainbow, statistically, is related to the timing of 5 September 2022 in some way. Maybe Aunty Pat with her cabbage batch died of heart attack in Tenerife on the same afternoon and everyone knew the McD's Happy Meal at the Westfield Mall was her favorite lunch on Wednesdays. :p

Problem is the more conditions you add, you may end up with a single event from which you can't calculate any probability at all.

Yes, and we must therefore stick to the relevant conditions and not add conditions out of pro-bias for mystical explanations just to rack up the 'amazing' improbability of the occurrence (which, if you or @Ann K think I'm doing, then you're unfairly assuming a hidden agenda that I simply don't have). But neither should we dishonestly ignore other relevant conditions to make those rainbows seem more common and mundane than they actually are. Common sense suggests that rainbows appearing at various locations in Britain during a choppy nationwide rain cloud cover is likely and a fairly common occurrence. The same common sense suggests that them appearing at any particular location of random choice under that same overall blanket of clouds is fairly unlikely. But all this is just flippant subjective ramblings which, I hope, we all wish to move forward from.

As I've said in post #78, epistemologically, we must guard against at least two kinds of confirmation bias which pose a risk to unbiased and objective inquiry:

(1) Not acknowledging even the seeming remarkability of rainbows appearing above the Queen's chief residences to observers exactly at the moment of the announcement of her death.

(2) Self-servingly using these appearances as an endorsement of some particular theology claiming privileged access to truth as opposed to every other worldview.
 
Last edited:

jplaza

Member
Let's assume you can calculate that probability of a rainbow appearing.

Let's assume you can define in a meaningful way the probability of a monarch dying.

Let's assume you find a way to combine both to calculate the combined probability of both things happening at the same time (Bayes theorem comes to mind)

Let's assume the result is a very low value. (Which I guess that's what you want to prove.)

The extraordinary thing is not 1-in-million-chances happening once.

The extraordinary is 1-in-a-million-chances happening 1 in ten times.
 

LilWabbit

Senior Member
Let's assume the result is a very low value. (Which I guess that's what you want to prove.)

Nope. Only trying to find out how to calculate the likelihood, which could just as well be high.

The extraordinary thing is not 1-in-million-chances happening once.

The extraordinary is 1-in-a-million-chances happening 1 in ten times.

Even if the recorded history of monarchs spans only 6,000 years?

Even a single co-occurrence would seem quite extraordinary if the mean historical frequency of rainbow appearances at the two locations on 8 September at 1830 hrs is 10,000+ years.
 

LilWabbit

Senior Member
The following critique keeps getting repeated so let me articulate the question and the response once more in hopefully clearer terms. The answer provides the main rationale for this thread whilst not implying any particular conclusion.

Question: Are there rational reasons to consider the rainbow appearances at issue on this thread in any way more extraordinary than any other rainbows in history?

Answer: Seemingly yes. Both the exact timing and the location of the rainbow appearances share a seemingly obvious reference point to one person. In this case the Queen.


1. Seemingly obvious reference point to the Queen in the timing of the rainbow: The public announcement of her death at 1830 hrs on 8 September 2022.

2. Seemingly obvious reference point to the Queen in the location of the rainbow: Her two chief residences.

The timing and location of most rainbow appearances throughout known history don't have such seemingly obvious reference points to one person. Hence the event is at least seemingly extraordinary and seemingly non-arbitrary. But not necessarily so upon closer scientific scrutiny should it be demonstrated as just another random coincidence that was likely to happen sooner or later either with this monarch or some other one. This could be done, for instance, by demonstrating that many days on an average September boast a rainbow appearance at those locations and times.

Or, in addition, by demonstrating that the same rainbows were seen far and wide in places with no obvious relation to the Queen.

However, in the latter case (which doesn't seem well-supported by the weather maps we've thus far found), the timing would still remain a commonly identifiable reference point to the Queen even for those observing the rainbows at a greater distance. Plus the main crowds observing them would have still been at Buckingham Palace and the overall location non-trivially been London in England, the Queen's realm of reign, for a widely viewable rainbow appearance on the day of her death.

A high frequency of rainbows appearing at those times and locations on an average September would therefore make the strongest case against the extraordinariness of the event.

Rainbows appearing at other times or other locations are irrelevant unless their timing and location also share a seemingly obvious reference point to one person.
 
Last edited:
Top