Is DST Causing Confusion in Determining UTC for Historical Events?

[kepler]

Hi,

A peculiar situation ( every time DST ends ): DST ( = UTC+1 ) ended, for example, in 1979 - in Portugal - the 30th September at 2 AM - it lasted until 1h59m59s AM and, one second later, corresponded automatically to 1 AM ( UTC 0 ). Ok, that's fair. But now, let's supose that we know of an event - of any kind, important or not - that ocurred at 1h30m AM in that day. Which system do we apply to know the exact UTC of the event?
It could have happened during DST, 30 minutes before the change of time - so, since DST was active, it would have happened at UTC 0h30m AM ; but it could have happened after the change, 30 minutes after DST ended - UTC 1h30 AM ( since the correction was made ). So, what will it be? Shall we throw a coin?

This time issues give me an headache...

Kind regards,

Kepler

 

[ohwilleke]

You should do what any good scientist does, make your units clear when there is a chance of ambiguity.

 

[kepler]

yes, but...

I agree completely - but the problem is that the right construction of constants and variants decreases as we go into the past. Look at the example of the babylonian calendar in the Jesus epoch; it depended of witnesses of one phase of the Moon that traveled hours or days, by foot, to Jerusalem to tell their 'scientists' what was happening some miles from there. And it would be different in different towns. Easter was celebrated during several days because of those differences. Trying to understand or make a calendar rule to that period is insane ( you have to study the astronomical events hard, like Moon phases, and still be carefull... ).

We're not perfect; the computer was one of the most importants discoveries of the XX century, and still it only counted years until 99 ! ( two digits ) - wasn't it obvious that a serious calendar should manage such developed machine?

Or the failed astro mission because of some orbital and launch calculations were made based in miles and quilometers, both at the same time?

I agree with you, but we're humans - and make constant mistakes. In the past, present and - be sure - in the future.

Cheers,

Kepler

 

[Nereid]

In the case of time, there is an agreed, constant, unambiguous 'clock'. Ideally, all astronomers use this, and there is no confusion. Where it does arise, it is usually due to human error (e.g. insufficient accuracy in the description).

 

[ohwilleke]

DST has the virtue of being a non-cumulative problem. If you are looking at ancient astronomical events, you know that DST won't be an issue since it is self correcting. The maximium problem is one hour no matter how far back you look. Moreover, ancient measurements don't even have standard time zones (those were developed in response to railroads) and don't have an accuracy sufficient to make an hour of DST adjustment relevant.

 

[kepler]

well...

Forgive me to disagree. First I was only giving an example in one context. But let's supose we go far in the past; one hour, in astronomy, is a century - just like 10^(-6) is huge in curve fitting in Math. Have you ever seen how much the moon "moves itself" in one hour? 5 mihutes are sufficient to miss the veracity of an eclipse or not. The calculations of planetary positions depends on the analisys of eclipses ( see Stephenson and Morrison work), since they came from the extrapolation or tabbed values of delta-t. The extrapolation results from factors like not knowing if there's an hour in advance or not.

For this reasons - and others - the astronomers build a timeline continuous. Universal time is not used in planetary computions, but TDB is. The rest is research. From the calculation of some past eclipses we are able to determine historic events or periods of great interest to archeology and ancient history.

What simplifies this issue in articular, is that, normally, the ancients weren't concern with DST - like the equinox of spring ( only in 1582 the pope admited that there was a difference of a dozen days, aprox., and changed it - after 16 centuries...).

Don't try to simplify something it certainly isn't...

Kind regards,

Kepler

 

[turbo]

Forgive me to disagree. First I was only giving an example in one context. But let's supose we go far in the past; one hour, in astronomy, is a century - just like 10^(-6) is huge in curve fitting in Math. Have you ever seen how much the moon "moves itself" in one hour? 5 mihutes are sufficient to miss the veracity of an eclipse or not. The calculations of planetary positions depends on the analisys of eclipses ( see Stephenson and Morrison work), since they came from the extrapolation or tabbed values of delta-t. The extrapolation results from factors like not knowing if there's an hour in advance or not.

For this reasons - and others - the astronomers build a timeline continuous. Universal time is not used in planetary computions, but TDB is. The rest is research. From the calculation of some past eclipses we are able to determine historic events or periods of great interest to archeology and ancient history.

What simplifies this issue in articular, is that, normally, the ancients weren't concern with DST - like the equinox of spring ( only in 1582 the pope admited that there was a difference of a dozen days, aprox., and changed it - after 16 centuries...).

Don't try to simplify something it certainly isn't...

Kind regards,

Kepler

Dear kepler, let's assume that on some place on Earth, the supernova that formed the Crab Nebula became apparent on Earth at midnight on July the 4th ( a great day for fireworks!) in the year 1054. Given the state of timekeeping then, and the reliance on local time by astronomers of the day, how can DST possibly have an effect on the historical record? As ohwilleke pointed out, this is not a cumulative error, and an offset of one hour in 951 years is not a daunting problem. If a Chinese astronomer claimed to have observed the brightening at 02:11 local time 951 years ago, how would DST invalidate his observation? By the way, I'm being a bit silly here - the Chinese astronomer obviously would have rounded off his observation and claimed that the brightening occurred somewhat before "a quarter past the hour" using the ancient time-keeping standard that would be adopted by all the civilized world after pocket watches became widely available. OK, I'm still being silly - having a reference date is still pretty darned good, and the time-standard is irrelevant. Heck! I live in Maine, and it gets dark here almost 2 hours before it gets dark on the other side of the Eastern time zone - look at a map to see that. Geography and default inclusion in the EST zone mess us up - at least DST eases some of the effect in the summer.

 

[Hurkyl]

Some (most? all?) time zones have different units for the daylight time and standard time. For example, the eastern U.S. has "Eastern Daylight Time" and "Eastern Standard Time". (Though most people forget to use EDT like they're supposed to)

I thought that UTC didn't adjust for daylight savings time?

 

[kepler]

It doesn't; England does. Only by chance the meridian of Greenwich, to which UTC refers, is located there.
UTC, the previous GMT, is a time scale for "ruling" correspondences between the solar day in all locations on Earth. Even so, it's not accurate in astronomical terms; we take a day as the period between two transits of the sun in the same location - and we assume the value of 24 hours. But, at least 3 factors contribute to the innacuracy of this value: the fact that the Earth moves in an orbit around the sun from place, in that period; the fact of the rotation of the Earth is less than 24 hours ( aprox 4 minutes less ); and the fact of the lunar gravity "delays" the rotation of Earth because of the tides effect. There are more factors. But, all together, resume in huge time differences in the past, present and future. But that's a problem we have now; for the ancients, midday was always midday. The events are referred in Universal Time. But astronomers and historians must take in account the difference to know accurately when things really happen. For example: At this point we have a difference of about 64 secodes. In the year 1054 the dif. was of less than 22 minutes ( enough to miss an eclipse - one of the most important events in astrology and astronomy ). Of course, it all depends of the event we're talking about: one hour in the supernova example is not relevant. But 22 minutes are relevant, for instance, in viewing a comet in a particular day or not above the horizon, rising or setting. Comets were truly important in the observations of the astrologers. One war could be lost because the delay of one day - the moral would be lost :) - but it's a fact ! I have some piety of astrologers back then ( even far more than the XI century - in the years before Christ ). If they missed an event, or were wrong in their predictions, the king would chop his head off - in China, they would cut their head with a saw ( not a pretty way of dying...).
Either way, to resume: DST was not a problem up until now. The days were adapted according to midday. It's our problem now, to see if there's the need of make extra calculations to see the veracity or not of a particular event.

One hour, one minute counts...it's like your influence in the Cosmos: when you get up from a chair a star moves in the sky...

Kind regards,

Kepler