[tz] question re zone America/Punta_Arenas in Chile
tim at timtimeonline.com
Tue Mar 15 03:27:30 UTC 2022
On Fri, 11 Mar 2022 at 12:39, Alois Treindl via tz <tz at iana.org> wrote:
> I wonder on which basis the 1946/47 DST was left out for Puntas_Arenas?
> Is that documented somewhere?
Although it doesn't appear to have been documented in our extended
commentary, there's indeed a clue in the short line-comment that reads
It can sometimes get confusing looking at UNTIL lines, but the only place
the UT offsets of the two zones actually deviate in our 2021e data is the
roughly six-week period from through 1946-07-15 through 1946-08-31,
inclusive. Compare the extract from America/Punta_Arenas:
-5:00 - -05 1942 Aug 1
-4:00 - -04 1947 Apr 1
-5:00 - -05 1942 Aug 1
-4:00 - -04 1946 Jul 15
-4:00 1:00 -03 1946 Sep 1 # central Chile
-4:00 - -04 1947 Apr 1
So what happened on 1946-07-15? Well, sure enough, an emergency decree
which only applied to the central part of the country:
Núm. 3,891. Adelanta la hora oficial en el territorio nacional comprendido
entre las provincias de Coquimbo y Concepción, inclusives, por el plazo de
No. 3,891. Advancing the official time in the national territory between
the provinces of Coquimbo and Concepción, inclusive, for a period of 30 days
Specifically (my Google-assisted translation):
"In the national territory between the provinces of Coquimbo and
Concepción, inclusive, the Official Time will be advanced by sixty minutes,
for a period of 30 days, starting from 24:00 on July 14 [of the current
That would include Santiago, which advanced to -03, but did not reach so
far south as Magallanes, which would have stayed on -04.
The decree contemplated "[t]hat this advancement of the Official Time, even
though it has been proposed for the cities of Santiago and Valparaíso only,
must be agreed with that of other cities, due to the connection of various
activities that require it, such as, for example, the operation of rail
As the 30 days reached their expiry, the time change was extended:
Núm. 4,506. Prorroga hasta el 31 del mes en curso la vigencia del cambio de
hora oficial en el territoria nacional comprendido entre las provincias de
Coquimbo y Concepción, inclusives
No. 4,506. Extending until the 31st of the current month [August] the
validity of the official time change in the national territory between the
provinces of Coquimbo and Concepción, inclusive
Later in August, a bill reunified the country's clocks:
"The official time for the entire Republic will be from 1 September to 31
March of each year, the meridian of the Astronomical Observatory of Lo
Espejo, advanced by 42 minutes and 45 seconds, which corresponds to the
20th time zone, four hours west of Greenwich, and will be called Summer
Time. From 1 April to 31 August, it will be that of the Lo Espejo
Observatory Meridian delayed by 17 minutes and 15 seconds, which
corresponds to the 19th time zone, five hours west of Greenwich, and will
be called Winter Time."
Technically, this bill came into force "from the date of its publication",
but given that that was sufficiently close to the 1946-08-31 expiry of the
extended emergency act anyway, and the rest of the country would have
already been on -04, it seems just as likely that clocks changed in the
center of the country on 1946-08-31 24:00, as previously scheduled. Under
this law, the 1946/1947 observance of Summer Time at -04 lasted for all six
of its months, but our data show the observance of Winter Time at -05 was
cut short after about seven weeks on 1947-05-21 when -04 was adopted as the
new Standard Time.
This differs somewhat from our horaoficial.cl source
which implies the divergence *began* on 1946-09-01 rather than *ended* on
it, but the legal texts seem clear to me: Our 2021e data is right.
On Sat, 12 Mar 2022 at 14:44, Paul Eggert via tz <tz at iana.org> wrote:
> Thanks for pointing out what appears to be a transcription error when I
> created America/Punta_Arenas in 2016. I installed the attached proposed
> patch into the development version.
Unfortunately, I believe this means that the data change here should at
least be reverted. We can also better specify the transition dates and
expand commentary as necessary.
And while it's before our 1970 cutoff for zone creation, those tracking
expanded pre-1970 data should be keen to only put the correct areas on -03
for these few weeks in 1946. It really was only the middle bit on -03, not
the south, and not the north.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the tz