On October 17, 2003, "Rodrigo Severo" <rodrigo.lists@fabricadeideias.com> wrote:
4. Fernando de Noronha island had wrong timezone info as it should include all DSTs for the Pernambuco state after 1988 as Fernando de Noronha island became part of the Pernambuco state after the brazilian contitution of 1988. If someone has any doubt about it, please check decree 96676 (pre 1988) which clearly includes Pernambuco state, Fernando de Noronha and Atlantic islands. Compare with decree 98077 (pos 1988) which clearly includes Pernambuco state and Atlantic islands (doesn't mention Fernando de Noronha because it already mentioned Pernambuco).
I agree with the underlying facts stated, but I have some doubts about the conclusions drawn from them. My concern is that the oceanic islands are so unimportant that they may have been negligently omitted from the decrees. They're in a different time zone from the mainland, which suggests that they might get special treatment. To get more certainty, I searched the Internet for direct testimony about the observation of summer time the oceanic islands. I wasn't able to reach any definite conclusions, because there was too much contradictory information. I wanted to place my findings on record in case anyone else can use them or add to them. Those who don't care about the intricate details may stop reading here. Some background: The phrase "ilhas oceânicas" (oceanic islands; the word "Atlantic" is not used in the decrees) can mean any islands off the coast of Brazil. There are many oceanic islands that are no more than about 10 km. from the mainland. The time zone decrees, however, are probably referring to the five groups that are over 100 km. offshore. Trindade and Martin Vaz are administratively part of Espírito Santo state; Atol das Rocas is in Rio Grande do Norte; Fernando de Noronha and Penedos de São Pedro e São Paulo are in Pernambuco. Fernando de Noronha and the Penedos were a separate territory, called Fernando de Noronha, from 1942-09-02 to 1989-01-01. According to just about all sources, all of these remote islands have UTC-2 for standard time, while the standard times in mainland Brazil are UTC-3 or earlier. They lie between 28° and 34° longitude West, near the central meridian the UTC-2 strip. To save a little space, I'll use the standard abbreviations ES for Espírito Santo, FN for Fernando de Noronha, PE for Pernambuco, and RN for Rio Grande do Norte. I think Sr. Severo's point is that FN was not mentioned in the second decree because it had become part of PE. He believes that after 1988, whenever PE is mentioned, FN is included. That may be so, but I think there's another possible explanation. I think "oceanic islands" always includes FN. The reason that FN is mentioned explicitly in the first decree is that it was then an "entidade da federação", ranking equally with the states. For this analysis, I started from the hypothesis that the decrees mean exactly what they say. If that's the case, the oceanic islands observe summer time on the same schedule as the states they belong to, unless the decree mentions them specifically. I also hypothesized that the oceanic islands all had UTC-2 as their standard time. Then I checked to see if any of the Web sites I found disagree with the hypothesis. Here are the results, starting with the most official sites. http://pcdsh01.on.br/ Web site of the Time Service Division of the National Observatory of Brazil. This is the source for the decrees we've been talking about. According to the decrees, PE and RN both observed summer time every year that it was observed in Brazil up until 1990. They resumed observing it in the summers of 1999-2002, except that they went off it early in the summer of 2000-2001, along with several other states. As for ES, it has always observed summer time whenever any part of Brazil did. The oceanic islands aren't explicitly mentioned in any of the decrees after number 98077. The only problem here is that mentioning the oceanic islands is redundant. http://pcdsh01.on.br/MapaHORVER03.htm The map on this page clearly shows that none of the oceanic islands observed summer time in 2003-2004, not even Trindade and Martin Vaz. The hypothesis says that they should observe it along with their state (ES). http://pcdsh01.on.br/Fusbr.htm http://pcdsh01.on.br/Fusbrhv.htm Tables on these pages state that none of the oceanic islands observe summer time, except for FN, which is supposed to be in UTC-3 in the winter and UTC-2 in the summer. This is probably a mistake, because no reliable source ever shows FN observing UTC-3 at any time. For Trindade and Martin Vaz, we have the same conflict as before. http://www.aneel.gov.br/arquivos/pdf/resumo2001.pdf http://www.aneel.gov.br/arquivos/pdf/periodos2001.pdf http://www.aneel.gov.br/arquivos/pdf/horadotados2001.pdf http://www.aneel.gov.br/area.cfm?id_area=65 These pages are from the Brazilian National Electrical Energy Agency. They show that FN and the oceanic islands don't observe summer time. This probably refers to 2002-2003. Same problem with Trindade and Martin Vaz. http://www.estado.estadao.com.br/edicao/pano/96/02/08/provs502.html This appears to be an official Web site for São Paulo State. The page is dated February 8, 1996. It talks about the energy savings attributable to summer time for 1995-1996. It says that, after summer time ended, FN and other oceanic islands remained (ficam) 60 minutes ahead of Brasilia time. This would seem to imply that they observed summer time along with Brasilia. If so, that would conflict with the hypothesis. But there is an ambiguity in Portuguese. While the primary meaning of "ficar" is "to remain", it is often used to mean simply "to be". The statement may not really say anything about summer time in the oceanic islands. http://www2.uol.com.br/JC/_1999/2712/tu2312d.htm This article from the Recife "Jornal do Commercio" for December 23, 1999 implies that FN was observing summer time that year. That fits the hypothesis. http://www.formosaonline.com.br/geonline/textos/geografia/cartografia03.htm This educational site says that FN and other oceanic islands do not observe summer time; no date is given. If the page is up-to-date, and refers to the period since February, 2002, it fits the hypothesis except for Trindade and Martin Vaz. http://groups.yahoo.com/group/fernando-de-noronha/message/41 http://groups.yahoo.com/group/fernando-de-noronha/message/42 These are two consecutive messages from a thread regarding airline schedules in October, 2000, the month when several states decided to go off summer time a week or two after they went on it. The first message says that FN is a special case, and it follows PE time but has its own time zone. If that makes any sense at all, it must mean that FN is always one hour ahead of PE. The second message says that departures from "Natal/FEN" will occur an hour earlier than the time printed on the ticket, starting on 2000-10-22, when Natal stopped participating in summer time. "Natal/FEN" probably means that the information applies both to Natal, the capital of RN state, and FN. The decrees say that RN state went off summer time on 2000-10-22, but PE did it a week earlier, on 2000-10-15. The only conflict with our hypothesis is whether FN turned back the clock on the 22nd or the 15th of October. http://www.flying.scuba.nom.br/info_aspec_geog.htm This tourism site says that FN doesn't observe summer time; there is no clue as to its date. No conflict, assuming the date is recent. http://www.pisa.tur.br/roteiro_novo/roteiros/noronha/nornha_8dias.htm This tourism site says that FN did not observe summer time in 2002-2003. No conflict. http://www.jferreira.com.br/turismo/informacoes/dicas/005.htm This tourism site says that FN does not observe summer time. No date is given. No conflict, assuming a recent date. http://www.getaway.com.br/Diferen%C3%A7as.htm This tourism site says that FN did not observe summer time in 1999-2000, but PE did. If our hypothesis is right, FN should have observed summer time that year. http://www.flytime.com.br/horario.htm This tourism site says that FN did not observe summer time in 2000-2001, but the other oceanic islands did. The hypothesis says that FN should have observed it, and Trindade and Martin Vaz shouldn't. http://www.espaco-do-ceu.com.br/Hora%20de%20Ver%E3o.htm This astrology site has a table summarizing the decrees. It differs in minor ways from the decrees. In the 1988-1990 period, it doesn't mention FN, but leaves us to conclude that it is covered as part either of PE or the oceanic islands. It mentions that PE observed summer time in 2000-2002, but adds FN explicitly for 2001-2002. If the omission of FN in 2000-2001 is significant, that would contradict the hypothesis. www.radioescotismo.com.br/Fusos.htm This page, devoted to Brazilian radio scouts, says that FN did observe summer time and other oceanic islands did not. It has a copyright date of 2001 and a "last modified" of August 26, 2002, so it probably refers to 2000-2001 or 2001-2002. In either of those years, the hypothesis says that all of the oceanic islands should have observed it. http://www.novomilenio.inf.br/santos/h0086.htm Novo Milênio is an electronic newspaper. This page is a reprint of an article from the Santos newspaper "A Tribuna" for January 30, 1978. It says that standard time for the oceanic islands (specifically, FN, Atol das Rocas, and Trindade) was UTC-2 at some prior time, but now (in 1978) it was UTC-3. This would conflict with our hypothesis, but I think it must be a mistake. I have copies of Almanaque Abril, the Brazilian reference book, for 1978 and 1985. They both say that the oceanic islands had a standard time of UTC-2. The tz ftp site has an archive of old e-mails to the tz mailing list. Stefan Johansson wrote on 1999-10-02 that he had a newspaper clipping from O Estado de S. Paulo of that date, regarding the new decree for 1999-2000. Without quoting from the article, he summarizes his conclusions in a table. On one row of the table, FN is shown as not observing summer time that year; on another row, PE (except FN) does observe it. This is inconsistent with the hypothesis. Gwillim Law