Corrections to surface station data

This is a list of errors that relate to the surface station data. All of the tables as they were at the start of February 2009 can be found here for temperature and pressure and here for winds. Any suspected errors should be reported to Steve Colwell

The GTS values for Mawson station were loaded into the GTS archive multiple times and this resulted in misleading percentages being displayed for 2006 and 2007. The error has been corrected and this has made no difference to the values that were displayed on the READER website, though the percentages that were displayed have changed. The correction also means that the values for May 2006 (89%), June 2006 (88%), October 2006 (88%), May 2007 (82%) and June 2007 (82%) will no longer appear in the downloadable text file as there is insufficient reliable data to calculate an accurate mean.

A duplicate copy of GTS values for most stations for 13th November 2007 at 06UTC and 9th August 2008 at 00UTC has been found in the GTS archive. This has been corrected but it is not expected that it will make any difference to the monthly mean values that are already on the READER website. It may make a small difference to the 00 and 12 tables.

An error has been identified with the Bellingshausen data for February 2003 where multiple copies of the data for 00UTC on the 1st existed in the database. This has been corrected and it has made a change to the data: the monthly mean temperature has changed from 2.0 to 2.1, station level pressure from 986.0 to 985.4 sea level pressure from 987.9 to 987.3 and wind speed from 11.6 to 12.8

It was reported by Nicholas Lewis that the McMurdo 12Z temperature file contains some values that are obviously pressure values for August 1994, February 1995, December 1995, December 1997 and March 1998, also that there is a discrepancy between the averages calculated using data from the 00, 06 12 and 18 observation tables and the values displayed in the All tables. The values in the All table are correct and the values in the other tables will be corrected.

Nicholas Lewis has raised a concern that the temperature values in 1975 from Adelaide Island seem low, we agree with this and the values will be checked against the original values in the BAS archives.

(20 Feb 2009) Incorrect temperature values for Adelaide Island for 1975 have been corrected. Incorrect values for February 1961, August - December 1961, January 1962, February 1962 and January - May 1976 have been removed.

(3 Mar 2009) Incorrect January 1988 Gough Island temperature of 1.0 has been replaced with correct value of 15.2.

(3 Mar 2009) GTS processing has been improved to ensure that duplicate observations are not used.

(3 Mar 2009) Where a value appears in blue in the All table any red values in the corresponding 00, 06, 12 and 18 tables have been removed.

(5 Aug 2009) AWS temperature values from 1997 onwards and pressure values from 2001 onwards for Scott Base have been replaced with the climatological means issued from Scott Base. The values from the AWS have been added to the AWS data page.

(6 Jan 2011) Temperature data for Halley from June 2008 onwards has been updated with values calculated from the PRT (platinum resistance thermometer) rather than from the HMP combined temperature and humidity probe. This has increased the monthly values by between 0.1 and 0.3 deg C. A copy of the old values can be found here

(7 Mar 2013) Values for Byrd station have been updated with observations obtained from the NCDC

(1 Apr 2014) Values for Syowa have been adjusted by 3 hours to make them UTC. Temperatures from before the adjustment can be found here