Metadata about the BAS ozone data

General

A paper describing the instrument meta data and observations is in preparation.  

The data should be credited to J D Shanklin, British Antarctic Survey, Madingley Road, Cambridge, England. CB3 0ET

Station names:  

Faraday, Antarctica.  65°15'S, 64°16'W.  Height approximately 16m.  Originally known as Argentine Islands (which is actually a group of islands off the coast of the Antarctic Peninsula) the station became known as Faraday in the mid 1970s.  In 1996 it was handed over to the Ukraine and its name changed to Vernadsky.  In publications, the name Faraday should be used when referring to data prior to 1996 and Faraday/Vernadsky thereafter.
Halley, Antarctica.  75°35'S, 26°13'W.  Height approximately 33m.  Please note that the station name is Halley and this should be used in all publications.  The geographical feature known as Halley Bay (which the station was called when first set up in 1956) no longer exists.  The station is on a floating ice-shelf and moves over time.  The position given is the current one, but the actual location is likely to be within 15km of this.  Halley 6 become operational in 2012 February.  The station was moved about 15km to a new location in 2017 (75°34'S, 25°49'W), but became a summer only station due to the risk of calving on the ice-shelf.
King Edward Point, South Georgia.  54°16'S, 36°30'W.  Height approximately 5m.  The station at King Edward Point is near Grytviken, in Cumberland Bay, South Georgia.  Observations ceased due to military action in 1982 April.
Rothera, Antarctica.  67°34'S, 68°08'W.  Height approximately 32m.  The name Rothera should be used in publications.  The SAOZ instrument can make observations throughout the year.

Observing season

Approximate value of mu at Local Apparent Noon

Instruments

Datasets - ozone

Observations at all stations are recorded in UTC.  All observations at Halley and Vernadsky are made with the Dobson ozone spectrophotometer.  Ozone values from refurbished or recently installed Dobsons are likely to undergo frequent revision until the instrument constants are well determined.  Ozone values from the old instruments are also likely to undergo revisions following both the on-site intercomparisons, and intercomparisons at Hoenpeissenberg.  All ozone values are given in Dobson Units.

Halley Observations with Dobson 103  (with valve electronics and mechanical commutator) ceased at Halley 5 on 2005 December 26 and observations with Dobson 73 (with modern electronics) commenced on 2005 December 24.  Observations with Dobson 73 ceased at Halley 5 on 2012 February 13 and observations began with Dobson 31 at Halley 6, some 20km away, on 2012 February 4.  The short overlap periods allow the instrument constants to be slightly adjusted if necessary.  The station was moved in 2016/17.  Observations were made with Dobson 73, operating in automated mode, for a short period from 2018 January 25 to 2018 February 8.  Dobson 31 ran from 2017 December 7 to 2018 February 26.

Vernadsky.  Observations with Dobson 31 (with valve electronics and mechanical commutator) ceased at Vernadsky on 2005 March 29 and observations with Dobson 123 (with modern electronics) commenced on 2005 March 27.    The short overlap period allowed the instrument constants for both instruments to be slightly adjusted.  During the overlap periods a mean of both instruments is given.  

Current year and recent Provisional daily mean ozone values
The provisional daily mean data for years since 1972 is provided in a format for generating daily WMO CREX messages.  The file has nine columns, giving the month (MM), day (DD), modified Julian date [approximately days since 1900] (JJJJJ), mean ozone value in Dobson Units (XXX), standard deviation of the individual values (SD), number of readings (N), average mu value (MU), mean hour of observation [add 12 to get UT] (HOUR), time-span in minutes from first to last observation (SPAN).
Current year and recent Provisional individual ozone values
The provisional individual ozone values for years since 1972 is provided in annual files. Each file has ten columns, giving the month, day, hour, minute (in UT), modified Julian date [approximately days since 1900] (JJJJJ), the mu value, the solar zenith distance, a code for the observation type (see below), the standard Dobson code for the observation type (see below) and the computed ozone value in Dobson Units.
Daily
The daily data sets were divided into seasonal blocks starting on August 1st each year and ending on 30th April.  They contained the provisional ozone data for 1972 - 2012  obtained using the Dobson ozone spectrophotometer. The definitive instrument and zenith calibration values have not yet been determined. A daily mean value was given, using all the observations for that day, assigning equal weight to each. The final daily values may differ from those given here, but should not differ from them by more than 5%.  A few moon observations which have been carried out outside this period are not included in this data set, but are available on request. A value of 0 indicates that no measurement was made on that day. Some experimental zenith observations at low solar elevations are included from April 1993 onwards, these have lower accuracies than normal observations. The Halley data for 1987/88 includes the results from ozonesonde flights.  These files have been replaced by the more complete files described above.
Monthly
The monthly data sets run from August to April.  Data for 1957 - 1972 was published in BAS Scientific Report No 90 in 1975 and used ozone absorption coefficients determined by Powell, which are the same as those of Vigroux for AD, but are otherwise different.  These ozone values have been approximately corrected to the Bass-Pauer coefficients using the WMO recommended factor of 0.9743.  Note that this correction factor is for the US Standard Atmosphere, however the Antarctic atmosphere is some way from this and the correction could be improved.  Data after this period is reduced using the standard Bass-Pauer coefficients as published in Komhyr et al.
Rothera
The first SAOZ instrument at Rothera measured total column nitrogen dioxide and ozone from 1997 to 2007. The values given here are preliminary and only show general trends in ozone amounts.  In particular, the calibration of the instrument is such that it reads about 10% lower than a Dobson at values around 100 DU.  A new SAOZ type instrument was installed during the 2006/7 season and it became the operational instrument from 2008 January 1.  The instrument constants for the new SAOZ were revised in mid July 2008, and previously published values were increased by about 2%.  Following a station power cut in mid June 2011, some data was lost and when data recording resumed an incorrect time was recorded.  This has been corrected, however the data between 2011 June 16 and 2011 June 29 may contain residual errors.  The instrument failed on 2017 January 6 and was replaced with the spare.

There is also a Bentham spectro-radiometer at Rothera, which can be used to compute ozone levels.  The original instrument was destroyed in a fire on 2001 September 28, however a replacement was commissioned and became operational on 2003 March 3.  For further information on this data set and permission to use it please contact Paul Geissler at BAS.  All values have been updated to the TOMS 8 algorithm.  The uv data is available from the WOUDC.

Dobson ozone observation codes:

First code: 1 - ADDS, 2 - CDDS, 3 - CDFS, 4 - CDFM, 5 - ADZ, 6 - CDZ, 7 - CC'Z, 8 - CDFFS, 9 - CDDS high mu, 10 - CDZ high mu
Second code: Standard LS form where:
L = 0 - AD, 2 - CD, 6 - CD focussed, 9 - CDA, 16 - CD filter focussed
S = 0 - Direct sun, 1 - Direct moon, 2 - Zenith blue, 3 - Zenith high cloud, 4 - Zenith middle cloud, 5 - Zenith low cloud (inc Ns), 7 - Zenith fog, 8 - Stratospheric cloud

Datasets - temperature

Temperature values shown in the graphs are given in °C.  The 70 hPa pressure level is close to the height of the maximum ozone concentration, however this is not always reached by balloons, particularly during the winter.  The 100 hPa level is therefore used as a reference.  In cases where the sonde reached at least 125 hPa the temperature may be extrapolated to 100 hPa.  The Peninsula data uses a mean of stations whenever more than one is available.  Whilst these approximations are appropriate for these ozone studies, the data are not suitable for pure climate studies.  The READER data set is our quality controlled reference dataset for temperature.  Over the last 30 years the mean 100 hPa temperature has declined in most months.  The most notable change is in November at Halley.

Halley  Radiosondes are released daily at Halley.  The nominal station elevation is currently 20m (2018), but was 30m.  The processing system was upgraded from Vaisala MW31 to MW41 in 2018 January.

Antarctic Peninsula   The record is a composite from several stations and the balance has changed with time, so this data set should not be used for definitive climate studies.    The Polarstern was stationed in the western Weddell Sea from early December 2004 to early January 2005 and launched one or two sondes each day; she was again in the area in January and February 2013; when in the area her routine flights are included.  There have been a few ozone sonde flights from Ferraz, but these have not been included in the record as they were not distributed on the GTS.  No sondes are currently launched at Bellingshausen or Vernadsky.  There are flights from Marambio as part of an ozone sonde programme, however these did not always get onto the GTS in the early days of the program.  A radiosonde programme started at Frei in 2017 December, with a few flights from O'Higgins.  There has been a radiosonde programme of several flights per week at Rothera since 2003 March, increasing to daily flights for the QUOBI project from mid June to mid October 2003.  Strong winds on 2004 October 2 & 3 damaged the hangar at Rothera and this significantly restricted the conditions under which sondes could be launched.  Sonde type changed from RS80 to RS92 in 2007.  A new launch facility become operational in 2007, and the processing system was upgraded from an MW15 to an MW31 in 2008 January, though this failed during the winter and flight control reverted to the MW15.  Vaisala RS92 radiosondes are currently released four times a week at Rothera and these form the primary contribution to the record.  The system has been upgraded from MW31 to MW41.