Update, 22 Jan 2017: as a result of this comparison of data sources, some data values used to compute the CRU SOI have been updated:
The original sources for the Tahiti and Darwin sea-level pressure (SLP) data used to calculate the CRU SOI are documented in the original references below. Since their publication, the SOI series has been extended by appending monthly SLP observations for Tahiti and Darwin sourced mostly from monthly CLIMAT reports, but occasionally from Met Services or CPC if the data for a particular month were not included in a CLIMAT report.
The Tahiti and Darwin SLP series assembled by CRU have been compared with data obtained from three other sources, all downloaded on 13 Jan 2017:
CRU data run from 1855 (Tahiti, with some missing values in 1855 and 1860; some of the early values have been infilled using neighbouring stations, see references below) and from 1866 (Darwin). BoM data run from 1876. CPC data run from 1951. Met Office CLIMAT data run from 2000.
The CRU data are compared with the other three sources for the full period of their overlap. The comparisons are listed in Comparison of Tahiti and Darwin SLP sources.
Available values are listed for each month. When a source differs from the CRU value, the difference is also listed (other source minus CRU value) preceded by "=D". The upper part of the comparison file lists all those months where at least one other source differs by at least 0.3 hPa from the CRU value. The lower part of the comparison file lists all years and all months, regardless of any differences or the size of any differences.
There are 4 months since January 2000 when the CRU value differs from the CLIMAT value by 0.1 hPa but matches the CPC value. These represent cases where the CLIMAT value was not available at the time and the CPC value was used to update the CRU series instead. One month shows a larger difference: Feb 2006, the CRU value is 0.3 hPa lower than all three sources. Since the CRU value should match at least one of these sources, it is likely that a transcription error affected one digit. The CRU value has now been updated to match the CLIMAT value.
Prior to 2000 when only BoM and CPC are available for comparison (and only BoM prior to 1951), there are occasional differences that exceed 0.3 hPa. Most occur prior to 1916. These reflect differences between the original sources (there are multiple possible reasons, including differences in how missing values are dealt with, different approaches to creating a homogeneous series, calculation of daily values from different times of observation during the day) and therefore these values are left as originally published in the CRU papers listed below (see those for sources). BoM describe their source as "TAHITI MSLP is provided by Météo France interregional direction for French Polynesia".
There are 16 months since January 2000 when the CRU value differs from the CLIMAT value but matches the CPC value (or the BoM value). These differences are no larger than 0.2 hPa and represent cases where the CLIMAT value was not available at the time and the CPC (or occasionally the BoM) value was used to update the CRU series instead. Two months show larger differences: Mar 2006, the CRU value is 1.0 hPa lower than CLIMAT, 0.9 hPa lower than CPC and 0.8 hPa lower than BoM; and Sep 2011, the CRU value is 0.3 hPa higher than CLIMAT and 0.2 hPa higher than both CPC and BoM. Since the CRU value should match at least one of these sources, it is likely that a transcription error affected one digit. In both cases, the CRU value has now been updated to match the CLIMAT value.
Prior to 2000, there are very few differences that exceed 0.3 hPa between the CRU values and either the BoM or CPC values. However the two largest differences in the comparison do occur for Darwin: Feb 1883, the CRU value is 3.6 hPa higher than BoM; Jul 1979, the CRU value matches BoM but is 3.5 hPa higher than CPC. The latter value has been checked against the daily observations at synoptic hours 0900 and 1500 and this confirms that the CRU and BoM values are correct, and the CPC value is in error for Jul 1979 (personal communication, Blair Trewin, BoM, 15/11/2006). All values prior to 2000 are left as originally published in the CRU papers listed below.
The three monthly values updated following this comparison of sources are not in the standardisation period (1951-1980), so they do not affect the calculation of the SOI overall. Only the three months are affected (Feb and Mar 2006, and Sep 2011). The standard deviation of SLP is about 1 hPa in these three months, so the changes represent about 0.3 SD (Feb 2006 and Sep 2011) and 1 SD (Mar 2006) for one of the two stations used to compute the SOI.
The resulting SOI changes are:
Data files prior to this update are available for comparison: