HadSST.3.1.1.0 Release Notes

Two bugs were identified in HadSST.3.1.0.0. These have been fixed in HadSST.3.1.1.0.

• Reports from ships given the generic callsign MASKSTID, which denotes a ship whose call sign has been masked in the ICOADS Real Time feed, were not flagged as having generic callsigns. This led to all reports with this callsign being dealt with as a single ship in the uncertainty calculation and consequently, the uncertainty was incorrect. MASKSTID is now included in the list of generic call signs. This leads to a small reduction in the measurement and sampling uncertainty from 2007 to the present and in the total estimated uncertainty.
• The line of code for reading in WMO publication 47 metadata for May 2014 was incorrect, so a default file was being used. In this case, the default file was set to be WMO Publication 47 for May 2014, so the bug had no effect on the output. Nevertheless, it has been corrected.

The net effect of these bug fixes is to reduce the measurement and sampling uncertainty for the period 2007 to present. The sea-surface temperatures themselves are unaffected. Despite the bug fix, SST measurement and sampling uncertainties have actually incresed slightly over the past few years. This effect is most pronounced in the Southern Hemisphere and is most likely due to the drop in numbers of drifting buoys reporting since 2010.


Figure 1: Measurement and sampling uncertainty in monthly global average from 1850 to July 2014. The black line shows the estimate from HadSST.3.1.1.0. The red line shows the estimate from HadSST.3.1.0.0.

Maintained by: John Kennedy Last updated: 14/08/2014 Expires: 14/08/2015 Hosted by: Met Office Hadley Centre

© Crown Copyright