In message , Martin Rowley
writes
... I managed to locate my original post from a couple of years ago: I
think the data are still pertinent.


quote
Following some comments regarding the changes in sea surface
temperatures in recent years, I thought I would dig out the maps
published in the 'Weather in Home Waters' series and compare with
current values (actually 15th February ... roughly in the coldest point
of the annual cycle).

The averaging period used in that publication is based on period: 1854
to 1958 (punched card dataset: US Navy); There is always a problem doing
this exercise in that the data are not homogenous, and with SST
measurement in particular, arguments surround how accurate or otherwise
the data are; indeed the publication I took the long-period averages
from discuss the differences between the 'bucket' method (recommended)
and the 'intake' method.

These maps are apparently based primarily on *intake* temperatures,
which the author states had (at this time) been found to be on average
some 0.5degC *warmer* than 'real' surface SST values (intake water is
taken some way below the waterline, but is subject to warming on being
drawn through & past the ship's internal pipework, into the engine-room
area). Taking the worst case scenario then, the 'historic' map might be
regarded as a little too warm, and the differences found below perhaps
greater. That of course assumes that current values are accurate!

On the Weather Ships there were 3 ways of measuring sea temperatu

1. The canvas bucket method.
2. A tap on a pipe in the engine room. You turned on the tap and held
a thermometer bulb in the flow.
3. A thermistor somewhere in the intake pipework connected to a
readout in the Met Office at the stern of the ship.

They all gave different readings, with the bucket usually being the
lowest. There was a theory that the water in the bucket was cooled by
evaporation before the reading settled. I was never convinced by that.

The "rule" was that the bucket method should be used whenever possible
but stormy weather often precluded that. Also, some observers were less
keen than others about using the bucket method every hour on a filthy
wet night in December :-). What often happened, but not always, was that
the bucket method was used at a main synoptic hour and a thermistor
reading was taken at the same time. Thermistor readings, adjusted by the
difference found at the main synoptic hour, were then used at the
intervening hours until the bucket method was used again at the next
main synoptic hour. In stormy weather, when it was not feasible to use
the bucket method, the thermistor readings were used unadjusted.

The manual readings using the tap in the engine room were logged once
daily but were not normally used in the SYNOPs.

Sea temperatures in the weather ship archive are therefore a random mix
of bucket and intake figures.

Norman.
(delete "thisbit" twice to e-mail)
--
Norman Lynagh Weather Consultancy
Chalfont St Giles 85m a.s.l.
England