On 08/07/2015 12:46, Eskimo Will wrote:

"Norman" wrote in message
...
Eskimo Will wrote:


"Norman" wrote in message
...
Metman2012 wrote:

snip

No doubt there have been changes which may have been significant,
but
then don't the climatologists make allowances for this. I know
that even
such mundane things as SST measurement had to be adjusted by the
method
and material of actually measuring the sea temp - the buckets
used to get
the water had different characteristics and work was done to
homogenize
the results.

Just my 2 pennyworth


As you imply, the sea surface temperature measurement can vary
significantly dependent on what method is used for the measurement.
Unfortunately, the method used is not usually recorded. For
example, on
the Ocean Weather Ships (at least on the British ones) any one of the
following could be used:

- canvas bucket (usually used only in relatively benign
weather)
- rubber bucket (usually used only when under way)
- thermistor loctated somewhere in the engine-room sea water
intake - direct measurement of the engine-room sea water
intake
(turn on a tap and stick a thermometer into the water)

The first 2 methods sampled the water at, or very close to the
surface. The
other 2 sampled the water a few metres below the surface. It was up
to the
individual observer to decide which method was used at each
observation. I
don't know if any comparative tests were ever made. I can't recall any
during my time on the Weather Ships.


So the years of research put into homogenizing the SST dataset was a
waste of
time as vital information would be missing?


I don't know how the homogenisation could be done if the method used
for each
SST temperature measurement was not known.


That is my point. *If* they, (David Parker et al), have made assumptions
about how various ships made measurements over the years then the
integrity of the extremely important global homogenized SST dataset used
to make judgements on how temperature has changed over the decades may
be in doubt. This needs to be followed up as the implications could be
massive.

Will
It was implied to me that they did know how the measurements were taken
(or at least some of them), but this was in the 80s so I forget which
researcher told me. He overheard two of us being somewhat rude about the
data and corrected us (very gracefully) about what was done. I accept
that if the method wasn't recorded then that would cause problems for
homogenization. My colleague is now a chief in CFO/NMC or whatever it's
called now (.