There's a piece in today's Times - the Thunderer column - that talks
about the latest furore over our July temperature "record".
A temperature of 38.1C was recorded in Tonbridge, Kent, on July 22,
1868, listed in The English Climate by HH Lamb has now been struck from
the records because it apparently was not recorded to modern standards.

I have no idea of the author's scientific background but he states that
more records are being broken now because of the presence of more AWS
sites. He also puts forward the argument that "compared with old
thermometers, modern recording equipment is capable of registering very
brief increases in temperature of a few seconds duration".
I am not sure that this argument carries much weight.

Comments anyone.

Scott W wrote:
There's a piece in today's Times - the Thunderer column - that talks
about the latest furore over our July temperature "record".
A temperature of 38.1C was recorded in Tonbridge, Kent, on July 22,
1868, listed in The English Climate by HH Lamb has now been struck from
the records because it apparently was not recorded to modern standards.

I have no idea of the author's scientific background but he states that
more records are being broken now because of the presence of more AWS
sites. He also puts forward the argument that "compared with old
thermometers, modern recording equipment is capable of registering very
brief increases in temperature of a few seconds duration".
I am not sure that this argument carries much weight.

Comments anyone.

The author is right about the presence of more sites so that a
high temperature is less likely to be missed but I know nothing about
the response time of AWS thermometers. Even if as short as he claims I
can't see this making a great deal of difference. The Tonbridge
"record" was recorded in a Glaisher stand which is like a Stevenson
Screen with the door open. This can make a considerable difference as
any conscientious weather recorder knows.

Tudor Hughes, Warlingham, Surrey,.

He also puts forward the argument that "compared with old
thermometers, modern recording equipment is capable of
registering very
brief increases in temperature of a few seconds duration".
I am not sure that this argument carries much weight.

Comments anyone.

I would have thought that any old max-min thermometer was
'capable of ....'. Whether such a thermometer is of acceptable
accuracy is an entirely different question.

Anne

Scott W wrote:

I have no idea of the author's scientific background but he states that
more records are being broken now because of the presence of more AWS
sites. He also puts forward the argument that "compared with old
thermometers, modern recording equipment is capable of registering very
brief increases in temperature of a few seconds duration".
I am not sure that this argument carries much weight.

Comments anyone.

I would have thought the response time of an automatic thermometer to a
short-period change would be much shorter than a standard mercury
thermometer? I'm no expert, I'm sure others on here with
instrumentation knowledge would say more. But if this is the case, for
the same day, if there are short-term fluctuations in temperature on
top of the daily "sine wave" of temperature, then surely the modern
thermometers with a better response time will record higher
temperatures?

Going on to a side-topic, the same sort of argument in terms of spatial
distribution of sensors, observations, is being put forward on the
current "is global warming affecting hurricane numbers" which is so
pertinent to the industry I work in (assessing natural catastrophic
risk for insurance). Clearly hurricane numbers have risen since 1995 as
the atlantic thermohaline circulation has switched into another "warm
phase" of SSTs, but there is the general issue of whether storms going
back to even as recently as the 1970s and 1980s were "undersampled"
owing to lack of observations. One of the tenets of global warming on
hurricane activity is that numbers may not necessarily increase, but
their intensity does.

Now that the Hurricane Hunter planes scan hurricanes more frequently,
we're suddenly seeing all these records going - longest time for a
storm as an intense hurricane, smallest eyewall diameter, lowest
latitude for an intense hurricane - which certainly suggests that
there's an element of better recording of hurricanes that may give the
false impression that global warming is leading them to be more
intense.

Richard

"Richard Dixon" wrote in message
ps.com...

Scott W wrote:

I have no idea of the author's scientific background but he states that
more records are being broken now because of the presence of more AWS
sites. He also puts forward the argument that "compared with old
thermometers, modern recording equipment is capable of registering very
brief increases in temperature of a few seconds duration".
I am not sure that this argument carries much weight.

Comments anyone.

I would have thought the response time of an automatic thermometer to a
short-period change would be much shorter than a standard mercury
thermometer? I'm no expert, I'm sure others on here with
instrumentation knowledge would say more. But if this is the case, for
the same day, if there are short-term fluctuations in temperature on
top of the daily "sine wave" of temperature, then surely the modern
thermometers with a better response time will record higher
temperatures?

I cannot add anything in respect of the physics of instrumentation;
however, it should go almost without saying that AWSs have been
designed to emulate as closely as possible the output of manual
stations(1). Many (most?) AWSs in the national network have control
mercury-in-glass thermometers in the screen alongside the electronic
sensors, and the output from the two types, once properly
calibrated, is rarely more than 0.2 degC apart. That's actually
close to the accuracy of m-i-g thermometers themselves. Often
(and I speak from experience) the two give identical readings.

Sure, they're *capable* of quicker response times, but the
capability is not used - not needed - when emulated m-i-g
thermometers.

It is also true that a properly calibrated and maintained AWS
removes the possibility of accidental or deliberate mis-reading
by humans, which would tend to remove some extreme
readings from the corpus of data.

Any implied suggestion that there are *more* official stations
because of the arrival of AWSs should also be stamped on.
There are many fewer stations in the climate network now
compared with 25 or 50 years ago, although their distribution
may be geographically rather more even.

The Tonbridge max was achieved, not on a Glaisher stand, but
in a boxed screen, completely open at the bottom, and above a
gravel pathway, all in a smallish kitchen garden enclosed by a
brick wall. Although the observations were made accurately and
assiduously by Dr Fielding, and the dataset is probably inherently
consistent, the record fails on so many counts in terms of a
standardised network it is surprising it survived so long.

(1) This aspiration manifestly fails in respect of sunshine duration
as we have noted frequently on usw

Philip

"Richard Dixon" wrote :

Going on to a side-topic, the same sort of argument in terms of spatial
distribution of sensors, observations, is being put forward on the
current "is global warming affecting hurricane numbers" which is so
pertinent to the industry I work in (assessing natural catastrophic
risk for insurance). Clearly hurricane numbers have risen since 1995 as
the atlantic thermohaline circulation has switched into another "warm
phase" of SSTs, but there is the general issue of whether storms going
back to even as recently as the 1970s and 1980s were "undersampled"
owing to lack of observations. One of the tenets of global warming on
hurricane activity is that numbers may not necessarily increase, but
their intensity does.

Now that the Hurricane Hunter planes scan hurricanes more frequently,
we're suddenly seeing all these records going - longest time for a
storm as an intense hurricane, smallest eyewall diameter, lowest
latitude for an intense hurricane - which certainly suggests that
there's an element of better recording of hurricanes that may give the
false impression that global warming is leading them to be more
intense.

Absolutely, Richard.

One or two commentators have pointed out many times over recent
years that the hurricane data set is unlikely to be historically consistent.
Several storms in the last two years have been called hurricanes
which earlier analysts wouldn't have batted an eyelid at ... notably
in the Azores region, and that one that developed south of Portugal.
We may also note that some storms are re-graded long after the event
after very close and lengthy scrutiny ...

It makes one wonder just how extraordinary a season 1933 was!

Philip

Philip Eden wrote:
"Richard Dixon" wrote in message

I would have thought the response time of an automatic thermometer to a
short-period change would be much shorter than a standard mercury
thermometer? I'm no expert, I'm sure others on here with
instrumentation knowledge would say more. But if this is the case, for
the same day, if there are short-term fluctuations in temperature on
top of the daily "sine wave" of temperature, then surely the modern


I can only speak from about 25 years ago when developing early AWSs at
Beaufort Park, but I can't remember any attempt to filter readings from
the standard Pt100 sensors to match the thermal time constants of mig
thermometers.

Martin

Philip

JPG wrote:
Philip Eden wrote:
"Richard Dixon" wrote in message

I would have thought the response time of an automatic thermometer to a
short-period change would be much shorter than a standard mercury
thermometer? I'm no expert, I'm sure others on here with
instrumentation knowledge would say more. But if this is the case, for
the same day, if there are short-term fluctuations in temperature on
top of the daily "sine wave" of temperature, then surely the modern


I can only speak from about 25 years ago when developing early AWSs at
Beaufort Park, but I can't remember any attempt to filter readings from
the standard Pt100 sensors to match the thermal time constants of mig
thermometers.

Martin

Philip

The response time of any sensor can be increased simply by
making it bigger or surrounding it with a certain amount of insulation.
I presume this has been done with AWS's and there would be no need to
filter the readings.
As for the Tonbridge record, I must be thinking of the
record at Greenwich in 1911, which I am fairly certain *was* measured
in a Glaisher stand.

Tudor Hughes, Warlingham, Surrey, 556 ft, wishing it were 3556 ft.

"Tudor Hughes" wrote :

The response time of any sensor can be increased simply by
making it bigger or surrounding it with a certain amount of insulation.
I presume this has been done with AWS's and there would be no need to
filter the readings.
As for the Tonbridge record, I must be thinking of the
record at Greenwich in 1911, which I am fairly certain *was* measured
in a Glaisher stand.

Yes, that one was! And it's not beyond the bounds of possibility
that it was turned less assiduously that it should have been (Glaisher
stands, being fully open at the front, and partially open at the sides,
had to be turned regularly during the course of the day especially
during high summer to keep direct radiation from the thermometers)
on 9 Aug 1911.

Tudor Hughes, Warlingham, Surrey, 556 ft, wishing it were 3556 ft.

I know what you mean ...

Philip

"Philip Eden" philipATweatherHYPHENukDOTcom wrote in message
...

"Tudor Hughes" wrote :

Tudor Hughes, Warlingham, Surrey, 556 ft, wishing it were 3556 ft.

I know what you mean ...

Philip

Even at 8deg N here and above 3,000ft it is not nice with high humidity and
a temperature 28degC !

Joe
--
Bangalore, India