During the late seventies and early eighties our lab looked
after a small weather station on behalf of the Met Office. We logged rainfall,
snow, temperature, sunshine hours and every now and then a chap from the Met
Office would collect the data.
All data was hand written of course and ever since climate
change came to be such a hot topic I’ve often wondered how reliable it was. In
my view those far off days have something to tell us about historical data and
the fact that it was collected and transcribed by people, not automated
instruments. Historical protocols and historical behaviour – a minefield of
unknowns.
To record daily maximum and minimum temperatures, we used a
simple max/min thermometer housed in a wooden Stevenson screen.
Every day someone from the lab would read the two temperatures, write them down
and reset the thermometer.
If we missed a day, which happened occasionally for a
variety of reasons, then the Met Office chap would nag us about it when he
collected the data, look up a temperature record of the nearest station to ours
and insert the readings into our record. He once told me that this was standard
procedure – they didn’t accept incomplete data.
Yet at the time the data was fit for purpose, although that
doesn’t mean it was fit for a far more tightly specified purpose dreamed up
decades later.
In those days nobody knew that such temperature records
would one day be used to justify global political decisions on energy policy.
Nobody knew that long term temperature changes of less than one degree
centigrade would acquire such dramatic significance.
Not that our station was ever likely to figure in these
games I hasten to add. It closed some time ago. I’m merely dredging up some
memories to highlight the tricky nature of historical temperature data.
Stripping off some of the gloss you might say. There is a lot of that in
climate science.
For example, our thermometer was never recalibrated. I’m
sure it was checked before being installed, but even simple thermometers change
over time and today it would be regularly calibrated against a certified
standard. Ours wasn’t - ever.
Apart from the unknown condition of the thermometer, how
many errors were made by people who took the readings and wrote them onto sheets
of paper come wind, rain or snow?
In my experience, scientists are reluctant to take
cognisance of human error even for highly uncertain factors such as historical
and somewhat loosely defined protocols. Yet the historical global temperature
record and our evidence of recent warming relies on such data.
Were the protocols and equipment used my lab capable of
detecting a small temperature rise over a century?
One degree? No.
Two degrees? Doubtful.
Three degrees? Maybe.
Of course this is merely my opinion. I don’t actually know
and neither does anyone else. Nobody can go back and calibrate our thermometer,
review the protocol we followed and audit the way we followed it. There are some
things we could do such as comparing our record to the record of nearby
thermometers, but is that sufficient to detect small long term changes?
Taking the wider view, are we able to estimate such changes
from long historical records based on protocols not designed for that purpose? Always
assuming written protocols were used of course - and what about calibration facilities? How many were calibrated against the equivalent of NPL standards? Some? A few? None?
Yet in terms of time span, manual surface temperature records
derived from a range of old and possibly dubious measurement protocols account
for at least two thirds of our surface temperature record for the past century.
Note – this
post gives an excellent insight into the pitfalls of temperature
measurement.
Figure 1: Distributions and ranges for climate sensitivity from different lines of evidence. The circle indicates the most likely value. The thick coloured bars indicate likely value (more than 66% probability). The thin coloured bars indicate most likely values (more than 90% probability). Dashed lines indicate no robust constraint on an upper bound. The IPCC likely range (2 to 4.5°C) and most likely value (3°C) are indicated by the vertical grey bar and black line, respectively.
Basic 
Intermediate
.JPG)
The skeptic argument...