Chronology Current Month Current Thread Current Date
[Year List] [Month List (current year)] [Date Index] [Thread Index] [Thread Prev] [Thread Next] [Date Prev] [Date Next]

Re: [Phys-L] a mere 38 years ago



On 05/21/2013 09:11 PM, John Mallinckrodt wrote:

http://aerosol.ucsd.edu/classes/sio217a/sio217afall08-myth1970.pdf

That's a nice piece of work. I was not previously aware of it.

OTOH I did not previously have much need of it, since it refutes
a myth, and I nobody had ever tried to sell me this particular
myth before.

Executive summary:

++ To call into question the climate science of today based on
something that allegedly happened in the 1970s is ridiculous.
Even *if* a nontrivial mistake had been made, there is not the
slightest reason to think that a similar mistake could be made
today, given the greater quantity and quality of data, the
greater computing power, the more-detailed models, and the
greater cross-checking between independent models.

++ To call into question the climate science of today based on
a mistake /that never actually happened/ is beyond ridiculous.

As documented by Peterson et al., at *no* time in the 1970s
(or later) was there any scientific consensus in favor of
anthropogenic global cooling.

=====================
Here are some observations that may help put things in perspective.

-- In 1975, if you wanted to do research on weather (let alone
climate) you needed a supercomputer. The Cray-1 was announced
in 1975, but NCAR didn't get one until 1977. In 1975 they did
have a CDC 7600. It had a MTBF (mean time between failure) of
less than a day. It had 65 k words of primary memory and 512 k
words of secondary memory (64 bit words). It ran at supercomputer
speed, with a peak performance of 36 megaflops. Such machines
cost on the order of 5 million dollars and weighed about 5 metric
tons. They came with *no* software. NCAR had to write their own
operating system.

Such machines were classified as "munitions" and exports were
severely restricted.

In contrast, the beat-up old laptop that I use today does
floating-point orders of magnitude faster than the CDC 7600.
The phone that I carry around has 7000 times more memory.

-- Climate science is hard work. Everybody involved knew
this in the 1970s and before. They understood the importance
of massive computing power. They also knew the importance of
quality and quantity of data. The so-called Butterfly Effect
had been observed in a weather simulation back in 1961 ... so
they knew not to jump to conclusions.

-- If you wanted to write a paper on the subject, you couldn't
do it on a personal computer. The Apple II was not introduced
until 1977. The IBM PC was not introduced until 1981.

-- If you wanted to summarize or visualize the results, you
couldn't do it using a spreadsheet program. VisiCalc was
not introduced until 1979.

-- Scientific programming was done mostly in Fortran. PL/I
existed but was never widely used for this purpose. The
earliest version of the C language appeared an 1972, but
the language specification was still in flux as of 1975.
The K&R book did not come out until 1977. The work that
led to C++ did not begin until 1979. The original (1986)
version of the _Numerical Recipes_ book used Fortran.

-- The Nimbus-3 satellite had been launched in 1969. It
was the first satellite to provide data that was seriously
useful for weather forecasting and research. This satellite
died and was replaced by others, so that by 1975 there was
a five-year-long data record. Of course there was weather
data available before this, but coverage of the oceans and
of underdeveloped areas was very spotty. Satellites provide
systematic global coverage.

-- The Mariner 9 probe arrived at Mars in 1971. Its nominal
objective was to take pictures of the surface, but when it
got there, a planet-wide dust storm was in progress. So
the scientists did the only sensible thing; they studied
the storm.

The idea of global cooling due to particulates was not new,
but this called attention to it.

Nobody in the scientific community thought that particulates
were the only factor to be considered when building models
of weather and climate.

-- The term "nuclear winter" was not coined until shortly
before publication of the TTAPS paper in 1983. So this
term could not have colored scientific thinking -- or even
the blithering of the popular press -- during the 1970s.
Also, again, nobody with any sense thought that particulates
were the only factor to be considered, or even the only
anthropogenic factor.

====================

Bottom line: What would you expect climate scientists to have
done in the 1970s? I would have expected the to gradually figure
things out.

++ I reckon they did a good job, especially considering how
little they had to work with.

++ To call into question the climate science of today based on
something that allegedly happened in the 1970s is ridiculous.
Even if a relevant mistake had been made, there is not the
slightest reason to think that a similar mistake could be
made today, given the greater quantity and quality of data,
the greater computing power, the more-detailed models, and
the greater cross-checking between independent models.

++ To call into question the climate science of today based on
a mistake /that never actually happened/ is beyond ridiculous.