Thanks
for your patience. The full model has just been released, containing some 44
datasets, 33 sheets, 90+graphs, and 15,000 lines of code. (It's never
finished!)
It's
an excel sheet. Hopefully easier to get
more eyes and brains on the project. Download the 20Mb file and see the details
on the post link here.
This
big piece of work is offered freely to the world. Thanks to everyone who has
helped make it possible. We really do appreciate your donations and support! Read on
There
was a large drop in the trend of total solar irradiation around 2004. We found
that there is a delay of one sunspot cycle, about 11 years on average, between
changes in trends in solar radiation and the corresponding change in
temperature. That means things could cool in about 2015 (= 2004 + 11), but
because this solar cycle is a bit longer, the cooling is more likely to start
in 2017 (= 2004 + 13). Models based on CO2 predict warming, but the
solar-driven-model points sharply down. We'll hopefully know in a few years
whether the notch-delay solar theory is right. Unlike the IPCC models, this one
is falsifiable.
It's
a brave man who tries to predict the weather. We are just following the data
(as flawed as it is). Will the sun stay quiet? Will the world cool? Read on
The
model gets the biggest test we can give it at the moment, and you can see here
how well you think it "predicts" the last 200 odd years. The fit is
really very reasonable, though there is a divergence in the 1960's and 70s when
the model predicts the world should have warmed but it didn't.
There
are likely several reasons why the world could have cooled then instead of
warming, but the surprise in the model is that the data that seems to explain
it the best were the atmospheric bomb tests. We'll be saying more about this
soon. It seems so improbable, but we have the references for studies which
suggest that the 503 bombs with an extraordinary 440Mt total yield would
definitely have a significant cooling effect. The big question is "how
much". David has marked the contributing factors, as the models "sees"
it on every graph. Nothing is hidden. Read on
Because
this is such a large and controversial work it's treading on a few toes, and generating
quite the storm. A few people have stepped in the public debate before they
were ready. This is unfortunate, and clearing up those details has taken some
time, but the whole debate and links to critics are available from my site.
In
comments on Watts Up, Leif Svalgaard and Willis Eschenbach misread a graph and
a strange Bermuda Triangle moment occurred, where a silly idea grew legs. So
just in case there is anyone out there who thought for a moment that data was
"Falsified" or "almost fraud", we nail that idea to the
wall here. Everything was described on the graph. If only people read more
carefully. Read on
We
debunk more strange ideas. Leif Svalgaard didn't realize there was a fall in
11-year smoothed TSI data around 2003. We graphed his very own solar data to
show it's there too. Could it be that few people have noticed the sharp
downturn in 11-year smoothed solar trends? Read on
Some
people have suggested that the Notch filter doesn't mean anything because the
transfer function would still find a notch even if the temperature data was
just white noise. This is true, but meaningless. Any maths function can find a
relationship with two unrelated things. What matters are the assumptions you
make and the questions you are trying to ask. We show that the transfer
function applies if, and only if, there is some causal link between total solar
irradiance and the temperature of the Earth. That seems like a pretty safe bet
to us. Read on
Look,
it's a big paper. Lubos missed the central concept, and didn't read on, nor did
he notice David specifically explained his mistake in an email 3 months ago. We
appreciate feedback, but wish people would read more carefully. We reply to all
his points here. Read on
Cheers
from Jo
PS:
Join us on twitter or on facebook.
No comments:
Post a Comment