Total Pageviews

Wednesday, November 27, 2019

Reality vs. Climate Compiuter Games -- Backtesting designed to make the climate models look good ... fails to do that !

SUMMARY:
The UN Intergovernmental 
Panel on Climate Change 
            (IPCC) 
attributes most change 
in global average 
temperatures to 
carbon dioxide, 
practically ignoring
the major greenhouse 
gas water vapor, and
almost completely
ignoring 4.5 billion
years of natural 
climate change
NOT caused by CO2.

The IPCC also guesses 
future growth of CO2 
in the atmosphere.

They don't 
use the 
term guess -- 
they talk about  
"Relative 
Concentration 
Pathways"
    (RCPs)

RCPs relate to 
Earth’s "energy 
imbalance", 
in watts per 
square meter.

The IPCC’s 
5th Assessment 
Report 
(AR5, 2013)
defines RCPs 
in increasing 
order of energy 
imbalances, 
as RCP
2.0W/m2; 
4.5 W/m2; 
6.0 W/m2; and 
8.5 W/m2. 

The extreme growth 
of CO2 RCP 8.5 yields 
scary temperature 
predictions that 
climate scientists 
and politicians love.

The less extreme
RCP 6.0 makes more 
sense, but still makes
the climate model
predictions look bad.

The more modest 
RCP 4.5 recognizes
that many nations
are trying to slow
CO2 emissions 
growth.

Using RCP 4.5 
should make
climate model 
predictions 
seem very
reasonable.

If the model predictions
using RCP 4.5 are not
close to reality, then
the models are wrong.

Here's the results 
of several models 
assuming RCP 4.5.

The good news is 
few predictions 
are near triple
of reality.

With RCP 4.5,
all predictions
are more than 
double of reality, 
but most are not 
far above double.

I guess that's 
close enough
for government
climate "science"?



DETAILS:
John Christy, of the 
Earth System Science 
Center at the University 
of Alabama in Huntsville, 
compared some 
US climate model 
simulations using the 
KMNI Climate Explorer, 
from the Royal Netherlands 
Meteorological Institute 
            (KNMI), 
the Dutch national 
weather service.


10 simulations using 
two NCAR / UCAR models 
       (CCSM4 and CCSM1) 
funded by the National 
Science Foundation (NSF)


3 simulations using 
three Geophysical Fluid 
Dynamics Laboratory (GFDL) 
models, 


1 simulation using 
a model funded by NOAA


34 simulations using 
the two Goddard Institute 
for Space Studies (GISS) 
models funded by NASA. 


The temperature 
observations
are the Tropical 
Mid-Tropospheric 
Temperatures 
from 1979 to 2018 
    (40 years) 
taken from 
four different 
balloon datasets; 
the average 
of three different 
satellite datasets; 
and the world-wide 
reanalysis datasets 
used daily for 
calibrating 
weather models. 


Starting in 1995, 
the models began
to greatly 
overestimate 
the warming
of the tropical 
mid- troposphere. 


The overestimate of 
temperature trends 
for the RCP 4.5 scenario 
is just over twice 
the actual warming 
that happened.



TEMPERATURE  
REALITY:
From balloon and 
satellite data, 
the average trend 
in increasing 
atmospheric
temperatures is 
+0.12 degrees C. 
per decade. 



TEMPERAURE  FROM
CLIMATE   MODELS:

Two NASA - GISS models:
+ 0.25 C. per decade
+ 0.26 C. per decade


Three GFDL models:
+ 0.40 C. per decade
+ 0.32 C. per decade
+ 0.33 C. per decade


Two NCAR / UCAR models:
+ 0.26 C. per decade 
+ 0.24 C. per decade. 


Even under this 
low 4.5 RCP
CO2 emissions 
scenario, none 
of the models 
come close to what 
actually happened
in the atmosphere.