Tuesday, May 26, 2015

The Evolution of Models - 10

Fig. 1
In this series, I have been discussing sea level rise (SLR) as analyzed by "computer models."

I call them software models.

Those models have an impact on your life, an impact that they should not have, not because SLR software is bad, but because these models that are being used for policy have not been prepared, or at least adjusted, properly.

An example is:
"sea level rise could be 4 ft. by 2100"
(President Obama). Most experts agree that "4 ft. by 2100" is a known low ball figure produced by models that consistently underestimate SLR.

It is associated with another phrase that causes people to wonder why such a phrase ("worse than previously thought") is spoken and written so often (The Phrase is Back: "Worse Than Previously Thought").

Fig. 2
The use of the phrase is still with us, even though it shouldn;t be (Sleeping Giant in the Arctic, Climate Change Could Overwhelm California).

So, today I want to explain, using the words of a scientist "on the ground," why that is so.

The video at the bottom of the page is a clip from the recent Economist Arctic Summit 2015 (I can't imagine a more horrible place for a scientist to have to go).

Fig. 3
The well meaning scientist has to talk to an audience of blood sucking neoCapitalists, as they sharpen their teeth in preparation for the "negotiations to come."

The hapless scientist hints that he wants funding for research projects, while all the blood sucker investors want is clearly some more and more of that "blood from the Earth" (The Fleets & Terrorism Follow The Oil).

The only question asked at the end of the scientist's presentation was "can we capture the methane venting from the sea floor and sell it?" (paraphrased).

But I digress.

What caught my ear during his tortured presentation in front of those vultures was his explanation of "primitive models" (meaning old original SLR software).

He intimated that those "primitives" began to be developed in the late 1960's (~half a century ago).

Fig. 4
I have noted some of his statements in that regard (concerning SLR models) and the point in the video when he made them, so that you can take note if you watch the short video (if, on the other hand, you don't want to watch the video, the notes will give you the gist of the current folly).

Regular readers know that during this "The Evolution of Models" series, I have developed an SLR program that is data driven, and it can project the statements made by Dr. James Hansen, who predicted then what is happening now, quite a while ago:
The increasing Greenland mass loss ... can be fit just as well by exponentially increasing annual mass loss, a behavior that Hansen (2005, 2007) argues could occur because of multiple amplifying feedbacks as an ice sheet begins to disintegrate. A 10-year doubling time would lead to 1 meter sea level rise by 2067 ... 2045 ... for 5-year doubling time and 2055 ... for a 7-year doubling time.
(The Evolution of Models - 7). I added the calculations for a 3 year (Fig. 1) doubling, which Dr. Hansen had not mentioned (3yr doubling = 2035).

Since he had been exposed to the software that consistently underestimates SLR, I added the 3 year doubling graph (Fig. 1).

Dr. Hansen had given the years it would take for a 1m / 3ft. SLR with doubling in 5 year (Fig. 2). 7 year (Fig. 3), and 10 year (Fig. 4) increments ("when" is determined by rate of acceleration).

The Dredd Blog software log files and .csv file printout match Dr. Hansen's expectations: 5yr doubling = 2045, 7yr doubling = 2055, and 10yr doubling = 2067.

The next post in this series is here, the previous post in this series is here.

01:30 Warming is concentrated in the Arctic ... this pattern has been evident for decades
02:45 These same models that began development in the late 60's always had an Arctic amplified warming because the physics of that reflectivity is simple.
02:55 Removal of reflective cover absorb more sunlight. The primitive models got that and its of course validated by observation.
04:32 Within our lifetimes we will bear witness to this story unfolding rather rapidly.
04:40 The red line shows observations of sea ice area. It depicts how sea ice is retreating at twice the rate [doubling] that our best model projections show.
05:05 The models still don't get the true fidelity of climate, the rapidity of sea ice loss ...
05:30 ... retreating at a rate four times [doubling * 2] what model projections currently get ... the models are if anything underestimating ...
07:53 Just in the past ten years science reveals more sensitive response than was previously encoded in models.

Researcher sweating it out at an Economic Summit:

1 comment:

  1. Life and death decisions made with defective models. It figures.