Article

2014-05-03 Science & Technology

Archived material is restricted to Rantburg regulars and members. If you need access email fred.pruitt=at=gmail.com with your nick to be added to the members list. There is no charge to join Rantburg as a member.

Posted by Grunter 2014-05-03 00:00|| || Front Page|| [9 views ] Top

#1 Good find. Very interesting skim. Going to take a while and some effort to read it in depth to see what's there.

Posted by OldSpook 2014-05-03 00:22|| 2014-05-03 00:22|| Front Page Top

#2 The only thing wrong with the global climate models is their inability to predict the temperature. It's not their fault. They suffer the delusion that CO2 is related to temperature.

Let's just say they lack 'predictive skill' and give them a t-shirt anyway. Just for showing up.

Posted by SteveS 2014-05-03 01:31|| 2014-05-03 01:31|| Front Page Top

#3 initially bountiful food sources. Porpoises and deep ocean fish like tuna, virtually disappeared from the islanders diet.

Lobster has disappeared from mine. Had nothing to do with the climate.

Posted by Besoeker 2014-05-03 05:08|| 2014-05-03 05:08|| Front Page Top

#4 The only thing wrong with the global climate models is their inability to predict the temperature

Apparently, they can't run in reverse to match empirical data before their 'start' dates as well.

Posted by P2kontheroad 2014-05-03 08:23|| 2014-05-03 08:23|| Front Page Top

#5 Ditto Besoeker.

As a 30+ year software dude with much modeling experience I can honestly say that computer models are the alpha & omega of the trope "GIGO".

Of course it's easier when you control the creation of the garbage going in. The thing in this debate that twists my knickers the most is the media supported myth that computer modeling is on a par with multi-dimensional astro physics and occult ritual.

Anyone who has ever created a spread-sheet to track their retirement funds or anything else has created a computer model. Put in a 50% increase in you income weekly and your retirement nest egg grows quickly. Type in the real .05% rate and not much happens. Shouldn't be hard to understand for most non-LIV people.

Posted by AlanC 2014-05-03 08:48|| 2014-05-03 08:48|| Front Page Top

#6 I don't read the obits or look at the weather report. If I wake up in the morning, it's all GOOD !

Posted by Besoeker 2014-05-03 09:35|| 2014-05-03 09:35|| Front Page Top

#7 As a 30+ year software dude with much modeling experience

AlanC, I've seen suggestions from other modelling types that we may be seeing second and third order effects -- or higher -- of the interactions of complex adaptive systems, and that if so climate may never be reducible to an algorithm. Thoughts?

Posted by trailing wife 2014-05-03 11:44|| 2014-05-03 11:44|| Front Page Top

#8 TW, this is a good example of why the media is so befuddled. I don't claim to be an expert in climate science and THAT is where all discussion has to happen if anything worthwhile is to come from said discussion.

Modelling complex systems, adaptive or not, is hard. My modelling had mostly to do with predicting application system performance to justify the 7 or 8 figure purchase plans for new hardware. Those systems were relatively simple but sufficiently complex to be subject to all sorts of sneaky errors (not to mention that the desired results were known in advance).

Modelling is a representative of the 3 types of lies: Lies, Damn Lies and Statistics. From what I saw in the release of those Climategate e-mails and code fragments there were two big flashing red lights in the code. 1) there were no comments in the Fortran and 2) there was an undocumented simple external file that was used to massage some of the calculations. Could they have been totally innocent? Sure. BUT until all of the code and data is made public we'll never know.

Now, after getting all of that off my chest 8^)

Yes, Climate may never (my NSHO agrees) be reducible to an algorithm or system thereof. At least not until every wing-flap of every butterfly in china can be quantified and the output of every volcano & undersea vent can be predicted.

Posted by AlanC 2014-05-03 12:22|| 2014-05-03 12:22|| Front Page Top

#9 My take away from this article is that oceanic and climate scientists make the same mistake that most scientists do in favoring incrementalism over a cycle of stability marked by occasional and random dramatic change. In this case, volcanoes. Incrementalism is an easier story to tell and it is easier to model. There just isn't any support for this approach in observed nature.

Posted by Iblis 2014-05-03 12:32|| 2014-05-03 12:32|| Front Page Top

#10 re: #9 Iblis, you nailed it. Incrementalism is much easier to model; and there isn't any support for this in observed nature.

Similar issues seen in static vs. dynamic economic modeling. One is easy, the other is real.

Posted by AlanC 2014-05-03 13:59|| 2014-05-03 13:59|| Front Page Top

#11 (not to mention that the desired results were known in advance).

:)

Posted by Shipman 2014-05-03 14:29|| 2014-05-03 14:29|| Front Page Top

#12 Nature is discontinuous. Processes "jump" sometimes. Nothing is smooth in nature if you look at it closely enough. Yet climate scientists continue to assume their climate models use continuous functions, because that's the only way they know how to think: linearly.

Scientists always fall for correlation, assuming it to be causation; Classical fallacy - post hoc ergo proper hoc.

Need to ask themselves "What if Im wrong?". thats the first sign of a real thinker.

Posted by OldSpook 2014-05-03 15:43|| 2014-05-03 15:43|| Front Page Top

#13 In other words: Shit Happens! And you can't model shit happening.

Posted by CrazyFool 2014-05-03 15:53|| 2014-05-03 15:53|| Front Page Top

#14 Let us ignore the obvious that global warmists, climate changists and global cooling predictors have an agenda and a conclusion that needs supported (preferably with data or charts).

Predicting weather years in advance is a challenge when you do not have significant data from the past. We really only have accurate temperature data from the last 100 years of the 2.5 billion or so years of temperature. Weatherman are marginally successful 3 to 5 days out let alone 20 years out. In addition it is a stretch to consider industrial production a factor on weather when that has really been around just 150 years.

Consider this. Each day the temperature changes 20 - 30 degrees from the lows to the highs for the day. Should I really care that instead of being 45 - 65 today it is going to be 44 - 66 in 20 years? I am not sure I want to give up life's luxuries such a car travel and electronic gizmos as well as trillions of dollars of green investment to progressives to solve that 'problem'.

Posted by Airandee 2014-05-03 17:19|| 2014-05-03 17:19|| Front Page Top

#15 It's much worse than GIGO

You can put near perfect in * near perfect model and loop a few times say 1.5% error per week.

After a year you might as well be flipping a coin.

Computer models are the most expensive dice ever invented.

Posted by Bright Pebbles 2014-05-03 18:28|| 2014-05-03 18:28|| Front Page Top

#16 BP you are, of course, correct. There's a reason they call it Chaos Theory. However your concern is really only valid if the experimenters/modellers are honest.

This is true of all statistical analytics. If you're not honest you can make the answer anything you want.

Posted by AlanC 2014-05-03 19:14|| 2014-05-03 19:14|| Front Page Top

23:21 OldSpook
23:00 Pearl Borgia1889
21:51 Silentbrick
20:53 Shipman
20:51 Shipman
20:49 Shipman
20:47 Shipman
20:02 DarthVader
19:54 Redneck Jim
19:38 Redneck Jim
19:24 Redneck Jim
19:22 Barbara
19:14 AlanC
19:12 Redneck Jim
18:29 Thing From Snowy Mountain
18:28 Bright Pebbles
18:01 Besoeker
18:01 SteveS
17:57 JFM
17:56 Shipman
17:55 Besoeker
17:48 Shipman
17:44 g(r)omgoru
17:40 Besoeker