humanities.philosophy.objectivism

TC wrote:
> On Mar 25, 3:44 am, TommCatt wrote:
>
>> There are several things wrong with the graph you keep referring to with
>> such pride. Most notably, the fact that the "forecast" is, in fact, not
>> a forecast at all.
>
> I don't believe I claimed it was a forecast. It is a postcast or a
> retrodiction. But the goodness of fit of modeled versus observed
> based on only five parameters is significant.

No it's not. When you consider that climate is driven by hundreds of
parameters, many with much more effect than the five shown, to make a
good fit based only on those five indicates mathematical sleight-of-hand.

>> ... Take the daily price chart over the last year of
>> any randomly chosen stock and any decent programmer can model it to a
>> good deal of accuracy.
>
> Can they model it as a function of greenhouse gas concentration,
> solar input, ozone concentration, volcanic particulate concentration
> and sulfate aerosol concentration?

Yes. When you already know what the answer should be, you can model your
function based on any input -- whether or not the inputs have anything
really to do with the modeled effect.

> .............
>> the match between "modeled" and "observed" is /too/ good. There is no
>> obvious correlation to the jumble of causative factors shown at the
>> bottom.
>
> Why would you expect a correlation between human activity
> - greenhouse gases, or solar activity, or volcanic acitivity?

Well...because they're listed.

>> The only time there is a connection is when co2 and temperature
>> took off together after 1960. However, temperature spiked between 1910
>> and 1940 with no such correlation.
>
> Oh. You mean you can't visualize the 5-dimensional correlation
> between the inputs and the model output. Most humans can't.
> That is why there are computers and statistics.

It saves a lot of time and electronic ink to just say, "I don't know."

> But which is it? Is the match too good or do you think the
> miismatch on decadal scales like around 1940 bad?

The modeled line is too good for the inputs shown.

>> What /is/ difficult is projecting the forecast into the future. When
>> that happens with any degree of accuracy, get back to us.
>
> You'll have to wait for next year. But wait, the future will be
> past next year, so you'll never be satisfied.

Save the straw man arguments for novices. When you have a model for,
say, the year 2005 that was generated /before/ 2005 rather than 2006 or
later, then you have something.

>> However, there is some amount of prediction that can be made.
>> Unfortunately, your chart is much too short in duration to be useful.
>> See the chart inhttp://www.solcomhouse.com/globalwarming.htmfor a
>> chart that goes back 410 000 years.
>
> So what is it that you want prediction or postdiction?

Prediction.

> Is postdiction from the distance past OK, but not from the last
> century?

I wasn't postdicting (is that even a word?) -- I was predicting. And
yes, it is a bit safer to do when you look back far enough to detect
cycles. Examining only a century's worth of data is essentially trying
to plot a trend from a single data point. Consider the time period from
any temperature maximum to the next minimum. The Earth was cooling. The
Earth was cooling a great deal -- with glaciers swallowing entire
continents. Yet there are many centuries during that time that steadily
grew warmer. Just looking at the that century would lead to the quite
faulty conclusion that the Earth was in the middle of a warming period.

>> From this we can see somewhat of a
>> cycle. Temperature reached a peak about 410 000, 330 000, 240 000 and
>> 130 000 years ago. Assuming the present time is the next peak, we can
>> make the following general conclusions:
>
> This graph does not include the recent past. Industrial era CO2
> levels above 280 ppm are not depicted.

Right. And we have another 2 to 4 degrees to go, no matter what the co2
levels are or will be.

>> 1) The span between the peaks are 80 000, 90 000, 110 000 and 130 000
>> years. These peaks seem to be occurring at longer intervals.
>
> Yes standard ice age paleoclimatology. What is your point?

Just pointing out some of the more obvious cycles.

>> 2) The temperature maximum at each previous peak is 2 to 4 degrees C
>> above the present temperature. If past performance is any indication,
>> the current temperature will continue to rise for at least another 2
>> degrees and possibly another 4. No matter what.
>
> Actually to my eye the current period is anomalous. Previous warm
> periods have spiked quite sharply. The current era has plateaued
> at a relatively constant value for 10,000 years.

And this means what to you? By no possible stretch of anthropomorphic
imagination can Man have affected climate over the last 10 000 years.
(For one thing, if you did, you would have to admit that Man's impact
seemed to be a dampening of the naturally occurring severity. Something
the GW hysterics are not going to admit.)

> As I said in another post
> Perhaps rational discussion will reveal that our carbon emissions area
> a good thing.

Or, more likely, completely insignificant. But you don't actually
believe what you just wrote because, judging from your posts, you have
completely bought into the whole "human produced co2 is harmful" idea.

> But as long as both pro and con are hysterical .....

Oh, that /is/ hysterical...8-)

TommCatt
--
The chemist's creed: If you're not part of the solution, you're part of
the precipitate.