Wednesday, June 08, 2011

Riddle me this: A Global Warming Question

The much discussed hockey stick graph is a graphic depiction of one thousand years of weather data, only a small part of which data is actually observed data. The quality and adjustment of this data has been much debated. 

CONDITIONS: If the Earth is approximately 4.5 billion years old as determined from the oldest minerals so far discovered, and; the earliest life forms, “protists” a complex single cell with nucleus, appeared approximately 1.8 billion years ago, and; we assume that life forms require “Climate” in its broadest definition to exist, and; we establish the average data measurement period as one year of average high/low temperatures, and; we elect to use the total life of the existence of “Climate” as the overall population of data points, and; we have gathered a sample of 1000 years of "Climate" data:

Q1. Is 1000 years a valid sample size to produce a statistically valid prediction of future “Climate”? (1000 years yields a sample of only 5.56E-07 with a multitude of variables, many unknowable from geological records, most of which are inferred rather than observed.)

Q2. Since Chaos Theory posits sensitive dependence on initial conditions in order to accurately forecast future events, what initial conditions as of 1.8 Bn years ago were used in the climate change forecasting models and to what extent can these models cleanly replicate known historical climate records or even general climate conditions during a geological period between -1.8Bn years and -1000 years.  For instance, the Pleistocene Epoch?

Q3. If you start a climate sample at year 1000 and do a linear regression trend line, the first 1.79999 billion years necessary to determine the actual trajectory of the linear regression is missing.  Therefore, of what value is a climate graph depicting the 1000 years between 1000 A.D. and 2000 A.D. in predicting large scale effects of temperature changes on the planet?

It is therefore evident that the trend line from the 1000 year sample would in fact tell us very little about future “Climate”.  In fact, the absence of "initial conditions" and trend lines for the preceding 1.79999 Bn years would skew the scale of change relative to the elapsed time.

In short, predictions of the consequences of Global Warming from any cause, as reflected in the “Hockey Stick Graph” whether accurate or not, are fatally compromised.

No comments:

Post a Comment