sábado, 18 de setembro de 2010

WHAT IS NOISE, WHAT IS REAL? HOW IS VERIFIED THE CONCENTRATION PEAKS OF CO2 AND METHANE IN ICE CORES?

The presence of CO2 or Methane in ice cores extracted from Antarctic or Greenland is the principal way to verify part of Earth environmental history. The two more important projects at this moment are considered Vostok and Epica projects. Both are a multinational projects. In the Vostok Project  a deep hole of 3.623 meters was made on a glacial lake in a glacial URSS base in Antarctica. In the European Project for Ice Coring in Antarctica (Epica) a hole of 3.734 meters deep was made in the C Dome of Concordia Station ( Coordinates 75°06’S; 123°21’E and  elevation of 3.233 meters over sea level), concluded in January 17, 2006.

The extracted samples of these holes are carefully divided in segments of tenths of centimeters. These segments had his age dated by convenient methods with controlled uncertainty to attend the project objectives.

With these samples dated, many radioisotope analyses (deuterium) were done to calculate the concentration of CO2, Methane and N2O and verify the temperature.

By considering the absolute dating process correct and the concentration values of CO2, Methane and N2O in the samples, it´s possible to estimate macro-variations of whether in the last 650.000 years.

Now we go to the point. Is it correct to say that the concentration of CO2 is now the biggest of all times, by considering only these datas? Certainly not!

If we look the original data (which is open for everybody) on page of NOAA at: http://www.ncdc.noaa.gov/paleo/icecore/antarctica/domec/domec_epica_data.html
It´s possible that a little, but important aspect are not consider by the people. The interval time scale to observe the palioclimatology variations and the time scale used in the analogue studied events.

 Considering all determination corrects (?), without any problem in chronology process or other related procedures, we see that  the elapsed time between the base and the top of each ice sample is roughly 450 years ( more exactly 300  to 550 years ). As a minimal interval, it is necessary to work with a average of 300 years.

What is the problem in all that? Simple: means properties which varies during 300 years can’t be used to estimate peaks in 30 to 40 years. This is a basic signal analysis problem, the sample rate theorem known by Nyquist-Shanonn. When we desire to reconstruct a signal with “B” frequency spectrum range we need to guaranty “2B” frequency sample of the original signal.
 
For example, if we want to reconstruct the behavior of any variable in an historic period of 50 years, a sample with 25 years is required.

This can be considered filigree, but this care with data is essential and we can’t forget these details when we are talking about extremes values of a signal.

If we are analyzing a maximal point which had happened millions years ago, and once this value is a peak value, is mandatory an inflection point on the curve data. If we had not the adequate sample frequency, the result will be certainly atennueted.

I hope so I’m wrong, because I don’t understand why dozens of researchers use these “ice cores” means data without considerations about the incompatibility of acquisition sample and data utilization for their conclusions. 

To ilustrate the problem, see the following figure.



Nenhum comentário:

Postar um comentário