Climate Prediction through Statistical Methods

TL;DR

This paper explores how statistical analysis of temperature fluctuations can classify climate types, linking historical climate shifts with present temperature data without relying on traditional indicators.

Contribution

It introduces a novel application of Detrended Fluctuation Analysis to classify climates based solely on temperature fluctuation patterns.

Findings

Temperature fluctuations can classify climate types.

Statistical connection between present and historic climate shifts.

Temperature data alone suffices for climate classification.

Abstract

Climate change is a reality of today. Paleoclimatic proxies and climate predictions based on coupled atmosphere-ocean general circulation models provide us with temperature data. Using Detrended Fluctuation Analysis, we are investigating the statistical connection between the climate types of the present and these local temperatures. We are relating this issue to some well-known historic climate shifts. Our main result is that the temperature fluctuations with or without a temperature scale attached to them, can be used to classify climates in the absence of other indicators such as pan evaporation and precipitation.