Predictive Analytics of Climate Trends

Abstract

The accumulation of data in the weather observation will be continued steadily to infinity. To perform weather predictions, a huge amount of data needs to be processed and assimilated to create of acceptable models.

This paper presents a method of estimating a long term (more than 100 years) temperature and precipitation sets for predicting of a future trends. The method uses average solar cycle length as a uniform sampling basis for assessment of weather features. Long-term observation of climatic variables shows that in the theory of weather chaos exists a certain order. The order is expressed as a long-term increasing temperature trends and in rising of ocean levels. Here we demonstrate how the proposed approach can be highly effective in varied large-scale applications involving regional temperature trends in Caucasus region.