Quantitative Estimates of the Impact of the Most Important Factors on Global Climate Change over the Past 150 Years

Abstract

We propose using multiple regression to create a statistical model for describing climate change under the influence of specified climate-forming factors. This model provides not only estimates of the temporal evolution of global temperature, but also a set of corresponding confidence intervals with a high level of statistical significance (probability). Eliminating the linear trend of climatic temperature series (CRUTEM) and atmospheric CO₂ concentration allows an objective quantitative assessment of the impact of natural factors on climate change. The global CRUTEM temperature responds quasi-synchronously to the fluctuations in the average surface temperature of the North Atlantic (AMO index); however, to changes in solar activity (Wolf numbers), it does so with a delay of approximately 15 years. The linear trend of increasing CO₂ concentration in the atmosphere explains almost all the interannual variability and reflects the linear trend of global temperature, but covers only part of its interannual variability.