Analyzing climate data from 1850 to 2015

Free whitepaper

Analyzing climate data from 1850 to 2015

This white paper explains how to conduct a complete time series data analysis from descriptive analysis and data visualization to forecasting using the XLSTAT software.




The aim of this analysis is to describe, visualize and forecast temperature changes on a global scale. To do this, we use XLSTAT time series analysis. Analyzing temporal data is a specific branch of data and statistical analysis. This type of data is called *time series* data. We will present various methods that can be used to analyze this type of data. The data used here are monthly air and ocean temperatures from 1850 to 2015. These data can be downloaded for free at: http://berkeleyearth.org/data/

Table of contents

  • Introduction
  • Visualizing time series data
  • Descriptive analysis of time series data
  • Decomposition of Seasonality in a time series
  • The Holt-Winters Method
  • Testing for stationarity
  • Time Series differencing
  • Test for Stationarity on the Series Following Differencing
  • ACF and PACF on the Series Following Differencing for the ARIMA Model
  • ARIMA Model
  • Conclusion

Writer profil

Amaury Labenne

Senior consultant in statistics

Senior consultant in statistics, Amaury Labenne holds a PhD in applied mathematics. While writing his doctoral thesis on dimension reduction methods, he taught statistics and their uses at university. In 2016, Amaury joined the Addinsoft R&D team, of which he was then in charge until 2020. He participated to the development and improvement of XLSTAT major statistical features. Amaury Labenne is now a senior data scientist consultant. He also provides training on data analysis and statistics. Very pedagogical, he knows how to adapt his courses to his audience. He favors teaching methods based on examples and real business applications rather than on theoretical and mathematical explanations.

Fill in the form to download the document

French | Japanese | English