Read the article

Analyzing climate data from 1850 to 2015

by Amaury Labenne, Data Scientist

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/

Summary:

  1. Introduction
  2. Visualizing time series data
  3. Descriptive analysis of time series data
  4. Decomposition of Seasonality in a time series
  5. The Holt-Winters Method
  6. Testing for stationarity
  7. Time Series differencing
  8. Test for Stationarity on the Series Following Differencing
  9. ACF and PACF on the Series Following Differencing for the ARIMA Model
  10. ARIMA Model
  11. Conclusion