AMSI header image

Climate Statistics









Prof John Boland
The University of South Australia


In this course we will be studying techniques in the statistical analysis of climate variables. We will do it in a themed manner – within the areas of energy and water resource management.

There has been increasing interest in the study and application of statistical methods relating to climate variables. The most obvious area has been that of investigations to do with climate change (see [1] for example). This course will focus on some of the tools used in an area that could be said to have gained greater prominence because of the need to on one hand mitigate against the effects of climate change and on the other to adapt to them. One of the effective measures being taken to mitigate against climate change is the ever increasing uptake of renewable energy technologies. The best way to get most out of these initiatives is to fully analyse their potential contribution. This requires use of varied statistical analysis and time series tools, as well as optimisation techniques. To understand the impacts of climate change on such things as available water resources, one needs to be able to fully characterise their present state, once again requiring sophisticated statistical methods.

Apart from developing the tools necessary for the investigations mentioned above, we will discover how to use statistical methods to help us adapt to climate change. For example, two recent research programs specifically deal with this in two areas of impact. The National Climate Change Adaption Research Facility (NCCARF) is funding several projects, including that of understanding how to design buildings to be thermally robust in 2050. I am leading a section of that project to do with altering the weather data sets presently used for house energy ratings to incorporate the projected effects of climate change. We will discuss the proposed methods. A similar study funded by the South Australian government is trying to identify the impacts of climate change on water resources in that state. Before beginning this task, we need to perform sophisticated analysis of the present climate, particularly to do with statistical modelling of rainfall. These are some examples of the areas we will focus on in the course.


  • Time Series Analysis of Climate Variables
    • The components of a time series model.
    • Additive and multiplicative models.
    • Spectral decomposition.
    • Box-Jenkins models.
    • Forecasting techniques, including probabilistic forecasts.
    • GARCH and other volatility models.
  • Spatial-Temporal Statistics
    • Principal Component Analysis
    • Measures of Spatial Correlation including Correlative Coherence.
  • Statistical Coherence of Models on Different Time Scales – examples from wind farm output and rainfall.
  • Energy Meteorology - Statistical Modelling of Climate Variables for Electricity Generation.
  • Wavelets and their application to climate variables, such as the Southern Oscillation Index and wind farm output.
  • Long range dependence.
  • If time allows, we will talk about modelling of climate variables using Artificial Neural Networks, Numerical Weather Predictors and the associated ensemble modelling for variance estimation.

Contact hours

28 hours of hours spread over four weeks, plus consultation as required.


A basic understanding of statistical decision making, construction of statistical models, partial differential equations and integral transforms.


  • [1] I. Allison, M. Bird, J. Church, M. England, I. Enting , D. Karoly, M. Raupach, J. Palutikof and S. Sherwood,  The Science of Climate Change: Questions and Answers,   Australian Academy of Science, 2010.
  • Tsay, R.S., Analysis of Financial Time Series, 2nd Edition, John Wiley & Sons, 2005.
  • Chatfield, C., The Analysis of Time Series, An Introduction, Chapman and Hall, 1989.

About John Boland

John Boland is Professor of Environmental Mathematics at the University of South Australia. His research applies statistical analysis and mathematical modelling to problems in climate science, renewable energy utilisation and water resource management.