Temperature Fluctuation Project

This website is aimed at sharing all the materials related to Temperature Fluctuations project by J.K. Dagsvik, M. Fortuna and S.H. Moen, a multi-years research project aimed at investigating the stationary and fractal nature of temperature fluctuations. The idea and the main results are summarised in the paper ”How does the temperature vary over time? Evidence on the stationary and fractal nature of temperature fluctuations”, during publication in Journal of the Royal Statistical Society, series A

Read the paper

The idea behind

The idea behind our project, in a more intutive approach, and a quick summary of the main results + press materials and other resources

Get the code

R functions supporting the Temperature Fluctuations paper: data exploration, stationarity and FGN testing, estimators performance

Download the data

Temperature time series from all around the world, available as monthly figures and two-millennia reconstructed temperature by Moberg et al.


This section contains all the results and the appendices of the Temperature fluctuations paper. All the results may be reproduced via their free code and data, see above sections Data and Code

Temperature fluctuation paper - Main Results

This section contains all the results included in the Temperature fluctuations paper

Continue reading

Appendix B - Summary data information

Information on the weather stations, lenght of the related time series and missing data

Continue reading

Appendix C - Estimation and test results

H estimation by Whittle and Wavelet estimators. Chi square statistics and stationarity tests

Continue reading

Appendix D - Properties of the estimators

Properties of different estimators of the FGN models, with boostrap simulations

Continue reading

Appendix E - Temperature stations infographics

Infographics on 96 weather stations, with temperature fluctuations, autocorrelation and statistical tests

Continue reading

Appendix F - Selection procedure

Details on the selection procedure of the available temperature time series

Continue reading