Application of statistical control charts to climate change detection in Masvingo city, Zimbabwe

Abstract

This paper outlines an evaluation of Cumulative Sum (CUSUM) and Exponentially Weighted Moving Average (EWMA) control charting techniques to determine if the climate of Masvingo station in Zimbabwe has changed over the past years. In particular, the climate indicator considered in this work is the annual air temperatures for Masvingo station for the period 1952 to 2001. The two control charts were then applied to the data and both techniques proved effective in detecting climate change. However, the CUSUM chart detected a significant shift in mean annual air temperatures in 1983 while the EWMA chart detected the shift in 1993. The two methods suggested that the origin of the shift was in 1981. An upward shift in the mean level and a subsequent test in the mean difference points to the fact that the mean annual temperature is going up. This means that the temperatures are getting warmer by the years. The use of these two techniques confirms the existence of climate change in Masvingo city, Zimbabwe.