Application to Interval Temporal Change

Robert Gilmore Pontius, Clark University


This chapter applies the concepts of the previous chapter to compare maps concerning a continuous phenomenon, specifically sea surface temperature. The X variable derives from February 1982. We consider two Y variables at August 1982 and February 2010. This chapter gives equations to stratify the analysis. The global data have two strata: the northern hemisphere and the southern hemisphere. The metrics, scatter plots, and maps each reveal information that the other two forms of presentation do not reveal. Mean Deviation reveals global cooling during the half-year interval and global warming during the 28-year interval. The components of Mean Absolute Deviation quantify how most of the change between seasons is Allocation deviation across hemispheres, while most of the change between years is Quantity deviation. Correlation and the scatter plots show a stronger linear association between times during the 28-year interval than during the half-year interval. The maps show the spatial distribution of the temperature deviations, which the scatter plots and most of the metrics fail to indicate.