Doing this on a year - to - year basis shows NO apparent correlation with the
absolute «globally and annually averaged land and
sea surface temperature anomaly» (i.e. HadCRUT3), but does show a weak correlation with the
CHANGE in
temperature from the previous year, for example: