On the correlation of nonlinear variables containing secular trend variations: numerical experiments

Abstract

Due to global warming, the general circulation, underlying surfaces characteristics, and geophysical and meteorological elements all show evident secular trends. This paper points out that when calculating the correlation of two variables containing their own obvious secular trends, the interannual correlation characteristics between the two variables may be distorted (overestimated or underestimated). Numerical experiments in this paper show that if two variables have opposite secular trends, the correlation coefficient between the two variables is reduced (the positive correlation is underestimated, or the negative correlation is overestimated); and if the two variables have the same sign of secular trends, the correlation coefficient between the two variables is increased (the positive correlation is overestimated, or the negative correlation is underestimated). Numerical experiments also suggest that the effect of secular trends on the interannual correlation of the two variables is interchangeable, that is to say, as long as the values of the two trends are not changed, the two variables interchange their positions, and the effect of the secular trends on the interannual correlation coefficient of the two variables remains the same. If the two variables have the same-(opposite-) sign trends, the effect of secular trends on the interannual correlation coefficient is more (less) distinctive. A meteorological example is given.