Abstract
This study analytically shows that two time series(xi, yi)-related 5 statistical quantities such as mean values (x, y), variances(Sx2, Sy2), covariance C or slope value of a correlation straight line passing through the origin a0 can be derived from another 5 statistical quantities such as slope value a and intercept value b of a correlation straight line, bias ν, correlation coefficient ρ and root-mean-square error σ without using the original time series data and vice versa. The validity is confirmed by almost perfect agreement between the above-mentioned statistical quantities estimated using the observation-based and the MSM(Meso Scale Model)-based wind speed data samples over 1 year at 9 ocean buoy stations deployed around offshore areas of Korea. On the contrary, substitution into the derived relationships of either of two sets of the 5 statistical quantities such as a, b, ν, ρ, σ provided by Heo et al.1) does not yield either the observation-related mean value x or the variance Sx2 to be essentially coincident with each other in the two cases. This may be a problem to be solved in the concerned paper.
