Mahalanobis Distance, in short MD, is a generalized measure using all multivariate data for a certain simple objective measure. Weather forcasting is one of the ideal areas to use MD. The critical steps to use MD are
・Define Mahalanobis space to construct a generalized unit measure from those multivariate data
・Evaluate the magnitude of error of the method using S/N ratis.
・Choose reasonable measurements for simplication.
In this article, how to construct MD to predict the intensity of rain is explained as an introductory case.