I was thinking before joining this group to compute a daily index for water, it could be the coefficients of the SUPPORT VECTORS for a SVR algorithm performed on some precipitation density in certain regions.

Or the internal weight for a Neural Network global approximator (See Hornik paper I posted earlier).

So the index is not just hodge podge of numbers set by someone in some business/political establishment, rather the index has a meaningful concept and it is algorithmic.