the error variance used in divand about divand.jl

[Alexander-Barth]

If you are familiar with the Data Assimilation notation (http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.361.9263&rep=rep1&type=pdf) , this parameter is essential diag(R) ./ diag(P) where R is the error covariance of the observation (assumed to be diagonally if espilon2 is a scalar or vector) and P is the error covariance of the background estimate.

Your interpretation is indeed correct. In practise one does often not have directly the matrix R (or your e1, e2, ...). So instead of trying to estimate e1^2,... and s^2 , we are trying to estimate directly their ratio, because only their ratio affect the end-result.

The value of this parameter depends on your case, but when deriving ocean climatologies for physical variables values of 1 or 0.1 are quite typical.