Hurst was was ahead of the game in his analysis of river flows over 50 years ago but most climate «scientists» seem to assume that statistical methods appropriate for Gaussian distributions can be used for their data. (wattsupwiththat.com)
«(assuming Gaussian errors in observable parameters, which is approximately equivalent to a uniform «prior» in feedback parameter)» (judithcurry.com)
I would agree with that, if there was some a priori knowledge that in fact the plotted function was known to actually be a straight line buried in Gaussian random noise. (wattsupwiththat.com)