Pk. Kitanidis, Generalized covariance functions associated with the Laplace equation and their use in interpolation and inverse problems, WATER RES R, 35(5), 1999, pp. 1361-1367
This work examines which generalized covariance function when used in the s
tochastic approach produces the flattest possible estimate of an unknown fu
nction that is consistent with the data. Such an estimate is the plainest p
ossible continuous function, thus in a sense eliminating details that are i
rrelevant or unsupported by data. The answer is found from the solution of
the following variational problem: Determine the function that reproduces t
he data, has the smallest gradient (in the square norm sense), and has a gr
adient that vanishes at large distances from the observations. The generali
zed covariance functions are shown to be the Green's functions for the free
-space Laplace equation: the linear distance, in one dimension; the logarit
hmic distance in two dimensions; and the inverse distance in three dimensio
ns. It is demonstrated that they are appropriate covariance functions for i
ntrinsic random fields, a modification is proposed to facilitate numerical
implementation, and a couple of examples are presented to illustrate the ap
plicability of the methodology.