A method and model for modeling a characteristic C that is distributed within a domain. A provided base equation expresses C as a function f of a variable V through use of N+1 parameters C.sub.0, C.sub.1, . . . , C.sub.N in the form C=f(C.sub.0, C.sub.1, . . . , C.sub.N, V), wherein N.gtoreq.1, and wherein C.sub.0, C.sub.1, . . . , C.sub.N are subject to uncertainty. A probability density function (PDF) is provided for describing the probability of occurrence of C.sub.0 in accordance with the uncertainty. Subsidiary equations expressing C.sub.1, . . . , C.sub.N in terms of C.sub.0 are provided. A value of C may be sampled by: providing a value V'' of V; picking a random value C.sub.0R of C.sub.0 from the PDF; computing values C.sub.1R, . . . , C.sub.NR of C.sub.1, . . . , C.sub.N, respectively, by substituting C.sub.0R into the subsidiary equations; and calculating C by substituting C.sub.0R, C.sub.1R, . . . , C.sub.NR and V'' into the base equation.