Efficient sensitivity/uncertainty analysis using the combined stochastic response surface method and automated differentiation: Application to environmental and biological systems
Ss. Isukapalli et al., Efficient sensitivity/uncertainty analysis using the combined stochastic response surface method and automated differentiation: Application to environmental and biological systems, RISK ANAL, 20(5), 2000, pp. 591-602
Estimation of uncertainties associated with model predictions is an importa
nt component of the application of environmental and biological models. "Tr
aditional" methods for propagating uncertainty, such as standard Monte Carl
o and Latin Hypercube Sampling, however, often require performing a prohibi
tive number of model simulations, especially for complex, computationally i
ntensive models. Here, a computationally efficient method for uncertainty p
ropagation, the Stochastic Response Surface Method (SRSM) is coupled with a
nother method, the Automatic Differentiation of FORTRAN (ADIFOR). The SRSM
is based on series expansions of model inputs and outputs in terms of a set
of "well-behaved" standard random variables. The ADIFOR method is used to
transform the model code into one that calculates the derivatives of the mo
del outputs with respect to inputs or transformed inputs. The calculated mo
del outputs and the derivatives at a set of sample points are used to appro
ximate the unknown coefficients in the series expansions of outputs. A fram
ework for the coupling of the SRSM and ADIFOR is developed and presented he
re. Two case studies are presented, involving (1) a physiologically based p
harmacokinetic model for perchloroethylene for humans, and (2) an atmospher
ic photochemical model, the Reactive Plume Model. The results obtained agre
e closely with those of traditional Monte Carlo and Latin hypercube samplin
g methods, while reducing the required number of model simulations by about
two orders of magnitude.