Premium
A method for parameter sensitivity analysis in differential equation models
Author(s) -
Kohberger Robert C.,
Scavia Donald,
Wilkinson John W.
Publication year - 1978
Publication title -
water resources research
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.863
H-Index - 217
eISSN - 1944-7973
pISSN - 0043-1397
DOI - 10.1029/wr014i001p00025
Subject(s) - sensitivity (control systems) , eigenvalues and eigenvectors , parameter space , mathematics , differential equation , quadratic equation , state space , differential (mechanical device) , scale (ratio) , model parameter , regression analysis , mathematical optimization , mathematical analysis , statistics , physics , engineering , geometry , quantum mechanics , electronic engineering , thermodynamics
A numeric method for analyzing global parameter sensitivity about a fixed point in parameter space for differential equation models is presented. The method is suitable for large‐scale, multiresponse systems which may not be in steady state. By using a quadratic model, the relationship between several global response characteristics and parameter perturbations is examined. Sensitivity relationships are defined with both backward elimination regression model selection procedures and eigenvalue‐eigenvector analyses. An example of the method is given using an ecosystem model consisting of 14 coupled differential equations.