GLOBAL SENSITIVITY ANALYSIS
Global nsitivity analysis is the study of how the uncertainty in the output of a model (numerical or otherwi) can be apportioned to different sources of uncertainty in the model input”. Global could be an unnecessary specification here, were it not for the fact that most analysis met in the literature are local or one-factor-at-a-time.
All models have u for global nsitivity analysis. Applications worked out by the Joint Rearch Centre group for Applied Statistics include: Atmospheric chemistry, transport emission modelling, fish population dynamics, composite indicators, hydrocarbon exploration models, macroeconomic modelling, radioactive waste management.
Prescriptions have been issued for global nsitivity analysis of models when the ud for policy analysis. In Europe, the European Commission recommends nsitivity analysis in the context of the extended impact asssment guidelines and handbook (2002). Similar recommendation in the United States EPA’s White Paper on model u acceptability (1999).
The EC handbook for extended impact asssment, a working document by the European Commission, 2002, states: “A good nsitivity analysis should conduct analys over the full range of plausible values of key parameters and their interactions, to asss how impacts change in respon to changes in key parameters”. The EPA paper (1999) is less prescriptive, but insists on the need for uncertainty and nsitivity analysis.
We list below what are the desirable properties of an ideal global nsitivity analysis method.
1. Cope with the influence of scale and shape. The influence of the input should incorporate the effect of the range of input variation and the form of its probability density function (pdf). It matters whether the pdf of an input factor is uniform or normal, and what are the distribution parameters.
2. Include multidimensional averaging. In a local approach to SA one computes partial derivatives, as discusd above. This is the effect of the variation of a factor when all oth
ers are kept constant at the central (nominal) value. A global method should instead evaluate the effect of a factor while all others are varying as well.
3. Be model independent. The method should work regardless of the additivity or linearity of the model. A global nsitivity measure must be able to appreciate the so-called interaction effect, especially important for non-linear, non-additive models. The ari when the effect of changing two factors is different from the sum of their individual effects.
4. Be able to treat grouped factors as if they were single factors. This property of synthesis is esntial for the agility of the interpretation of the results. One would not want to be confronted with a SA made of den tables of nsitivity measures.
南洋学院
| Property 1
Scale and shape | Property 2
Multi-dimensional averaging | Property 2
Model independence | Property 2
Grouping of factors |
Derivatives, Local methods | N | 朱志刚 N | N | Y |
Regression method, e.g standardid regression coefficients | Y | Y | N | N |
Morris | N / Y | Y 蜗牛简笔画 | Y | 怎样重启路由器Y |
Variance bad methods | Y | Y | Y | Y |
Monte Carlo Filtering父母课堂读后感 | Y | Y | 川贝炖梨 Y | N |
Table Properties of nsitivity measures |
| | | | |
A few words about the output Y of interest. In our experience, the target of interest should not be the model output per , but the question that the model has been called to answer. To make an example, if a model predicts contaminant distribution over space and time, it is the total area where a given threshold is exceeded at a given time which would play as output of interest, or the total health effects per time unit.
One should ek from the analys conclusions of relevance to the question put to the model,
as oppod to relevant to the model, e.g.
Uncertainty in emission inventories [in transport] are driven by variability in driving habits more
than from uncertainty in engine emission data.
七律元宵节In transport with chemical reaction problems, uncertainty in the chemistry dominates over uncertainty in the inventories.
Engineered barrier count less than geological barriers in radioactive waste migration.
This remark on the output of interest clearly applies to model u, not to model building, where the analyst might have interest in studying a variety of intermediate outputs.
Suggested References
Campolongo, F., Saltelli A,. Jenn, N.R., Wilson, J., and Hjorth, J., 1999a, The role of multipha chemistry in the oxidation of dimethylsulphide (DMS). A latitude dependent analysis, Journal of Atmospheric Chemistry, 32, 327-356.
松柏拼音 Campolongo, F., Tarantola, S., and Saltelli, A., 1999b, Tackling quantitatively large dimensionality problems, Computer Physics Communications, 117, 75-85.
