SlideShare a Scribd company logo
1 of 21
Eawag: Swiss Federal Institute of Aquatic Science and Technology
Sensitivity Analysis
9th EAWAG Summer School in Environmental System Analysis 2017
Andreas Scheidegger
12.06.2017
What are all the knobs for?
Sensitivity Analysis
Investigate how a model output responds to
changes in model parameters and/or model inputs.
Factors
(SA does not make a difference of inputs and parameters)
Goals of Sensitivity Analysis
• Better understanding of the model and its mechanisms
• Sanity check: does the model behave as expected?
• Identification of influential and non-influential model
parameters
The model is investigated, not the underlying system!
Change Input
Sensitivity Analysis
?
Sensitivity Analysis
Change Parameter
?
Sensitivity Analysis
Change Parameter more
?
Sensitivity Analysis
Change input, but use different parameter
?
Sensitivity Analysis
Change input, but use different parameter
Local vs. Regional Sensitivity Analysis
Two approaches to sensitivity analysis:
Local Sensitivity Analysis
Regional (global) Sensitivity Analysis
Investigation of the sensitivity of model results on parameters at a given
reference point in parameter space
The results depend on the choice of the reference point and the
parameter ranges
The results do depend on the choice of the parameter distribution
Investigation of the sensitivity of model results to the parameters based
on a probability distribution of the model parameters
Local Sensitivity Analysis
Two approaches to sensitivity analysis:
Regional Sensitivity Analysis
The results do no longer depend on the choice of a single reference
point but on the choice of the parameter distribution
Investigation of the sensitivity of model results to the parameters based
on a probability distribution of the model parameters
Local Sensitivity Analysis
Investigation of the sensitivity of model results on parameters at a given
reference point in parameter space
The results depend on the choice of the reference point and the
parameter ranges
Local Sensitivity Analysis
Deterministic model function:
The sensitivity of a model result to a parameter depends on:
functional relationship provided by the model equations
selected ranges/distributions of parameters
Local Sensitivity Analysis
A “natural” measure: slope (= derivative) of the model function with
respect to the component of the parameter vector
Local Sensitivity Analysis
Sensitivity rankings: sensitivities for each model output can be difficult to evaluate. It can
be useful to average the squares of the sensitivity functions at
different values of the index i
average sensitivity of the
model for a given parameter j
Local Sensitivity Analysis
Local sensitivity analysis can provide useful insights into the
model mechanisms
It is computationally relatively inexpensive
Nonlinearities of the model are not taken into account
Parameter interactions are not observed
Regional Sensitivity Analysis
Two approaches to sensitivity analysis:
Local Sensitivity Analysis
Investigation of the sensitivity of model results on parameters at a given
reference point in parameter space
The results depend on the choice of the reference point and the
parameter ranges
Regional (global) Sensitivity Analysis
The results do no longer depend on the choice of a single reference
point but on the choice of the parameter distribution
Investigation of the sensitivity of model results to the parameters based
on a probability distribution of the model parameters
Regional Sensitivity Analysis – Variance decomposition
Variance-based techniques are based on a decomposition of the variance of the
model output into contributions due to different parameters.
A fraction of the variance of a model result i is due to the distribution of a parameter j
Regional Sensitivity Analysis– Variance decomposition
Variance-based sensitivity analysis is based on the comparison of the total variance
of the model output with a “reduced” variance when keeping one parameter fixed
The degree of variance reduction is a
measure of the contribution of the fixed
parameter to the total output variance
To remove the dependence of the conditional variance on where we fixed the
parameter, we take the expected value of these conditional variances with
respect to the marginal distribution of the selected parameter
Regional Sensitivity Analysis
Fourier amplitude sensitivity testing (FAST)
1. Change all inputs with in
different frequencies
2. Analyze frequency spectra of
outputs
Image source: https://homepages.thm.de/~hg54/mmk_2006/script/multimedia/multimedia.htm
Sensitivitypackage
Typical Workflow
Generate
parameter/input
combinations
Compute indices
loop over all parameter/input combinations
Regional sensitivity analysis
• Change all parameter together
• Computationally more
expensive
• Considers interactions
• Describes model behavior
across a parameter region
Sensitivity analysis
• Learn about a model, not a system
• Gives hints which parameter can be identified well from data
Summary
Local sensitivity analysis
• Change one parameter at the time
• Computationally cheap
• Simple interpretation
• No interactions
• Results only valid for one point in
parameter space

More Related Content

What's hot

Antimicrobial preservatives
Antimicrobial  preservativesAntimicrobial  preservatives
Antimicrobial preservativesMohiteSwapnali
 
computer aided biopharmaceutical characterization :gastrointestinal absorptio...
computer aided biopharmaceutical characterization :gastrointestinal absorptio...computer aided biopharmaceutical characterization :gastrointestinal absorptio...
computer aided biopharmaceutical characterization :gastrointestinal absorptio...Affrin Shaik
 
Advanced Drug delivery systems
Advanced Drug delivery systemsAdvanced Drug delivery systems
Advanced Drug delivery systemsFarzana Sultana
 
ACTIVE TRANSPORT- hPEPT1,ASBT,OCT,OATP, BBB-Choline Transporter.pptx
ACTIVE TRANSPORT- hPEPT1,ASBT,OCT,OATP, BBB-Choline Transporter.pptxACTIVE TRANSPORT- hPEPT1,ASBT,OCT,OATP, BBB-Choline Transporter.pptx
ACTIVE TRANSPORT- hPEPT1,ASBT,OCT,OATP, BBB-Choline Transporter.pptxPawanDhamala1
 
Tumour targeting and Brain specific drug delivery
Tumour targeting and Brain specific drug deliveryTumour targeting and Brain specific drug delivery
Tumour targeting and Brain specific drug deliverySHUBHAMGWAGH
 
Computer aided formulation development
Computer aided formulation developmentComputer aided formulation development
Computer aided formulation developmentSiddu K M
 
computers in clinical development
 computers in clinical development computers in clinical development
computers in clinical developmentSUJITHA MARY
 
virtual trial FED and fasted state.pptx
virtual trial FED and fasted state.pptxvirtual trial FED and fasted state.pptx
virtual trial FED and fasted state.pptxMrRajanSwamiSwami
 
sales forecasting and its method
sales forecasting and its method sales forecasting and its method
sales forecasting and its method Nirmal Maurya
 
COMPUTER SIMULATIONS IN PHARMACOKINETICS & PHARMACODYNAMICS
COMPUTER SIMULATIONS  IN  PHARMACOKINETICS & PHARMACODYNAMICSCOMPUTER SIMULATIONS  IN  PHARMACOKINETICS & PHARMACODYNAMICS
COMPUTER SIMULATIONS IN PHARMACOKINETICS & PHARMACODYNAMICSsagartrivedi14
 
Problem of variables
Problem of variablesProblem of variables
Problem of variablessuresh gautam
 
Descriptive versus mechanistic modelling
Descriptive versus mechanistic modellingDescriptive versus mechanistic modelling
Descriptive versus mechanistic modellingSayeda Salma S.A.
 
Gastrointestinal absorption simulation using in silico methodology; by Dr. Bh...
Gastrointestinal absorption simulation using in silico methodology; by Dr. Bh...Gastrointestinal absorption simulation using in silico methodology; by Dr. Bh...
Gastrointestinal absorption simulation using in silico methodology; by Dr. Bh...bhupenkalita7
 
Drug Distribution ,Drug Excretion, Active Transport; P–gp, BCRP, Nucleoside T...
Drug Distribution ,Drug Excretion, Active Transport; P–gp, BCRP, Nucleoside T...Drug Distribution ,Drug Excretion, Active Transport; P–gp, BCRP, Nucleoside T...
Drug Distribution ,Drug Excretion, Active Transport; P–gp, BCRP, Nucleoside T...RushikeshPalkar1
 

What's hot (20)

Budget and cost control slide share
Budget and cost control slide shareBudget and cost control slide share
Budget and cost control slide share
 
Antimicrobial preservatives
Antimicrobial  preservativesAntimicrobial  preservatives
Antimicrobial preservatives
 
Aptamers as Drug of future
Aptamers as Drug of futureAptamers as Drug of future
Aptamers as Drug of future
 
Nanoparticle
NanoparticleNanoparticle
Nanoparticle
 
computer aided biopharmaceutical characterization :gastrointestinal absorptio...
computer aided biopharmaceutical characterization :gastrointestinal absorptio...computer aided biopharmaceutical characterization :gastrointestinal absorptio...
computer aided biopharmaceutical characterization :gastrointestinal absorptio...
 
Sales forecasting
Sales forecastingSales forecasting
Sales forecasting
 
Advanced Drug delivery systems
Advanced Drug delivery systemsAdvanced Drug delivery systems
Advanced Drug delivery systems
 
ACTIVE TRANSPORT- hPEPT1,ASBT,OCT,OATP, BBB-Choline Transporter.pptx
ACTIVE TRANSPORT- hPEPT1,ASBT,OCT,OATP, BBB-Choline Transporter.pptxACTIVE TRANSPORT- hPEPT1,ASBT,OCT,OATP, BBB-Choline Transporter.pptx
ACTIVE TRANSPORT- hPEPT1,ASBT,OCT,OATP, BBB-Choline Transporter.pptx
 
Tumour targeting and Brain specific drug delivery
Tumour targeting and Brain specific drug deliveryTumour targeting and Brain specific drug delivery
Tumour targeting and Brain specific drug delivery
 
Computer aided formulation development
Computer aided formulation developmentComputer aided formulation development
Computer aided formulation development
 
CADD UNIT- 1.pptx
CADD UNIT- 1.pptxCADD UNIT- 1.pptx
CADD UNIT- 1.pptx
 
computers in clinical development
 computers in clinical development computers in clinical development
computers in clinical development
 
virtual trial FED and fasted state.pptx
virtual trial FED and fasted state.pptxvirtual trial FED and fasted state.pptx
virtual trial FED and fasted state.pptx
 
sales forecasting and its method
sales forecasting and its method sales forecasting and its method
sales forecasting and its method
 
COMPUTER SIMULATIONS IN PHARMACOKINETICS & PHARMACODYNAMICS
COMPUTER SIMULATIONS  IN  PHARMACOKINETICS & PHARMACODYNAMICSCOMPUTER SIMULATIONS  IN  PHARMACOKINETICS & PHARMACODYNAMICS
COMPUTER SIMULATIONS IN PHARMACOKINETICS & PHARMACODYNAMICS
 
Problem of variables
Problem of variablesProblem of variables
Problem of variables
 
Descriptive versus mechanistic modelling
Descriptive versus mechanistic modellingDescriptive versus mechanistic modelling
Descriptive versus mechanistic modelling
 
Gastrointestinal absorption simulation using in silico methodology; by Dr. Bh...
Gastrointestinal absorption simulation using in silico methodology; by Dr. Bh...Gastrointestinal absorption simulation using in silico methodology; by Dr. Bh...
Gastrointestinal absorption simulation using in silico methodology; by Dr. Bh...
 
Drug Distribution ,Drug Excretion, Active Transport; P–gp, BCRP, Nucleoside T...
Drug Distribution ,Drug Excretion, Active Transport; P–gp, BCRP, Nucleoside T...Drug Distribution ,Drug Excretion, Active Transport; P–gp, BCRP, Nucleoside T...
Drug Distribution ,Drug Excretion, Active Transport; P–gp, BCRP, Nucleoside T...
 
Aquasomes
AquasomesAquasomes
Aquasomes
 

Similar to Sensitivity analysis

non linerity SA.pptx
 non linerity SA.pptx non linerity SA.pptx
non linerity SA.pptxPawanDhamala1
 
A Systems Approach to the Modeling and Control of Molecular, Microparticle, a...
A Systems Approach to the Modeling and Control of Molecular, Microparticle, a...A Systems Approach to the Modeling and Control of Molecular, Microparticle, a...
A Systems Approach to the Modeling and Control of Molecular, Microparticle, a...ejhukkanen
 
Metabolomic Data Analysis Workshop and Tutorials (2014)
Metabolomic Data Analysis Workshop and Tutorials (2014)Metabolomic Data Analysis Workshop and Tutorials (2014)
Metabolomic Data Analysis Workshop and Tutorials (2014)Dmitry Grapov
 
validation and verification part 2.pptx
validation and verification part 2.pptxvalidation and verification part 2.pptx
validation and verification part 2.pptxubaidullah75790
 
From sensor readings to prediction: on the process of developing practical so...
From sensor readings to prediction: on the process of developing practical so...From sensor readings to prediction: on the process of developing practical so...
From sensor readings to prediction: on the process of developing practical so...Manuel Martín
 
AIAA-MAO-DSUS-2012
AIAA-MAO-DSUS-2012AIAA-MAO-DSUS-2012
AIAA-MAO-DSUS-2012OptiModel
 
CAE344 ESA UNIT I.pptx
CAE344 ESA UNIT I.pptxCAE344 ESA UNIT I.pptx
CAE344 ESA UNIT I.pptxssuser7f5130
 
DSUS_MAO_2012_Jie
DSUS_MAO_2012_JieDSUS_MAO_2012_Jie
DSUS_MAO_2012_JieMDO_Lab
 
PEMF-1-MAO2012-Ali
PEMF-1-MAO2012-AliPEMF-1-MAO2012-Ali
PEMF-1-MAO2012-AliMDO_Lab
 
Measurement System Analysis
Measurement System AnalysisMeasurement System Analysis
Measurement System AnalysisRonald Shewchuk
 
Measurement Systems Analysis - Variable Gage R&R Study Metrics, Applications ...
Measurement Systems Analysis - Variable Gage R&R Study Metrics, Applications ...Measurement Systems Analysis - Variable Gage R&R Study Metrics, Applications ...
Measurement Systems Analysis - Variable Gage R&R Study Metrics, Applications ...Gabor Szabo, CQE
 
02trainingmaterialformsa (1) 111030062223-phpapp02
02trainingmaterialformsa (1) 111030062223-phpapp0202trainingmaterialformsa (1) 111030062223-phpapp02
02trainingmaterialformsa (1) 111030062223-phpapp02Junelly Grace Catalan-Tecson
 
02training material for msa
02training material for msa02training material for msa
02training material for msa營松 林
 
26738157 sampling-design
26738157 sampling-design26738157 sampling-design
26738157 sampling-designMounzer BOUBOU
 

Similar to Sensitivity analysis (20)

non linerity SA.pptx
 non linerity SA.pptx non linerity SA.pptx
non linerity SA.pptx
 
A Systems Approach to the Modeling and Control of Molecular, Microparticle, a...
A Systems Approach to the Modeling and Control of Molecular, Microparticle, a...A Systems Approach to the Modeling and Control of Molecular, Microparticle, a...
A Systems Approach to the Modeling and Control of Molecular, Microparticle, a...
 
Metabolomic Data Analysis Workshop and Tutorials (2014)
Metabolomic Data Analysis Workshop and Tutorials (2014)Metabolomic Data Analysis Workshop and Tutorials (2014)
Metabolomic Data Analysis Workshop and Tutorials (2014)
 
validation and verification part 2.pptx
validation and verification part 2.pptxvalidation and verification part 2.pptx
validation and verification part 2.pptx
 
From sensor readings to prediction: on the process of developing practical so...
From sensor readings to prediction: on the process of developing practical so...From sensor readings to prediction: on the process of developing practical so...
From sensor readings to prediction: on the process of developing practical so...
 
AIAA-MAO-DSUS-2012
AIAA-MAO-DSUS-2012AIAA-MAO-DSUS-2012
AIAA-MAO-DSUS-2012
 
CAE344 ESA UNIT I.pptx
CAE344 ESA UNIT I.pptxCAE344 ESA UNIT I.pptx
CAE344 ESA UNIT I.pptx
 
Research
ResearchResearch
Research
 
DSUS_MAO_2012_Jie
DSUS_MAO_2012_JieDSUS_MAO_2012_Jie
DSUS_MAO_2012_Jie
 
PEMF-1-MAO2012-Ali
PEMF-1-MAO2012-AliPEMF-1-MAO2012-Ali
PEMF-1-MAO2012-Ali
 
Linear regression analysis
Linear regression analysisLinear regression analysis
Linear regression analysis
 
MSA (GR&R)
MSA (GR&R)MSA (GR&R)
MSA (GR&R)
 
Measurement System Analysis
Measurement System AnalysisMeasurement System Analysis
Measurement System Analysis
 
Process dynamic control
Process dynamic controlProcess dynamic control
Process dynamic control
 
Measurement Systems Analysis - Variable Gage R&R Study Metrics, Applications ...
Measurement Systems Analysis - Variable Gage R&R Study Metrics, Applications ...Measurement Systems Analysis - Variable Gage R&R Study Metrics, Applications ...
Measurement Systems Analysis - Variable Gage R&R Study Metrics, Applications ...
 
02trainingmaterialformsa (1) 111030062223-phpapp02
02trainingmaterialformsa (1) 111030062223-phpapp0202trainingmaterialformsa (1) 111030062223-phpapp02
02trainingmaterialformsa (1) 111030062223-phpapp02
 
02training material for msa
02training material for msa02training material for msa
02training material for msa
 
MI Mod 1.ppt
MI Mod 1.pptMI Mod 1.ppt
MI Mod 1.ppt
 
26738157 sampling-design
26738157 sampling-design26738157 sampling-design
26738157 sampling-design
 
om
omom
om
 

More from Andreas Scheidegger

Bayesian End-Member Mixing Model
Bayesian End-Member Mixing ModelBayesian End-Member Mixing Model
Bayesian End-Member Mixing ModelAndreas Scheidegger
 
Formulation of model likelihood functions
Formulation of model likelihood functionsFormulation of model likelihood functions
Formulation of model likelihood functionsAndreas Scheidegger
 
Recurrent Neuronal Network tailored for Weather Radar Nowcasting
Recurrent Neuronal Network tailored for Weather Radar NowcastingRecurrent Neuronal Network tailored for Weather Radar Nowcasting
Recurrent Neuronal Network tailored for Weather Radar NowcastingAndreas Scheidegger
 
Kalibrierung von Kanalnetzmodellen mit binären Messdaten
Kalibrierung von Kanalnetzmodellen mit binären MessdatenKalibrierung von Kanalnetzmodellen mit binären Messdaten
Kalibrierung von Kanalnetzmodellen mit binären MessdatenAndreas Scheidegger
 
Experimental design approach for optimal selection and placement of rain sensors
Experimental design approach for optimal selection and placement of rain sensorsExperimental design approach for optimal selection and placement of rain sensors
Experimental design approach for optimal selection and placement of rain sensorsAndreas Scheidegger
 
Bayesian assimilation of rainfall sensors with fundamentally different integr...
Bayesian assimilation of rainfall sensors with fundamentally different integr...Bayesian assimilation of rainfall sensors with fundamentally different integr...
Bayesian assimilation of rainfall sensors with fundamentally different integr...Andreas Scheidegger
 
New information sources for rain fields
New information sources for rain fieldsNew information sources for rain fields
New information sources for rain fieldsAndreas Scheidegger
 

More from Andreas Scheidegger (8)

Bayesian End-Member Mixing Model
Bayesian End-Member Mixing ModelBayesian End-Member Mixing Model
Bayesian End-Member Mixing Model
 
Formulation of model likelihood functions
Formulation of model likelihood functionsFormulation of model likelihood functions
Formulation of model likelihood functions
 
Review of probability calculus
Review of probability calculusReview of probability calculus
Review of probability calculus
 
Recurrent Neuronal Network tailored for Weather Radar Nowcasting
Recurrent Neuronal Network tailored for Weather Radar NowcastingRecurrent Neuronal Network tailored for Weather Radar Nowcasting
Recurrent Neuronal Network tailored for Weather Radar Nowcasting
 
Kalibrierung von Kanalnetzmodellen mit binären Messdaten
Kalibrierung von Kanalnetzmodellen mit binären MessdatenKalibrierung von Kanalnetzmodellen mit binären Messdaten
Kalibrierung von Kanalnetzmodellen mit binären Messdaten
 
Experimental design approach for optimal selection and placement of rain sensors
Experimental design approach for optimal selection and placement of rain sensorsExperimental design approach for optimal selection and placement of rain sensors
Experimental design approach for optimal selection and placement of rain sensors
 
Bayesian assimilation of rainfall sensors with fundamentally different integr...
Bayesian assimilation of rainfall sensors with fundamentally different integr...Bayesian assimilation of rainfall sensors with fundamentally different integr...
Bayesian assimilation of rainfall sensors with fundamentally different integr...
 
New information sources for rain fields
New information sources for rain fieldsNew information sources for rain fields
New information sources for rain fields
 

Recently uploaded

Davis plaque method.pptx recombinant DNA technology
Davis plaque method.pptx recombinant DNA technologyDavis plaque method.pptx recombinant DNA technology
Davis plaque method.pptx recombinant DNA technologycaarthichand2003
 
Harmful and Useful Microorganisms Presentation
Harmful and Useful Microorganisms PresentationHarmful and Useful Microorganisms Presentation
Harmful and Useful Microorganisms Presentationtahreemzahra82
 
The dark energy paradox leads to a new structure of spacetime.pptx
The dark energy paradox leads to a new structure of spacetime.pptxThe dark energy paradox leads to a new structure of spacetime.pptx
The dark energy paradox leads to a new structure of spacetime.pptxEran Akiva Sinbar
 
Topic 9- General Principles of International Law.pptx
Topic 9- General Principles of International Law.pptxTopic 9- General Principles of International Law.pptx
Topic 9- General Principles of International Law.pptxJorenAcuavera1
 
FREE NURSING BUNDLE FOR NURSES.PDF by na
FREE NURSING BUNDLE FOR NURSES.PDF by naFREE NURSING BUNDLE FOR NURSES.PDF by na
FREE NURSING BUNDLE FOR NURSES.PDF by naJASISJULIANOELYNV
 
Pests of Blackgram, greengram, cowpea_Dr.UPR.pdf
Pests of Blackgram, greengram, cowpea_Dr.UPR.pdfPests of Blackgram, greengram, cowpea_Dr.UPR.pdf
Pests of Blackgram, greengram, cowpea_Dr.UPR.pdfPirithiRaju
 
Scheme-of-Work-Science-Stage-4 cambridge science.docx
Scheme-of-Work-Science-Stage-4 cambridge science.docxScheme-of-Work-Science-Stage-4 cambridge science.docx
Scheme-of-Work-Science-Stage-4 cambridge science.docxyaramohamed343013
 
User Guide: Capricorn FLX™ Weather Station
User Guide: Capricorn FLX™ Weather StationUser Guide: Capricorn FLX™ Weather Station
User Guide: Capricorn FLX™ Weather StationColumbia Weather Systems
 
Grafana in space: Monitoring Japan's SLIM moon lander in real time
Grafana in space: Monitoring Japan's SLIM moon lander  in real timeGrafana in space: Monitoring Japan's SLIM moon lander  in real time
Grafana in space: Monitoring Japan's SLIM moon lander in real timeSatoshi NAKAHIRA
 
Neurodevelopmental disorders according to the dsm 5 tr
Neurodevelopmental disorders according to the dsm 5 trNeurodevelopmental disorders according to the dsm 5 tr
Neurodevelopmental disorders according to the dsm 5 trssuser06f238
 
Behavioral Disorder: Schizophrenia & it's Case Study.pdf
Behavioral Disorder: Schizophrenia & it's Case Study.pdfBehavioral Disorder: Schizophrenia & it's Case Study.pdf
Behavioral Disorder: Schizophrenia & it's Case Study.pdfSELF-EXPLANATORY
 
Pests of castor_Binomics_Identification_Dr.UPR.pdf
Pests of castor_Binomics_Identification_Dr.UPR.pdfPests of castor_Binomics_Identification_Dr.UPR.pdf
Pests of castor_Binomics_Identification_Dr.UPR.pdfPirithiRaju
 
LIGHT-PHENOMENA-BY-CABUALDIONALDOPANOGANCADIENTE-CONDEZA (1).pptx
LIGHT-PHENOMENA-BY-CABUALDIONALDOPANOGANCADIENTE-CONDEZA (1).pptxLIGHT-PHENOMENA-BY-CABUALDIONALDOPANOGANCADIENTE-CONDEZA (1).pptx
LIGHT-PHENOMENA-BY-CABUALDIONALDOPANOGANCADIENTE-CONDEZA (1).pptxmalonesandreagweneth
 
Is RISC-V ready for HPC workload? Maybe?
Is RISC-V ready for HPC workload? Maybe?Is RISC-V ready for HPC workload? Maybe?
Is RISC-V ready for HPC workload? Maybe?Patrick Diehl
 
Sulphur & Phosphrus Cycle PowerPoint Presentation (2) [Autosaved]-3-1.pptx
Sulphur & Phosphrus Cycle PowerPoint Presentation (2) [Autosaved]-3-1.pptxSulphur & Phosphrus Cycle PowerPoint Presentation (2) [Autosaved]-3-1.pptx
Sulphur & Phosphrus Cycle PowerPoint Presentation (2) [Autosaved]-3-1.pptxnoordubaliya2003
 
User Guide: Pulsar™ Weather Station (Columbia Weather Systems)
User Guide: Pulsar™ Weather Station (Columbia Weather Systems)User Guide: Pulsar™ Weather Station (Columbia Weather Systems)
User Guide: Pulsar™ Weather Station (Columbia Weather Systems)Columbia Weather Systems
 
(9818099198) Call Girls In Noida Sector 14 (NOIDA ESCORTS)
(9818099198) Call Girls In Noida Sector 14 (NOIDA ESCORTS)(9818099198) Call Girls In Noida Sector 14 (NOIDA ESCORTS)
(9818099198) Call Girls In Noida Sector 14 (NOIDA ESCORTS)riyaescorts54
 
Environmental Biotechnology Topic:- Microbial Biosensor
Environmental Biotechnology Topic:- Microbial BiosensorEnvironmental Biotechnology Topic:- Microbial Biosensor
Environmental Biotechnology Topic:- Microbial Biosensorsonawaneprad
 
BIOETHICS IN RECOMBINANT DNA TECHNOLOGY.
BIOETHICS IN RECOMBINANT DNA TECHNOLOGY.BIOETHICS IN RECOMBINANT DNA TECHNOLOGY.
BIOETHICS IN RECOMBINANT DNA TECHNOLOGY.PraveenaKalaiselvan1
 
Call Girls In Nihal Vihar Delhi ❤️8860477959 Looking Escorts In 24/7 Delhi NCR
Call Girls In Nihal Vihar Delhi ❤️8860477959 Looking Escorts In 24/7 Delhi NCRCall Girls In Nihal Vihar Delhi ❤️8860477959 Looking Escorts In 24/7 Delhi NCR
Call Girls In Nihal Vihar Delhi ❤️8860477959 Looking Escorts In 24/7 Delhi NCRlizamodels9
 

Recently uploaded (20)

Davis plaque method.pptx recombinant DNA technology
Davis plaque method.pptx recombinant DNA technologyDavis plaque method.pptx recombinant DNA technology
Davis plaque method.pptx recombinant DNA technology
 
Harmful and Useful Microorganisms Presentation
Harmful and Useful Microorganisms PresentationHarmful and Useful Microorganisms Presentation
Harmful and Useful Microorganisms Presentation
 
The dark energy paradox leads to a new structure of spacetime.pptx
The dark energy paradox leads to a new structure of spacetime.pptxThe dark energy paradox leads to a new structure of spacetime.pptx
The dark energy paradox leads to a new structure of spacetime.pptx
 
Topic 9- General Principles of International Law.pptx
Topic 9- General Principles of International Law.pptxTopic 9- General Principles of International Law.pptx
Topic 9- General Principles of International Law.pptx
 
FREE NURSING BUNDLE FOR NURSES.PDF by na
FREE NURSING BUNDLE FOR NURSES.PDF by naFREE NURSING BUNDLE FOR NURSES.PDF by na
FREE NURSING BUNDLE FOR NURSES.PDF by na
 
Pests of Blackgram, greengram, cowpea_Dr.UPR.pdf
Pests of Blackgram, greengram, cowpea_Dr.UPR.pdfPests of Blackgram, greengram, cowpea_Dr.UPR.pdf
Pests of Blackgram, greengram, cowpea_Dr.UPR.pdf
 
Scheme-of-Work-Science-Stage-4 cambridge science.docx
Scheme-of-Work-Science-Stage-4 cambridge science.docxScheme-of-Work-Science-Stage-4 cambridge science.docx
Scheme-of-Work-Science-Stage-4 cambridge science.docx
 
User Guide: Capricorn FLX™ Weather Station
User Guide: Capricorn FLX™ Weather StationUser Guide: Capricorn FLX™ Weather Station
User Guide: Capricorn FLX™ Weather Station
 
Grafana in space: Monitoring Japan's SLIM moon lander in real time
Grafana in space: Monitoring Japan's SLIM moon lander  in real timeGrafana in space: Monitoring Japan's SLIM moon lander  in real time
Grafana in space: Monitoring Japan's SLIM moon lander in real time
 
Neurodevelopmental disorders according to the dsm 5 tr
Neurodevelopmental disorders according to the dsm 5 trNeurodevelopmental disorders according to the dsm 5 tr
Neurodevelopmental disorders according to the dsm 5 tr
 
Behavioral Disorder: Schizophrenia & it's Case Study.pdf
Behavioral Disorder: Schizophrenia & it's Case Study.pdfBehavioral Disorder: Schizophrenia & it's Case Study.pdf
Behavioral Disorder: Schizophrenia & it's Case Study.pdf
 
Pests of castor_Binomics_Identification_Dr.UPR.pdf
Pests of castor_Binomics_Identification_Dr.UPR.pdfPests of castor_Binomics_Identification_Dr.UPR.pdf
Pests of castor_Binomics_Identification_Dr.UPR.pdf
 
LIGHT-PHENOMENA-BY-CABUALDIONALDOPANOGANCADIENTE-CONDEZA (1).pptx
LIGHT-PHENOMENA-BY-CABUALDIONALDOPANOGANCADIENTE-CONDEZA (1).pptxLIGHT-PHENOMENA-BY-CABUALDIONALDOPANOGANCADIENTE-CONDEZA (1).pptx
LIGHT-PHENOMENA-BY-CABUALDIONALDOPANOGANCADIENTE-CONDEZA (1).pptx
 
Is RISC-V ready for HPC workload? Maybe?
Is RISC-V ready for HPC workload? Maybe?Is RISC-V ready for HPC workload? Maybe?
Is RISC-V ready for HPC workload? Maybe?
 
Sulphur & Phosphrus Cycle PowerPoint Presentation (2) [Autosaved]-3-1.pptx
Sulphur & Phosphrus Cycle PowerPoint Presentation (2) [Autosaved]-3-1.pptxSulphur & Phosphrus Cycle PowerPoint Presentation (2) [Autosaved]-3-1.pptx
Sulphur & Phosphrus Cycle PowerPoint Presentation (2) [Autosaved]-3-1.pptx
 
User Guide: Pulsar™ Weather Station (Columbia Weather Systems)
User Guide: Pulsar™ Weather Station (Columbia Weather Systems)User Guide: Pulsar™ Weather Station (Columbia Weather Systems)
User Guide: Pulsar™ Weather Station (Columbia Weather Systems)
 
(9818099198) Call Girls In Noida Sector 14 (NOIDA ESCORTS)
(9818099198) Call Girls In Noida Sector 14 (NOIDA ESCORTS)(9818099198) Call Girls In Noida Sector 14 (NOIDA ESCORTS)
(9818099198) Call Girls In Noida Sector 14 (NOIDA ESCORTS)
 
Environmental Biotechnology Topic:- Microbial Biosensor
Environmental Biotechnology Topic:- Microbial BiosensorEnvironmental Biotechnology Topic:- Microbial Biosensor
Environmental Biotechnology Topic:- Microbial Biosensor
 
BIOETHICS IN RECOMBINANT DNA TECHNOLOGY.
BIOETHICS IN RECOMBINANT DNA TECHNOLOGY.BIOETHICS IN RECOMBINANT DNA TECHNOLOGY.
BIOETHICS IN RECOMBINANT DNA TECHNOLOGY.
 
Call Girls In Nihal Vihar Delhi ❤️8860477959 Looking Escorts In 24/7 Delhi NCR
Call Girls In Nihal Vihar Delhi ❤️8860477959 Looking Escorts In 24/7 Delhi NCRCall Girls In Nihal Vihar Delhi ❤️8860477959 Looking Escorts In 24/7 Delhi NCR
Call Girls In Nihal Vihar Delhi ❤️8860477959 Looking Escorts In 24/7 Delhi NCR
 

Sensitivity analysis

  • 1. Eawag: Swiss Federal Institute of Aquatic Science and Technology Sensitivity Analysis 9th EAWAG Summer School in Environmental System Analysis 2017 Andreas Scheidegger 12.06.2017
  • 2. What are all the knobs for?
  • 3. Sensitivity Analysis Investigate how a model output responds to changes in model parameters and/or model inputs. Factors (SA does not make a difference of inputs and parameters)
  • 4. Goals of Sensitivity Analysis • Better understanding of the model and its mechanisms • Sanity check: does the model behave as expected? • Identification of influential and non-influential model parameters The model is investigated, not the underlying system!
  • 8. Sensitivity Analysis Change input, but use different parameter ?
  • 9. Sensitivity Analysis Change input, but use different parameter
  • 10. Local vs. Regional Sensitivity Analysis Two approaches to sensitivity analysis: Local Sensitivity Analysis Regional (global) Sensitivity Analysis Investigation of the sensitivity of model results on parameters at a given reference point in parameter space The results depend on the choice of the reference point and the parameter ranges The results do depend on the choice of the parameter distribution Investigation of the sensitivity of model results to the parameters based on a probability distribution of the model parameters
  • 11. Local Sensitivity Analysis Two approaches to sensitivity analysis: Regional Sensitivity Analysis The results do no longer depend on the choice of a single reference point but on the choice of the parameter distribution Investigation of the sensitivity of model results to the parameters based on a probability distribution of the model parameters Local Sensitivity Analysis Investigation of the sensitivity of model results on parameters at a given reference point in parameter space The results depend on the choice of the reference point and the parameter ranges
  • 12. Local Sensitivity Analysis Deterministic model function: The sensitivity of a model result to a parameter depends on: functional relationship provided by the model equations selected ranges/distributions of parameters
  • 13. Local Sensitivity Analysis A “natural” measure: slope (= derivative) of the model function with respect to the component of the parameter vector
  • 14. Local Sensitivity Analysis Sensitivity rankings: sensitivities for each model output can be difficult to evaluate. It can be useful to average the squares of the sensitivity functions at different values of the index i average sensitivity of the model for a given parameter j
  • 15. Local Sensitivity Analysis Local sensitivity analysis can provide useful insights into the model mechanisms It is computationally relatively inexpensive Nonlinearities of the model are not taken into account Parameter interactions are not observed
  • 16. Regional Sensitivity Analysis Two approaches to sensitivity analysis: Local Sensitivity Analysis Investigation of the sensitivity of model results on parameters at a given reference point in parameter space The results depend on the choice of the reference point and the parameter ranges Regional (global) Sensitivity Analysis The results do no longer depend on the choice of a single reference point but on the choice of the parameter distribution Investigation of the sensitivity of model results to the parameters based on a probability distribution of the model parameters
  • 17. Regional Sensitivity Analysis – Variance decomposition Variance-based techniques are based on a decomposition of the variance of the model output into contributions due to different parameters. A fraction of the variance of a model result i is due to the distribution of a parameter j
  • 18. Regional Sensitivity Analysis– Variance decomposition Variance-based sensitivity analysis is based on the comparison of the total variance of the model output with a “reduced” variance when keeping one parameter fixed The degree of variance reduction is a measure of the contribution of the fixed parameter to the total output variance To remove the dependence of the conditional variance on where we fixed the parameter, we take the expected value of these conditional variances with respect to the marginal distribution of the selected parameter
  • 19. Regional Sensitivity Analysis Fourier amplitude sensitivity testing (FAST) 1. Change all inputs with in different frequencies 2. Analyze frequency spectra of outputs Image source: https://homepages.thm.de/~hg54/mmk_2006/script/multimedia/multimedia.htm
  • 21. Regional sensitivity analysis • Change all parameter together • Computationally more expensive • Considers interactions • Describes model behavior across a parameter region Sensitivity analysis • Learn about a model, not a system • Gives hints which parameter can be identified well from data Summary Local sensitivity analysis • Change one parameter at the time • Computationally cheap • Simple interpretation • No interactions • Results only valid for one point in parameter space