Note: Front Cover -- Sensitivity Analysis in Earth Observation Modelling -- Sensitivity Analysis in Earth Observation Modelling -- Copyright -- Dedication -- Contents -- List of Contributors -- Preface -- ABOUT THE COVER -- 1 - INTRODUCTION TO SA IN EARTH OBSERVATION (EO) -- 1 - OVERVIEW OF SENSITIVITY ANALYSIS METHODS IN EARTH OBSERVATION MODELING -- 1. INTRODUCTION -- 1.1 DEFINING THE MODEL OUTPUTS AND INPUTS FOR SENSITIVITY ANALYSIS -- 1.1.1 Defining Factor (or Parametric) Uncertainty -- 2. LOCAL SENSITIVITY ANALYSIS -- 2.1 CORRELATION ANALYSIS -- 2.2 REGRESSION ANALYSIS -- 3. GLOBAL SENSITIVITY ANALYSIS -- 3.1 ONE-AT-A-TIME SENSITIVITY ANALYSIS METHODS -- 3.2 THE MORRIS METHOD FOR FACTOR SCREENING -- 3.3 VARIANCE-BASED SENSITIVITY ANALYSIS -- 3.4 SAMPLING METHODS FOR GLOBAL SENSITIVITY ANALYSIS -- 3.4.1 Random Sampling -- 3.4.2 Stratified Sampling and the Latin Hypercube -- 3.4.3 Sampling for Sensitivity Indices -- 3.5 SURROGATE MODELS FOR GLOBAL SENSITIVITY ANALYSIS -- 3.5.1 Generalized Linear Modeling -- 3.5.2 Neural Networks -- 3.5.3 Direct Sensitivity Analysis of Surrogate Models -- 3.6 POLYNOMIAL CHAOS -- 3.7 GAUSSIAN PROCESS AND BAYES LINEAR EMULATION -- 4. GRAPHICAL METHODS FOR GLOBAL SENSITIVITY ANALYSIS -- 4.1 SCATTER PLOTS -- 4.2 PLOTTING THE RESPONSE SURFACE -- 4.3 PLOTTING THE SENSITIVITY INDICES -- 5. CONCLUSIONS -- REFERENCES -- 2 - MODEL INPUT DATA UNCERTAINTY AND ITS POTENTIAL IMPACT ON SOIL PROPERTIES -- 1. INTRODUCTION -- 2. A WORLD OF MODELS - HOW CAN THEY BE CLASSIFIED? -- 3. CAN WE TRUST MODELS? - MODEL ACCURACY AND THEIR SENSITIVITY TO INPUT DATA UNCERTAINTY -- 4. SELECTING THE MOST APPROPRIATE MODEL -- 5. WHY AND HOW TO ACCOUNT FOR MODELING UNCERTAINTIES CAUSED BY DIFFERENT INPUT DATA SOURCES -- 6. ASSESSING SENSITIVITY OF ENVIRONMENTAL MODELS. , 7. HOW SOIL TEXTURE MEASURED WITH VISIBLE-NEAR-INFRARED SPECTROSCOPY AFFECTS HYDROLOGICAL MODELING: A CASE STUDY -- 7.1 STUDY SITES AND INSTRUMENTS -- 7.2 SOIL SAMPLES -- 7.3 CHEMOMETRICS -- 7.4 IMPACT OF CHEMOMETRIC METHOD ON SOIL PREDICTION -- 7.5 DIFFERENT INSTRUMENTS, DIFFERENT SOIL PREDICTIONS? WHAT WAS FINALLY THE BEST SOIL PREDICTION ACCURACY? -- 7.6 WHAT DOES THIS FINALLY MEAN FOR OUR ENVIRONMENTAL MODELING? -- 8. WHAT DID WE LEARN? -- REFERENCES -- 2 - LOCAL SA METHODS: CASE STUDIES -- 3 - LOCAL SENSITIVITY ANALYSIS OF THE LANDSOIL EROSION MODEL APPLIED TO A VIRTUAL CATCHMENT -- 1. INTRODUCTION -- 2. MATERIALS AND METHODS -- 2.1 MODEL DESCRIPTION -- 2.2 SENSITIVITY ANALYSIS -- 2.2.1 Parameters -- 2.2.2 Virtual Catchment -- 2.2.3 Sensitivity Calculation -- 3. RESULTS AND DISCUSSION -- 3.1 LINEAR HILLSLOPE -- 3.1.1 Aggregated Parameters -- 3.2 COMPLEX HILLSLOPES -- 4. CONCLUDING REMARKS -- Acknowledgments -- REFERENCES -- 4 - SENSITIVITY OF VEGETATION PHENOLOGICAL PARAMETERS: FROM SATELLITE SENSORS TO SPATIAL RESOLUTION AND TEMPORAL CO ... -- 1. INTRODUCTION -- 2. MONITORING VEGETATION PHENOLOGY -- 3. SENSITIVITY ANALYSIS -- 4. SENSITIVITY OF REMOTELY SENSED PHENOLOGICAL PARAMETERS -- 4.1 SATELLITE SENSOR -- 4.2 VEGETATION INDEX -- 4.3 SPATIAL RESOLUTION -- 4.4 COMPOSITE PERIOD, SMOOTHING, AND FILTERING -- 4.4.1 Composite Period -- 4.4.2 Smoothing Techniques -- 5. CASE STUDY -- 5.1 STUDY AREA -- 5.2 DATA AND METHODOLOGY -- 5.3 RESULTS AND DISCUSSION -- 6. CONCLUSION -- REFERENCES -- 5 - RADAR RAINFALL SENSITIVITY ANALYSIS USING MULTIVARIATE DISTRIBUTED ENSEMBLE GENERATOR∗ -- 1. INTRODUCTION -- 2. DATA AND METHODS -- 2.1 STUDY AREA AND DATA SOURCES -- 2.2 THE MULTIVARIATE DISTRIBUTED ENSEMBLE GENERATOR -- 2.3 THE XINANJIANG MODEL -- 3. METHODOLOGY -- 3.1 EXPERIMENTAL DESIGN -- 3.2 VERIFICATION METHOD -- 4. RESULTS AND DISCUSSION. , 4.1 IMPLEMENTATION OF ENSEMBLE FLOW GENERATION -- 4.2 IMPACT OF ERROR DISTRIBUTION ON MODEL OUTPUT -- 4.3 IMPACT OF SPATIOTEMPORAL DEPENDENCE ON MODEL OUTPUT -- 5. CONCLUSIONS -- REFERENCES -- 6 - FIELD-SCALE SENSITIVITY OF VEGETATION DISCRIMINATION TO HYPERSPECTRAL REFLECTANCE AND COUPLED STATISTICS -- 1. INTRODUCTION -- 2. BACKGROUND ON SPECTRAL DISCRIMINATION OF VEGETATION -- 2.1 PARAMETRIC VERSUS NONPARAMETRIC STATISTICAL TESTS -- 2.1.1 Other Discrimination Methods -- 2.2 UNALTERED VERSUS PROCESSED HYPERSPECTRAL REFLECTANCE -- 2.3 CASE STUDIES FOR EFFECTS OF TYPE OF REFLECTANCE AND STATISTICAL TEST ON THE VEGETATION DISCRIMINATION RESULTS -- 3. SENSITIVITY OF SPECTRAL DISCRIMINATION OF VEGETATION TO THE TYPE OF REFLECTANCE AND STATISTICAL TEST -- 3.1 HYPERSPECTRAL DATA AND METHOD DESCRIPTION -- 3.2 SENSITIVITY OF VEGETATION SPECTRAL DISCRIMINATION TO REFLECTANCE TYPE AND STATISTICAL METHOD -- 3.3 SENSITIVITY OF VEGETATION SPECTRAL DISCRIMINATION TO THE NUMBER OF OBSERVATIONS -- 4. FINAL REMARKS -- REFERENCES -- 3 - GLOBAL (OR VARIANCE)-BASED SA METHODS: CASE STUDIES -- 7 - A MULTIMETHOD GLOBAL SENSITIVITY ANALYSIS APPROACH TO SUPPORT THE CALIBRATION AND EVALUATION OF LAND SURFACE MODELS -- 1. INTRODUCTION -- 2. MODEL AND METHODS -- 2.1 REGIONAL SENSITIVITY ANALYSIS -- 2.2 VARIANCE-BASED SENSITIVITY ANALYSIS -- 2.3 THE PAWN DENSITY-BASED METHOD -- 2.4 THE JULES MODEL -- 2.5 THE SANTA RITA CREOSOTE STUDY SITE -- 2.6 EXPERIMENTAL SETUP: DEFINITION OF INPUT FACTORS AND OUTPUTS -- 2.7 DEFINITION OF THE RANGE OF VARIATION OF THE INPUT FACTORS -- 3. RESULTS -- 3.1 RESULTS OF REGIONAL SENSITIVITY ANALYSIS -- 3.2 RESULTS OF VARIANCE-BASED SENSITIVITY ANALYSIS -- 3.3 RESULTS OF PAWN -- 3.4 OVERALL SENSITIVITY ASSESSMENT FROM THE MULTIMETHOD APPROACH -- 4. CONCLUSIONS -- Acknowledgments -- REFERENCES. , 8 - GLOBAL SENSITIVITY ANALYSIS FOR SUPPORTING HISTORY MATCHING OF GEOMECHANICAL RESERVOIR MODELS USING SATELLITE I ... -- 1. INTRODUCTION -- 2. CASE STUDY -- 2.1 SURFACE DEFORMATION AT THE KB-501 WELL OF IN SALAH SITE -- 2.2 THREE-DIMENSIONAL HYDROMECHANICAL MODEL OF KB-501 -- 3. METHODS -- 3.1 VARIANCE-BASED GLOBAL SENSITIVITY ANALYSIS -- 3.2 PRINCIPLES OF METAMODELING -- 3.3 INTRODUCTION TO KRIGING METAMODEL -- 3.4 DESCRIPTION OF THE WORKFLOW -- 4. APPLICATION -- 4.1 REDUCING THE NUMBER OF UNCERTAINTY INPUT PARAMETERS -- 4.2 CALIBRATION OF THE RESERVOIR YOUNG'S MODULUS -- SUMMARY AND FUTURE WORK -- Acknowledgments -- REFERENCES -- 9 - ARTIFICIAL NEURAL NETWORKS FOR SPECTRAL SENSITIVITY ANALYSIS TO OPTIMIZE INVERSION ALGORITHMS FOR SATELLITE-BAS ... -- 1. INTRODUCTION -- 2. DATA AND METHODS -- 2.1 ARTIFICIAL NEURAL NETWORKS: OVERVIEW -- 2.1.1 Artificial Neural Networks for the Inversion of Satellite Measurements of Spectral Radiation for the Observation of the Ear ... -- 2.2 NEURAL NETWORK-BASED TECHNIQUES TO REDUCE THE INPUT VECTOR DIMENSIONALITY -- 2.2.1 Extended Pruning -- 2.2.2 Autoassociative Neural Networks -- 2.3 SAMPLE DATA SET -- 2.3.1 Sulfate Aerosols and Their Extinction Coefficient -- 2.3.2 Thermal Infrared Satellite Pseudo-Observations -- 3. RESULTS -- 3.1 TRAINING AND TESTING THE MAXIMUM DIMENSIONALITY NEURAL NETWORK -- 3.2 SELECTION OF THE INPUT WAVELENGTHS AND SPECTRAL SENSITIVITY ANALYSIS -- 3.3 COMPARING THE PERFORMANCES OF REDUCED DIMENSIONALITY NEURAL NETWORK -- 4. CONCLUSIONS -- Acknowledgments -- REFERENCES -- 10 - GLOBAL SENSITIVITY ANALYSIS FOR UNCERTAIN PARAMETERS, MODELS, AND SCENARIOS -- 1. INTRODUCTION -- 2. MORRIS METHOD -- 3. SOBOL' METHOD -- 3.1 FIRST-ORDER AND TOTAL-EFFECT SENSITIVITY INDICES -- 3.2 MONTE CARLO IMPLEMENTATION AND TWO APPROXIMATION METHODS. , 3.2.1 Sparse-Grid Collocation for Evaluating Mean and Variance -- 3.2.2 Distributed Evaluation of Local Sensitivity Analysis -- 4. SOBOL' METHOD FOR MULTIPLE MODELS AND SCENARIOS -- 4.1 HIERARCHICAL FRAMEWORK FOR UNCERTAINTY QUANTIFICATION -- 4.2 GLOBAL SENSITIVITY INDICES FOR SINGLE MODEL AND SINGLE SCENARIO -- 4.3 GLOBAL SENSITIVITY INDICES FOR MULTIPLE MODELS BUT SINGLE SCENARIO -- 4.4 GLOBAL SENSITIVITY INDICES FOR MULTIPLE MODELS AND MULTIPLE SCENARIOS -- 5. SYNTHETIC STUDY WITH MULTIPLE SCENARIOS AND MODELS -- 5.1 SYNTHETIC CASE OF GROUNDWATER REACTIVE TRANSPORT MODELING -- 5.2 UNCERTAIN SCENARIOS, MODELS, AND PARAMETERS -- 5.3 TOTAL SENSITIVITY INDEX FOR HEAD UNDER INDIVIDUAL MODELS AND SCENARIOS -- 5.4 TOTAL SENSITIVITY INDEX FOR HEAD UNDER MULTIPLE MODELS BUT INDIVIDUAL SCENARIOS -- 5.5 TOTAL SENSITIVITY INDEX FOR HEAD UNDER MULTIPLE MODELS AND MULTIPLE SCENARIOS -- 5.6 TOTAL SENSITIVITY INDEX FOR ETHENE CONCENTRATION -- 6. USING GLOBAL SENSITIVITY ANALYSIS FOR SATELLITE DATA AND MODELS -- 7. CONCLUSIONS AND PERSPECTIVES -- Acknowledgments -- REFERENCES -- 4 - OTHER SA METHODS: CASE STUDIES -- 11 - SENSITIVITY AND UNCERTAINTY ANALYSES FOR STOCHASTIC FLOOD HAZARD SIMULATION -- 1. INTRODUCTION -- 2. BASIC PRINCIPLES OF STOCHASTIC APPROACH TO FLOOD HAZARD -- 2.1 STOCHASTIC SIMULATION OF RESERVOIR INFLOWS -- 2.1.1 Storm Seasonality -- 2.1.2 Precipitation Magnitude-Frequency Relationship -- 2.1.3 Temporal and Spatial Distribution of Storms -- 2.1.4 Air Temperature and Freezing Level Temporal Patterns -- 2.1.5 The 1000-mb Air Temperature Simulation -- 2.1.6 Air Temperature Lapse Rates -- 2.1.7 Freezing Level -- 2.1.8 Watershed Model Antecedent Conditions Sampling -- 2.1.9 Initial Reservoir Level -- 2.2 SIMULATION OF RESERVOIR OPERATION-FLOOD ROUTING -- 2.3 SIMULATION PROCEDURE. , 3. UNCERTAINTY ASSOCIATED WITH STOCHASTICALLY DERIVED FLOOD QUANTILES.