Operational flood forecasting, warning and response for multi-scale flood risks in developing cities

Description

The aim of this book is to contribute to understanding risk knowledge and to forecasting components of early flood warning, particularly in the environment of tropical high mountains in developing cities. This research covers a challenge, taking into account the persistent lack of data, limited resources and often complex climatic, hydrologic and hydraulic conditions. In this research, a regional method is proposed for assessing flash flood susceptibility and for identifying debris flow predisposition at the watershed scale. An indication of hazard is obtained from the flash flood susceptibility analysis and continually, the vulnerability and an indication of flood risk at watershed scale was obtained. Based on risk analyses, the research follows the modelling steps for flood forecasting development. Input precipitation is addressed in the environment of complex topography commonly found in mountainous tropical areas. A distributed model, a semi-distributed model and a lumped model were all used to simulate the discharges of a tropical high mountain basin with a páramo upper basin. Performance analysis and diagnostics were carried out in order to identify the most appropriate model for the study area for flood early warning. Finally, the Weather Research and Forecasting (WRF) model was used to explore the added value of numerical weather models for flood early warning in a páramo area. Content: 1 Introduction 1.1 Background 1.2 Scope of the thesis 1.3 Outline of the thesis 2 Regional debris flow susceptibility analysis in mountainous peri-urban areas through morphometric and land cover indicators 2.1 Introduction 2.2 Methods and Data 2.2.1 Study Area 2.2.2 Methodology 2.2.2.1 Development of the morphometric indicator 2.2.2.2 Development of the land cover indicator 2.2.2.3 Development of a composite susceptibility index 2.3 Results 2.3.1 Estimation of the morphometric indicator for the study area 2.3.1.1 Morphometric indicator model 2.3.1.2 Assessment of appropriateness of the morphometric indicator 2.3.2 Land cover indicator 2.3.3 Combination of indicators to obtain a final susceptibility index 2.4 Discussion 2.4.1 Morphometric indicator 2.4.2 Debris flow propagation 2.4.3 Land cover indicator, composite susceptibility index and comparison of results 2.5 Conclusions 3 Regional prioritisation of flood risk in mountainous areas 3.1 Introduction 3.2 Conceptualization of Vulnerability 3.3 Methods and Data 3.3.1 Study Area 3.3.2 Methodology 3.3.2.1 Delineation of exposure areas 3.3.2.2 Choice of indicators and principal component analysis for vulnerability assessment 3.3.2.3 Sensitivity of the vulnerability indicator 3.3.2.4 Categories of recorded damage in the study area 3.3.2.5 Prioritization of watersheds 3.4 Results 3.4.1 Exposure Areas 3.4.2 Socio-economic fragility indicators 3.4.3 Lack of Resilience and coping capacity indicators 3.4.4 Physical exposure indicators 3.4.5 Vulnerability indicator 3.4.6 Prioritization of watersheds according to the qualitative risk indicator and comparison with damage records 3.4.7 Sensitivity analysis of the vulnerability indicator 3.5 Discussion 3.5.1 Exposure areas 3.5.2 Representativeness and relative importance of indicators 3.5.3 Sensitivity of the vulnerability indicator 3.5.4 Usefulness of the prioritization indicator 3.6 Conclusions 4 Spatial interpolation for real-time rainfall field estimation in areas with complex topography 4.1 Introduction 4.2 Methods and Data 4.2.1 Study Area 4.2.2 Precipitation data 4.2.3 Geostatistical interpolation procedure 4.2.3.1 Interpolation techniques 4.2.3.2 Topographic parameters as secondary variables 4.2.3.3 Cross validation and statistical criteria of comparison 4.2.3.4 Conditional Simulations 4.3 Results 4.3.1 Exploratory data analysis 4.3.2 Classification of daily datasets 4.3.3 Variogramanalysis 4.3.4 Analysis of performance of the interpolators for the individual storms 4.3.5 Analysis of performance of the interpolators using the climatological variograms and applicability of the climatological variograms for individual event rainfall field generation 4.3.6 Analysis of secondary variables 4.3.7 Analysis of uncertainty in estimates of storm volumes 4.4 Discussion 4.4.1 Characteristics of the rainfall fields 4.4.2 Performance of the climatological variograms and applicability to the generation of individual event rainfall fields 4.4.3 Choice between KED and OK 4.4.4 Volumetric comparison 4.5 Conclusions 5 Hydrological model assessment for flood early warning in a tropical high mountain basin 102 5.1 Introduction 5.2 Study Area 5.3 Methods 5.3.1 Modelling set up and calibration 5.3.1.1 Description of themodels 5.3.1.2 Hydrometeorological forcing 5.3.1.3 Model Configuration and Calibration 5.3.2 Performance analysis and diagnostics 5.3.3 Analysis of precipitation input uncertainty and comparison of models 5.4 Results 5.4.1 Model calibration 5.4.1.1 KGE for HECHMSSMA, TOPMODEL and TETIS 5.4.2 Comparison of water balance fluxes 5.4.3 Signature measures from the flow duration curve (FDC) 5.4.4 Rainfall ensemble analysis, input precipitation uncertainty 5.4.5 Comparison ofmodel ensembles 5.5 Discussion 5.5.1 Model calibration and performance 5.5.1.1 Water balance fluxes and hydrometeorological forcing 5.5.1.2 Pixel size and flux variation for the TOPMODEL and TETIS 5.5.1.3 HECHMSSMA calibration results and fluxes 5.5.1.4 Flow duration curve and signatures 5.5.2 Comparison of discharge ensembles 5.6 Conclusions 6 Streamflow forecasts fromWRF precipitation for flood early warning in tropical mountain areas 6.1 Introduction 6.2 Methods and data 6.2.1 Study Area 6.2.2 WRF model data and observed rainfall fields 6.2.3 Methodology 6.2.3.1 Generation of Precipitation Forecasts 6.2.3.2 Verification of forecasts 6.3 Results 6.3.1 Bias correction of precipitation forecasts through DBS 6.3.2 Quantile regressionmodel 6.3.3 Verification of precipitation forecasts 6.3.4 Verification of deterministic precipitation forecasts and ensemble mean 6.3.5 Verification of deterministic discharge forecasts and ensemble mean 6.3.6 Verification of probabilistic forecasts 6.3.7 Discussion 6.3.7.1 Evaluating precipitation forecasts from the WRF model 6.3.7.2 Evaluating discharge forecast 6.4 Conclusions ... 158 7 Conclusions and Recommendations 7.1 Conclusions 7.1.1 Regional Flood risk analysis 7.1.2 Hydrometeorological inputs 7.1.3 Hydrological models for flood early warning 7.2 Added value of the numerical weather prediction model WRF in the flood forecasting system 7.3 Recommendations