Claire-Marie Duluc

Spatial Global Sensitivity Analysis of High Resolution classified topographic data use in 2D urban flood modelling

This paper presents a spatial Global Sensitivity Analysis (GSA) approach in a 2D shallow water equations based High Resolution (HR) flood model. The aim of a spatial GSA is to produce sensitivity maps which are based on Sobol index estimations. Such an approach allows to rank the effects of uncertain HR topographic data input parameters on flood model output. The influence of the three following parameters has been studied: the measurement error, the level of details of above-ground elements representation and the spatial discretization resolution. To introduce uncertainty, a Probability Density Function and discrete spatial approach have been applied to generate 2,000 DEMs. Based on a 2D urban flood river event modelling, the produced sensitivity maps highlight the major influence of modeller choices compared to HR measurement errors when HR topographic data are used. The spatial variability of the ranking is enhnaced. Sobol index maps produced by a spatial GSA rank in space the weight of each uncertain parameter on the variability of output parameter of interest.The weight of the modeller choices, with respect to the influence of HR dataset accuracy is enhanced.Added value is for modeller to better understand limits of his model.Requirements and limits for this approach are related to subjectivity of choices and to computational cost.

Feedback from Uncertainties Propagation Research Projects Conducted in Different Hydraulic Fields: Outcomes for Engineering Projects and Nuclear Safety Assessment

Hydraulic river models are applied for various purposes such as safety against flooding, navigation, or ecological rehabilitation. Much effort has been put into the development of sophisticated numerical model systems. These numerical models are based on a deterministic approach and the results are presented in terms of measurable quantities (water depths, flow velocities, etc.). However, the modeling of river processes involves numerous uncertainties associated both to the numerical structure of the model, to the knowledge of the physical parameters which force the system, and to the randomness inherent to the natural phenomena. As a consequence, dealing with uncertainties can be a difficult task for both practitioners (Iooss in Journal de la Societe Francaise de Statistique 152(1):1–23, 2011, [1]) and new guidance (ASN in protection of basic nuclear installations against external flooding, 2013 [2]). In the context of nuclear safety, the Institute for Radioprotection and Nuclear Safety (IRSN) assesses studies conducted by the operators for different reference flood situations (local rain, small or large watershed flooding, sea levels, etc.), in agreement with the recommendation reported by the guide ASN n°. 13 (ASN in protection of basic nuclear installations against external flooding, 2013 [2]). The guide provides some recommendations to deal with uncertainties, by proposing a specific conservative approach to cover hydraulic modeling uncertainties. Especially, the most influencing parameter of the numerical model is identified and an unfavorable value is taken in order to cover a whole set of parameters. Depending on the situation, the influencing parameter might be the Strickler coefficients, levee behaviors, simplified topographic assumptions, etc. Obviously, identifying the most influencing parameter and giving it a penalizing value is challenging and usually questionable. In this context, IRSN conducted cooperative (Compagnie Nationale du Rhone, I-CiTy laboratory of Polytech’Nice, Atomic Energy Commission) research activities since 2011 in order to investigate feasibility and benefits of Uncertainties Analysis (UA) and Global Sensitivity Analysis (GSA) when applied to hydraulic modeling. A specific methodology, presented in Sect. 2, was tested by using the computational environment Promethee, which allows carrying out uncertainties propagation study. This methodology was applied with various numerical models and in different contexts (Sect. 3), as river flooding on the Rhone River (Nguyen et al. in La Houille Blanche 5:55–62, 2015 [3]) and on the Garonne River (in the context of the Garonne river test case), for the studying of local rainfall (Abily et al. Environ Model Softw 77:183–195, 2016 [4]) or for tsunami generation, in the framework of the ANR-research project TANDEM. The feedback issued from these previous studies is analyzed (technical problems, limitations, interesting results, etc.) in Sect. 4 and the perspectives and a discussion on how a probabilistic approach of uncertainties should improve the actual deterministic methodology for risk assessment (also for other engineering applications) is finally given.

Global Sensitivity Analysis with 2D Hydraulic Codes: Application on Uncertainties Related to High-Resolution Topographic Data

Technologies such as aerial photogrammetry allow production of 3D topographic data including complex environments such as urban areas. Therefore, it is possible to create High-Resolution (HR) Digital Elevation Models (DEM) incorporating thin above-ground elements influencing overland flow paths. Although this category of “big data” has a high level of accuracy, there are still errors in measurements and hypothesis under DEM elaboration. Moreover, operators look for optimizing spatial discretization resolution in order to improve flood model computation time. Errors in measurement, errors in DEM generation, and operator choices for inclusion of this data within 2D hydraulic model, might influence the results of flood model simulations. These errors and hypothesis may influence significantly the flood modeling results variability. The purpose of this study is to investigate uncertainties related to (i) the own error of high-resolution topographic data and (ii) the modeler choices when including topographic data in hydraulic codes. The aim is to perform a Global Sensitivity Analysis (GSA) which goes through a Monte-Carlo uncertainty propagation, to quantify impact of uncertainties, followed by “Sobol” indices computation, to rank influence of identified parameters on result variability. A process using a coupling of an environment for parametric computation (Promethee) and a code relying on 2D shallow water equations (FullSWOF_2D) has been developed (P-FS tool). The study has been performed over the lower part of the Var river valley using the estimated hydrograph of a 1994 flood event. HR topographic data has been made available for the study area, which is 17.5 km2, by Nice municipality. Three uncertain parameters were studied: the measurement error (var. E), the level of details of above-ground element representation in DEM (buildings, sidewalks, etc.) (var. S), and the spatial discretization resolution (grid cell size for regular mesh) (var. R). Parameter var. E follows a probability density function, whereas parameters var. S and var. R are discrete operator choices. Combining these parameters, a database of 2,000 simulations has been produced using P-FS tool implemented on a high-performance computing structure. In our study case, the output of interest is the maximal water surface reached during simulations. A stochastic sampling on the produced result database has allowed to perform a Monte-Carlo approach. Sensitivity index have been produced at given points of interest, enhancing the relative weight of each uncertain parameters on variability of calculated overland flow. Perspectives for Sobol index maps production are brought to light.

Use of Standard 2D Numerical Modeling Tools to Simulate Surface Runoff Over an Industrial Site: Feasibility and Comparative Performance Survey Over a Test Case

Intense pluvial generated surface flow over an industrial facility represents a flood risk requiring an appropriated approach for risk assessment. Runoff over industrial site might have flow regime changes, wild flooding/drying extend, as well as small water deep properties. This makes standard bidimensional (2D) numerical surface flow models use particularly challenging. Indeed, numerical treatment of these properties might not be specifically supported by models. Furthermore, it gets close by their traditional application domain limits. Accordingly, an assessment of this group of numerical tool use for such a purpose needs to be in detailed studied to evaluate feasibility, performance, and relevance of their use in this context. This chapter aims to focus on common 2D numerical modeling tools use for application over an industrial plant test case to simulate surface runoff scenarios. Feasibility of such an approach is hereby studied. Performances and relevance of this attempt are evaluated. Our test case has specificities of real industrial plants in terms of domain extend, topography, and surface drainage structures. Tested scenarios state a uniform net 100 mm 1 h long rainfall event in a context of storm water sewer pipe failure. Selected tested models were a 2D finite differences diffusive wave model and an array of different 2D shallow water equation [2D shallow water equations (SWEs)]–based models. Comparison has been conduced over computed maximal water depth and water deep evolution. Results reveal a feasibility of these tools application for the studied specific purpose. They underline the necessity of a highly fine spatial and temporal discretization. Tested categories of average 2D SWEs–based models show in a large extend similar results in water depth calculation. Used indicator of results reliability estimation did not point out major critical aspects in calculation. Limits inherent of these categories of models use for this domain of application are underlined. Relevance of this approach is raised up.

Uncertainties of a 1D Hydraulic Model with Levee Breaches: The Benchmark Garonne

In a fluvial environment, the main role of levees is to canalise water downstream of rivers and to reduce the risk of flooding in nearby areas. Levee failure can be either structural or hydraulic. Structural failure occurs where a breach in a flood defence system leads to the inundation of the protected area whereas hydraulic failure refers to flooding before the designed protection level is attained and without prior damage to the flood defence system. Nowadays, hydrodynamic modelling codes are able to perform hydraulic failure such as overflowing by means of any appropriate weir equation, however, only a few allow to simulate structural failure. HEC-RAS can do both and enables to model levee breaches with a simple but flexible parametric module. The aim of our study is to evaluate the capacity of a 1D hydraulic model to represent levees breaches and subsequent flooding. To do so, a 1D storage area model is built with HEC-RAS and calibrated using data provided by the ‘Benchmark Garonne’ project initiated by EDF. The study case is based on the 1981 historical flood event of the Garonne River between Tonneins and La Reole (Sect. 2). The model is introduced and compared to two other hydraulic models used in the benchmark (Sect. 3). Two sensitivity analyses with respect to sets of hydraulic parameters and levee breach parameters are carried out (Sect. 4). Results expressed as maximum water levels show that the main channel roughness coefficient and the final breach width are the most influencing model parameters, respectively. Levee breaches appear to be a non-negligible source of uncertainty in hydraulic modelling, comparable to uncertainties arising from model structure or model calibration. In order to improve our modelling approach, a ground survey and a literature survey is conducted to collect data about the breaches that occurred in the study area, in particular, during the 1981 flood (Sect. 5). Historical evidence shows that a significant number of breaches occurred since 1875.

Analysis of the risk associated to coastal flooding hazards: A new historical extreme storm surges dataset for Dunkirk, France

This paper aims to demonstrate the technical feasibility of a historical study devoted to French nuclear power plants (NPPs) which can be prone to extreme coastal flooding events. It has been shown in the literature that the use of historical information (HI) can significantly improve the probabilistic and statistical modeling of extreme events. There is a significant lack of historical data on coastal flooding (storms and storm surges) compared to river flooding events. To address this data scarcity and to improve the estimation of the risk associated with coastal flooding hazards, a dataset of historical storms and storm surges that hit the Nord-Pasde-Calais region during the past five centuries was created from archival sources, examined and used in a frequency analysis (FA) in order to assess its impact on frequency estimations. This work on the Dunkirk site (representative of the Gravelines NPP) is a continuation of previous work performed on the La Rochelle site in France. Indeed, the frequency model (FM) used in the present paper had some success in the field of coastal hazards and it has been applied in previous studies to surge datasets to prevent coastal flooding in the La Rochelle region in France. In a first step, only information collected from the literature (published reports, journal papers and PhD theses) is considered. Although this first historical dataset has extended the gauged record back in time to 1897, serious questions related to the exhaustiveness of the information and about the validity of the developed FM have remained unanswered. Additional qualitative and quantitative HI was extracted in a second step from many older archival sources. This work has led to the construction of storm and coastal flooding sheets summarizing key data on each identified event. The quality control and the cross-validation of the collected information, which have been carried out systematically, indicate that it is valid and complete in regard to extreme storms and storm surges. Most of the HI collected is in good agreement with other archival sources and documentary climate reconstructions. The probabilistic and statistical analysis of a dataset containing an exceptional observation considered as an outlier (i.e., the 1953 storm surge) is significantly improved when the additional HI collected in both literature and archives is used. As the historical data tend to be extreme, the right tail of the distribution has been reinforced and the 1953 “exceptional” event does not appear as an outlier any more. This new dataset provides a valuable source of information on storm surges for future characterization of coastal hazards.