HAL-Ecole des Ponts ParisTech
Not a member yet
42743 research outputs found
Sort by
Elasticity-based mesh morphing technique with application to reduced-order modeling
No full textInternational audienceThe aim of this article is to introduce a new methodology for constructing morphings between shapes that have identical topology. The morphings are obtained by deforming a reference shape, through the resolution of a sequence of linear elasticity equations, onto every target shape. In particular, our approach does not assume any knowledge of a boundary parametrization. Furthermore, we demonstrate how constraints can be imposed on specific points, lines and surfaces in the reference domain to ensure alignment with their counterparts in the target domain after morphing. Additionally, we show how the proposed methodology can be integrated in an offline and online paradigm, which is useful in reduced-order modeling involving variable shapes. This framework facilitates the efficient computation of the morphings in various geometric configurations, thus improving the versatility and applicability of the approach. The methodology is illustrated on the regression problem of the drag and lift coefficients of airfoils of non-parameterized variable shapes
Short and Simple Confidence Intervals When the Directions of Some Effects Are Known
No full textInternational audienceWe introduce adaptive confidence intervals on a parameter of interest in the presence of nuisance parameters, such as coefficients on control variables, with known signs. Our confidence intervals are trivial to compute and can provide significant length reductions relative to standard ones when the nuisance parameters are small. At the same time, they entail minimal length increases at any parameter values. We apply our confidence intervals to the linear regression model, prove their uniform validity, and illustrate their length properties in an empirical application to a factorial design field experiment and a Monte Carlo study calibrated to the empirical application
The subtleties of three-dimensional radiative effects in contrails and cirrus clouds
No full textInternational audienceAbstract. The radiative effect of cirrus, contrails, and contrail cirrus affects the energy budget of the Earth and climate change. Those clouds, and especially contrails, are heterogeneous and their holes and sides exert three-dimensional radiative effects. This study uses the htrdr Monte Carlo radiative transfer code to investigate the sensitivity of the cloud radiative effect (CRE) to the geometrical dimensions and optical depth of optically thin ice clouds (cloud optical depth <4), with particular emphasis on three-dimensional radiative effects. When the Sun is at zenith, an increase in cloud optical depth causes a linear increase in shortwave (SW) CRE but a saturation of longwave (LW) CRE, causing the net CRE to change sign from positive to negative. The optical depth at which this change in sign occurs depends on the cloud geometry. 3D effects make the one-dimensional SW and LW CREs more positive for the Sun at zenith, reaching the same order of magnitude as the 1D CRE itself for clouds with high aspect ratios. The angular dependence of ice crystal scattering strongly increases shortwave CRE when solar zenith angle increases. 3D effects change sign from positive at zenith to negative at large zenith angles as the Sun's rays interact more with the cloud sides. Integrating instantaneous CRE and 3D effects over selected days of the year indicates compensation of SW with LW 3D effects for some cloud orientations, but 3D effects remain important in some cases. These results suggest that the 3D structure of cirrus and contrails needs to be considered to finely quantify their CRE and radiative forcing
Migration and Local Innovation: Evidence from Fine-Grained Data from OECD Countries
No full textDocuments de travail du Centre d'Economie de la Sorbonne 2025.21 - ISSN : 1955-611X, eISSN : 2968-6687Does the presence of migrants influence innovation at the local level? This paper answers this question using novel data containing fine-grained information on the migrant population and geo-coded data on patent locations for a large set of 19 OECD countries over the 1990-2014 period. We find that a one percentage point increase in the local migrant share increases patent applications by 2.5%. This effect is driven by more urbanised and economically developed localities, where innovation levels are already higher to begin with. However, this impact becomes insignificant when aggregating observations at larger geographical levels, suggesting that the effect of migration on innovation is concentrated in space and features high rates of spatial decay
Une méthode multi-échelle basée sur les données pour la simulation de l'endommagement anisotrope des structures hétérogènes
No full textInternational audienceA Data-Driven multiscale method for modeling anisotropic damage in heterogeneous structures is proposed. This approach, based on harmonic analysis of damage combined with numerical homogenization, uses the results of off-line damage calculations on RVEs as data for constructing the evolution laws of the internal variables and the anisotropic damage model. At the macroscopic scale, on-line computations are carried out, using the evolution laws and the anisotropic damage model constructed, to simulate crack nucleation and propagation in composite structures, without specific assumptions on the anisotropy related to RVEs geometry. A modified regularization technique is applied to guarantee mesh independence. An example of damage modeling of a heterogeneous structure with periodic micro-structure is presented. The results obtained from the homogenized model are validated by comparison with high-fidelity simulation.Une méthode multi-échelle basée sur les données pour modéliser l’endommagement anisotrope dans les structures hétérogènes est proposée. Cette approche, reposant sur l’analyse harmonique de l’endommagement combinée à l’homogénéisation numérique, utilise des résultats de calculs « Off-line » de fissuration sur des VERs comme données pour construire les lois d’évolution des variables internes etle modèle d’endommagement anisotrope. A l’échelle macroscopique, des calculs « On-line » sont effectués, en utilisant les lois d’évolution et le modèle d’endommagement anisotrope construits, pour simuler la nucléation et la propagation de fissures dans les structures composites, sans hypothèses spécifiquessur l’anisotropie liée à la géométrie des VERs. Une technique de régularisation modifiée est appliquéepour garantir l’indépendance du maillage. Un exemple de modélisation de l’endommagement d’une structure hétérogène à microstructure périodique est présenté. Les résultats obtenus à partir du modèle homogénéisé sont validés par comparaison avec une simulations haute fidélité
The price of youth: Income and Age-Hypergamy
No full textThis paper focuses on the role of age gaps and income in the mating process. We test the predictions of different theories of age heterogamy using a French exhaustive administrative database of newly formed same-sex and heterosexual couples. In both heterosexual and man-man couple formation, the age gap between the richer and poorer partner, and the likelihood of an extreme age gap, increase with the income of the male richer partner, especially when he belongs to the top income decile. By contrast, women, even when they rank high in the income distribution, are usually matched with an older partner. Women are found to mate with a younger partner in only one specific case: above their 60s, when they provide at least 75% of the household income or rank higher than their partner in the distribution of income in their respective age group. The marriage market thus seems to be best described by exchange theory, mitigated by a remnant of traditional norms. In a way, the price of youth depends on the sex of the buyer
Fonctionnement hydrologique des jardins de pluie - de l’évaluation in-situ à la modélisation pour une diversité de contextes
No full textNature-based solutions, such as bioretention systems, are in growing use for resilient urban runoff management. A bioretention system includes a vegetated depression to collect and infiltrate runoff, a substrate to provide water filtration and retention/evapotranspiration, and optionally a transition, drainage or storage layer. Bioretention can control runoff volume, delay flow peaks and enhance evapotranspiration (ET), thereby contributing to a more natural local water balance. However, its hydrological performance is likely to vary, depending on system design and local context. The objectives of this research are: i) to elucidate the impacts of various bioretention design characteristics on their hydrological functioning, with a special focus on cases with low exfiltration potential and under an oceanic climate, typically corresponding to the conditions found around Paris; and ii) to assess the reliability and robustness of HYDRUS-1D model in representing different designs and input parameters.Through a thorough literature review, this research began with a comprehensive overview of bioretention designs, experimental setups and modelling approaches considered in previous studies. The analysis also highlighted the following knowledge gaps: 1) underrepresentation of some local contexts and designs; 2) limited studies on long-term water balance monitoring; and 3) insufficient investigation on subsoil and shallow groundwater interactions.The PhD relied on continuous monitoring of three bioretention prototypes in Paris region and on field investigations to explore the efficiency of runoff volume reduction under unfavourable subsoil conditions (e.g., limited or forbidden exfiltration), the relative importance of different hydrologic processes and ways to enhance them. The field conditions include: 1) An unlined system with a relatively high hydraulic loading ratio (HLR) of 13 over clay soil, examining how IWS may enhance exfiltration; 2) Two lined systems, assessing whether a combination of low HLR (i.e.4), fine-textured substrate and IWS can enhance ET and act as a primary mechanism for volume reduction.Overall, the three systems achieved significant runoff volume reduction over the monitoring period (43% for JdB1, 48% for JdB2, 63% for SC), despite the lack of exfiltration or low permeability of the subsoil. For the unlined system, the volume reduction performance was largely controlled by the thickness of the IWS (extending IWS from 2 cm to 22 cm increased volume reduction from 27% to 55%). For the lined systems, it is controlled by ET, which was more than doubled in presence of an IWS. In the unlined system, a capillary barrier between the transition and drainage layers promoted soil water retention, yet ET remained lower than potential evapotranspiration (PET). Silt loam substrate in the lined systems supported vegetation but led to clogging and cracking issues which were mitigated by the IWS. A risk of water intrusion from perched lens was also identified when the bottom of the bioretention is connected to a low permeability subsoil.The final part of this research involved an evaluation of HYDRUS-1D for bioretention modelling. A sensitivity analysis assessed how varying levels of input knowledge impact the model robustness. The results on water balance, especially drainage volume, proved to be robust and accurate and were not significantly affected by uncertainties in input parameters. However, the soil moisture profile results were highly dependent on the bottom boundary condition and, to a lesser extent, on soil hydrodynamic properties. Neither of the two tested boundary conditions allowed for a good description of soil moisture, indicating that the model could not adequately capture the hydraulic behaviour of the system. Besides, for a HLR of 13, vegetation properties had a very limited impact on the water balanceLes solutions fondées sur la nature, telles que les systèmes de bio rétention, sont de plus en plus utilisées pour une gestion résiliente du ruissellement urbain. Un système de bio rétention est constitué d’une dépression végétalisée destinée à collecter et infiltrer les eaux de ruissellement, d’un substrat assurant la filtration et la rétention de l’eau ainsi que l’évapotranspiration et éventuellement d’une couche de transition, de drainage ou de stockage. La bio rétention permet de gérer le volume de ruissellement, de retarder les pics de débit et de favoriser l’évapotranspiration (ET), contribuant ainsi à un équilibre hydrique local plus naturel. Toutefois, ses performances hydrologiques peuvent varier selon le design du système et le contexte local. Les objectifs de cette recherche sont : i) d’élucider les effets de différentes caractéristiques de conception des systèmes de bio rétention sur leur fonctionnement hydrologique, avec un accent particulier sur les cas à faible potentiel d’exfiltration dans un climat océanique, typique des conditions observées en région parisienne ; et ii) d’évaluer la fiabilité et la robustesse du modèle HYDRUS-1D dans la représentation de conceptions variées et de paramètres d’entrée hétérogènes. Ce travail a débuté par une revue exhaustive de la littérature, fournissant une synthèse des types de conceptions, des dispositifs expérimentaux et des approches de modélisation considérées dans les études précédentes. L’analyse a également mis en évidence plusieurs lacunes dans les connaissances actuelles : 1) la sous-représentation de certains contextes locaux et designs 2) le nombre limité d’études portant sur la surveillance à long terme du bilan hydrique et 3) le manque d’investigations concernant les interactions entre le sol sous-jacent et les nappes phréatiques peu profondes. Cette thèse repose sur une surveillance continue de 3 prototypes de bio rétention en région parisienne et sur des investigations de terrain visant à évaluer l’efficacité de la réduction du volume de ruissellement dans des conditions pédologiques défavorables (par exemple, exfiltration limitée ou interdite), ainsi que l’importance relative des processus hydrologiques et les leviers d’optimisation. Les configurations de terrain incluent 1) un système non étanchéifié avec un ratio de charge hydraulique (HLR) élevé de 13, sur un sol argileux, analysant comment la zone de saturation intermittente (IWS) peut favoriser l’exfiltration 2) deux systèmes étanchéifiés, évaluant si la combinaison d’un faible HLR (i.e. 4), d’un substrat à texture fine et d’une IWS peut renforcer l’ET et constituer le principal mécanisme de réduction de volume. Dans l’ensemble, les trois systèmes ont permis une réduction significative du volume de ruissellement sur la période de surveillance (43 % pour JdB1, 48 % pour JdB2, 63 % pour SC), malgré l’absence d’exfiltration ou la faible perméabilité du sol. Pour le système non étanchéifié, la performance de réduction de volume est principalement déterminée par l’épaisseur de la zone saturée intermittente (IWS) : son augmentation de 2 cm à 22 cm a entraîné une amélioration de la réduction de 27 % à 55 %. Dans les systèmes étanchéifiés, la réduction est principalement liée à l’ET, qui a plus que doublé en présence d’une IWS. Dans le système non étanchéifié, une barrière capillaire entre la couche de transition et celle de drainage a favorisé la rétention d’eau dans le sol, bien que l’ET soit restée inférieure à l’évapotranspiration potentielle (ETP). Le substrat limono-argileux des systèmes étanchéifiés a soutenu la végétation, mais a également provoqué des problèmes de colmatage et de fissuration, atténués par la présence de l’IWS
Les mobilités liées au e-commerce
No full textIn response to the rapid growth of online commerce, significantly accelerated by the COVID-19 pandemic, the MOBS (Mobilities of E-Shopping) project aims to analyse how digital purchasing practices influence travel behaviour and household mobility in France. While the substitution hypothesis—that e-commerce replaces physical travel—is a central assumption in digital impact studies, the research reveals more complex interactions between online shopping and mobility practices, ranging from substitution to complementarity.The report is based on a nationwide survey conducted between 2023 and 2024, involving over 5,000 households. It pursues a dual objective: first, to understand the relationships between online shopping behaviours and associated mobility (frequency, transport modes, distances, etc.); and second, to assess the environmental footprint of these new forms of mobility created by e-commerce and decoupled households purchases.The survey covers a broad spectrum of online purchases (B2C and C2C; home delivery, collection from pick-up points, lockers, other intermediate points, or shops) and is designed to be representative of the metropolitan French population in both geographic and socio-demographic terms.Face à la croissance rapide du commerce en ligne, notamment accélérée par la pandémie de COVID-19, le projet MOBS (Mobilities of e-shopping) ambitionne d’analyser comment les pratiques d’achats numériques influencent les déplacements et la mobilité des ménages français. Si l’hypothèse de substitution selon laquelle le e-commerce remplacerait les déplacements physiques est une des hypothèses centrales faites sur les impacts du numérique, la recherche met en lumière des interactions entre achats en ligne et pratiques de mobilité plus complexes, pouvant aller de la substitution à la complémentaire.Le rapport se fonde sur une enquête nationale menée entre 2023 et 2024 auprès de plus de 5000 ménages dont l’objectif est double. Il s’agit d’abord de comprendre les liens entre comportements d’achat en ligne et mobilités associées (fréquences, modes de transport, distances, etc.). Enfin, le second objectif consiste à évaluer l’empreinte environnementale de des nouvelles mobilités générées par l’usage du e-commerce et des achats de plus en plus découplés.L’enquête couvre un ensemble vaste du spectre des achats en ligne (B2C et C2C ; livraisons à domicile, retrait en point relais, consignes et autres points intermédiaires ou magasins) et se veut représentative de la population française métropolitaine sur les plans géographique et socio-démographique
A tree-based Polynomial Chaos expansion for surrogate modeling and sensitivity analysis of complex numerical models
No full textThis paper introduces Tree-based Polynomial Chaos Expansion (Tree-PCE), a novel surrogate modeling technique designed to efficiently approximate complex numerical models exhibiting nonlinearities and discontinuities. Tree-PCE combines the expressive power of Polynomial Chaos Expansion (PCE) with an adaptive partitioning strategy inspired by regression trees. By recursively dividing the input space into hyperrectangular subdomains and fitting localized PCEs, Tree-PCE constructs a piecewise polynomial surrogate that improves both accuracy and computational efficiency. The method is particularly well-suited for global sensitivity analysis, enabling direct computation of Sobol' indices from local expansion coefficients and introducing a new class of sensitivity indices derived from the tree structure itself. Numerical experiments on synthetic and real-world models, including a 2D morphodynamic case, demonstrate that Tree-PCE offers a favorable balance between accuracy and complexity, especially in the presence of discontinuities. While its performance depends on the compromise between the number of subdomains and the degree of local polynomials, this trade-off can be explored using automated hyperparameter optimization frameworks. This opens promising perspectives for systematically identifying optimal configurations and enhancing the robustness of surrogate modeling in complex systems