Analyse de changements multiples : une approche probabiliste utilisant les réseaux bayésiens

La maintenance du logiciel est une phase très importante du cycle de vie de celui-ci. Après les phases de développement et de déploiement, c’est celle qui dure le plus longtemps et qui accapare la majorité des coûts de l'industrie. Ces coûts sont dus en grande partie à la difficulté d’effectuer des changements dans le logiciel ainsi que de contenir les effets de ces changements. Dans cette perspective, de nombreux travaux ont ciblé l’analyse/prédiction de l’impact des changements sur les logiciels. Les approches existantes nécessitent de nombreuses informations en entrée qui sont difficiles à obtenir. Dans ce mémoire, nous utilisons une approche probabiliste. Des classificateurs bayésiens sont entraînés avec des données historiques sur les changements. Ils considèrent les relations entre les éléments (entrées) et les dépendances entre changements historiques (sorties). Plus spécifiquement, un changement complexe est divisé en des changements élémentaires. Pour chaque type de changement élémentaire, nous créons un classificateur bayésien. Pour prédire l’impact d’un changement complexe décomposé en changements élémentaires, les décisions individuelles des classificateurs sont combinées selon diverses stratégies. Notre hypothèse de travail est que notre approche peut être utilisée selon deux scénarios. Dans le premier scénario, les données d’apprentissage sont extraites des anciennes versions du logiciel sur lequel nous voulons analyser l’impact de changements. Dans le second scénario, les données d’apprentissage proviennent d’autres logiciels. Ce second scénario est intéressant, car il permet d’appliquer notre approche à des logiciels qui ne disposent pas d’historiques de changements. Nous avons réussi à prédire correctement les impacts des changements élémentaires. Les résultats ont montré que l’utilisation des classificateurs conceptuels donne les meilleurs résultats. Pour ce qui est de la prédiction des changements complexes, les méthodes de combinaison "Voting" et OR sont préférables pour prédire l’impact quand le nombre de changements à analyser est grand. En revanche, quand ce nombre est limité, l’utilisation de la méthode Noisy-Or ou de sa version modifiée est recommandée.

Global hydrology modelling and uncertainty: running multiple ensembles with a campus grid.

Uncertainties associated with the representation of various physical processes in global climate models (GCMs) mean that, when projections from GCMs are used in climate change impact studies, the uncertainty propagates through to the impact estimates. A complete treatment of this ‘climate model structural uncertainty’ is necessary so that decision-makers are presented with an uncertainty range around the impact estimates. This uncertainty is often underexplored owing to the human and computer processing time required to perform the numerous simulations. Here, we present a 189-member ensemble of global river runoff and water resource stress simulations that adequately address this uncertainty. Following several adaptations and modifications, the ensemble creation time has been reduced from 750 h on a typical single-processor personal computer to 9 h of high-throughput computing on the University of Reading Campus Grid. Here, we outline the changes that had to be made to the hydrological impacts model and to the Campus Grid, and present the main results. We show that, although there is considerable uncertainty in both the magnitude and the sign of regional runoff changes across different GCMs with climate change, there is much less uncertainty in runoff changes for regions that experience large runoff increases (e.g. the high northern latitudes and Central Asia) and large runoff decreases (e.g. the Mediterranean). Furthermore, there is consensus that the percentage of the global population at risk to water resource stress will increase with climate change.

Statistical downscaling of river flows

An extensive statistical ‘downscaling’ study is done to relate large-scale climate information from a general circulation model (GCM) to local-scale river flows in SW France for 51 gauging stations ranging from nival (snow-dominated) to pluvial (rainfall-dominated) river-systems. This study helps to select the appropriate statistical method at a given spatial and temporal scale to downscale hydrology for future climate change impact assessment of hydrological resources. The four proposed statistical downscaling models use large-scale predictors (derived from climate model outputs or reanalysis data) that characterize precipitation and evaporation processes in the hydrological cycle to estimate summary flow statistics. The four statistical models used are generalized linear (GLM) and additive (GAM) models, aggregated boosted trees (ABT) and multi-layer perceptron neural networks (ANN). These four models were each applied at two different spatial scales, namely at that of a single flow-gauging station (local downscaling) and that of a group of flow-gauging stations having the same hydrological behaviour (regional downscaling). For each statistical model and each spatial resolution, three temporal resolutions were considered, namely the daily mean flows, the summary statistics of fortnightly flows and a daily ‘integrated approach’. The results show that flow sensitivity to atmospheric factors is significantly different between nival and pluvial hydrological systems which are mainly influenced, respectively, by shortwave solar radiations and atmospheric temperature. The non-linear models (i.e. GAM, ABT and ANN) performed better than the linear GLM when simulating fortnightly flow percentiles. The aggregated boosted trees method showed higher and less variable R2 values to downscale the hydrological variability in both nival and pluvial regimes. Based on GCM cnrm-cm3 and scenarios A2 and A1B, future relative changes of fortnightly median flows were projected based on the regional downscaling approach. The results suggest a global decrease of flow in both pluvial and nival regimes, especially in spring, summer and autumn, whatever the considered scenario. The discussion considers the performance of each statistical method for downscaling flow at different spatial and temporal scales as well as the relationship between atmospheric processes and flow variability.

Multimodel estimate of the global terrestrial water balance: setup and first results

Six land surface models and five global hydrological models participate in a model intercomparison project (WaterMIP), which for the first time compares simulation results of these different classes of models in a consistent way. In this paper the simulation setup is described and aspects of the multi-model global terrestrial water balance are presented. All models were run at 0.5 degree spatial resolution for the global land areas for a 15-year period (1985-1999) using a newly-developed global meteorological dataset. Simulated global terrestrial evapotranspiration, excluding Greenland and Antarctica, ranges from 415 to 586 mm year-1 (60,000 to 85,000 km3 year-1) and simulated runoff ranges from 290 to 457 mm year-1 (42,000 to 66,000 km3 year-1). Both the mean and median runoff fractions for the land surface models are lower than those of the global hydrological models, although the range is wider. Significant simulation differences between land surface and global hydrological models are found to be caused by the snow scheme employed. The physically-based energy balance approach used by land surface models generally results in lower snow water equivalent values than the conceptual degree-day approach used by global hydrological models. Some differences in simulated runoff and evapotranspiration are explained by model parameterizations, although the processes included and parameterizations used are not distinct to either land surface models or global hydrological models. The results show that differences between model are major sources of uncertainty. Climate change impact studies thus need to use not only multiple climate models, but also some other measure of uncertainty, (e.g. multiple impact models).

Climate impacts on river flow: projections for the Medway catchment UK with UKCP09 and CATCHMOD

The potential impact of climate change on areas of strategic importance for water resources remains a concern. Here, river flow projections for the River Medway, above Teston in southeast England are presented, which is just such an area of strategic importance. The river flow projections use climate inputs from the Hadley Centre Regional Climate Model (HadRM3) for the time period 1960–2080 (a subset of the early release UKCP09 projections). River flow predictions are calculated using CATCHMOD, the main river flow prediction tool of the Environment Agency (EA) of England and Wales. In order to use this tool in the best way for climate change predictions, model setup and performance are analysed using sensitivity and uncertainty analysis. The model's representation of hydrological processes is discussed and the direct percolation and first linear storage constant parameters are found to strongly affect model results in a complex way, with the former more important for low flows and the latter for high flows. The uncertainty in predictions resulting from the hydrological model parameters is demonstrated and the projections of river flow under future climate are analysed. A clear climate change impact signal is evident in the results with a persistent lowering of mean daily river flows for all months and for all projection time slices. Results indicate that a projection of lower flows under future climate is valid even taking into account the uncertainties considered in this modelling chain exercise. The model parameter uncertainty becomes more significant under future climate as the river flows become lower. This has significant implications for those making policy decisions based on such modelling results. Copyright © 2010 John Wiley & Sons, Ltd.

Sensitivity of terrestrial precipitation trends to the structural evolution of sea surface temperatures

Pronounced intermodel differences in the projected response of land surface precipitation (LSP) to future anthropogenic forcing remain in the Coupled Model Intercomparison Project Phase 5 model integrations. A large fraction of the intermodel spread in projected LSP trends is demonstrated here to be associated with systematic differences in simulated sea surface temperature (SST) trends, especially the representation of changes in (i) the interhemispheric SST gradient and (ii) the tropical Pacific SSTs. By contrast, intermodel differences in global mean SST, representative of differing global climate sensitivities, exert limited systematic influence on LSP patterns. These results highlight the importance to regional terrestrial precipitation changes of properly simulating the spatial distribution of large-scale, remote changes as reflected in the SST response to increasing greenhouse gases. Moreover, they provide guidance regarding which region-specific precipitation projections may be potentially better constrained for use in climate change impact assessments.

Análise de impacto baseada em rastreabilidade de artefatos para linhas de produto de software

Software Product Line (SPL) consists of a software development paradigm, whose main focus is to identify features common and variability among applications in a specific domain. An LPS is designed to attend all products requirements from its product family. These requirements and LPS may have changes over time due to several factors, such as evolution of product requirements, evolution of the market, evolution of SLP process, evolution of the technologies used to develop the products. To handle these changes, LPS should be modified and evolve in order to not become obsolete, and adapt itself to new requirements. The Changes Impact Analysis is an activity that understand and identify what consequences these changes are cause on LPS. Impact Analysis on LPS may be supported by traceability relationships, which identify relationships between artefacts created during all phases of software development. Despite the solutions of change impact analysis based on traceability for software, there is a lack of solutions for assessing the change impact analysis based on traceability for LPS, since existing solutions do not include estimates specific to the artefacts of LPS. Thus, this paper proposes a process of change impact analysis and an tool for assessing the change impact through traceability of artefacts in LPS. For this purpose, we specified a process of change impact analysis that considers artifacts produced during the development of LPS. We have also implemented a tool which allows estimating and identifying artefacts and products of LPS affected from changes in other products, changes in class, changes in features, changes between releases of LPS and artefacts related to changes in core assets and variability. Finally, the results were evaluated through metrics

Avaliando a robustez e manutenibilidade do comportamento excepcional de aplicações C#

Mainstream programming languages provide built-in exception handling mechanisms to support robust and maintainable implementation of exception handling in software systems. Most of these modern languages, such as C#, Ruby, Python and many others, are often claimed to have more appropriated exception handling mechanisms. They reduce programming constraints on exception handling to favor agile changes in the source code. These languages provide what we call maintenance-driven exception handling mechanisms. It is expected that the adoption of these mechanisms improve software maintainability without hindering software robustness. However, there is still little empirical knowledge about the impact that adopting these mechanisms have on software robustness. This work addresses this gap by conducting an empirical study aimed at understanding the relationship between changes in C# programs and their robustness. In particular, we evaluated how changes in the normal and exceptional code were related to exception handling faults. We applied a change impact analysis and a control flow analysis in 100 versions of 16 C# programs. The results showed that: (i) most of the problems hindering software robustness in those programs are caused by changes in the normal code, (ii) many potential faults were introduced even when improving exception handling in C# code, and (iii) faults are often facilitated by the maintenance-driven flexibility of the exception handling mechanism. Moreover, we present a series of change scenarios that decrease the program robustness

Predicting dependencies using domain-based coupling

Software dependencies play a vital role in programme comprehension, change impact analysis and other software maintenance activities. Traditionally, these activities are supported by source code analysis; however, the source code is sometimes inaccessible or difficult to analyse, as in hybrid systems composed of source code in multiple languages using various paradigms (e.g. object-oriented programming and relational databases). Moreover, not all stakeholders have adequate knowledge to perform such analyses. For example, non-technical domain experts and consultants raise most maintenance requests; however, they cannot predict the cost and impact of the requested changes without the support of the developers. We propose a novel approach to predicting software dependencies by exploiting the coupling present in domain-level information. Our approach is independent of the software implementation; hence, it can be used to approximate architectural dependencies without access to the source code or the database. As such, it can be applied to hybrid systems with heterogeneous source code or legacy systems with missing source code. In addition, this approach is based solely on information visible and understandable to domain users; therefore, it can be efficiently used by domain experts without the support of software developers. We evaluate our approach with a case study on a large-scale enterprise system, in which we demonstrate how up to 65 of the source code dependencies and 77% of the database dependencies are predicted solely based on domain information.

Gridded climate data from 5 Global Climate Models (GCM) of the Last Glacial Maximum (LGM) downscaled to 30 arc seconds for Europe, with links to NetCDF files

Studies on the impact of historical, current and future global change require very high-resolution climate data (less or equal 1km) as a basis for modelled responses, meaning that data from digital climate models generally require substantial rescaling. Another shortcoming of available datasets on past climate is that the effects of sea level rise and fall are not considered. Without such information, the study of glacial refugia or early Holocene plant and animal migration are incomplete if not impossible. Sea level at the last glacial maximum (LGM) was approximately 125m lower, creating substantial additional terrestrial area for which no current baseline data exist. Here, we introduce the development of a novel, gridded climate dataset for LGM that is both very high resolution (1km) and extends to the LGM sea and land mask. We developed two methods to extend current terrestrial precipitation and temperature data to areas between the current and LGM coastlines. The absolute interpolation error is less than 1°C and 0.5 °C for 98.9% and 87.8% of all pixels for the first two 1 arc degree distance zones. We use the change factor method with these newly assembled baseline data to downscale five global circulation models of LGM climate to a resolution of 1km for Europe. As additional variables we calculate 19 'bioclimatic' variables, which are often used in climate change impact studies on biological diversity. The new LGM climate maps are well suited for analysing refugia and migration during Holocene warming following the LGM.

Agile Construction and Evolution of Product-Line Architectures

Software Product Line Engineering (SPLE) has proved to have significant advantages in family-based software development, but also implies the up¬front design of a product-line architecture (PLA) from which individual product applications can be engineered. The big upfront design associated with PLAs is in conflict with the current need of "being open to change". However, the turbulence of the current business climate makes change inevitable in order to stay competitive, and requires PLAs to be open to change even late in the development. The trend of "being open to change" is manifested in the Agile Software Development (ASD) paradigm, but it is spreading to the domain of SPLE. To reduce the big upfront design of PLAs as currently practiced in SPLE, new paradigms are being created, one being Agile Product Line Engineering (APLE). APLE aims to make the development of product-lines more flexible and adaptable to changes as promoted in ASD. To put APLE into practice it is necessary to make mechanisms available to assist and guide the agile construction and evolution of PLAs while complying with the "be open to change" agile principle. This thesis defines a process for "the agile construction and evolution of product-line architectures", which we refer to as Agile Product-Line Archi-tecting (APLA). The APLA process provides agile architects with a set of models for describing, documenting and tracing PLAs, as well as an algorithm to analyze change impact. Both the models and the change impact analysis offer the following capabilities: Flexibility & adaptability at the time of defining software architectures, enabling change during the incremental and iterative design of PLAs (anticipated or planned changes) and their evolution (unanticipated or unforeseen changes). Assistance in checking architectural integrity through change impact analysis in terms of architectural concerns, such as dependencies on earlier design decisions, rationale, constraints, and risks, etc.Guidance in the change decision-making process through change im¬pact analysis in terms of architectural components and connections. Therefore, APLA provides the mechanisms required to construct and evolve PLAs that can easily be refined iteration after iteration during the APLE development process. These mechanisms are provided in a modeling frame¬work called FPLA. The contributions of this thesis have been validated through the conduction of a project regarding a metering management system in electrical power networks. This case study took place in an i-smart software factory and was in collaboration with the Technical University of Madrid and Indra Software Labs. La Ingeniería de Líneas de Producto Software (Software Product Line Engi¬neering, SPLE) ha demostrado tener ventajas significativas en el desarrollo de software basado en familias de productos. SPLE es un paradigma que se basa en la reutilización sistemática de un conjunto de características comunes que comparten los productos de un mismo dominio o familia, y la personalización masiva a través de una variabilidad bien definida que diferencia unos productos de otros. Este tipo de desarrollo requiere el diseño inicial de una arquitectura de línea de productos (Product-Line Architecture, PLA) a partir de la cual los productos individuales de la familia son diseñados e implementados. La inversión inicial que hay que realizar en el diseño de PLAs entra en conflicto con la necesidad actual de estar continuamente "abierto al cam¬bio", siendo este cambio cada vez más frecuente y radical en la industria software. Para ser competitivos es inevitable adaptarse al cambio, incluso en las últimas etapas del desarrollo de productos software. Esta tendencia se manifiesta de forma especial en el paradigma de Desarrollo Ágil de Software (Agile Software Development, ASD) y se está extendiendo también al ámbito de SPLE. Con el objetivo de reducir la inversión inicial en el diseño de PLAs en la manera en que se plantea en SPLE, en los último años han surgido nuevos enfoques como la Ingeniera de Líneas de Producto Software Ágiles (Agile Product Line Engineering, APLE). APLE propone el desarrollo de líneas de producto de forma más flexible y adaptable a los cambios, iterativa e incremental. Para ello, es necesario disponer de mecanismos que ayuden y guíen a los arquitectos de líneas de producto en el diseño y evolución ágil de PLAs, mientras se cumple con el principio ágil de estar abierto al cambio. Esta tesis define un proceso para la "construcción y evolución ágil de las arquitecturas de lineas de producto software". A este proceso se le ha denominado Agile Product-Line Architecting (APLA). El proceso APLA proporciona a los arquitectos software un conjunto de modelos para de¬scribir, documentar y trazar PLAs, así como un algoritmo para analizar vel impacto del cambio. Los modelos y el análisis del impacto del cambio ofrecen: Flexibilidad y adaptabilidad a la hora de definir las arquitecturas software, facilitando el cambio durante el diseño incremental e iterativo de PLAs (cambios esperados o previstos) y su evolución (cambios no previstos). Asistencia en la verificación de la integridad arquitectónica mediante el análisis de impacto de los cambios en términos de dependencias entre decisiones de diseño, justificación de las decisiones de diseño, limitaciones, riesgos, etc. Orientación en la toma de decisiones derivadas del cambio mediante el análisis de impacto de los cambios en términos de componentes y conexiones. De esta manera, APLA se presenta como una solución para la construcción y evolución de PLAs de forma que puedan ser fácilmente refinadas iteración tras iteración de un ciclo de vida de líneas de producto ágiles. Dicha solución se ha implementado en una herramienta llamada FPLA (Flexible Product-Line Architecture) y ha sido validada mediante su aplicación en un proyecto de desarrollo de un sistema de gestión de medición en redes de energía eléctrica. Dicho proyecto ha sido desarrollado en una fábrica de software global en colaboración con la Universidad Politécnica de Madrid e Indra Software Labs.

Efectos del cambio climático y usos del suelo sobre los recursos hídricos de la cuenca del río Tordera (Barcelona, España)

Se presentan los efectos del cambio global en la cuenca del río Tordera (España) para el periodo 2000-2050, escenarios climáticos A2 (medio-alto) definidos por el Panel Intergubernamental del Cambio Climático (IPCC, 200) y escenarios socioeconómicos (cambios previstos en la cuenca) denominados estable y tendencial. Los efectos sobre los recursos hídricos se han analizado de forma conjunta superficial-subterránea mediante una metodológica de tipo acoplado. Para establecer los impactos futuros sobre los recursos hídricos se ha seleccionado el Modelo de Circulación Global ECHAM5 (Max Planck Institute). Los resultados obtenidos indican una disminución de la precipitación del 11.3% y un aumento de la temperatura de 1ºC, respecto a los valores históricos de la zona. De acuerdo a la proyección futura (2050) sobre cambios en los recursos hídricos, la escorrentía superficial obtenida mediante simulación con el código HEC-HMS 3.4 experimenta una reducción del 31.8% respecto al valor histórico y la recarga natural, estimada mediante VISUAL-Balan, se reduce en un 11.7%. El balance en el acuífero deltaico simulado mediante MODFLOW 2009.1 Pro muestra igualmente una disminución de los parámetros del balance. Los cambios del uso del suelo previstos de acuerdo a la legislación vigente (escenarios socioeconómicos) no conducen a la generación de un impacto apreciable en los recursos hídricos; según los escenarios definidos la variación de precipitación y temperatura son los parámetros fundamentales del cambio previsto.

Modélisation de l'évolution hydroclimatique des flux et stocks d'eau verte et d'eau bleue du bassin versant de la Garonne

La gestion intégrée de la ressource en eau implique de distinguer les parcours de l’eau qui sont accessibles aux sociétés de ceux qui ne le sont pas. Les cheminements de l’eau sont nombreux et fortement variables d’un lieu à l’autre. Il est possible de simplifier cette question en s’attardant plutôt aux deux destinations de l’eau. L’eau bleue forme les réserves et les flux dans l’hydrosystème : cours d’eau, nappes et écoulements souterrains. L’eau verte est le flux invisible de vapeur d’eau qui rejoint l’atmosphère. Elle inclut l’eau consommée par les plantes et l’eau dans les sols. Or, un grand nombre d’études ne portent que sur un seul type d’eau bleue, en ne s’intéressant généralement qu’au devenir des débits ou, plus rarement, à la recharge des nappes. Le portrait global est alors manquant. Dans un même temps, les changements climatiques viennent impacter ce cheminement de l’eau en faisant varier de manière distincte les différents composants de cycle hydrologique. L’étude réalisée ici utilise l’outil de modélisation SWAT afin de réaliser le suivi de toutes les composantes du cycle hydrologique et de quantifier l’impact des changements climatiques sur l’hydrosystème du bassin versant de la Garonne. Une première partie du travail a permis d’affiner la mise en place du modèle pour répondre au mieux à la problématique posée. Un soin particulier a été apporté à l’utilisation de données météorologiques sur grille (SAFRAN) ainsi qu’à la prise en compte de la neige sur les reliefs. Le calage des paramètres du modèle a été testé dans un contexte differential split sampling, en calant puis validant sur des années contrastées en terme climatique afin d’appréhender la robustesse de la simulation dans un contexte de changements climatiques. Cette étape a permis une amélioration substantielle des performances sur la période de calage (2000-2010) ainsi que la mise en évidence de la stabilité du modèle face aux changements climatiques. Par suite, des simulations sur une période d’un siècle (1960-2050) ont été produites puis analysées en deux phases : i) La période passée (1960-2000), basée sur les observations climatiques, a servi de période de validation à long terme du modèle sur la simulation des débits, avec de très bonnes performances. L’analyse des différents composants hydrologiques met en évidence un impact fort sur les flux et stocks d’eau verte, avec une diminution de la teneur en eau des sols et une augmentation importante de l’évapotranspiration. Les composantes de l’eau bleue sont principalement perturbées au niveau du stock de neige et des débits qui présentent tous les deux une baisse substantielle. ii) Des projections hydrologiques ont été réalisées (2010-2050) en sélectionnant une gamme de scénarios et de modèles climatiques issus d’une mise à l’échelle dynamique. L’analyse de simulation vient en bonne part confirmer les conclusions tirées de la période passée : un impact important sur l’eau verte, avec toujours une baisse de la teneur en eau des sols et une augmentation de l’évapotranspiration potentielle. Les simulations montrent que la teneur en eau des sols pendant la période estivale est telle qu’elle en vient à réduire les flux d’évapotranspiration réelle, mettant en évidence le possible déficit futur des stocks d’eau verte. En outre, si l’analyse des composantes de l’eau bleue montre toujours une diminution significative du stock de neige, les débits semblent cette fois en hausse pendant l’automne et l’hiver. Ces résultats sont un signe de l’«accélération» des composantes d’eau bleue de surface, probablement en relation avec l’augmentation des évènements extrêmes de précipitation. Ce travail a permis de réaliser une analyse des variations de la plupart des composantes du cycle hydrologique à l’échelle d’un bassin versant, confirmant l’importance de prendre en compte toutes ces composantes pour évaluer l’impact des changements climatiques et plus largement des changements environnementaux sur la ressource en eau.

Analyse de sensibilité d’un multimodèle hydrologique

Les enjeux hydrologiques modernes, de prévisions ou liés aux changements climatiques, forcent l’exploration de nouvelles approches en modélisation afin de combler les lacunes actuelles et d’améliorer l’évaluation des incertitudes. L’approche abordée dans ce mémoire est celle du multimodèle (MM). L’innovation se trouve dans la construction du multimodèle présenté dans cette étude : plutôt que de caler individuellement des modèles et d’utiliser leur combinaison, un calage collectif est réalisé sur la moyenne des 12 modèles globaux conceptuels sélectionnés. Un des défis soulevés par cette approche novatrice est le grand nombre de paramètres (82) qui complexifie le calage et l’utilisation, en plus d’entraîner des problèmes potentiels d’équifinalité. La solution proposée dans ce mémoire est une analyse de sensibilité qui permettra de fixer les paramètres peu influents et d’ainsi réduire le nombre de paramètres total à caler. Une procédure d’optimisation avec calage et validation permet ensuite d’évaluer les performances du multimodèle et de sa version réduite en plus d’en améliorer la compréhension. L’analyse de sensibilité est réalisée avec la méthode de Morris, qui permet de présenter une version du MM à 51 paramètres (MM51) tout aussi performante que le MM original à 82 paramètres et présentant une diminution des problèmes potentiels d’équifinalité. Les résultats du calage et de la validation avec le « Split-Sample Test » (SST) du MM sont comparés avec les 12 modèles calés individuellement. Il ressort de cette analyse que les modèles individuels, composant le MM, présentent de moins bonnes performances que ceux calés indépendamment. Cette baisse de performances individuelles, nécessaire pour obtenir de bonnes performances globales du MM, s’accompagne d’une hausse de la diversité des sorties des modèles du MM. Cette dernière est particulièrement requise pour les applications hydrologiques nécessitant une évaluation des incertitudes. Tous ces résultats mènent à une amélioration de la compréhension du multimodèle et à son optimisation, ce qui facilite non seulement son calage, mais également son utilisation potentielle en contexte opérationnel.