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. / Integrated water resource management requires distinction between water paths that are directly available for society and those which are not. Water pathways, from precipitation to the oceans or the atmosphere, are highly variable from one place to another. The complexity of water pathways can be simplified by focusing on two main categories of water resources: blue water, which is the stock and flow moving into the hydrosystem that is directly available (e.g. rivers, lakes, aquifers and groundwater flow), and green water, which is the invisible flow of water vapor leaving the hydrosphere to the atmosphere. The latter includes the water used by forests, grasslands, rain fed crops, and the water in soils. However, many hydrological studies focus only on blue water, particularly the discharge or more rarely the ground water recharge, ignoring all green water components, therefore missing the overall picture. At the same time, climate change highlighted in recent years have been found to impact water pathway distributions by affecting different components of the hydrological cycle at the watershed scale. The study presented here exploits the SWAT hydrological model to assess the variation of different components of a hydrosystem facing climate change. The study area is the watershed of the Garonne River, where data is available. The first part of this work focused on refining the implementation of the model in order to better tackle the problem at hand. Particular attention has been paid to the use of gridded weather data (SAFRAN product) as well as to the simulation of snow present in the mountainous portion of the watershed. Calibration of the model parameters was tested through a differential split sampling method, based on calibration and validation using climatically contrasted periods, in order to test the robustness of the model. These steps led to a substantial improvement in the simulations performance over the calibration period (2000-2010) and demonstrated the robustness of the model within a climate change context. The improved SWAT model was next used to produce simulations over a hundred-year period (1960-2050), an analysis carried out in two steps: First, the past period (1960-2000) simulation, based on observed climatic data, was used to validate discharge simulations for which very good performance was obtained. Analysis of the different components of the hydrological cycle showed a strong impact on flows and stocks of green water, with a reduction of the water content in soil and a substantial increase in evapotranspiration. Blue water is mostly impacted in terms of snow stock and discharge flow, which both showed a substantial decrease. Secondly, hydrological projections were performed (2010-2050) based on a selection of climate scenarios and models, submitted to dynamic downscaling. Analysis of these projections partly confirmed the conclusions drawn from the historic period: i.e. a substantial impact on green water, with a decrease of the soil water content and an increase of potential evapotranspiration. The projections also revealed that the soil water content during the summer season is such that it reduces the actual evapotranspiration, highlighting possible future deficits of green water stocks. Furthermore, if the analysis of blue water components always presented a substantial decrease in the snowpack, discharge appears to increase during autumn and winter periods. These results indicate an "acceleration" of blue surface water components which is likely related to an increase in extreme rainfall events. In this study, an analysis of the variation of the main hydrological cycle components have been proposed at a watershed scales, confirming the importance of taking into account all these components to evaluate the climate change impact and more broadly environmental changes on water resources.

As áreas variáveis de afluência (AVAs) são dinâmicas, apresentando expansão das áreas saturadas durante os eventos de chuva, geralmente próximas aos cursos d\'água e, no momento que a chuva cessa, estas áreas saturadas se contraem. O escoamento superficial ocorre nessas áreas devido ao excesso de saturação, provocado pelo aumento do volume d\'água armazenado no perfil de solo e, extravasamento nas áreas com solos rasos, próxima aos rios. Dessa forma, faz-se necessário quantificar o processo de escoamento superficial para a correta delimitação das AVAs. A utilização dos modelos hidrológicos para essa finalidade teve início após legislação estadunidense que define níveis máximos permitidos para poluição difusa. Diversos modelos foram desenvolvidos para quantificar a entrada de poluentes nos corpos hídricos, entretanto não havia maneira precisa de localizar as áreas variáveis de afluência, sendo estas as mais propensas a carrear os contaminantes. Somente através da utilização de modelos hidrológicos distribuídos foi possível considerar o componente espacial, ou seja, a localização exata da ocorrência dos processos hidrológicos, e sua inter-relação com uso de solo e tipo de solo, permitindo testar diferentes cenários avaliando quais áreas convertidas em florestas contribuiriam para maior ganho de serviços ecossistêmicos relacionados à manutenção de recursos hídricos. Portanto, foram modelados 3 cenários: o cenário atual, o cenário AVA e o cenário Código Florestal. O primeiro representa a situação atual do uso do solo, e fornece base para comparação com outros cenários. A probabilidade de saturação para este cenário foi definida com uso do modelo hidrológico GSSHA, permitindo delimitar as áreas variáveis de afluência e criar o cenário AVA, o qual simula a restauração florestal em todas as áreas variáveis de afluência. Por último, foi modelado o cenário Código Florestal, que simula a restauração florestal das áreas de preservação permanentes (APPs), com a função de avaliar quais os impactos para a manutenção dos recursos hídricos caso seja cumprido o Código Florestal (Lei nº. 4.711/65) e sejam restauradas todas as áreas de preservação permanente. Os resultados mostram que a restauração das AVA, com alteração de apenas 4,04% da área total da bacia, aumentaria em 48% a infiltração da água no solo, eliminando a geração de escoamento superficial em áreas agrosilvopastoris e conseqüente carreamento de poluentes provenientes dessas áreas. A restauração das APPs representa uma alteração de 9,36% da área da bacia e promove a recuperação da dinâmica de expansão e contração das nascentes da bacia hidrográfica, que garante redução da vazão e atraso do pico de vazão, evitando respostas hidrológicas hortonianas na bacia hidrográfica. Ambos cenários apresentam benefícios para manutenção dos recursos hídricos. As áreas de preservação permanente apresentam papel significativo na proteção dos recursos hídricos, protegendo mais de 60% das AVAs e sendo de fácil delimitação. A utilização do índice topográfico como variável substituta à modelagem hidrológica apresentou correlação de ~0,33, que permite utilizar o índice para uma análise exploratória, porém insuficiente para delimitar as áreas variáveis de afluência. / The variables source areas (VSA) are dynamic, showing expansion of saturated areas during rain events, usually near to streams and, at the time the rain stops, these saturated areas contract. Runoff occurs in these areas due to saturation excess overland flow, caused by increased of stored volume water in the soil profile, and extravasation in areas with shallow soils, next to streams. Thus, it is necessary to quantify the process of runoff for the correct delineation of VSA. The use of hydrological models for this purpose began after U.S. law which sets maximum permitted levels for diffuse pollution. Several models have been developed to quantify the entry of pollutants in water bodies, however there was no accurate way to pinpoint variables source areas, which are the most likely to carrying contaminants. Only through the use of distributed hydrological models was possible to consider the spatial component, in other words, the exact location of the occurrence of hydrological processes and their interrelationship with land use and soil type, allowing you to test different scenarios by assessing which areas converted to forests contribute to greater gains in ecosystem services related to maintenance of water resources. Therefore, were evaluated three scenarios: the actual scenario, the VSA scenario and the Forest Code scenario. The first one represents the current state of land use and provides a basis for comparison with other scenarios. The probability of saturation for this scenario was defined using the hydrological model GSSHA, allowing to delimit variables source areas and to create the VSA scenario, which simulates forest restoration in all variables source areas. Finally, was modeled the Forestry Code scenario, which simulates forest restoration of permanent preservation areas (PPA), whose function is to assess the impacts for the maintenance of water resources if it complied the Forest Code (Law nº. 4.711/65) and restored all permanent preservation areas. The results show that the restoration of the VSA, with only a 4,04% change of the total area of the watershed, it would increase in 48% water infiltration into the soil, eliminating the generation of surface runoff and consequent carry pollutants from these areas. The restoration of the PPA represents a change of 9,36% of the watershed area and promotes the recovery of dynamic expansion and contraction of the headwaters of the watershed, which ensures reduction in flow rate and delay peak flow, avoiding answers hortonian in the hydrological basin. Both scenarios provide benefits for maintenance of water resources. The permanent preservation areas have significant role in protecting water resources, protecting more than 60% of VSA and being easy delimitation. The use of topographic index as surrogate parameter correlated to the hydrological modeling of ~ 0,33, which allows use the index to an exploratory analysis, but insufficient to delineate the variables source areas.

La formation du gypse (CaSO4.2H2O) dans les bassins marginaux méditerranéens au cours du Messinien est contrôlée par la restriction des bassins et par le cycle hydrologique local. La compétition entre l’évaporation et l’apport d’eau douce par les rivières, en association avec des échanges limites avec la Méditerranée, ont permis la mise en place de conditions chimiques favorables à la formation du gypse. La restriction des bassins entraine, de plus, l’augmentation de la consommation de l’oxygène par les microorganismes, l’anoxie, et la mise en place d’un cycle biogéochimique actif du soufre. Au cours de cette étude, j’utilise la composition isotopique stable du gypse en tant que traceur des cycles de l’eau et du soufre dans les bassins marginaux. Le but est de mieux comprendre les conditions hydrologiques et géochimiques qui ont mené à la précipitation du gypse. Ce problème se place dans un débat actuel de la communauté scientifique, particulièrement depuis que de récents travaux proposent qu’une partie du gypse dans les bassins marginaux aurait pu précipiter à partir d’une colonne d’eau à faible salinité (£ 35 PSU) - hypothèse qui peut sembler peu réaliste d’un point de vue géochimique. J’ai mené une analyse isotopique à haute-résolution des couches de gypse qui composent les alternances cycliques gypse-marnes dans les bassins messiniens de Caltanissetta (BC, Sicile) et du Piémont (BP, nord-ouest de l’Italie). Ces alternances gypse-marnes correspondraient à l’expression sédimentaire des cycles astronomiques de précession (~20 ka), pendant lesquels les bassins marginaux ont subi une alternance de conditions climatiques arides et humides. Le cycle hydrologique a été trace grâce aux mesures des compositions isotopiques de l’oxygène et de l’hydrogène de l’eau de cristallisation des gypses ; le cycle biogéochimique du soufre a quant à lui été trace en mesurent les compositions isotopiques du soufre et de l’oxygène des ions sulfates des gypses. J’ai pu observer que : (1) les isotopes de l’eau piégée dans les gypses sont nettement plus légers que ceux théoriquement attendus pour des gypses ayant précipite uniquement à partir d’une eau de mer, et (2) l’eau du BC est caractérisée par un plus fort déficit en deutérium - compare à sa teneur en 18O - que l’eau du BP. Combine a un modèle hydrologique numérique, ces observations impliquent que (1) la précipitation des gypses est réalisée sous l’influence d’un apport d’eau douce fluviatile important, particulièrement au nord du BP, et est donc caractérisée par de très faibles salinités (27-50 psu pour le BC et 10-42 psu dans le BP) ; (2) le contraste de déficit en deutérium entre les deux bassins résultant de la différence de teneur en deutérium des flux évaporais respectifs, est contrôlée par a une différence d’humidité atmosphérique : cela implique que l’atmosphère au-dessus du BC était plus sèche que celle au-dessus du BP. Nous pouvons alors proposer qu’un gradient latitudinal d’humidité relative similaire à l’actuel existait au Messinien, apportant la preuve d’un climat de type méditerranéen dans la région il y a 5.97 Ma. La composition isotopique des ions sulfate suggère globalement une formation de gypse sous influence marine. Cependant, des divergences observées avec la signature marine messinienne mettent en évidence un cycle biogéochimique du soufre actif, contrôlé par la réduction des sulfates et l’oxydation des sulfures. En particulier : (1) les sulfates du BP enrichis en 18O et 34S sont indicateurs d’une sulfato-reduction dans un système géochimique ouvert ou la perte de 32S est liée a la formation de minéraux soufres sédimentaires ; (2) l’enrichissement (BC) ou l’appauvrissement (BP) significatifs en 18O dans les échantillons alors que la teneur en 34S est la même que celle de l’eau de mer indique une reoxydation de sulfure dans un système géochimique ferme d’un bassin marginal soit évaporait (BC) soit dilue (BP). / The formation of gypsum (CaSO4.2H2O) in Messinian Mediterranean marginal basins is controlled by basin restriction and the local hydrological cycle. Acting together, evaporation, river input and restricted water exchange with the Mediterranean basin bring about the chemical conditions for gypsum formation. Basin restriction also leads to enhanced microbial oxygen consumption, anoxia, and the triggering of active biogeochemical sulfur cycling. In this work I use the stable isotopic composition of gypsum as a proxy of water and sulphur cycling in the marginal basins. The goal is to better understand the hydrological and geochemical conditions that lead to gypsum precipitation. This is an open question, especially since recent work has proposed that part of the gypsum in marginal basins precipitated from a low-salinity (£ 35 PSU) water column - a hypothesis that seems unrealistic based on simple geochemical considerations. I carried out a high-resolution isotopic study of gypsum layers composing gypsum-marl cycles in the Messinian Caltanissetta (Sicily) and Piedmont (north-western Italy) marginal basins (CB and PB, respectively). These gypsum-marl cycles are thought to be the sedimentary expression of astronomical precession cycles (~20 kyr), during which the marginal basins experienced a succession of arid and wet conditions. The hydrological cycle was tracked by measuring the oxygen and hydrogen isotope composition of the gypsum-bound water molecule; the biogeochemical sulfur cycle was tracked by measuring the sulfur and oxygen isotope composition of the gypsum sulfate ion. I observed that: (1) the isotopes of gypsum-bound water are considerably lighter than those expected for gypsum precipitated via evaporation of seawater, and (2) water in the Caltanissetta basin was characterized by a higher deuterium deficit - compared to its 18O content - than water in the Piedmont basin. In conjunction with a hydrological box-model, these observations imply that (1) gypsum precipitation takes place under the influence of large riverine freshwater fluxes, particularly in the North Piedmont basin, that result in very low salinities (27-50 psu in CB and 10-42 psu in PB) and (2) the contrast in deuterium deficit results from atmospheric humidity-drived difference in the deuterieum content of the evaporative flux, implying that the atmosphere over the CB was drier than that over the PB. Thus, a latitudinal relative humidity gradient similar to the modern one existed in the Messinian, providing evidence for a Mediterranean-like climate in the region 5.97 million years ago. The isotopic composition of the gypsum sulfate ion suggests that it originates from coeval sea water. Deviation from the Messinian marine signature, however, highlights an active biogeochemical sulfur cycle driven by sulfate reduction and sulfide oxidation. In particular, (1) 18O- and 34S-rich sulfate in the Piedmont basin indicates sulfate-reduction in a geochemically open system where 32S is lost to sedimentary sulfide minerals, and (2) significant 18O-enrichment (CB) or 18O-depletion (PB), in samples where the 34S concent is that of seawater, indicates re-oxidation of sulphide in a geochemically closed system of an evaporative (CB) or dilution (PB) marginal basin. A strong relation between the hydrological cycle and the biogeochemical cycle is thus highlighted in marginal Messinian basins

The spatial patterns of land surface and subsurface characteristics determine the spatial heterogeneity of hydrological processes. Soil depth is one of these characteristics and an important input parameter required by distributed hydrological models that explicitly represent spatial heterogeneity. Soil is related to topography and land cover due to the role played by topography and vegetation in affecting soil-forming processes. The research described in this dissertation addressed the development of statistical models that predict the soil depth pattern over the landscape; derivation of new topographic variables evaluated using both serial and parallel algorithms; and evaluation of the impacts of detailed soil depth representation on simulations of stream flow and soil moisture. The dissertation is comprised of three papers. In paper 1, statistical models were developed to predict soil depth pattern over the watershed based on topographic and land cover variables. Soil depth was surveyed at locations selected to represent the topographic and land cover variation at the Dry Creek Experimental Watershed, near Boise, Idaho. Explanatory variables were derived from a digital elevation model and remote sensing imagery for regression to the field data. Generalized Additive and Random Forests models were developed to predict soil depth over the watershed. The models were able to explain about 50% of the soil depth spatial variation, which is an important improvement over the soil depth extracted from the SSURGO national soil database. In paper 2, definitions of the new topographic variables derived in the effort to model soil depth, and serial and Message Passing Interface parallel implementations of the algorithms for their evaluation are presented. The parallel algorithms enhanced the processing speed of large digital elevation models as compared to the serial recursive algorithms initially developed. In paper 3, the impact of spatially explicit soil depth information on simulations of stream flow and soil moisture as compared to soil depth derived from the SSURGO soil database has been evaluated. The Distributed Hydrology Vegetation Soil Model was applied using automated parameter optimization technique with all input parameters the same except soil depth. Stream flow was less impacted by the detailed soil depth information, while simulation of soil moisture was slightly improved due to the detailed representation of soil depth.

Les systèmes de prévision des crues décrivent les transformations pluie-débit en se basant sur des représentations simplifiées. Ces représentations modélisent les processus physiques impliqués avec des descriptions empiriques, ou basées sur des équations de la mécanique classique. Les performances des modèles actuels de prévision des crues sont affectées par différentes incertitudes liées aux approximations et aux paramètres du modèle, aux données d’entrée et aux conditions initiales du bassin versant. La connaissance de ces incertitudes permet aux décideurs de mieux interpréter les prévisions et constitue une aide à la décision lors de la gestion de crue. L’analyse d’incertitudes dans les modèles hydrologiques existants repose le plus souvent sur des simulations de Monte-Carlo (MC). La mise en œuvre de ce type de techniques requiert un grand nombre de simulations et donc un temps de calcul potentiellement important. L'estimation des incertitudes liées à la modélisation hydrologique en temps réel reste donc une gageure. Dans ce projet de thèse, nous développons une méthodologie de prévision des crues basée sur les réseaux Bayésiens (RB). Les RBs sont des graphes acycliques dans lesquels les nœuds correspondent aux variables caractéristiques du système modélisé et les arcs représentent les dépendances probabilistes entre ces variables. La méthodologie présentée propose de construire les RBs à partir des principaux facteurs hydrologiques contrôlant la génération des crues, en utilisant à la fois les observations disponibles de la réponse du système et les équations déterministes décrivant les processus concernés. Elle est conçue pour prendre en compte la variabilité temporelle des différentes variables impliquées. Les dépendances probabilistes entre les variables (paramètres) peuvent être spécifiées en utilisant des données observées, des modèles déterministes existants ou des avis d’experts. Grâce à leurs algorithmes d’inférence, les RBs sont capables de propager rapidement, à travers le graphe, différentes sources d'incertitudes pour estimer leurs effets sur la sortie du modèle (ex. débit d'une rivière). Plusieurs cas d’études sont testés. Le premier cas d’étude concerne le bassin versant du Salat au sud-ouest de la France : un RB est utilisé pour simuler le débit de la rivière à une station donnée à partir des observations de 3 stations hydrométriques localisées en amont. Le modèle présente de bonnes performances pour l'estimation du débit à l’exutoire. Utilisé comme méthode inverse, le modèle affiche également de bons résultats quant à la caractérisation de débits d’une station en amont par propagation d’observations de débit sur des stations en aval. Le deuxième cas d’étude concerne le bassin versant de la Sagelva situé en Norvège, pour lequel un RB est utilisé afin de modéliser l'évolution du contenu en eau de la neige en fonction des données météorologiques disponibles. Les performances du modèle sont conditionnées par les données d’apprentissage utilisées pour spécifier les paramètres du modèle. En l'absence de données d'observation pertinentes pour l’apprentissage, une méthodologie est proposée et testée pour estimer les paramètres du RB à partir d’un modèle déterministe. Le RB résultant peut être utilisé pour effectuer des analyses d’incertitudes sans recours aux simulations de Monte-Carlo. Au regard des résultats enregistrés sur les différents cas d’études, les RBs se révèlent utiles et performants pour une utilisation en support d’un processus d'aide à la décision dans le cadre de la gestion du risque de crue. / Flood forecasting describes the rainfall-runoff transformation using simplified representations. These representations are based on either empirical descriptions, or on equations of classical mechanics of the involved physical processes. The performances of the existing flood predictions are affected by several sources of uncertainties coming not only from the approximations involved but also from imperfect knowledge of input data, initial conditions of the river basin, and model parameters. Quantifying these uncertainties enables the decision maker to better interpret the predictions and constitute a valuable decision-making tool for flood risk management. Uncertainty analysis on existing rainfall-runoff models are often performed using Monte Carlo (MC)- simulations. The implementation of this type of techniques requires a large number of simulations and consequently a potentially important calculation time. Therefore, quantifying uncertainties of real-time hydrological models is challenging. In this project, we develop a methodology for flood prediction based on Bayesian networks (BNs). BNs are directed acyclic graphs where the nodes correspond to the variables characterizing the modelled system and the arcs represent the probabilistic dependencies between these variables. The presented methodology suggests to build the RBs from the main hydrological factors controlling the flood generation, using both the available observations of the system response and the deterministic equations describing the processes involved. It is, thus, designed to take into account the time variability of different involved variables. The conditional probability tables (parameters), can be specified using observed data, existing hydrological models or expert opinion. Thanks to their inference algorithms, BN are able to rapidly propagate, through the graph, different sources of uncertainty in order to estimate their effect on the model output (e.g. riverflow). Several case studies are tested. The first case study is the Salat river basin, located in the south-west of France, where a BN is used to simulate the discharge at a given station from the streamflow observations at 3 hydrometric stations located upstream. The model showed good performances estimating the discharge at the outlet. Used in a reverse way, the model showed also satisfactory results when characterising the discharges at an upstream station by propagating back discharge observations of some downstream stations. The second case study is the Sagelva basin, located in Norway, where a BN is used to simulate the accumulation of snow water equivalent (SWE) given available weather data observations. The performances of the model are affected by the learning dataset used to train the BN parameters. In the absence of relevant observation data for learning, a methodology for learning the BN-parameters from deterministic models is proposed and tested. The resulted BN can be used to perform uncertainty analysis without any MC-simulations to be performed in real-time. From these case studies, it appears that BNs are a relevant decisionsupport tool for flood risk management.

Hydrologiska analyser blir ett allt vanligare tillvägagångssätt för att kunna planera inför eventuella översvämningar. Då naturkatastrofer som dessa kommer öka i framtiden i takt med att temperaturen höjs och nederbörden ökar så har Fagerstakommun påbörjat framtagning av skyfallskartering. Syftet med studien är att undersöka hur resultatet av skyfallskartering och hydrologisk modellering påverkas av dataunderlagetsupplösning. Ytterligare ett mål med arbetet är att undersöka om den nationella höjdmodellen (NH) lämpar sig att användas som underlag vid framtida skyfallskarteringar.Området som berörs av studien ligger beläget i Fagerstas centrala stadskärna och delades uppi ett område nordväst om Bergslagssjukhuset och ett i stadsdelen Västanfors. Fagersta kommun har genomfört en laserskanning som ligger till grund för dataunderlagen i analyserna. Laserpunkterna harförarbetats i Cyclone 3DRochberäknatsi FME Workbenchgenom TIN (Triangulated Irregular Network) och sedan har digitala höjdmodeller (eng. digital elevation model, DEM) skapats i upplösningarna 0,5 m, 1 m, 2 m, 5 m, 10 m, 25 m och 50 m. Även den nyligen framställda NH användes som ett underlag. Denna består av en DEM i upplösning 1 m. Skyfallskartering genomfördes i ArcMap 10.8 medmetoden att lokalisera lågpunkter som riskerar att svämmas över. Hydrologisk modellering tillämpades i syfte att framställa flödesvägar som visar var vattnet rinner vid översvämning. Resultaten jämfördes med hjälp av feature agreement statisticssom visade hur bra dataunderlagen överensstämde mot referensupplösningen på 0,5 m. Ytterligare valideringar gjordes mot en skyfallskartering som tillhandahölls av Fagersta kommun. Denna genomfördes år 2017 av SMHI på uppdrag av Länsstyrelsen och jämfördes mot 0,5 m, 1 m och NH 1 m.Resultat från skyfallskarteringen visade att NH överensstämmer väl med Fagerstas egna skyfallskartering. Det visade att utbredningen generellt ökar och fill-up värdet minskar vid en lägre upplösning. Tvärsnitt från lågpunkterna visade attdetta även generaliserar djupet. De hydrologiska modelleringarna styrker att lägre upplösning generaliserar området. En slutsats kan dras att upplösningarna på 10 m och uppåt inte böranvändas vid en mer noggrann analys. Däremot kan de vara tillräckliga som underlag vid översiktlig kartering. Ytterligare slutsats drogsatt NH lämpar sig som underlag vid skyfallskartering för de flesta ändamålen. Ett underlag med tätare punkttäthet bör dockanvändas vid analyser som kräver ett mer detaljerat resultat. / Hydrological analyses are becoming a more common tool for being able to prepare for eventual floods. As natural disasters such as these will increase in the future as higher temperatures and increased precipitation Fagersta municipality has begun producing cloudburst maps. The aim of this study is to investigate the effects that resolution has on hydrological modelling and cloudburst mapping. Another goal of the study is to examine if the Swedish national height model (NH) is suitable for use in forthcoming cloudburst analyses.The study area is located in the city center of Fagersta and is divided into two areas, one northwest of the hospital named Bergslagssjukhuset and the other over the district Västanfors. Fagersta municipality has carried out a laserscanning which is used to create the different data sets in the analyses. The measured points have been preworked in Cyclone 3DR and calculated in FME Workbench as a TIN (Triangulated Irregular Network). Digital elevation models (DEM) were later on created with the following resolutions;0,5 m, 1 m, 2 m, 5 m, 10 m, 25 m and 50 m. The recently produced NH was also used as a data set. It consists of a DEM with a resolution of 1 m. The cloudburst mapping was implemented in ArcMap 10.8 with the method to locate blue spots that run the risk of being flooded. Hydrological modelling was applied with the aim of producing stream networks which presents where water will flow in case of a flood. The results were compared by using feature agreement statistics which showed how comparable the data sets were to the reference resolution of 0,5 m. Additional validations were made by using Fagersta municipality's own cloudburst map which was carried out in 2017 by SMHI on behalf of Länsstyrelsen. The datawas compared to 0,5 m, 1 m and NH 1 m.Results from the cloudburst mapping showed that NH compares well with Fagerstas data set. It also showed that the spread of blue spots generally increases and the fill-up valuedecreases at a lower resolution. Cross sections from the blue spots showed that this also generalizes depth. The hydrological modellings strengthens that lower resolutions generalizes areas. A conclusion can be drawn that the resolutions of 10 m and up should not be used in a precise analysis. They can however be used for easy-reference mapping. Another conclusion was drawn that NH is suitable for use as a data set in cloudburst mapping for most purposes. A data set with a better point density should however be used in analyses which require a more detailed result.

Temporal gravimeter observations, used in geodesy and geophysics to study variation of the Earth’s gravity field, are influenced by local water storage changes (WSC) and – from this perspective – add noise to the gravimeter signal records. At the same time, the part of the gravity signal caused by WSC may provide substantial information for hydrologists. Water storages are the fundamental state variable of hydrological systems, but comprehensive data on total WSC are practically inaccessible and their quantification is associated with a high level of uncertainty at the field scale. This study investigates the relationship between temporal gravity measurements and WSC in order to reduce the hydrological interfering signal from temporal gravity measurements and to explore the value of temporal gravity measurements for hydrology for the superconducting gravimeter (SG) of the Geodetic Observatory Wettzell, Germany. A 4D forward model with a spatially nested discretization domain was developed to simulate and calculate the local hydrological effect on the temporal gravity observations. An intensive measurement system was installed at the Geodetic Observatory Wettzell and WSC were measured in all relevant storage components, namely groundwater, saprolite, soil, top soil and snow storage. The monitoring system comprised also a suction-controlled, weighable, monolith-filled lysimeter, allowing an all time first comparison of a lysimeter and a gravimeter. Lysimeter data were used to estimate WSC at the field scale in combination with complementary observations and a hydrological 1D model. Total local WSC were derived, uncertainties were assessed and the hydrological gravity response was calculated from the WSC. A simple conceptual hydrological model was calibrated and evaluated against records of a superconducting gravimeter, soil moisture and groundwater time series. The model was evaluated by a split sample test and validated against independently estimated WSC from the lysimeter-based approach. A simulation of the hydrological gravity effect showed that WSC of one meter height along the topography caused a gravity response of 52 µGal, whereas, generally in geodesy, on flat terrain, the same water mass variation causes a gravity change of only 42 µGal (Bouguer approximation). The radius of influence of local water storage variations can be limited to 1000 m and 50 % to 80 % of the local hydro¬logical gravity signal is generated within a radius of 50 m around the gravimeter. At the Geodetic Observatory Wettzell, WSC in the snow pack, top soil, unsaturated saprolite and fractured aquifer are all important terms of the local water budget. With the exception of snow, all storage components have gravity responses of the same order of magnitude and are therefore relevant for gravity observations. The comparison of the total hydrological gravity response to the gravity residuals obtained from the SG, showed similarities in both short-term and seasonal dynamics. However, the results demonstrated the limitations of estimating total local WSC using hydrological point measurements. The results of the lysimeter-based approach showed that gravity residuals are caused to a larger extent by local WSC than previously estimated. A comparison of the results with other methods used in the past to correct temporal gravity observations for the local hydrological influence showed that the lysimeter measurements improved the independent estimation of WSC significantly and thus provided a better way of estimating the local hydrological gravity effect. In the context of hydrological noise reduction, at sites where temporal gravity observations are used for geophysical studies beyond local hydrology, the installation of a lysimeter in combination with complementary hydrological measurements is recommended. From the hydrological view point, using gravimeter data as a calibration constraint improved the model results in comparison to hydrological point measurements. Thanks to their capacity to integrate over different storage components and a larger area, gravimeters provide generalized information on total WSC at the field scale. Due to their integrative nature, gravity data must be interpreted with great care in hydrological studies. However, gravimeters can serve as a novel measurement instrument for hydrology and the application of gravimeters especially designed to study open research questions in hydrology is recommended. / Zeitabhängigen Gravimetermessungen, die in der Geodäsie und der Geophysik eingesetzt werden, um Variationen des Erdschwerefelds zu messen, werden durch lokale Wasserspeicheränderungen beeinflusst und verursachen – aus dieser Perspektive – ein hydrologisches Störsignal in den Gravimetermessungen. Gleichzeitig bietet der Teil des Gravimetersignals, der durch Wasserspeicheränderungen hervorgerufen wird, das Potential wichtige Informationen über hydrologische Speicher zu gewinnen, da zwar Wasserspeicher eine grundlegende Zustandsgröße hydrologischer Systeme darstellt, jedoch ihre Quantifizierung mit einem hohen Maß an Unsicherheiten auf der Feldskala behaftet ist. Diese Studie untersucht die Beziehung zwischen zeitabhängigen Gravimetermessungen und Wasserspeicheränderungen, um die Gravimetermessungen von dem hydrologischen Störsignal zu bereinigen und um den Nutzen der Gravimetermessungen für die Hydrologie zu erkunden. Dies geschieht am Beispiel des Supraleitgravimeters (SG) des Geodätischen Observatoriums Wettzell in Deutschland. Ein 4D Vorwärtsmodel mit einer räumlich genesteten Diskretisierungsdomäne wurde entwickelt, um die lokalen hydrologischen Masseneffekte auf Gravimetermessungen zu simulieren. Des Weiteren wurde ein intensives Messsystem am Geodätischen Observatorium Wettzell installiert, um die Wasserspeicheränderungen in allen relevanten Speicherkomponenten, also im dem Grundwasser, in der ungesättigten Zone und im Schneespeicher zu messen. Das Monitoringsystem beinhaltete auch einen wägbaren, monolithischen Lysimeter mit Matrixpotentialübertragung, der es uns ermöglichte, zum ersten Mal einen Lysimeter direkt mit einem Gravimeter zu vergleichen. Die Lysimetermessungen wurden in Kombination mit komplementären hydrologischen Beobachtungen und einem 1D-Modell verwendet, um die Wasserspeicheränderungen auf der Feldskala zu bestimmen. Die Gesamtwasserspeicheränderungen wurden bestimmt, Unsicherheiten abgeschätzt und der hydrologische Masseneffekt auf Gravimetermessungen berechnet. Schlussendlich wurde ein einfaches, konzeptionelles, hydrologisches Modell mittels der Zeitreihen von dem SG, Bodenfeuchte- und Grundwassermessungen kalibriert und evaluiert. Das Modell wurde durch einen “Split-Sample-Test” evaluiert und basierend auf unabhängig bestimmten Wasserspeicheränderungen bestimmt auf Grundlage der Lysimetermessungen validiert. Die Simulation des hydrologischen Masseneffektes auf Gravimetermessungen zeigte, dass Wasserspeicheränderungen von einem Meter Höhe entlang der Topographie, einen Erdschwereeffekt von 52 µGal hervorriefen, während in der Geodäsie im Allgemeinen die gleiche Wassermassenvariation in flachem Terrain eine Erdschwereeffekt von nur 42 µGal (Bouguer-Platte) hervorruft. Der Einflussradius der lokalen Wasserspeicheränderungen kann auf 1000 m begrenzt werden, und 50 % bis 80 % des lokalen hydrologischen Erdschweresignals wird in einem Radius von 50 m um den Gravimeter generiert. Wasserspeichervariationen in der Schneedecke, im Oberboden, dem ungesättigten Saprolith und im gelüfteten Aquifer, sind allesamt wichtige Größen der lokalen Wasserbilanz. Mit der Ausnahme von Schnee beeinflussen alle Speicheränderungen die Gravimetermessungen in derselben Größenordnung und sind daher für die Gravimetermessungen von Bedeutung. Ein Vergleich des lokalen hydrologischen Gravitationseffektes mit den SG Residuen zeigte sowohl ereignisbezogene als auch saisonalen Übereinstimmungen. Weiterhin zeigten die Ergebnisse jedoch auch die Grenzen bei der Bestimmung der gesamten lokalen Wasserspeichervariationen mithilfe hydrologischer Punktmessungen auf. Die Ergebnisse des Lysimeter-basierten Ansatzes zeigten, dass SG Residuen mehr noch, als bisher aufgezeigt, durch lokale Wasserspeicheränderungen hervorgerufen werden. Ein Vergleich der Resultate mit anderen Methoden, die in der Vergangenheit zur Korrektur zeitabhängiger Erdschwerebeobachtungen durch Bestimmung des lokalen hydrologischen Masseneffekte verwendet wurden, zeigte, dass die unabhängige Berechnung von Wasserspeicheränderungen durch Lysimetermessungen erheblich verbessert werden kann und dass diese somit eine verbesserte Methode zur Bestimmung des lokalen hydrologischen Erdschwereeffekts darstellt. Die Installation eines Lysimeters ist somit im Zusammenhang mit einer Reduzierung des hydrologischen Störsignals und an Standorten, wo zeitabhängige Erdschwerebeobachtungen für geophysikalische Studien, die über die lokale Hydrologie hinausgehen verwendet werden, zu empfehlen. Aus hydrologischer Sicht zeigte diese Studie, dass die Verwendung von zeitabhängigen Gravimetermessungen als Kalibrierungsdaten die Modellergebnisse im Vergleich zu hydrologischen Punktmessungen verbesserten. Auf Grund ihrer Fähigkeit, über verschiedene Speicherkomponenten und ein größeres Gebiet zu integrieren, bieten Gravimeter verallgemeinerte Informationen über die Gesamtwasserspeicherveränderungen auf der Feldskala. Diese integrative Eigenschaft macht es notwendig, Erdschweredaten in hydrologischen Studien mit großer Vorsicht zu interpretieren. Dennoch können Gravimeter der Hydrologie als neuartiges Messinstrument dienen und die Nutzung von Gravimetern, die speziell für die Beantwortung noch offener Forschungsfragen der Hydrologie entwickelt wurden wird hier empfohlen.

Within a research project about future sustainable water management options in the Elbe River basin, quasi-natural discharge scenarios had to be provided. The semi-distributed eco-hydrological model SWIM was utilised for this task. According to scenario simulations driven by the stochastical climate model STAR, the region would get distinctly drier. However, this thesis focuses on the challenge of meeting the requirement of high model fidelity even for smaller sub-basins. Usually, the quality of the simulations is lower at inner points than at the outlet. Four research paper chapters and the discussion chapter deal with the reasons for local model deviations and the problem of optimal spatial calibration. Besides other assessments, the Markov Chain Monte Carlo method is applied to show whether evapotranspiration or precipitation should be corrected to minimise runoff deviations, principal component analysis is used in an unusual way to evaluate local precipitation alterations by land cover changes, and remotely sensed surface temperatures allow for an independent view on the evapotranspiration landscape. The overall insight is that spatially explicit hydrological modelling of such a large river basin requires a lot of local knowledge. It probably needs more time to obtain such knowledge as is usually provided for hydrological modelling studies. / Innerhalb eines Forschungsprojekts zu zukünftigen nachhaltigen Optionen der Wasserwirtschaft im Elbe-Einzugsgebiet mußten quasi-natürliche Abflußszenarien bereitgestellt werden. Zu diesem Zweck wurde das räumlich diskretisierte ökohydrologische Modell SWIM eingesetzt. Nach den von dem stochastischen Klimamodell STAR angetriebenen Szenariosimulationen würde die Region deutlich trockener werden. Allerdings ist das Hauptthema dieser Dissertation die Herausforderung, die Ansprüche an hohe Modelltreue auch für kleinere Teileinzugsgebiete zu erfüllen. Normalerweise ist die Qualität der Simulationen für innere Punkte geringer als am Gebietsauslaß. Vier Fachartikel-Kapitel und das Diskussionskapitel beschäftigen sich mit den Gründen für lokale Modellabweichungen und dem Problem optimaler räumlicher Kalibrierung. Unter anderem wird die Markovketten-Monte-Carlo-Methode angewendet, um zu zeigen, ob Verdunstung oder Niederschlag korrigiert werden sollte, um Abweichungen des Abflusses zu minimieren, die Hauptkomponentenanalyse wird auf eine unübliche Weise benutzt, um lokale Niederschlagsänderungen aufgrund von Landnutzungsänderungen zu untersuchen, und fernerkundete Oberflächentemperaturen erlauben eine unabhängige Sicht auf die Verdunstungslandschaft. Die grundlegende Erkenntnis ist, daß die räumlich explizite hydrologische Modellierung eines so großen Flußeinzugsgebiets eine Menge Vor-Ort-Wissen erfordert. Wahrscheinlich wird mehr Zeit benötigt, solches Wissen zu erwerben, als üblicherweise für hydrologische Modellstudien zur Verfügung steht.

This thesis details the development and calibration of a model created by coupling a land surface simulation model named CLASS with a hydrologic model named WATFLOOD. The resulting model, known as WatCLASS, is able to serve as a lower boundary for an atmospheric model. In addition, WatCLASS can act independently of an atmospheric model to simulate fluxes of energy and moisture from the land surface including streamflow. These flux outputs are generated based on conservation equations for both heat and moisture ensuring result continuity. WatCLASS has been tested over both the data rich BOREAS domains at fine scales and the large but data poor domain of the Mackenzie River at coarse scale. The results, while encouraging, point to errors in the model physics related primarily to soil moisture transport in partially frozen soils and permafrost. Now that a fully coupled model has been developed, there is a need for continued research by refining model processes and test WatCLASS's robustness using new datasets that are beginning to emerge. Hydrologic models provide a mechanism for the improvement of atmospheric simulation though two important mechanisms. First, atmospheric inputs to the land surface, such as rainfall and temperature, are transformed by vegetation and soil systems into outputs of energy and mass. One of these mass outputs, which have been routinely measured with a high degree of accuracy, is streamflow. Through the use of hydrologic simulations, inputs from atmospheric models may be transformed to streamflow to assess reliability of precipitation and temperature. In this situation, hydrologic models act in an analogous way to a large rain gauge whose surface area is that of a watershed. WatCLASS has been shown to be able to fulfill this task by simulating streamflow from atmospheric forcing data over multi-year simulation periods and the large domains necessary to allow integration with limited area atmospheric models. A second, more important, role exists for hydrologic models within atmospheric simulations. The earth's surface acts as a boundary condition for the atmosphere. Besides the output of streamflow, which is not often considered in atmospheric modeling, the earth's surface also outputs fluxes of energy in the form of evaporation, known as latent heat and near surface heating, known as sensible heat. By simulating streamflow and hence soil moisture over the land surface, hydrologic models, when properly enabled with both energy and water balance capabilities, can influence the apportioning of the relative quantities of latent and sensible heat flux that are required by atmospheric models. WatCLASS has shown that by improving streamflow simulations, evaporation amounts are reduced by approximately 70% (1271mm to 740mm) during a three year simulation period in the BOREAS northern old black spruce site (NSA-OBS) as compared to the use of CLASS alone. To create a model that can act both as a lower boundary for the atmosphere and a hydrologic model, two choices are available. This model can be constructed from scratch with all the caveats and problems associated with proving a new model and having it accepted by the atmospheric community. An alternate mechanism, more likely to be successfully implemented, was chosen for the development of WatCLASS. Here, two proven and well tested models, WATFLOOD and CLASS, were coupled in a phased integration strategy that allowed development to proceed on model components independently. The ultimate goal of this implementation strategy, a fully coupled atmospheric - land surface - hydrologic model, was developed for MC2-CLASS-WATFLOOD. Initial testing of this model, over the Saguenay region of Quebec, has yet to show that adding WATFLOOD to CLASS produces significant impacts on atmospheric simulation. It is suspected, that this is due to the short term nature of the weather simulation that is dominated by initial conditions imposed on the atmospheric model during the data assimilation cycle. To model the hydrologic system, using the domain of an atmospheric model, requires that methods be developed to characterize land surface forms that influence hydrologic response. Methods, such as GRU (Grouped Response Unit) developed for WATFLOOD, need to be extended to taken advantage of alternate data forms, such as soil and topography, in a way that allows parameters to be selected <I>a priori</I>. Use of GIS (Geographical Information System) and large data bases to assist in development of these relationships has been started here. Some success in creating DEMs, (Digital Elevation Model) which are able to reproduce watershed areas, was achieved. These methods build on existing software implementations to include lake boundaries information as a topographic data source. Other data needs of hydrologic models will build on relationships between land cover, soil, and topography to assist in establishing grouping of these variables required to determine hydrologic similarity. This final aspect of the research is currently in its infancy but provides a platform from which to explore for future initiatives. Original contributions of this thesis are centered on the addition of a lateral flow generation mechanism within a land surface scheme. This addition has shown a positive impact on flux returns to the atmosphere when compared to measured values and also provide increased realism to the model since measured streamflow is reproduced. These contributions have been encapsulated into a computer model known as WatCLASS, which together with the implementation plan, as presented, should lead to future atmospheric simulation improvements.

Drylands cover about 40% of the earth’s land surface and provide the basis for the livelihoods of 38% of the global human population. Worldwide, these ecosystems are prone to heavy degradation. Increasing levels of dryland degradation result a strong decline of ecosystem services. In addition, in highly variable semi-arid environments changing future environmental conditions will potentially have severe consequences for productivity and ecosystem dynamics. Hence, global efforts have to be made to understand the particular causes and consequences of dryland degradation and to promote sustainable management options for semi-arid and arid ecosystems in a changing world. Here I particularly address the problem of semi-arid savanna degradation, which mostly occurs in form of woody plant encroachment. At this, I aim at finding viable sustainable management strategies and improving the general understanding of semi-arid savanna vegetation dynamics under conditions of extensive livestock production. Moreover, the influence of external forces, i.e. environmental change and land reform, on the use of savanna vegetation and on the ecosystem response to this land use is assessed. Based on this I identify conditions and strategies that facilitate a sustainable use of semi-arid savanna rangelands in a changing world. I extended an eco-hydrological model to simulate rangeland vegetation dynamics for a typical semi-arid savanna in eastern Namibia. In particular, I identified the response of semi-arid savanna vegetation to different land use strategies (including fire management) also with regard to different predicted precipitation, temperature and CO2 regimes. Not only environmental but also economic and political constraints like e.g. land reform programmes are shaping rangeland management strategies. Hence, I aimed at understanding the effects of the ongoing process of land reform in southern Africa on land use and the semi-arid savanna vegetation. Therefore, I developed and implemented an agent-based ecological-economic modelling tool for interactive role plays with land users. This tool was applied in an interdisciplinary empirical study to identify general patterns of management decisions and the between-farm cooperation of land reform beneficiaries in eastern Namibia. The eco-hydrological simulations revealed that the future dynamics of semi-arid savanna vegetation strongly depend on the respective climate change scenario. In particular, I found that the capacity of the system to sustain domestic livestock production will strongly depend on changes in the amount and temporal distribution of precipitation. In addition, my simulations revealed that shrub encroachment will become less likely under future climatic conditions although positive effects of CO2 on woody plant growth and transpiration have been considered. While earlier studies predicted a further increase in shrub encroachment due to increased levels of atmospheric CO2, my contrary finding is based on the negative impacts of temperature increase on the drought sensitive seedling germination and establishment of woody plant species. Further simulation experiments revealed that prescribed fires are an efficient tool for semi-arid rangeland management, since they suppress woody plant seedling establishment. The strategies tested have increased the long term productivity of the savanna in terms of livestock production and decreased the risk for shrub encroachment (i.e. savanna degradation). This finding refutes the views promoted by existing studies, which state that fires are of minor importance for the vegetation dynamics of semi-arid and arid savannas. Again, the difference in predictions is related to the bottleneck at the seedling establishment stage of woody plants, which has not been sufficiently considered in earlier studies. The ecological-economic role plays with Namibian land reform beneficiaries showed that the farmers made their decisions with regard to herd size adjustments according to economic but not according to environmental variables. Hence, they do not manage opportunistically by tracking grass biomass availability but rather apply conservative management strategies with low stocking rates. This implies that under the given circumstances the management of these farmers will not per se cause (or further worsen) the problem of savanna degradation and shrub encroachment due to overgrazing. However, as my results indicate that this management strategy is rather based on high financial pressure, it is not an indicator for successful rangeland management. Rather, farmers struggle hard to make any positive revenue from their farming business and the success of the Namibian land reform is currently disputable. The role-plays also revealed that cooperation between farmers is difficult even though obligatory due to the often small farm sizes. I thus propose that cooperation needs to be facilitated to improve the success of land reform beneficiaries. / Semiaride (halbtrockene) Savannen bedecken große Teile der Erdoberfläche und sichern die Lebensgrundlage von vielen Millionen Menschen. Die häufigste Form der Landnutzung in diesen Trockengebieten ist die Produktion von Vieh in extensiver Weidelandbewirtschaftung. In Folge klimatischer Veränderungen und als Konsequenz aus der teils intensiven Beweidung dieser Trockengebiete kommt es häufig zur Degradierung derselben in Form einer Zunahme von ‚unerwünschter‘ holziger Vegetation auf Kosten von futterverwertbaren Gräsern. Dieser als Verbuschung bezeichnete Prozess hat schwere negative Auswirkungen auf die betroffenen Ökosysteme und ist die Ursache für einen zunehmenden Rückgang der ökonomischen Leistungsfähigkeit der betroffenen Betriebe. In meiner Dissertation befasse ich mich mit den Auswirkungen von Klimawandel und politischen Veränderungen auf die Savannenvegetation im südlichen Afrika und auf die Möglichkeiten für die Nutzung dieser Ökosysteme in Form von Viehwirtschaft. Hierbei möchte ich sowohl das allgemeine Verständnis der ökologischen Zusammenhänge verbessern, als auch Strategien für die nachhaltige Nutzung der Savannen identifizieren und bewerten. Da nicht nur ökologische, sondern auch ökonomische und politische Einflussfaktoren, wie zum Beispiel die umfangreichen Landumverteilungen im Rahmen der Bodenreform im südlichen Afrika auf die tatsächliche Landnutzung wirken, habe ich im Rahmen der Dissertation zudem untersucht, nach welchen Umwelt und Kapitalvariablen sich die Farmer, welche Ihr Land im Rahmen der Bodenreform zugeteilt bekommen haben, bei Ihren Entscheidungen richten. Methodisch verwende ich verschiedene Simulationsmodelle, welche zur Untersuchung der langfristigen Veränderungen von verschiedensten Szenarien (Klimawandel, Landnutzung) geeignet sind. Hierbei habe ich teilweise bestehende Modelle angepasst, aber auch ein neues Modell, welches zur Befragung von Farmern in Namibia verwendet wurde, entwickelt. Meine Dissertation führt im Wesentlichen zu vier Erkenntnissen: Erstens, zeigen meine Ergebnisse, welche große Bedeutung die spezifischen ökologischen Eigenschaften der Bäume und Sträucher in semiariden Savannen für die Vorhersage der Entwicklung dieser Systeme unter Klimawandel hat. Hierbei zeigte sich, dass insbesondere die Sensitivität der Keimlinge gegenüber Trockenheit und Feuer eine entscheidende Rolle spielt. Daraus folgt die zweite wesentliche Erkenntnis: Feuer eignet sich in herausragender Weise, um halbtrockene Savannen vor der Verbuschung zu bewahren. Drittens haben die Rollenspiele mit Farmern in Namibia gezeigt, dass deren Entscheidungen im Wesentlichen von finanziellen Schwierigkeiten und nicht von Umwelteinflüssen getrieben werden. Dennoch zeigten meine Ergebnisse, dass diese Farmer mit Ihrem derzeitigen Verhalten wahrscheinlich nicht zur weiteren Degradierung der Savannenvegetation beitragen. Die vierte, und mit am bedeutendste Erkenntnis aus meiner Arbeit ist, dass konservative Beweidungsstrategien mit geringen und konstanten Viehdichten notwendig sind um semiaride Savannen dauerhaft in ökologisch und ökonomisch nachhaltiger Weise zu Nutzen.