Refine
Year of publication
- 2013 (187) (remove)
Document Type
- Article (140)
- Doctoral Thesis (35)
- Review (5)
- Preprint (4)
- Other (2)
- Monograph/Edited Volume (1)
Is part of the Bibliography
- yes (187)
Keywords
- Climate change (4)
- Land use change (4)
- Iran (3)
- Northeast German Basin (3)
- Soil hydrology (3)
- climate change (3)
- erosion (3)
- exhumation (3)
- remote sensing (3)
- thermochronology (3)
Institute
- Institut für Geowissenschaften (187) (remove)
A new sedimentary sequence from Lago di Venere on Pantelleria Island, located in the Strait of Sicily between Tunisia and Sicily was recovered. The lake is located in the coastal infra-Mediterranean vegetation belt at 2 m a.s.l. Pollen, charcoal and sedimentological analyses are used to explore linkages among vegetation, fire and climate at a decadal scale over the past 1200 years. A dry period from ad 800 to 1000 that corresponds to the Medieval Warm Period' (WMP) is inferred from sedimentological analysis. The high content of carbonate recorded in this period suggests a dry phase, when the ratio of evaporation/precipitation was high. During this period the island was dominated by thermophilous and drought-tolerant taxa, such as Quercus ilex, Olea, Pistacia and Juniperus. A marked shift in the sediment properties is recorded at ad 1000, when carbonate content became very low suggesting wetter conditions until ad 1850-1900. Broadly, this period coincides with the Little Ice Age' (LIA), which was characterized by wetter and colder conditions in Europe. During this time rather mesic conifers (i.e. Pinus pinaster), shrubs and herbs (e.g. Erica arborea and Selaginella denticulata) expanded, whereas more drought-adapted species (e.g. Q. ilex) declined. Charcoal data suggest enhanced fire activity during the LIA probably as a consequence of anthropogenic burning and/or more flammable fuel (e.g. resinous Pinus biomass). The last century was characterized by a shift to high carbonate content, indicating a change towards drier conditions, and re-expansion of Q. ilex and Olea. The post-LIA warming is in agreement with historical documents and meteorological time series. Vegetation dynamics were co-determined by agricultural activities on the island. Anthropogenic indicators (e.g. Cerealia-type, Sporormiella) reveal the importance of crops and grazing on the island. Our pollen data suggest that extensive logging caused the local extinction of deciduous Quercus pubescens around ad1750.
The study of outcrop modeling is located at the interface between two fields of expertise, Sedimentology and Computing Geoscience, which respectively investigates and simulates geological heterogeneity observed in the sedimentary record. During the last past years, modeling tools and techniques were constantly improved. In parallel, the study of Phanerozoic carbonate deposits emphasized the common occurrence of a random facies distribution along single depositional domain. Although both fields of expertise are intrinsically linked during outcrop simulation, their respective advances have not been combined in literature to enhance carbonate modeling studies. The present study re-examines the modeling strategy adapted to the simulation of shallow-water carbonate systems, based on a close relationship between field sedimentology and modeling capabilities. In the present study, the evaluation of three commonly used algorithms Truncated Gaussian Simulation (TGSim), Sequential Indicator Simulation (SISim), and Indicator Kriging (IK), were performed for the first time using visual and quantitative comparisons on an ideally suited carbonate outcrop. The results show that the heterogeneity of carbonate rocks cannot be fully simulated using one single algorithm. The operating mode of each algorithm involves capabilities as well as drawbacks that are not capable to match all field observations carried out across the modeling area. Two end members in the spectrum of carbonate depositional settings, a low-angle Jurassic ramp (High Atlas, Morocco) and a Triassic isolated platform (Dolomites, Italy), were investigated to obtain a complete overview of the geological heterogeneity in shallow-water carbonate systems. Field sedimentology and statistical analysis performed on the type, morphology, distribution, and association of carbonate bodies and combined with palaeodepositional reconstructions, emphasize similar results. At the basin scale (x 1 km), facies association, composed of facies recording similar depositional conditions, displays linear and ordered transitions between depositional domains. Contrarily, at the bedding scale (x 0.1 km), individual lithofacies type shows a mosaic-like distribution consisting of an arrangement of spatially independent lithofacies bodies along the depositional profile. The increase of spatial disorder from the basin to bedding scale results from the influence of autocyclic factors on the transport and deposition of carbonate sediments. Scale-dependent types of carbonate heterogeneity are linked with the evaluation of algorithms in order to establish a modeling strategy that considers both the sedimentary characteristics of the outcrop and the modeling capabilities. A surface-based modeling approach was used to model depositional sequences. Facies associations were populated using TGSim to preserve ordered trends between depositional domains. At the lithofacies scale, a fully stochastic approach with SISim was applied to simulate a mosaic-like lithofacies distribution. This new workflow is designed to improve the simulation of carbonate rocks, based on the modeling of each scale of heterogeneity individually. Contrarily to simulation methods applied in literature, the present study considers that the use of one single simulation technique is unlikely to correctly model the natural patterns and variability of carbonate rocks. The implementation of different techniques customized for each level of the stratigraphic hierarchy provides the essential computing flexibility to model carbonate systems. Closer feedback between advances carried out in the field of Sedimentology and Computing Geoscience should be promoted during future outcrop simulations for the enhancement of 3-D geological models.
Based on a numerical model of the Northeast German Basin (NEGB), we investigate the sensitivity of the calculated thermal field as resulting from heat conduction, forced and free convection in response to consecutive horizontal and vertical mesh refinements. Our results suggest that computational findings are more sensitive to consecutive horizontal mesh refinements than to changes in the vertical resolution. In addition, the degree of mesh sensitivity depends strongly on the type of the process being investigated, whether heat conduction, forced convection or free thermal convection represents the active heat driver. In this regard, heat conduction exhibits to be relative robust to imposed changes in the spatial discretization. A systematic mesh sensitivity is observed in areas where forced convection promotes an effective role in shorten the background conductive thermal field. In contrast, free thermal convection is to be regarded as the most sensitive heat transport process as demonstrated by non-systematic changes in the temperature field with respect to imposed changes in the model resolution.
Reliable hydrological monitoring is the basis for sound water management in drained wetlands. Since statistical methods cannot be employed for unobserved or sparsely monitored areas, the primary design (first set-up) may be arbitrary in most instances. The objective of this paper is therefore to provide a guideline for designing the initial hydrological monitoring network. A scheme is developed that handles different parts of monitoring and hydrometry in wetlands, focusing on the positioning of surface water and groundwater gauges. For placement of the former, control units are used which correspond to areas whose water levels can be regulated separately. The latter are arranged depending on hydrological response units, defined by combinations of soil type and land use, and the chosen surface water monitoring sites. A practical application of the approach is shown for an investigation area in the Spreewald region in north-east Germany. The presented scheme leaves a certain degree of freedom to its user, allowing the inclusion of expert knowledge or special concerns. Based on easily obtainable data, the developed hydrological network serves as a first step in the iterative procedure of monitoring network optimisation. Copyright (c) 2013 John Wiley & Sons, Ltd.
A total of 271 pollen records were selected from a large collection of both raw and digitized pollen spectra from eastern continental Asia (70 degrees-135 degrees E and 18 degrees-55 degrees N). Following pollen percentage recalculations, taxonomic homogenization, and age-depth model revision, the pollen spectra were interpolated at a 500-year resolution and a taxonomically harmonized and temporally standardized fossil pollen dataset established with 226 pollen taxa, covering the last 22 cal lea. Of the 271 pollen records, 85% were published since 1990, with reliable chronologies and high temporal resolutions; of these, 50% have raw data with complete pollen assemblages, ensuring the quality of this dataset The pollen records available for each 500-year time slice are well distributed over all main vegetation types and climatic zones of the study area, making their pollen spectra suitable for paleovegetation and paleoclimate research. Such a dataset can be used as an example for the development of similar datasets for other regions of the world.
The Pleistocene archeological record in East Africa has revealed unusual accumulations of Acheulean handaxes at prehistoric sites. In particular, there has been intensive debate concerning whether the artifact accumulation at the Middle Pleistocene Olorgesailie (Southern Kenya Rift) and Kariandusi (Central Kenya Rift) sites were a result of fluvial reworking or of in situ deposition by hominids. We used a two-step approach to test the hypothesis of fluvial reworking. Firstly, the behavior of handaxes in water currents was investigated in a current flume and the flow threshold required to reorientate the handaxes was determined. The results of these experiments suggested that, in relatively high energy and non-steady flow conditions, handaxes will reorientate themselves perpendicular to the current direction. Secondly, an automated image analysis routine was developed and applied to archeological plans from three Acheulean sites, two at Olorgesailie and one at Kariandusi, in order to determine the orientations of the handaxes. A Rayleigh test was then applied to the orientation data to test for a preferred orientation. The results revealed that the handaxes at the Upper Kariandusi Site and the Olorgesailie Main Site Mid Trench had a preferential orientation, suggesting reworking by a paleocurrent. The handaxes from the Olorgesailie Main Site H/6A, however, appeared to be randomly oriented and in situ deposition by the producers therefore remains a possibility.
In soils and sediments there is a strong coupling between local biogeochemical processes and the distribution of water, electron acceptors, acids and nutrients. Both sides are closely related and affect each other from small scale to larger scales. Soil structures such as aggregates, roots, layers or macropores enhance the patchiness of these distributions. At the same time it is difficult to access the spatial distribution and temporal dynamics of these parameter. Noninvasive imaging techniques with high spatial and temporal resolution overcome these limitations. And new non-invasive techniques are needed to study the dynamic interaction of plant roots with the surrounding soil, but also the complex physical and chemical processes in structured soils. In this study we developed an efficient non-destructive in-situ method to determine biogeochemical parameters relevant to plant roots growing in soil. This is a quantitative fluorescence imaging method suitable for visualizing the spatial and temporal pH changes around roots. We adapted the fluorescence imaging set-up and coupled it with neutron radiography to study simultaneously root growth, oxygen depletion by respiration activity and root water uptake. The combined set up was subsequently applied to a structured soil system to map the patchy structure of oxic and anoxic zones induced by a chemical oxygen consumption reaction for spatially varying water contents. Moreover, results from a similar fluorescence imaging technique for nitrate detection were complemented by a numerical modeling study where we used imaging data, aiming to simulate biodegradation under anaerobic, nitrate reducing conditions.
This study aims to establish, evaluate, and apply a modern pollen-climate transfer function from the transition zone between arctic tundra and light-needled taiga in Arctic Siberia. Lacustrine samples (n = 96) from the northern Siberian lowlands of Yakutia were collected along four north-to-south transects crossing the arctic forest line. Samples span a broad temperature and precipitation gradient (mean July temperature, T-July: 7.5-18.7 degrees C; mean annual precipitation, P-ann: 114-315 mm/yr). Redundancy analyses are used to examine the relationship between the modern pollen signal and corresponding vegetation types and climate. Performance of transfer functions for T-July and P-ann were cross-validated and tested for spatial autocorrelation effects. The root mean square errors of prediction are 1.67 degrees C for T-July and 40 mm/yr for P-ann. A climate reconstruction based on fossil pollen spectra from a Siberian Arctic lake sediment core spanning the Holocene yielded cold conditions for the Late Glacial (1-2 degrees C below present T-July). Warm and moist conditions were reconstructed for the early to mid Holocene (2 degrees C higher T-July than present), and climate conditions similar to modern ones were reconstructed for the last 4000 years. In conclusion, our modern pollen data set fills the gap of existing regional calibration sets with regard to the underrepresented Siberian tundra-taiga transition zone. The Holocene climate reconstruction indicates that the temperature deviation from modern values was only moderate despite the assumed Arctic sensitivity to present climate change.
Pseudotachylyte veins frequently associated with mylonites and ultramylonites occur within migmatitic paragneisses, metamonzodiorites, as well as felsic and mafic granulites at the base of the section of the Hercynian lower crust exposed in Calabria (Southern Italy). The crustal section is tectonically superposed on lower grade units. Ultramylonites and pseudotachylytes are particularly well developed in migmatitic paragneisses, whereas sparse fault-related pseudotachylytes and thin mylonite/ultramylonite bands occur in granulite-facies rocks. The presence of sillimanite and clinopyroxene in ultramylonites and mylonites indicates that relatively high-temperature conditions preceded the formation of pseudotachylytes. We have analysed pseudotachylytes from different rock types to ascertain their deep crustal origin and to better understand the relationships between brittle and ductile processes during deformation of the deeper crust. Different protoliths were selected to test how lithology controls pseudotachylyte composition and textures. In migmatites and felsic granulites, euhedral or cauliflower-shaped garnets directly crystallized from pseudotachylyte melts of near andesitic composition. This indicates that pseudotachylytes originated at deep crustal conditions (> 0.75 GPa). In mafic protoliths, quenched needle-to-feather-shaped high-alumina orthopyroxene occurs in contact with newly crystallized plagioclase. The pyroxene crystallizes in garnet-free and garnet-bearing veins. The simultaneous growth of orthopyroxene and plagioclase as well as almandine, suggests lower crustal origin, with pressures in excess of 0.85 GPa. The existence of melts of different composition in the same vein indicates the stepwise, non-equilibrium conditions of frictional melting. Melt formed and intruded into pre-existing anisotropies. In mafic granulites, brittle faulting is localized in a previously formed thin high-temperature mylonite bands. migmatitic gneisses are deformed into ultramylonite domains characterized by s-c fabric. Small grain size and fluids lowered the effective stress on the c planes favouring a seismic event and the consequent melt generation. Microstructures and ductile deformation of pseudotachylytes suggest continuous ductile flow punctuated by episodes of high-strain rate, leading to seismic events and melting.
Environmental parameters such as rainfall, temperature and relative humidity can affect the composition of higher plant leaf wax. The abundance and distribution of leaf wax biomarkers, such as long chain n-alkanes, in sedimentary archives have therefore been proposed as proxies reflecting climate change. However, a robust palaeoclimatic interpretation requires a thorough understanding of how environmental changes affect leaf wax n-alkane distributions in living plants. We have analysed the concentration and chain length distribution of leaf wax n-alkanes in Acacia and Eucalyptus species along a 1500 km climatic gradient in northern Australia that ranges from subtropical to arid. We show that aridity affected the concentration and distribution of n-alkanes for plants in both genera. For both Acacia and Eucalyptus n-alkane concentration increased by a factor of ten to the dry centre of Australia, reflecting the purpose of the wax in preventing water loss from the leaf. Furthermore, Acacian-alkanes decreased in average chain length (ACL) towards the arid centre of Australia, whereas Eucalyptus ACL increased under arid conditions. Our observations demonstrate that n-alkane concentration and distribution in leaf wax are sensitive to hydroclimatic conditions. These parameters could therefore potentially be employed in palaeorecords to estimate past environmental change. However, our finding of a distinct response of n-alkane ACL values to hydrological changes in different taxa also implies that the often assumed increase in ACL under drier conditions is not a robust feature for all plant species and genera and as such additional information about the prevalent vegetation are required when ACL values are used as a palaeoclimate proxy.
Convexities in the longitudinal profiles of actively incising rivers are typically considered to represent the morphologic signal of a transient response to external perturbations in tectonic or climatic forcing. Distinguishing such knickzones from those that may be anchored to the channel network by spatial variations in rock uplift, however, can be challenging. Here, we combine stream profile analysis, Be-10 watershed-averaged erosion rates, and numerical modeling of stream profile evolution to evaluate whether knickzones in the Abukuma massif of northeast Japan represent a temporal or spatial change in rock uplift rate in relation to forearc shortening. Knickzones in channels that drain the eastern flank of the Abukuma massif are characterized by breaks in slope-area scaling and separate low-gradient, alluvial upper-channel segments from high-gradient, deeply-incised lower channel segments. Average erosion rates inferred from Be-10 concentrations in modern sediment below knickzones exceed erosion rates above knickzones by 20-50%. Although profile convexities could be interpreted as a transient response to an increase in rock uplift rate associated with slip on the range-bounding fault, geologic constraints on the initiation of fault slip and the magnitude of displacement cannot be reconciled with a recent, spatially uniform increase in slip rate. Rather, we find that knickzone position, stream profile gradients, and basin averaged erosion rates are best explained by a relatively abrupt spatial increase in uplift rate localized above a flat-ramp transition in the fault system. These analyses highlight the importance of considering spatially non-uniform uplift in the interpretation of stream profile evolution and demonstrate that the adjustment of river profiles to fault displacement can provide constraints on fault geometry in actively eroding landscapes. (C) 2013 Elsevier B.V. All rights reserved.
We review the geodynamic evolution of the Aegean-Anatolia region and discuss strain localisation there over geological times. From Late Eocene to Present, crustal deformation in the Aegean backarc has localised progressively during slab retreat. Extension started with the formation of the Rhodope Metamorphic Core Complex (Eocene) and migrated to the Cyclades and the northern Menderes Massif (Oligocene and Miocene), accommodated by crustal-scale detachments and a first series of core complexes (MCCs). Extension then localised in Western Turkey, the Corinth Rift and the external Hellenic arc after Messinian times, while the North Anatolian Fault penetrated the Aegean Sea. Through time the direction and style of extension have not changed significantly except in terms of localisation. The contributions of progressive slab retreat and tearing, basal drag, extrusion tectonics and tectonic inheritance are discussed and we favour a model (I) where slab retreat is the main driving engine, (2) successive slab tearing episodes are the main causes of this stepwise strain localisation and (3) the inherited heterogeneity of the crust is a major factor for localising detachments. The continental crust has an inherited strong heterogeneity and crustal-scale contacts such as major thrust planes act as weak zones or as zones of contrast of resistance and viscosity that can localise later deformation. The dynamics of slabs at depth and the asthenospheric flow due to slab retreat also have influence strain localisation in the upper plate. Successive slab ruptures from the Middle Miocene to the late Miocene have isolated a narrow strip of lithosphere, still attached to the African lithosphere below Crete. The formation of the North Anatolian Fault is partly a consequence of this evolution. The extrusion of Anatolia and the Aegean extension are partly driven from below (asthenospheric flow) and from above (extrusion of a lid of rigid crust).
The occurrence of neritic microbial carbonates is often related to ecological refuges, where grazers and other competitors are reduced by environmental conditions, or to post-extinction events (e.g. in the Late Devonian, Early Triassic). Here, we present evidence for Middle Jurassic (Bajocian) microbial mounds formed in the normal marine, shallow neritic setting of an inner, ramp system from the High Atlas of Morocco. The microbial mounds are embedded in cross-bedded oolitic facies. Individual mounds show low relief domal geometries (up to 3 m high and 4.5 m across), but occasionally a second generation of mounds exhibits tabular geometries (<1 m high). The domes are circular in plan view and have intact tops, lacking evidence of current influence on mound preferred growth direction or distribution patterns, or truncation. The mound fades consists almost entirely of non-laminated, micritic thrombolites with branching morphologies and fine-grained, clotted and peloidal fabrics. Normal marine biota are present but infrequent. Several lines of evidence document that microbial mound growth alternates with time intervals of active ooid shoal deposition. This notion is of general significance when compared with modern Bahamian microbialites that co-exist with active sub-aquatic dunes. Furthermore, the lack of detailed studies of Middle Jurassic, normal marine shallow neritic microbial mounds adds a strong motivation for the present study. Specifically, Bajocian mounds formed on a firmground substratum during transgressive phases under condensed sedimentation. Furthermore, a transient increase in nutrient supply in the prevailing mesotrophic setting, as suggested by the heterotrophic-dominated biota, may have controlled microbial mound stages.
Agricultural soils have a high potential for sequestration of atmospheric carbon due to their volume and several promising management options. However, there is a remarkable lack of information about the status quo of organic carbon in agricultural soils. In this study a comprehensive data set of 384 cropland soils and 333 grassland soils within the state of Bavaria in southeast Germany was analyzed in order to provide representative information on total amount, regional distribution and driving parameters of soil organic carbon (SOC) and nitrogen (N) in agricultural soils of central Europe. The results showed that grassland soils stored higher amounts of SOC (11.8 kg m(-2)) and N (0.92 kg m(-2)) than cropland soils (9.0 and 0.66 kg m(-2), respectively) due to moisture-induced accumulation of soil organic matter (SOM) in B horizons. Surprisingly, no distinct differences were found for the A horizons since tillage led to a relocation of SOM with depth in cropland soils. Statistical analyses of driving factors for SOM storage revealed soil moisture, represented by the topographic wetness index (TWI), as the most important parameter for both cropland and grassland soils. Climate effects (mean annual temperature and precipitation) were of minor importance in agricultural soils because management options counteracted them to a certain extent, particularly in cropland soils. The distribution of SOC and N stocks within Bavaria based on agricultural regions confirmed the importance of soil moisture since the highest cropland SOC and N stocks were found for tertiary hills and loess regions, which exhibited large areas with potentially high soil moisture content in extant floodplains. Grassland soils showed the highest accumulation of SOC and N in the Alps and Pre-Alps as a result of low temperatures, high amounts of precipitation and high soil moisture content in areas of glacial denudation. Soil class was identified as a further driving parameter for SOC and N storage in cropland soils. In total, cropland and grassland soils in Bavaria store 242 and 134 Mt SOC as well as 19 and 12 Mt N down to a soil depth of 1 m or the parent material, respectively.
In this paper a change-point detection method is proposed by extending the singular spectrum transformation (SST) developed as one of the capabilities of singular spectrum analysis (SSA). The method uncovers change points related with trends and periodicities. The potential of the proposed method is demonstrated by analysing simple model time series including linear functions and sine functions as well as real world data (precipitation data in Kenya). A statistical test of the results is proposed based on a Monte Carlo simulation with surrogate methods. As a result, the successful estimation of change points as inherent properties in the representative time series of both trend and harmonics is shown. With regards to the application, we find change points in the precipitation data of Kenyan towns (Nakuru, Naivasha, Narok, and Kisumu) which coincide with the variability of the Indian Ocean Dipole (IOD) suggesting its impact of extreme climate in East Africa.
In probabilistic seismic-hazard analysis, epistemic uncertainties are commonly treated within a logic-tree framework in which the branch weights express the degree of belief of an expert in a set of models. For the calculation of the distribution of hazard curves, these branch weights represent subjective probabilities. A major challenge for experts is to provide logically consistent weight estimates (in the sense of Kolmogorovs axioms), to be aware of the multitude of heuristics, and to minimize the biases which affect human judgment under uncertainty. We introduce a platform-independent, interactive program enabling us to quantify, elicit, and transfer expert knowledge into a set of subjective probabilities by applying experimental design theory, following the approach of Curtis and Wood (2004). Instead of determining the set of probabilities for all models in a single step, the computer-driven elicitation process is performed as a sequence of evaluations of relative weights for small subsets of models. From these, the probabilities for the whole model set are determined as a solution of an optimization problem. The result of this process is a set of logically consistent probabilities together with a measure of confidence determined from the amount of conflicting information which is provided by the expert during the relative weighting process. We experiment with different scenarios simulating likely expert behaviors in the context of knowledge elicitation and show the impact this has on the results. The overall aim is to provide a smart elicitation technique, and our findings serve as a guide for practical applications.
Water management and environmental protection is vulnerable to extreme low flows during streamflow droughts. During the last decades, in most rivers of Central Europe summer runoff and low flows have decreased. Discharge projections agree that future decrease in runoff is likely for catchments in Brandenburg, Germany. Depending on the first-order controls on low flows, different adaption measures are expected to be appropriate. Small catchments were analyzed because they are expected to be more vulnerable to a changing climate than larger rivers. They are mainly headwater catchments with smaller ground water storage. Local characteristics are more important at this scale and can increase vulnerability. This thesis mutually evaluates potential adaption measures to sustain minimum runoff in small catchments of Brandenburg, Germany, and similarities of these catchments regarding low flows. The following guiding questions are addressed: (i) Which first-order controls on low flows and related time scales exist? (ii) Which are the differences between small catchments regarding low flow vulnerability? (iii) Which adaption measures to sustain minimum runoff in small catchments of Brandenburg are appropriate considering regional low flow patterns? Potential adaption measures to sustain minimum runoff during periods of low flows can be classified into three categories: (i) increase of groundwater recharge and subsequent baseflow by land use change, land management and artificial ground water recharge, (ii) increase of water storage with regulated outflow by reservoirs, lakes and wetland water management and (iii) regional low flow patterns have to be considered during planning of measures with multiple purposes (urban water management, waste water recycling and inter-basin water transfer). The question remained whether water management of areas with shallow groundwater tables can efficiently sustain minimum runoff. Exemplary, water management scenarios of a ditch irrigated area were evaluated using the model Hydrus-2D. Increasing antecedent water levels and stopping ditch irrigation during periods of low flows increased fluxes from the pasture to the stream, but storage was depleted faster during the summer months due to higher evapotranspiration. Fluxes from this approx. 1 km long pasture with an area of approx. 13 ha ranged from 0.3 to 0.7 l\s depending on scenario. This demonstrates that numerous of such small decentralized measures are necessary to sustain minimum runoff in meso-scale catchments. Differences in the low flow risk of catchments and meteorological low flow predictors were analyzed. A principal component analysis was applied on daily discharge of 37 catchments between 1991 and 2006. Flows decreased more in Southeast Brandenburg according to meteorological forcing. Low flow risk was highest in a region east of Berlin because of intersection of a more continental climate and the specific geohydrology. In these catchments, flows decreased faster during summer and the low flow period was prolonged. A non-linear support vector machine regression was applied to iteratively select meteorological predictors for annual 30-day minimum runoff in 16 catchments between 1965 and 2006. The potential evapotranspiration sum of the previous 48 months was the most important predictor (r²=0.28). The potential evapotranspiration of the previous 3 months and the precipitation of the previous 3 months and last year increased model performance (r²=0.49, including all four predictors). Model performance was higher for catchments with low yield and more damped runoff. In catchments with high low flow risk, explanatory power of long term potential evapotranspiration was high. Catchments with a high low flow risk as well as catchments with a considerable decrease in flows in southeast Brandenburg have the highest demand for adaption. Measures increasing groundwater recharge are to be preferred. Catchments with high low flow risk showed relatively deep and decreasing groundwater heads allowing increased groundwater recharge at recharge areas with higher altitude away from the streams. Low flows are expected to stay low or decrease even further because long term potential evapotranspiration was the most important low flow predictor and is projected to increase during climate change. Differences in low flow risk and runoff dynamics between catchments have to be considered for management and planning of measures which do not only have the task to sustain minimum runoff.
Scarcity of hydrological data, especially streamflow discharge and groundwater level series, restricts the understanding of channel transmission losses (TL) in drylands. Furthermore, the lack of information on spatial river dynamics encompasses high uncertainty on TL analysis in large rivers. The objective of this study was to combine the information from streamflow and groundwater level series with multi-temporal satellite data to derive a hydrological concept of TL for a reach of the Middle Jaguaribe River (MJR) in semi-arid north-eastern Brazil. Based on this analysis, we proposed strategies for its modelling and simulation. TL take place in an alluvium, where river and groundwater can be considered to be hydraulically connected. Most losses certainly infiltrated only through streambed and levees and not through the flood plains, as could be shown by satellite image analysis. TL events whose input river flows were smaller than a threshold did not reach the outlet of the MJR. TL events whose input flows were higher than this threshold reached the outlet losing on average 30% of their input. During the dry seasons (DS) and at the beginning of rainy seasons (DS/BRS), no river flow is expected for pre-events, and events have vertical infiltration into the alluvium. At the middle and the end of the rainy seasons (MRS/ERS), river flow sustained by base flow occurs before/after events, and lateral infiltration into the alluvium plays a major role. Thus, the MJR shifts from being a losing river at DS/BRS to become a losing/gaining (mostly losing) river at MRS/ERS. A model of this system has to include the coupling of river and groundwater flow processes linked by a leakage approach.
Even though quite different in occurrence and consequences, from a modeling perspective many natural hazards share similar properties and challenges. Their complex nature as well as lacking knowledge about their driving forces and potential effects make their analysis demanding: uncertainty about the modeling framework, inaccurate or incomplete event observations and the intrinsic randomness of the natural phenomenon add up to different interacting layers of uncertainty, which require a careful handling. Nevertheless deterministic approaches are still widely used in natural hazard assessments, holding the risk of underestimating the hazard with disastrous effects. The all-round probabilistic framework of Bayesian networks constitutes an attractive alternative. In contrast to deterministic proceedings, it treats response variables as well as explanatory variables as random variables making no difference between input and output variables. Using a graphical representation Bayesian networks encode the dependency relations between the variables in a directed acyclic graph: variables are represented as nodes and (in-)dependencies between variables as (missing) edges between the nodes. The joint distribution of all variables can thus be described by decomposing it, according to the depicted independences, into a product of local conditional probability distributions, which are defined by the parameters of the Bayesian network. In the framework of this thesis the Bayesian network approach is applied to different natural hazard domains (i.e. seismic hazard, flood damage and landslide assessments). Learning the network structure and parameters from data, Bayesian networks reveal relevant dependency relations between the included variables and help to gain knowledge about the underlying processes. The problem of Bayesian network learning is cast in a Bayesian framework, considering the network structure and parameters as random variables itself and searching for the most likely combination of both, which corresponds to the maximum a posteriori (MAP score) of their joint distribution given the observed data. Although well studied in theory the learning of Bayesian networks based on real-world data is usually not straight forward and requires an adoption of existing algorithms. Typically arising problems are the handling of continuous variables, incomplete observations and the interaction of both. Working with continuous distributions requires assumptions about the allowed families of distributions. To "let the data speak" and avoid wrong assumptions, continuous variables are instead discretized here, thus allowing for a completely data-driven and distribution-free learning. An extension of the MAP score, considering the discretization as random variable as well, is developed for an automatic multivariate discretization, that takes interactions between the variables into account. The discretization process is nested into the network learning and requires several iterations. Having to face incomplete observations on top, this may pose a computational burden. Iterative proceedings for missing value estimation become quickly infeasible. A more efficient albeit approximate method is used instead, estimating the missing values based only on the observations of variables directly interacting with the missing variable. Moreover natural hazard assessments often have a primary interest in a certain target variable. The discretization learned for this variable does not always have the required resolution for a good prediction performance. Finer resolutions for (conditional) continuous distributions are achieved with continuous approximations subsequent to the Bayesian network learning, using kernel density estimations or mixtures of truncated exponential functions. All our proceedings are completely data-driven. We thus avoid assumptions that require expert knowledge and instead provide domain independent solutions, that are applicable not only in other natural hazard assessments, but in a variety of domains struggling with uncertainties.