Browsing by Author "Bárdossy, András (Prof. Dr. rer. nat. Dr.-Ing. habil.)"

Now showing 1 - 5 of 5

Open Access
Abflusskonzentration in mesoskaligen Einzugsgebieten unter Berücksichtigung des Sickerraumes
(2006) Rojanschi, Vlad; Bárdossy, András (Prof. Dr. rer. nat. Dr.-Ing. habil.)
Die physikalisch-basierte Modellierung der Abflusskonzentrationsprozesse in einem Einzugsgebiet wird üblicherweise auf die Modellierung von drei strickt von einander abgetrennten Teilen, dem Boden- (Wurzel-) Raum, dem Grundwasserraum und den Oberflächengewässern, reduziert. Dieses Schema vernachlässigt allerdings, dass sich zwischen dem Boden- und dem Grundwasserraum ein weiterer Bereich befindet, der hier als Sickerraum bezeichnet wird und der aus demjenigen Teil der ungesättigten Zone besteht, der nicht dem Bodenraum zugerechnet wird. Für größere Einzugsgebiete treten im Sickerraum, der bis zu mehreren hundert Metern mächtig sein kann, auch kleinere schwebende gesättigte Bereiche auf, was dazu führt, dass die Strömungsrichtung im Sickerraum nicht nur vertikal, sondern auch horizontal ist. Die Vernachlässigung der dort stattfindenden Prozesse schränkt die Aussagekraft des Gesamtmodells deutlich ein. Deshalb wurde die Modellierung der Abflusskonzentration unter Berücksichtigung der hydrologischen Rolle des Sickerraumes als Hauptthema dieser Arbeit ausgewählt. Untersuchungseinheiten sind das Einzugsgebiet der Oberen Donau (bis zum Pegel Passau-Achleiten, 77.000 km2), das ein Forschungschwerpunkt des BMBF-Projektverbundes GLOWA ist, und zwei ausgewählte Teileinzugsgebiete(Ammer und Naab). Der erste Schritt ist eine umfassende Untersuchung über die Anwendbarkeit von Ganglinienanalyseverfahren für die Abtrennung des Grundwasser- und Sickerraumabflusses von der gemessenen Gesamtabflussganglinie. Ein numerisches Programm, das erstmals zwölf relevante Ganglinienanalyseverfahren in einem einheitlichen Rahmen implementiert, wurde hier entwickelt und auf Ganglinien aus dem gesamten Gebiet der Oberen Donau angewandt. Die Analyse der Ergebnisse, ihrer Abhängigkeit von der Raum und Zeitskala, sowie ihrer Verbindung zu den Gebietseigenschaften führte zu neuen Erkenntnissen über die Verfahren. Eine Schätzung des Grundwasser- und Sickerraumabflusses konnte damit für jedes Teileinzugsgebiet berechnet werden. Die Analyse zeigt aber auch, dass die Verfahren mit Inkonsistenzen und Willkürlichkeiten behaftet sind, was nicht zu einer Anwendung ihrer Ergebnisse für quantitative Aussagen ermutigt. Im zweiten Schritt wurde ein neues Modellkonzept, das die explizite Betrachtung des Sickerraumes ermöglicht und damit die Modellierungslücke zwischen dem Bodenwasserhaushalts- und dem Grundwassermodell schließt, entwickelt, implementiert und auf das Gebiet der Ammer angewandt. Nicht nur die Modellgüte, sondern auch die Unsicherheit der Modellergebnisse und bei der Bestimmung der Modellparameter, die generalisierte und einzelne Sensitivität des Modells im Parameterraum, sowie die Wechselbeziehungen zwischen den Modellparametern wurden ausführlich untersucht. Mehrere Modellversionen mit unterschiedlichen Graden an Konzeptualisierung wurden dabei verglichen. Trotz der allgemein guten Anpassung der Modellergebnisse an die Modelldaten, konnten anhand der inversen Modellierung auf Grund der strukturellen Unsicherheit des Modells und der Eingangsdaten keine gut bestimmten Parameterwerte für den Sickerraum berechnet werden. Das führte dazu, dass interne Modellergebnisse wie der Grundwasserabfluss und der Sickerraumabfluss auch von einer großen Unsicherheit behaftet waren. Die allgemeine Erkenntnis ist, dass nur die Modellergebnisse, die anhand von Messdaten direkt geprüft werden können, als validiert und aussagekraftig gelten sollten. Um das Problem der strukturellen Unsicherheit zu lösen, wurde in einem dritten Schritt die Methode der inversen Modellierung erweitert und verbessert. Ein Regionalisierungsverfahren, dass die Modellparameter mit den Gebietseigenschaften mit Hilfe von linearen Beziehungen verbindet, wurde in den Kalibrierungsprozess direkt integriert. Der Ansatz wurde auf die Einzugsgebiete der Ammer und der Naab angewandt und lieferte gute Modellergebnisse und führte gleichzeitig zu einer viel geringeren strukturellen Unsicherheit des Modells. Durch die Interpretation der linearen Beziehungen konnten auch Schlüsse über die physikalische Plausibilität des Modells gezogen werden.
Open Access
Application of a non-parametric classification scheme to catchment hydrology
(2008) He, Yi; Bárdossy, András (Prof. Dr. rer. nat. Dr.-Ing. habil.)
Classification has been considered a fundamental step towards improved catchment hydrology science. Catchments classification has been traditionally carried out via Linnaeus-type cluster analysis, mainly represented by hierarchical approaches and methods based on partitioning of hydrological data set. This paper proposes a new scheme where the classification procedure is based on similarity interpreted as distances between catchments. The similarity or distance is defined under the following premises: 1. similar catchments behave similarly; 2. similarity can be described with catchments' characteristics; and 3. hydrological models are able to capture catchments' similarity. If many sets of model parameters lead to similar model performance for two catchments, they are considered as similar catchments. To implement the proposed scheme, two procedures, namely multidimensional scaling (MDS) and local variance reduction (LVR), are undertaken to construct a configuration of n catchments' characteristics in Euclidean space using information about similar performance between the catchments. The MDS is used to determine the appropriate dimension of the Euclidean space and the LVR is used to obtain the transformation matrix and the coordinates in the transformed Euclidean space. This scheme avoids the idea of parametric regression-based regionalization approaches where a regression function is pre-defined between model parameters and catchment descriptors. In the aforementioned approach, the function that is selected is usually subjective and arbitrary and one can also argue that a priori function is neither able to represent the highly complex hydrological processes nor consider the interdependences amongst model parameters. The proposed scheme is initially tested with a research version of the HBV-IWS model on a number of catchments within the Rhine Basin. Additionally a modified Xinanjiang model is applied to the same catchments to check if the assumption of invariant catchment similarity holds true. Invariant catchment similarity here assumes the catchments genuinely carry their similarities independent of the model used for simulation. This test is also a backstop measure to determine if the models under consideration are capturing the underlying simplified hydrological processes in a rational manner. The scheme will be extended to regional calibration of rainfall runoff models as well as regional drought or flood studies once similarity within catchments has been established. The proposed scheme will eventually contribute to the PUB (Predictions in Ungauged Basins) initiative.
Open Access
Event-based flood estimation using a random forest algorithm for the regionalization in small catchments
(Stuttgart : Eigenverlag des Instituts für Wasser- und Umweltsystemmodellierung der Universität Stuttgart, 2022) Pavía Santolamazza, Daniela; Bárdossy, András (Prof. Dr. rer. nat. Dr.-Ing. habil.)
The hydrological cycle is a complex system, composed of multiple variables, which in most cases are not measured. This is one of the reasons why it is a challenge to have models that adequately represent the expected discharges. The PUB initiative reinforces the need on having models that capture the different catchment interactions and represent various catchment processes. These models are more robust and thus can be more reliable to transfer to the ungauged catchments. In recent years, the field of hydrological research has focused on understanding and explaining the different processes present in catchments. Nevertheless, few applications that include pre- cipitation, the main responsible of runoff change,are found.Further understanding of the temporal and spatial dependence of the meteorological event triggering the floods is needed. In this study, an analysis of the meteorological event triggering the floods was carried out. The concept of entropy was used to characterize the temporal distribution of precipitation. It was found that the precipitation temporal entropy is a better indicator of hydrograph shape than the duration or the intensity. Further, the geographical interdependence of the amount of precipitation and the temporal precipitation entropy causing the floods was described looking at the association of sta- tions triples. This suggested that, up until a given quantile, flood events are more likely caused by precipitation events of total coverage. However, for larger quantile values, it is observed that as the quantile increases the probability of observing joint occurrence in space decreases. The tem- poral distribution of precipitation events causing the floods showed to be more associated in space than the amount of precipitation triggering the floods. Nonetheless, this temporal distribu- tion is not constant over all flood events, what can be attributed to d ifferent flood mechanisms. The Antecedent Precipitation Index (API) was used to explain the soil moisture content. The em- pirical distribution of (API) at the time of a flood was compared with empirical distributions of unconditioned (API) data series. T o this end, the Wilcoxon statistic and the Kolmogorov -Smirnov distance were used to compare the empirical distributions. The re sults showed that the soil mois- ture triggering the floods is not an annual extreme, rather a value close to the monthly maximum (API). Further, it was observed that the longer memory of the catchment gives more information about the occurrence of the flood. Additionally, in order to estimate the catchment reaction at the time of a flood, a regiona lization of the flood wave hydrographs was carried out. T o this end, three methods of defining the simi- larity of the floods were considered. In all three methods, the similarity matrices were generated using the random forest algorithm. The novelty of this procedure was the use of a supervised random forest to describe the similarity of the floods events. It was supervised given that the algorithm was trained to estimate a target variable. The proximity matrix was obtained by calcu- lating the joint occurrence of floods in the random forest space. For evaluating the estimation the hydrograph peak and the time to peak were used. In all three methods, the same tendencies were observed, an overestimation of the peak and an underestimation of the time to peak. However, the bias was observed to be smaller when an ensemble of similarity matrices was used as com- pared to having a single similarity matrix. Moreover, an approach using an unsupervised random forest was compared to the supervised one. It was found that the unsupervised random forest yields larger estimation errors. Finally, to estimate the volume of the flood event a rainfall-runoff model was modified to represent the study region. The model chosen in this study was EPIC. The model was calibrated to be more representative of the study region. To this end, the estimation errors in the space of the model parameters were studied. This allowed to find the model parameters that can better represent the study area. The values obtained were considered reasonable. For example, it is observed that the longer memory of the catchment is more representative of the study catchments, which are the same results as when analyzing the meteorological phenomenon causing the floods. Further, the values obtained for the regional constant, parameter modifying the initial abstraction of the catchment, were found to be smaller than the original ones obtained for United States catchments, which agrees with other studies in European catchments.
Open Access
Statistical modeling of precipitation for agricultural planning in the Volta Basin of West Africa
(2009) Laux, Patrick; Bárdossy, András (Prof. Dr. rer. nat. Dr.-Ing. habil.)
Water availability plays a vital role in the promotion of economic growth and reduction of poverty in the Volta Basin. Due to the increasing population pressure, there is a rapidly increasing demand for water. Climate change additionally impacts water availability and may aggravate water scarcity in the future. Agriculture is the major water consuming sector in the Volta Basin. High rainfall variability of the past often led to shortages in food supply and famines, especially in the Sahel. In such regions, where rainfall is limited to only few months per year, the exact determination of the rainy seasons’ onset is of crucial interest for farming management. Every year, farmers are faced with the question when to start sowing. Do the first rainfalls after the dry season resemble the onset of the rainy season or not? The farmers' seeds and effort will be lost if no significant rainfall follows within the following weeks after sowing. If they do not sow but the first rains were already part of the rainy season, valuable time for agricultural production is lost. Apparently, the benefits of a solution to this problem would be enormous. Therefore, one major goal of this thesis is to find relevant indicators to reliably predict the rainy season's onset. To archive this goal, two different strategies were followed within the context of the doctoral thesis: - The first strategy tackles the problem using solely the measured rainfall time series. Linear discriminant analysis and linear regression analysis were used to predict the onset of the rainy season. - The second strategy uses large-scale meteorological reanalysis fields from GCM output. A multi objective fuzzy logic-based classification algorithm is used to link large-scale meteorological conditions with the onset of the rainy season. Apart from the onset of the rainy season, different rainfall characteristics are analyzed, which are important for agricultural management. These are rainfall occurrence probability, rainfall amount and dry spell probability and are presented as risk maps, suitable for agricultural decision support in the Volta Basin. Drought analysis on regional scale is conducted. The Effective Drought Index is used to derive drought duration, drought intensity, and drought interarrival time. A bivariate Copula approach is therefore used to model the regional return periods of droughts using jointly drought duration and drought intensity. For the investigation of the impact of climate change on these agro-meteorological characteristics, local rainfall information on daily time scale is required. Especially in regions with weak infrastructure like the Volta Basin, all the applied methodologies are hampered by the fact that only little and incomplete meteorological information is available. Statistical downscaling of coarse resolved global circulation models (GCMs) in companion with stochastic rainfall simulation is applied alternatively to avoid these shortcomings. The performance of two different statistical approaches, a simple weather generator and a more sophisticated combined weather pattern classification and simulation approach are evaluated in the context of that doctoral thesis. For the latter, the performance is depending on various factors like e.g. the choice of the predictor(s), the location and size of the domain etc. The impact of climate change on rainfall variability in the Volta Basin is elaborated in terms of future rainy season onset dates and weather pattern frequencies using the A1B scenario driven ECHAM5 model output. For the future time slice 2011-2040, a drastic delay in the onset of the rainy season is expected. Wet and droughty weather patterns are expected to increase in the two northernmost regions of the Volta Basin. Finally, the impact of rainfall variability on crop yield is investigated via multiple linear regression analysis. Rainfall variability has been assessed in terms of the annual rainfall amount, the number of rainy days, the onset, cessation and length of the rainy season and the annual mean of the monthly averaged Effective Drought Index. Regression models, which explain up to 80% of the total variance of crop yield, could be established. It is found, that the annual precipitation amount is the dominant factor to estimate crop yield.
Open Access
Water balance in a poorly gauged basin in West Africa using atmospheric modelling and remote sensing information
(2008) Wagner, Sven; Bárdossy, András (Prof. Dr. rer. nat. Dr.-Ing. habil.)
Sustainable water resources management under increasing water demands and changing climate conditions is a central, socio-political challenge, in particular in climate sensitive regions. Decisions in sustainable water resources management require scientifically sound information of the current water resources and fluxes and future water availability. The first objective of this work is to provide estimations of the current water resources and fluxes in a poorly gauged basin, the White Volta basin in West Africa. This is a central task to support water management authorities and stakeholders in operational irrigation, water supply and running hydro-power strategies. To allow investigations in ungauged or poorly gauges basins, these instruments and methods should be applicable world wide, cost-effective and preferably public domain. In poorly gauged basins without automatic data recorders and online transmission other meteorological data sources for near real time estimations of the terrestrial water balance have to be used to overcome the temporal delay and/or the insufficient spatial resolution. Therefore, a joint atmospheric-hydrological modelling system with MM5 and WaSiM is developed which is able to provide near real time water balance estimations within 48 h. Additionally to meteorological modelling results and observation data, a TRMM product, which is available with approximately one month delay, is applied as precipitation data source. Besides meteorological driving data, land surface properties are essential input data for distributed hydrological modelling. Land surface properties information is usually taken from standard literature values and incorporated into hydrological modelling through tables depending on the land use. The second objective of this work is to increase the level of detail in the spatial and temporal dimension of land surface properties in hydrological modelling using satellite derived land surface properties and to investigate the impact on hydrological modelling results. In this study, products of the MODerate resolution Imaging Spectroradiometer (MODIS) remote sensing system for albedo and leaf area index LAI are imported into the hydrological model and investigated. For sustainable decisions in water resources management, additionally to the modelling result itself, the reliability or uncertainty of the result has to be quantified. Due to the fact that the spatial variability of rainfall is often termed as the major source of error in investigations of rainfall-runoff processes and modelling, the propagation of uncertainties, resulting from the calculation of areal precipitation from point measurements in water balance estimations, are investigated as third objective. Therefore, different spatial interpolation methods, including external drift kriging, for areal precipitation are applied, and their impact on water balance estimates is analysed. Furthermore, geostatistical simulations using the turning band method for areal precipitation are performed in order to investigate the propagation of uncertainties in water balance estimations. These results provide ranges of the temporal and spatial distribution of water balance variables as consequence of uncertainties from the calculation of areal precipitation from station data.