One challenge that faces hydrologists in water resources planning is to predict the catchment???s response to a given rainfall. Estimation of parameter uncertainty (and model uncertainty) allows assessment of the risk in likely applications of hydrological models. Bayesian statistical inference, with computations carried out via Markov Chain Monte Carlo (MCMC) methods, offers an attractive approach to model specification, allowing for the combination of any pre-existing knowledge about individual models and their respective parameters with the available catchment data to assess both parameter and model uncertainty. This thesis develops and applies Bayesian statistical tools for parameter estimation, comparison of model performance and hierarchical model aggregation. The work presented has three main sections. The first area of research compares four MCMC algorithms for simplicity, ease of use, efficiency and speed of implementation in the context of conceptual rainfall-runoff modelling. Included is an adaptive Metropolis algorithm that has characteristics that are well suited to hydrological applications. The utility of the proposed adaptive algorithm is further expanded by the second area of research in which a probabilistic regime for comparing selected models is developed and applied. The final area of research introduces a methodology for hydrologic model aggregation that is flexible and dynamic. Rigidity in the model structure limits representation of the variability in the flow generation mechanism, which becomes a limitation when the flow processes are not clearly understood. The proposed Hierarchical Mixtures of Experts (HME) model architecture is designed to do away with this limitation by selecting individual models probabilistically based on predefined catchment indicators. In addition, the approach allows a more flexible specification of the model error to better assess the risk of likely outcomes based on the model simulations. Application of the approach to lumped and distributed rainfall runoff models for a variety of catchments shows that by assessing different catchment predictors the method can be a useful tool for prediction of catchment response.

One challenge that faces hydrologists in water resources planning is to predict the catchment???s response to a given rainfall. Estimation of parameter uncertainty (and model uncertainty) allows assessment of the risk in likely applications of hydrological models. Bayesian statistical inference, with computations carried out via Markov Chain Monte Carlo (MCMC) methods, offers an attractive approach to model specification, allowing for the combination of any pre-existing knowledge about individual models and their respective parameters with the available catchment data to assess both parameter and model uncertainty. This thesis develops and applies Bayesian statistical tools for parameter estimation, comparison of model performance and hierarchical model aggregation. The work presented has three main sections. The first area of research compares four MCMC algorithms for simplicity, ease of use, efficiency and speed of implementation in the context of conceptual rainfall-runoff modelling. Included is an adaptive Metropolis algorithm that has characteristics that are well suited to hydrological applications. The utility of the proposed adaptive algorithm is further expanded by the second area of research in which a probabilistic regime for comparing selected models is developed and applied. The final area of research introduces a methodology for hydrologic model aggregation that is flexible and dynamic. Rigidity in the model structure limits representation of the variability in the flow generation mechanism, which becomes a limitation when the flow processes are not clearly understood. The proposed Hierarchical Mixtures of Experts (HME) model architecture is designed to do away with this limitation by selecting individual models probabilistically based on predefined catchment indicators. In addition, the approach allows a more flexible specification of the model error to better assess the risk of likely outcomes based on the model simulations. Application of the approach to lumped and distributed rainfall runoff models for a variety of catchments shows that by assessing different catchment predictors the method can be a useful tool for prediction of catchment response.

When developing a Watershed Protection Plan (WPP) or a Total Maximum Daily Load (TMDL), it is often difficult to accurately assess the pollutant load for a watershed because not enough water quality monitoring data are available. According to the Texas Commission on Environmental Quality (TCEQ), there are 274 bacteria impairments in Texas water bodies out of 386 impaired water bodies. Bacteria water quality data are often more sparse than other types of water quality data, which hinders the development of WPPs or TMDLs. The Spatially Explicit Load Enrichment Calculation Tool (SELECT) was used to develop watershed protection plans for four rural watersheds in Texas that are impaired due to E. coli bacteria. SELECT is an automated Geographical Information System (GIS) tool that can assess pathogen loads in watersheds using spatial factors such as land use, population density, and soil type. WPPs were developed for four rural Texas watersheds: Buck Creek, Lampasas River, five sub watersheds of the Little Brazos River, and Geronimo Creek. A spatial watershed model was developed to simulate bacteria concentrations in streams resulting from non point sources using SELECT combined with a simple rainfall-runoff model and applied to the Geronimo Creek watershed. The watershed model applies a rainfall-driven loading function to the potential E. coli loads calculated by the output of SELECT. The simulated runoff volumes and E. coli concentrations from the model were compared to actual monthly E. coli data collected at two sampling sites near the outlet of a subwatershed. The results show how SELECT methodology was applied to each watershed and adapted based on stakeholder concerns and data availability. The highest potential contributors were identified and areas of concern were highlighted to more effectively apply best management practices (BMPs). The runoff volumes were predicted with very good agreement (E = 0.95, RSR = 0.21 to 0.22) for both sampling sites. The predicted E. coli concentrations did not agree with measured concentrations for both sites using eight different methods. The results indicate that the model does not include significant factors contributing to the transport of E. coli bacteria but can be modified to include these factors.

The requirements of ecological studies and water resource management plans are driving demand for hydrological models of the rivers of the arid zone. Knowledge of the hydrology of Australia’s arid zone is poor, yet is critical in understanding the ecology of the region. The research presented in this thesis seeks to address some shortcomings in our understanding of the hydrology of the Australian arid zone. In particular, the research examines the requirements for modelling the flow regime of arid zone rivers, concentrating on the rivers of the Lake Eyre Basin (LEB). The LEB has exceptionally low annual runoff of 3.5 mm, its major rivers develop over extremely low gradients and are characterised by very wide floodplains and complex anastomosing flow paths in their mid to lower reaches. This research was driven by both a practical and theoretical impetus. Practically, hydrological data were required at the water body scale for a large number of sites across three river systems of the LEB, for use in a study, known as ARIDFLO, of the ecological responses to hydrological conditions. Because of the remoteness of these sites and the paucity of gauging stations on these rivers, modelling of the rivers was the only method for delivering the required discharge data. Theoretically, the challenge was set for creating hydrological models for some extraordinarily complex river systems, in terms of their size, catchment characteristics and flow regime variability. (For complete abstract open document)

Title: Application of the Snowmelt Runoff Model for snow accumulation and snowmelt modelling in experimental catchments Bystřice and Zlatý Brook in the Krušné Mountains Modelling of hydrological processes is a dynamically developing part of hydrology. The Snowmelt Runoff Model (SRM) was applied for modelling the runoff in two experimental catchments Bystřice and Zlatý Brook. The aim of this thesis is to set up and calibrate SRM model and to evaluate methods and procedures used for runoff simulations. The SRM model was used for modelling snow accumulation and snowmelt in two selected catchments in the Krušné Mountains. The snow depths and snow water equivalents are measured since 2009 at selected locations situated in catchments. Calibration and validation of the model was based on continual time series of precipitation, air temperature and discharge measured 2009. Hydrological years 2009 and 2010 were used for model calibration and hydrological years 2011 and 2012 were used for model validation. Sensitivity analysis, which quantifies the effect of individual model parameters on the simulating proces, was carried out based on results. Recession coefficient and runoff coefficient belong to the most sensitive parameters with highest impact on runoff simulations. Model calibration was successful, which...

The aim of The Diploma Thesis was to compile a conceptual rainfall-runoff model, that would be eligible to model discharge in conditions of climate changes. After thorough verifications of possible variants, user program Runoff Prophet that is eligible to simulate discharge in closing profile of any river basin was compiled within this paper. Runoff Prophet is deterministic lumped model with monthly computation time step and from the hydrologic phenomena it takes soil moisture, evapotranspiration, groundwater flow and the watercourse flow into account. Its calibration is based on the differential evolution principle with Nash–Sutcliffe model efficiency coefficient as the calibration criterion. Developed software was tested on Vír I. catchment basin and the results of this probe were evaluated from viewpoints of air temperature, precipitation and discharge characteristics in the Dalečín measurement river cross section in distant future according to A1B SRES climate scenario, implemented in LARS-WG weather generator.

This thesis applies the runoff process in ostrovská Bystřice river catchment in Ore Mountains. For this purpose we used runoff precipitation model HEC-HMS. Our specific goal is to model the soil runoff, which is observed on eleven runoff precipitation events in the period 2009 - 2013. We analyzed basic physical characteristics of soils, which occur in observed river basin. The results were afterwards used to set up parameters of studied methods. We found out, that the moisture from antecedent precipitation influences modeling of soil runoff. The antecedent precipitation conditions the change of basic physical characteristic. We disregarded the influence of evapotranspiration and effect of vegetation on soil runoff. For the parameterization we have chosen two methods, which describe soil runoff. It was SCS CN and Green Ampt methods. These two methods have been compared on sample of resulting events. The methods were manually and automatically calibrating. The results showed on the insignificant difference between both observed methods. No better significant predicative capacity was manifested for one or other methods. Key worlds: soil runoff modeling, SCS CN, Green Ampt, HEC-HMS, infiltration

This work deals with the modelling of runoff processes in the experimental Bystřice river catchment in the Krušné Mountains. Rainfall-runoff model WaSiM (Topmodel version) was used for this purpose.Objective of this study was modelling snow accumulation and snowmelt in winter periods 2009/2010 and 2010/2011. Sensitivity analysis of TOPMODEL parameters was performed using Monte Carlo and GLUE methodology. Wasim was calibrated manually for 25 and 250 m grid scales and daily timestep. Results of simulations in both spatial scales different spatial scales differed significantly. Better performance of modelling in finer scale wasn't proved in validation of the model. Results of the simulations pointed out uncertainty in model calibration. Rasters of modeled snow water equivalent were also analyzed in selected days representing periods of snow accumulation and snowmelt. The goal was to examine the functioning of used snow model (combination of temperature index method and energy balance method) within the WaSiM frame. Finer scale raster proved to be practical for this purpose It was found that the spatial heterogeneity in the distribution of snow is determined by evapotranspiration and also by effect of radiation correction (in the case of rainfall).Influence of interception and vegetation effects on...

Project Completion Report, OWRT Project No. A-049-ARIZ / Agreement No. 14-31-0001-4003 / Project Dates: July 1, 1973 - June 30, 1974. / Acknowledgement: The work upon which this report is based was supported by funds provided by the United States Department of the Interior, Office of Water Research and Technology, as authorized under the Water Resources Research Act of 1964. / A non-linear reservoir model is used to represent the rainfall-runoff relationships for thunderstorms on the urban watersheds of Tucson, Arizona. Two types of computer programs are developed: a calibration program to obtain a best -fit calculated hydrograph; and a verification program to generate storm hydrographs given the watershed characteristics and a hyetograph. Calibration reveals the relationship of the model parameters, namely, (f) the inflow coefficient, (a) the constant coefficient, and (TL) the time lag, to the total rainfall, drainage area, channel length, and infiltration capacity of the watershed. The average discrepancy between the predicted hydrograph and the actual hydrograph for Tucson urban watersheds is 20 -25 percent.

主催：JST/CREST，Vrije University, ALTERRA, IBPC