This study shows a comprehensive simulation of water and sediment fluxes from the catchment to the reach scale. We describe the application of a modelling cascade in a well researched study catchment through connecting stateof-the-art public domain models in ArcGIS. Three models are used consecutively: (1) the hydrological model SWAT to evaluate water balances, sediment input from fields and tile drains as a function of catchment characteristics; (2) the onedimensional hydraulic model HEC-RAS to depict channel erosion and sedimentation along a 9 km channel onedimensionally; and (3) the two-dimensional hydraulic model AdH for simulating detailed substrate changes in a 230 m long reach section over the course of one year. Model performance for the water fluxes is very good, sediment fluxes and substrate changes are simulated with good agreement to observed data. Improvement of tile drain sediment load, simulation of different substrate deposition events and carrying out data sensitivity tests are suggested as future work. Main advantages that can be deduced from this study are separate representation of field, drain and bank erosion processes; shown adaptability to lowland catchments and transferability to other catchments; usability of the model’s output for habitat assessments.
Development of any area often leads to more intensive land use and increase in the generation of pollutants. Modeling these changes is critical to evaluate emerging changes in land use and their effect on stream water quality. The objective of this study was to assess the impact of spatial patterns in land use and population density on the water quality of streams, in case of data scarcity, in the Chugoku district of Japan. The study employed artificial neural network (ANN) technique to assess the relationship between the total phosphorous (TP) in river water and the land use in 21 river basins in the district, and the model was able to reasonably estimate the TP in the stream water. Uncertainty analysis of ANN estimates was performed using the Monte Carlo framework, and the results indicated that the ANN model predictions are statistically similar to the characteristics of the measured TP values. It was observed that any reduction in forested area or increase in agricultural land in the watersheds may cause the increase of TP concentration in the stream. Therefore, appropriate watershed management practices should be followed before making any land use change in the Chugoku district. and Rozvoj územia často súvisí so zintenzívnením využívania krajiny a produkciou znečistenia. Dôležité je modelovanie týchto zmien a ich vplyvu na kvalitu vody v tokoch. Cieľom štúdie je určiť vplyv priestorových zmien pri využívaní krajiny a zmeny hustoty osídlenia na kvalitu vody v tokoch v čase nedostatku vody v oblasti Chugoku, Japonsko. Pri riešení sa využívajú umelé neurónové siete (artificial neural network -ANN), prostredníctvom ktorých sa určuje vzťah medzi celkovým obsahom fosforu (TP) v toku a využívaním kajiny v 21 povodiach oblasti; tento model je schopný vypočítať TP v tokoch. Analýza neurčitosti výsledkov dosiahnutých pomocou ANN bola vykonaná metódou Monte Carlo; výsledky analýzy naznačujú, že predpovede pomocou metódy ANN sú štatisticky podobné meraným hodnotám TP. Bolo zistené, že redukcia lesnatosti a zvýšenie plochy poľnohospodársky využívanej pôdy v povodí môže viesť k zvýšeniu koncentrrácie TP v toku. Je preto potrebné pred zmenou vo využívaní krajiny prijať zodpovedajúce opatrenia v manažmente krajiny, ktoré budú minimalizovať negatívne dôsledky zmien využívania krajiny.
Extensive lowland floodplains cover substantial parts of the glacially formed landscape of Northern Germany. Stream power is recognized as a force of formation and development of the river morphology and an interaction system between channel and floodplain. In order to understand the effects of the river power and flood power, HEC-RAS models were set up for ten river sections in the Upper Stör catchment, based on a 1 m digital elevation model and field data, sampled during a moderate water level period (September, 2011), flood season (January, 2012) and dry season (April, 2012). The models were proven to be highly efficient and accurate through the seasonal roughness modification. The coefficients of determination (R2 ) of the calibrated models were 0.90, 0.90, 0.93 and 0.95 respectively. Combined with the continuous and long-term data support from SWAT model, the stream power both in-channel and on the floodplain was analysed. Results show that the 10-year-averaged discharge and unit stream power were around 1/3 of bankfull discharge and unit power, and the 10-year-peak discharge and unit stream power were nearly 1.6 times the bankfull conditions. Unit stream power was proportional to the increase of stream discharge, while the increase rate of unit in-channel stream power was 3 times higher than that of unit stream power on the floodplain. Finally, the distribution of the hydraulic parameters under 10-years-peak discharge conditions was shown, indicating that only 1-10% of flow stream was generated by floodplain flow, but 40-75% volume of water was located on the floodplain. The variation of the increasing rate of the stream power was dominated by the local roughness height, while the stream power distributed on the floodplain mainly depended on the local slope of the sub-catchment.