In the lowlands of Nepal (Terai), the WHO drinking water guideline concentration of 10 μg/L for arsenic (As) is frequently exceeded. Since their introduction in 2006, iron-assisted bio-sand filters (Kanchan filters) are widely used to treat well water in Nepal. The filters are constructed on the basis of As-removal with corroding zero-valent iron (ZVI), with water flowing through a filter bed of iron nails placed above a sand filter. According to several studies, the performance of Kanchan filters varies greatly and depends on the size of the iron nails, filter design, water composition and operating conditions, leading to concerns about their actual efficiency. This study examined 38 Kanchan household filters for which insufficient As-removal was reported, to evaluate the reasons for limited removal efficiency and to define measures for improved performance. The measured arsenic removal ranged from 6.3% to 98.5 %. The most relevant factors were the concentrations of As and Fe in the raw water, with the best removal efficiency observed for water with low As (124 µg/l) and high Fe (4.94 mg/l). Although the concentrations of other elements, pH, flow rates, and contact time with ZVI also played a role, the combined evidence indicated that the reactivity of the frequently drying nail beds between filtrations was insufficient for efficient Asremoval. Optimized filters with added top layers of sand and raised water outlets with flow restrictions to keep nails permanently immersed and to increase contact times, should be able to achieve higher and more consistent arsenic removal efficiencies.

Pressure on water resources has reached unprecedented levels during the last decades because of climate change, industrialization, and population growth. As a result, vulnerability to inappropriate water availability and/or quality is increasing worldwide. In this paper, a Soil & Water Assessment Tool (SWAT) model of the Carp river watershed located in the city of Ottawa, Ontario was calibrated and validated. The model was then used to evaluate the individual and coupled impacts of urbanization and climate change on water quantity (discharge) and quality (nitrogen and phosphorus loads). While most of the watershed is currently rural, the headwaters will undergo rapid urbanization in the future, and there are concerns about possible negative impacts on water quantity and quality. Seven scenarios were developed to represent various watershed configurations in terms of land use and climate regime. Future climate time series were obtained by statistically downscaling the outputs of nine regional climate models, ran under representative concentration pathways (RCP)4.5 and RCP8.5. The impacts were evaluated at the main outlet and at the outlet of an upstream sub-watershed that would be most affected by urbanization. Results show that climate change and urbanization's impacts vary greatly depending on the spatial scale and geographic location. Globally, the annual average discharge will increase between 6.75 and 9.34% by 2050, while changes in annual average nitrogen and phosphorus loads will vary between −1.20 and 24.84%, and 19.15 and 23.81%, respectively. Local impacts in sub-watersheds undergoing rapid urbanization would be often much larger than watershed-scale impacts.

Water quality modeling is an important issue for both engineers and scientists. The QUAL2K model is a simulation tool that has been used widely for this purpose. Uncertainty and sensitivity analysis is a major step in water quality modeling. This article reports application of Monte Carlo analysis and classification tree sensitivity analysis in the modeling of the Zayandehrood River. First the model was calibrated and validated using two sets of data. Then, three input values (stream flow, roughness and decay rate) were considered for both analyses. The Monte Carlo analysis was conducted using triangular distribution of probability of occurrence for the input parameters. The classification tree analysis classifies outcome values into non-numeric categories. Considering the relationships between the input parameters in the classification tree analysis is the most important advantage of it. The analyses demonstrated the key input variables for three points of the river. The dissolved oxygen levels were mainly sensitive to the decay rate coefficient along the river.

No abstract