Flow Velocity Effects on Fe(III) Clogging during Managed Aquifer Recharge Using Urban Storm Water

AbstractThis study focuses on the Barind tract, a drought prone area situated in the north-west region of Bangladesh where inadequate rainfall and limited surface water have created high dependence on groundwater for irrigation and other purposes, leading to significant declines in groundwater level. Managed aquifer recharge (MAR) offers a potential solution to restore groundwater levels. This study sets out to identify the opportunities and challenges for implementing MAR in the Barind tract. To accomplish this aim, different data sets including bore log lithology, rainfall, groundwater levels, information about re-excavated ponds, dighis, kharies, beels, check dams, rubber dams, dug wells and other necessary information were collected from the Barind Multipurpose Development Authority (BMDA) and other sources and analyzed. Major opportunities for MAR are identified for about 2000 km of re-excavated kharies (canals) containing about 750 check dams, more than 3000 re-excavated ponds, a number of beels (comparatively large marshes) and other water bodies which are used to conserve runoff storm water for supplementary irrigation. The conserved water can be used for groundwater recharge and subsequently abstracted for irrigation. Furthermore, roof-top rain water from buildings can also be used for groundwater recharge purposes. In contrast, the major challenges include the high turbidity of storm water runoff leading to clogging of MAR structures, inadequacy of conventional direct surface methods of recharge due to the presence of a 15 m or more thick upper clay layer with limited percolation capacity, and lack of practical knowledge on MAR. Therefore, overcoming the challenges for MAR application is a prerequisite to maximize the opportunities from MAR that can support the sustainable use of groundwater resources.

Download Full-text

Evaluating a Sewershed Urban Storm Water Model for Variability in Parameter Sensitivity and Resolution Effects

10.31224/osf.io/u5tsz ◽

2020 ◽

Author(s):

Zhaokai Dong ◽

Daniel Bain ◽

Murat Akcakaya ◽

Carla Ng

Keyword(s):

Pipe Flow ◽

Peak Flow ◽

Model Performance ◽

Parameter Sensitivity ◽

Geometric Parameters ◽

Storm Water ◽

Parameter Estimates ◽

Kappa Statistics ◽

Thiessen Polygon ◽

Urban Storm

A high-quality parameter set is essential for reliable stormwater models. Model performance can be improved by optimizing initial parameter estimates. Parameter sensitivity analysis is a robust way to distinguish the influence of parameters on model output and efficiently target the most important parameters to modify. This study evaluates efficient construction of a sewershed model using relatively low-resolution (e.g., 30 meter DEM) data and explores model sensitivity to parameters and regional characteristics using the EPA’s Storm Water Management Model (SWMM). A SWMM model was developed for a sewershed in the City of Pittsburgh, where stormwater management is a critical concern. We assumed uniform or log-normal distributions for parameters and used Monte Carlo simulations to explore and rank the influence of parameters on predicted surface runoff, peak flow, maximum pipe flow and model performance, as measured using the Nash–Sutcliffe efficiency metric. By using the Thiessen polygon approach for sub-catchment delineations, we substantially simplified the parameterization of the areas and hydraulic parameters. Despite this simplification, our approach provided good agreement with monitored pipe flow (Nash–Sutcliffe efficiency: 0.41 – 0.85). Total runoff and peak flow were very sensitive to the model discretization. The size of the polygons (modeled subcatchment areas) and imperviousness had the most influence on both outputs. The imperviousness, infiltration and Manning’s roughness (in the pervious area) contributed strongly to the Nash-Sutcliffe efficiency (70%), as did pipe geometric parameters (92%). Parameter rank sets were compared by using kappa statistics between any two model elements to identify generalities. Within our relatively large (9.7 km^2) sewershed, optimizing parameters for the highly impervious (>50%) areas and larger pipes lower in the network contributed most to improving Nash–Sutcliffe efficiency. The geometric parameters influence the water quantity distribution and flow conveyance, while imperviousness determines the subcatchment subdivision and influences surface water generation. Application of the Thiessen polygon approach can simplify the construction of large-scale urban storm water models, but the model is sensitive to the sewer network configuration and care must be taken in parameterizing areas (polygons) with heterogenous land uses.

Download Full-text