Sustaining natural levels of base flow is critical to maintaining ecological function as stream catchments are urbanized. Stream base flow responds variably to urbanization. Base flow or water tables rise in some locations, fall in others, or remain constant. This variable response is the result of the array of natural (e.g., physiographic setting and climate) and anthropogenic (e.g., urban development and infrastructure) factors that influence hydrology. Perhaps because of this complexity, few simple tools exist to assist managers to predict baseflow change in their local urban area. We address this management need by presenting a decision-support tool that can be used to predict the likelihood and direction of baseflow change based on the natural vulnerability of the landscape and aspects of urban development. When the tool indicates a likely increase or decrease, managers can use it for guidance toward strategies that can reduce or increase groundwater recharge, respectively. An equivocal result from application of the tool suggests the need for a detailed water balance. The tool is embedded in an adaptive-management framework that encourages managers to define their ecological objectives, assess the vulnerability of their ecological objectives to changes in water-table height, and monitor baseflow responses to urbanization. We tested our framework with 2 different case studies: Perth, Western Australia, Australia and Baltimore, Maryland, USA. Together, these studies show how predevelopment water-table height, climate, and geology together with aspects of urban infrastructure (e.g., stormwater practices, leaky pipes) interacted such that urbanization led to rising (Perth) and falling (Baltimore) base flow. Greater consideration of subsurface components of the water cycle will help to protect and restore the ecology of urban fresh waters.

A novel form of urbanization, low impact development (LID), aims to engineer systems that replicate natural hydrologic functioning, in part by infiltrating stormwater close to the impervious surfaces that generate it. We sought to statistically evaluate changes in a base flow regime because of urbanization with LID, specifically changes in base flow magnitude, seasonality, and rate of change. We used a case study watershed in Clarksburg, Maryland, in which streamflow was monitored during whole‐watershed urbanization from forest and agricultural to suburban residential development using LID. The 1.11‐km²watershed contains 73 infiltration‐focused stormwater facilities, including bioretention facilities, dry wells, and dry swales. We examined annual and monthly flow during and after urbanization (2004–2014) and compared alterations to nearby forested and urban control watersheds. We show that total streamflow and base flow increased in the LID watershed during urbanization as compared with control watersheds. The LID watershed had more gradual storm recessions after urbanization and attenuated seasonality in base flow. These flow regime changes may be because of a reduction in evapotranspiration because of the overall decrease in vegetative cover with urbanization and the increase in point sources of recharge. Precipitation that may once have infiltrated soil, been stored in soil moisture to be eventually transpired in a forested landscape, may now be recharged and become base flow. The transfer of evapotranspiration to base flow is an unintended consequence to the water balance of LID. © 2016 The Authors Hydrological Processes Published by John Wiley & Sons Ltd.

Urban water balances are generally unknown, yet they are necessary for assessing water availability in an urbanizing world and for understanding the effects of urbanization on the hydrologic cycle. We assess the spatial and temporal variability of water balances of 65 watersheds in the Baltimore, MD, metropolitan area during Water Years 2001–2009. Each water balance term is quantified independently and includes both natural (precipitation, evapotranspiration, streamflow) and piped (sewer infiltration and inflow [I&I], lawn irrigation, water supply pipe leakage) watershed inflows and outflows. The analysis also compares Gravity Recovery and Climate Experiment storage data with changes in storage calculated using the residual between watershed inflows and outflows. We find that when considering only natural water balance terms, the water balance residual (inflows minus outflows) increases with urbanization, largely as a result of the decrease in evapotranspiration as modeled by the land surface model Global Land Data Assimilation System/Noah. During wet years, the difference between urban and rural natural water balance residuals narrows because of increased urban streamflow. Excess water of the natural water balance in urban areas is largely exported by I&I into wastewater collection pipes; for some urban watersheds this excess is greater than gauged annual streamflow. I&I also outweighs piped inputs from lawn irrigation and water supply pipe leakage in the Baltimore area analysis. The net effect of piped flows on the urban water balance is a watershed export ranging between 300 and 465 mm/yr, underscoring the importance of interactions between engineered water and wastewater infrastructure and natural water balance components.

The impact of urban development on surface flow has been studied extensively over the last half century, but effects on groundwater systems are still poorly understood. Previous studies of the influence of urban development on subsurface storage have not revealed any consistent pattern, with results showing increases, decreases, and negligible change in groundwater levels. In this paper, we investigated the effects of four key features that impact subsurface storage in urban landscapes. These include reduced vegetative cover, impervious surface cover, infiltration and inflow (I&I) of groundwater and storm water into wastewater pipes, and other anthropogenic recharge and discharge fluxes including water supply pipe leakage and well and reservoir withdrawals. We applied the integrated groundwater‐surface water‐land surface model ParFlow.CLM to the Baltimore metropolitan area. We compared the base case (all four features) to simulations in which an individual urban feature was removed. For the Baltimore region, the effect of infiltration of groundwater into wastewater pipes had the greatest effect on subsurface storage (I&I decreased subsurface storage 11.1% relative to precipitation minus evapotranspiration after 1 year), followed by the impact of water supply pipe leakage and lawn irrigation (combined anthropogenic discharges and recharges led to a 7.4% decrease) and reduced vegetation (1.9% increase). Impervious surface cover led to a small increase in subsurface storage (0.56% increase) associated with decreased groundwater discharge as base flow. The change in subsurface storage due to infiltration of groundwater into wastewater pipes was largest despite the smaller spatial extent of surface flux modifications, compared to other features.