Characterizing hydrologic alterations following urbanization through time and across space among U.S. cities

Background/Question/Methods

Processes leading to stream impairment differ among cities because of regional differences in physical environment and development history. Therefore, reducing negative urban impacts can benefit from comparative study. Urban infrastructure patterns and growth rates vary spatially and temporally within and among cities, with the resulting hydrologic impacts of urbanization also varying, and compromising aquatic ecosystem integrity in different ways. Here we evaluate the long-term hydrologic impacts of urbanization at two temporal scales. Seven cities with contrasting development histories were selected: Atlanta (GA), Baltimore (MD), Boston (MA), Pittsburgh (PA), Phoenix (AZ), Portland (OR), and Raleigh (NC). We characterized hydrologic responses across urbanization gradients by (1) contrasting short-term hydrologic trends from 2000-2012 across the cities; and (2) examining long-term (>40 yrs) hydrologic shifts in a subset of catchments within these cities. Urbanization gradients were determined using a suite of landscape metrics including road, housing, and population densities. Hydrologic change was characterized using USGS stream flow records and a suite of indices of hydrologic alteration, including the days flow is below/above 25^th percentile flow (high/low pulse events) and number of flow switches between rising and falling periods (flow reversals). 

Results/Conclusions

Across each city’s urbanization gradient, high and low flow frequency and number of flow reversals generally increased with urbanization, while the duration of low and high flow periods decreased with urbanization. Changes in baseflow with urbanization were less consistent among cities. Comparisons of relationships between high flow frequency and road density suggest the hydrologic impacts of urbanization were strongest in Raleigh, Atlanta, and Baltimore. Times series analysis revealed long-term shifts in the frequency of both high-flow events and reversals in urbanizing catchments. For example, a Pittsburgh basin that urbanized rapidly from 1960-1970, demonstrated an increase in the frequency of high pulses, from an average of 16 to 26 per year. Similar shifts in high pulse frequency occurred in other cities where urbanization occurred less abruptly, but the magnitude of shifts varied across cities. Inter-city variation likely depends on precipitation flux and seasonality, rate of urbanization, and pattern of urbanization. Our results suggest additional landscape metrics, such as infrastructure connectivity, development growth rate, and neighborhood age, can clarify how urbanization degrades aquatic ecosystems through changes in ecologically relevant hydrologic events.