EPA's Report on the Environment
High and Low Stream Flows
Flow is a critical aspect of the physical structure of stream ecosystems (Poff and Allan, 1995; Robinson et al., 2002). High flows shape the stream channel and clear silt and debris from the stream, and some fish species depend on high flows for spawning. Low flows define the smallest area available to stream biota during the year. In some cases, the lowest flow is no flow at all—particularly in arid and semi-arid regions where intermittent streams are common. Riparian vegetation and aquatic life in intermittent streams have evolved to complete their life histories during periods when water is available; however, extended periods of no flow can still impact their survival (Fisher, 1995). Changes in flow can be caused by dams, water withdrawals, ground water pumping (which can alter base flow), changes in land cover (e.g., deforestation or urbanization), and weather and climate (Calow and Petts, 1992).
This indicator, originally developed by the Heinz Center (2008), describes trends in stream flow volumes based on daily flow data collected by the U.S. Geological Survey’s (USGS’s) nationwide network of stream flow gauging sites from 1961 to 2015.
The first part of this indicator describes trends in high flow volume, low flow volume, and variability of flow in streams throughout the contiguous 48 states, relative to a baseline period of 1941-1960. Data were collected at two sets of USGS stream gauging stations: a set of approximately 700 “reference” streams that have not been substantially affected by dams and diversions and have had little change in land use over the measurement period, and a separate set of approximately 1,100 “non-reference” streams that reflect a variety of conditions (the exact number of sites with sufficient data varies from one metric to another). The indicator is based on each site’s annual 3-day high flow volume, 7-day low flow volume, and variability (computed as the difference between the 1st and 99th percentile 1-day flow volumes in a given year, divided by the median 1-day flow). Annual values for each metric were examined using a rolling 5-year window to reduce the sensitivity to anomalous events. For each site, the median value for the 5-year window was compared to the median value for the 1941-1960 baseline period. The indicator shows the proportion of sites where high flow, low flow, or variability of flow was more than 30 percent higher or 30 percent lower than the baseline. It also shows differences of more than 60 percent.
This indicator also examines no-flow periods in streams in grassland and shrubland areas of the contiguous 48 states. Data represent 244 USGS “reference” and “non-reference” stream gauging sites in watersheds with at least 50 percent grass or shrub cover, as defined by the 2001 National Land Cover Database (NLCD) (MRLC Consortium, 2007). The indicator reports the percentage of these streams with at least one no-flow day in a given year, averaged over a rolling 5-year window. Results are displayed for all grassland/shrubland streams, as well as for three specific ecoregion divisions (Bailey, 1995). This indicator also reports on the duration of no-flow periods. For a subset of 133 grassland/shrubland streams that had at least one no-flow day during the study period, the duration of the maximum no-flow period in each year was averaged over a rolling 5-year window and compared with the average no-flow duration for the same site during the 1941-1960 baseline period. A no-flow period more than 14 days longer than the baseline was described as a “substantial increase”; a no-flow period more than 14 days shorter than the baseline was classified as a “substantial decrease.”
What the Data Show
In an average year during the period of record, roughly 20 percent of streams had increases in high flow volume of more than 30 percent, relative to the 1941-1960 baseline (Exhibit 1). A similar percentage had decreases of more than 30 percent (Exhibit 1). Large fluctuations in high flow volume are apparent over time, with both sets of trends suggesting relatively wet periods in the early 1980s and mid-1990s and relatively dry periods around 1990 and the early 2000s. Reference and non-reference stream sites show similar patterns, although larger decreases in high flow volume were more common in the non-reference streams.
Since the early 1960s, more streams have shown increases in low flow volumes than have shown decreases, relative to the 1941-1960 baseline period (Exhibit 2). Among the many streams with larger low flows are a few (1 to 6 percent in an average year) with increases of more than 600 percent. Fluctuations over time are apparent, and while not as pronounced as the shifts in high flow (Exhibit 1), they generally tend to mirror the same relatively wet and dry periods since 1980. Reference and non-reference streams show similar low flow patterns over time, but reference sites are less likely to have experienced decreases in low flow.
Except for a few brief periods in the mid-1960s and again around 1980, decreased flow variability has been much more common than increased variability (Exhibit 3). Reference and non-reference streams have shown similar patterns in variability over time, although reference streams were slightly less likely to experience changes overall.
In areas with primarily grass or shrub cover, roughly 15 to 20 percent of stream sites typically have experienced periods of no flow in a given year (Exhibit 4). Overall, the number of streams experiencing no-flow periods has declined slightly since the 1960s. Streams in the California/Mediterranean ecoregion have shown the greatest decrease in no-flow frequency, but they still experience more no-flow periods than streams in the other two major grassland/shrubland ecoregion divisions. Among grassland/shrubland streams that have experienced at least one period of no flow since 1941, more streams have shown a substantial decrease in the duration of no-flow periods (relative to the 1941-1960 baseline) than a substantial increase (Exhibit 5).
The 1941-1960 baseline period was chosen to maximize the number of available reference sites and should provide a sufficiently long window to account for natural variability (Heinz Center, 2008); however, it does not necessarily reflect “undisturbed” conditions. Many dams and waterworks had already been constructed by 1941, and other anthropogenic changes (e.g., urbanization) were already widespread.
Although the sites analyzed here are spread widely throughout the contiguous U.S., gauge placement by USGS is not a random process. Gauges are generally placed on larger, perennial streams and rivers, and changes seen in these larger systems may differ from those seen in smaller streams and rivers.
- This indicator does not characterize trends in the timing of high and low stream flows, which can affect species migration, reproduction, and other ecological processes.
Stream flow measurements can be obtained from the USGS National Water Information System (NWIS) database (USGS, 2016) (http://waterdata.usgs.gov/nwis). This indicator is based on stream flow records that USGS extracted from NWIS and provided to EPA.
For More Information
This page provides links to non-EPA websites that provide additional information about this topic. You will leave the EPA.gov domain, and EPA cannot attest to the accuracy of information on that non-EPA page. Providing links to a non-EPA website is not an endorsement of the other site or the information it contains by EPA or any of its employees. Also, be aware that the privacy protection provided on the EPA.gov domain (see Privacy and Security Notice) may not be available at the external link.
You will need the free Adobe Reader to view some of the files on this page. See EPA's PDF page to learn more.