This indicator examines three important measures of streamflow conditions that occur over the course of a year. Figure 1 looks at the driest conditions each year, which are commonly calculated by averaging the lowest seven consecutive days of streamflow over the calendar year. Figure 2 examines high flow conditions, which are commonly calculated as the highest average flow over three consecutive days. Hydrologists use this measure because a three-day averaging period has been shown to effectively characterize runoff associated with storms and peak snowmelt throughout a range of watershed areas.
Figure 3 shows changes in the timing of spring runoff over time. This measure is limited to 78 stream gauges in areas where at least 30 percent of annual precipitation falls as snow. Scientists look at the total volume of water that passes by a gauge between January 1 and May 31, then determine the date when half of that water has gone by. This date is called the winter-spring center of volume date. A long-term trend toward an earlier date suggests that spring snowmelt is happening earlier. Figure 3 covers a shorter period of time than Figures 1 and 2 (50 years instead of 70 years) because of data availability.
- In the past 70 years, seven-day low flows have generally increased in the Northeast and Midwest (in other words, on the driest days, streams are carrying more water than before) in parts of the Southeast and the Pacific Northwest. Overall, more sites have seen increases than decreases. The graph shows the influence of widespread droughts—for example, during the 1950s and 1960s (see fig. 1).
- Three-day high flow trends vary from region to region across the country. For example, streams in the Northeast have generally seen an increase or little change in high flows since 1940, while some West Coast streams have seen a decrease and others have seen an increase (see Figure 2). Overall, more sites have seen increases than decreases.
- More than half of the streams measured show winter-spring runoff happening at least 2 days earlier than it did in the past. The most dramatic change has occurred in the Northeast(see Figure 3).
Measurements were taken in areas where streamflow is not highly affected by human influences, including land cover change. However, changes in land cover and land use over time could still influence streamflow trends at some streams. The gauges used for this indicator are not evenly distributed across the country.
Streamflow data were collected by the U.S. Geological Survey. These data came from a set of watersheds with minimal human impacts, which have been classified as reference gages. Daily average streamflow data are stored in the National Water Information System and are publicly available at: http://waterdata.usgs.gov/nwis/
Figure 1. This map shows changes in the minimum amount of water carried by rivers and streams
across the United States from 1940 to 2009. Minimum streamflow is based on the consecutive
seven-day period with the lowest average flow during a given calendar year. The timeline shows the
median 7-day low flow streamflow for all sites for each year, which means half of the streams
measured had higher flow than the amount shown, and half had lower flow. The graph shows
streamflow in terms of inches of runoff. The solid line represents a five-year moving average.
Figure 2. This map shows changes in the maximum amount of water carried by rivers and streams across
the United States from 1940 to 2009. Maximum streamflow is based on the consecutive three-day period
with the highest average flow during a given calendar year. The timeline shows the median three-day
high streamflow for all sites for each year, which means half of the streams measured had higher flow
than the amount shown, and half had lower flow. The graph shows streamflow in terms of inches of
runoff. The solid line represents a five-year moving average.
Figure 3. This map shows changes in the timing of peak spring flow carried by rivers and streams from
1940 to 2009. This analysis focuses on parts of the United States where streamflow is strongly influenced
by snowmelt. It is based on the winter-spring center of volume, which is the date when half of the
streamflow that occurred between January 1st and May 31st of each year has passed. The timeline
shows the median date for all sites for each year, which means half of the streams measured had their
center of volume occur before this date, and half occurred after. The solid line represents a five-year