Annual runoff in the Nation’s rivers and streams during water year 2017 (10.86 inches) was higher than the long-term (1930–2017) mean annual runoff of 9.33 inches for the contiguous United States (fig. 1). Nationwide, the 2017 streamflow ranked 10th highest out of the 88 years.

Figure 1. Annual runoff in the United States, 1930-2017

Record low streamflow levels were reported in Alaska (fig. 2). Streamflow was below normal in Hawaii, Georgia, New Jersey, Connecticut, and Tennessee. Streamflow was above normal in Illinois, Vermont, Utah, Montana, Washington, New York, Iowa, Indiana, Idaho, and Puerto Rico and the Virgin Islands. Streamflow was much above normal in Minnesota, Oregon, Nevada, Michigan, California, Wisconsin, and Wyoming. Most States had streamflow in the normal range.

Figure 2. Statewide streamflow ranks of the United States for 2017 relative to 1930-2017 mean annual streamflow.[**For Puerto Rico-Virgin Islands, 74 years of available data were used and the rank was adjusted accordingly]

The United States (including Puerto Rico and the Virgin Islands and the District of Columbia) is divided into 21 large drainages, or water-resources regions (fig. 3). These water-resources regions are based on surface topography and contain the drainage area of a major river; the combined drainage areas of a series of rivers, such as the Texas-Gulf region, which includes several rivers draining into the Gulf of Mexico; or the area of an island or island group. Water-resources regions provide a coherent, watershed-based framework for depicting streamflow variations.

Figure 3. Water-resources regions of the United States.

Streamflow was reported at record low levels in the Alaska region (fig. 4). Below normal streamflow was reported in the Hawaii and Tennessee regions. Streamflow was above normal in the Souris-RedRainy, Caribbean, and Upper Mississippi regions. Much above normal streamflow was reported in the Great Basin, Pacific Northwest, California, and Great Lakes regions.

Figure 4. Regional streamflow ranks in the United States for 2017 relative to 1930-2017 mean annual streamflow.[**For Puerto Rico-Virgin Islands, 74 years of available data were used and the rank was adjusted accordingly]

Autumn (October– December 2016) streamflow was much below normal in Alabama, Alaska, Georgia, New Jersey, and Mississippi (fig. 5). Streamflow was below normal in Connecticut, Massachusetts, the District of Columbia, Maine, New Hampshire, Rhode Island, and Tennessee. Above normal streamflow was reported in North Carolina, Oregon, Wyoming, California, Washington, South Dakota, North Dakota, Puerto Rico and the Virgin Islands, and Wisconsin. Much above normal streamflow was reported in Iowa, Montana, and Minnesota. Nationwide, autumn-season streamflow ranked 52d highest out of 88 years.

Figure 5. Autumn (October-December 2016) statewide ranks in the United States relative to 1930-2017 mean annual streamflow. [**For Puerto Rico-Virgin Islands, 74 years of available data were used and the rank was adjusted accordingly]

Winter (January–March 2017) streamflow was much below normal in North Carolina, the District of Columbia, Alaska, and Mississippi (fig. 6). Streamflow was below normal in Connecticut, New Jersey, Tennessee, Virginia, Maryland, South Carolina, Missouri, Alabama, Oklahoma, Arkansas, Delaware, Florida, Georgia, Kentucky, Rhode Island, Hawaii, and Louisiana. Above normal streamflow was reported in Vermont, Puerto Rico and the Virgin Islands, New York, Arizona, Washington, New Mexico, North Dakota, Iowa, Michigan, and Utah. Streamflow was much above normal in Colorado, Montana, Oregon, Idaho, Wisconsin, California, Minnesota, Nevada, and Wyoming. Nationwide, winter-season streamflow ranked 32d highest out of 88 years.

Figure 6. Winter (January-March 2017) statewide ranks in the United States relative to 1930-2017 mean annual streamflow. [**For Puerto Rico-Virgin Islands, 74 years of available data were used and the rank was adjusted accordingly]

Spring (April–June 2017) streamflow was much below normal only in Alaska (fig. 7). Below normal streamflow was reported in Arizona and Georgia. Above normal streamflow was reported in Oklahoma, Iowa, Pennsylvania, Mississippi, Ohio, Alabama, Vermont, Kansas, Massachusetts, Montana, Arkansas, Nevada, New Mexico, Rhode Island, Oregon, Louisiana, Washington, New York, North Carolina, Maine, Virginia, California, and New Hampshire. Streamflow much above normal was reported in Illinois, Idaho, Indiana, Michigan, Wisconsin, Missouri, and Wyoming. Nationwide, spring-season streamflow ranked fourth highest out of 88 years.

Figure 7. Spring (April-June 2017) statewide ranks in the United States relative to 1930-2017 mean annual streamflow. [**For Puerto Rico-Virgin Islands, 74 years of available data were used and the rank was adjusted accordingly]

Summer (July–September 2017) streamflow was much below normal only in Alaska (fig. 8). Below normal streamflow was reported in Hawaii, Arizona, Maine, and Washington. Above normal streamflow was observed in Missouri, Kentucky, Oklahoma, Wyoming, Nevada, Tennessee, Maryland, Vermont, New York, Ohio, Alabama, Arkansas, Wisconsin, the District of Columbia, and Indiana. Much above normal streamflow was reported in Florida, Michigan, Mississippi, California, Puerto Rico and the Virgin Islands, Louisiana, New Mexico, and Texas. Nationwide, summer-season streamflow ranked seventh highest out of 88 years.

Figure 8. Summer (July-September 2017) statewide ranks in the United States relative to 1930-2017 mean annual streamflow. [**For Puerto Rico-Virgin Islands, 74 years of available data were used and the rank was adjusted accordingly]

Assuming that individual streamgages act independently of each other, it is expected that the average streamflow at 5 percent of the streamgages will be high (greater than the 95th percentile) and 5 percent will be low (less than the 5th percentile) in any given month. The percentages of streamgages reporting high streamflow in 8 months of water year 2017 (October 2016 and January, February, March, April, May, July, and August 2017) were higher than expected (8, 10, 13, 12, 9, 11, 7, and 6 percent, respectively; fig. 9). In contrast, there were 3 months (November, February, and March) with a greaterthan-expected percentage of streamgages with low flows (7, 8, and 7 percent, respectively).

Figure 9. Percentage of streamgages with very high and very low monthly streamflow, October 1999-September 2017.