USGS Home
Contact USGS
Search USGS

U.S. GEOLOGICAL SURVEY
Scientific Investigations Report 2007–5159

Back to Table of Contents

Field Survey of the 2003 Flood

The HWMs and channel cross sections of the 2003 flood were surveyed in July–August 2004 at the same location used in the slope-area measurement made in 1923 (J.E. Stewart, U.S. Geological Survey, written commun., 1923). The site is a fairly straight reach of the Skagit River located downstream of the Dalles Gorge bedrock constriction, which is just downstream of the gaging station Skagit River near Concrete, Washington (station No. 121940000, fig. 1). A total station survey instrument was used to survey the cross sections and HWMs. During the cross-section survey, 2×2-inch wooden hubs were established near the site as survey control points based on Washington Department of Transportation survey monuments on the nearby highway. Elevation for the field survey was adjusted to the gage datum.

HWMs were particularly difficult to find near cross section 1 (labeled XS1 in fig. 1) and marks that were found were rated poor. Most of the marks were scour lines, flood-deposited sand, mud lines or small debris piles. Cross section 1 is near a rapidly expanding section just downstream of the Dalles Gorge, where it is difficult to assess the amount of energy loss due to expansion. Slope-area calculations are best applied to reaches where bed friction losses dominate and are less accurate when applied to reaches with expansions. Because only limited, poor quality HWMs could be located and the difficulties in estimating the energy loss due to expansion, XS1 was not used and only XS2 and XS3 were used in a two-section slope-area calculation to estimate the n value. Peak-flow discharge was determined from the stage-discharge rating for the gaging station supported by a current-meter measurement of 138,000 ft³/s on the day of the 2003 peak discharge. The water-surface elevation was well defined by HWMs at XS2, but there was about 2 ft of scatter of the good and excellent HWMs at XS3. Because of the uncertainty of the water-surface elevation at XS3, only a range of plausible n values (0.024 to 0.032) could be calculated for the 2003 flood.

Comparisons of channel cross sections and historical photographs (Mastin and Kresch, 2005) show that the channel has changed little in this particular reach of the Skagit River since 1923 except for the channel island (or channel bar at lower stages) crossed by XS2. This island/bar was likely barren and void of vegetation during the 1920s due to the series of large floods from 1897 to 1921. A series of historical photographs beginning in 1937 (historical photographs on file at the U.S. Geological Survey, Washington Water Science Center) shows a barren island/bar that has gradually increased in the amount of established vegetation. This island/bar is now densely forested.

Back to Table of Contents