9331 - Laserfiche WebLink

Home Browse Search

1986; Lyons et aI., 1992; Grams, 1997) than other inter-regional streams (Burkham, 1972; Van Steeter and Pitlick, 1998). These studies have been based on analysis of historical ground and aerial photographs, some dendrogeomorphic and sedimentologic analyses, and sediment mass balance calculations. However, the temporal res- olution of the timing and rate of narrowing of the postdam channel has only been as precise as the length of the time intervals between aerial photo- graph series. The studies of the Green River have demon- strated that significant channel narrowing occurred during the 1930s and 1940s when saltcedar (Tamarix spp.) colonized channel bars at a time when flood magnitudes were low (Graf, 1978), but the precise time when narrowing began and the duration of this period of channel adjustment is un- known. Graf (1978) suggested that the channel reached a stable width following saltcedar inva- sion, but the duration of this period of equilibrium is unknown. A renewed period of narrowing oc- curred after Flaming Gorge Dam was completed in October 1962 (Andrews, 1986), but the magni- tude of postdam narrowing in relation to previous channel adjustments is unknown. Grams (1997) showed that narrowing also occurred within debris fan-affected canyons of the eastern Uinta Moun- tains; the other studies were conducted in alluvial reaches with wide or narrow valleys. Lyons et al. (1992) used aerial photographs to measure chan- nel widening in alluvial reaches caused by large floods that occurred in the mid-1980s. In general, these studies suggest that the magnitude of chan- nel narrowing caused by saltcedar invasion was greatest in the downstream parts of the river sys- tem, that the magnitude of channel narrowing caused by Flaming Gorge Dam has been greatest in alluvial reaches near the dam, and that the onset of dam-induced narrowing began more recently in reaches further from the dam. We sought to further clarify the rate, timing, and processes of channel narrowing of the Green River during the twentieth century by analyzing the 2600 measurements of channel shape made by the U.S. Geological Survey at its gaging sta- tion near the town of Green River, Utah (station number 09315000), located 475 Ian downstream from Flaming Gorge Dam (Fig. 1). This is a site where the relative influences of saltcedar inva~ sion and dam-induced hydrologic change can be distinguished, because saltcedar established itself here about 30 years prior to completion of Flam- ing Gorge Dam. Clover and Jotter (1944) identi- fied saltcedar near the town of Green River in 1938, and Christensen (1962) concluded that saltcedar were "present," but not "widespread," along the Green River between 1933 and 1939. We supplemented the analysis of gaging data 1758 ALLRED AND SCHMIDT with analysis of historical ground and aerial pho- tographs, measurements of sediment transport at a high discharge, and sedimentologic and den- drogeomorphic analyses. GREEN RIVER NEAR GREEN RIVER, UTAH The study area is the wide alluvial valley of the Green River between the Book Cliffs to the north and the Green River Desert to the south (Fig. 1). Between the Book Cliffs and the town of Green River, the river valley occurs within the Creta- ceous Mancos Shale and is 4 Ian wide in places (Sable, 1956). The topographic depression along the base of the Book Cliffs that follows the strike of the Mancos Shale is locally known as the Gun- nison Valley; the axis of this valley is perpendicu- lar to the course of the Green River. Downstream from the town, the alluvial valley of the Green River is much narrower, and it is incised into the Jurassic Morrison Formation. About 5 Ian down- stream from the present cableway of the U.S. Ge- ological Survey, the Mancos Shale is downfaulted to river level and valley width is more similar to that upstream from the town. Gravel terraces occur at the margins of the al- luvial valley, and these are a source of gravel to the Green River. Other gravel may be transported downstream from Grey Canyon, which occurs upstream from the Book Cliffs. Most of the study reach has a gravel bed that is mantled in places by sand and silt. Channel width exceeds 100 m at most locations within the reach. Channel slope in the study area is variable but is typically between 0.0003 and 0.0005. Several islands within the study reach have persisted since at least 1916 (Herron, 1917). The Green River near Green River, Utah, car- ries a large suspended load (loms et al., 1964). Typical concentrations during spring runoff range from 1000 mg/! to 3000 mgll. Higher loads are often carried during summer rain events when suspended sediment concentrations as high as 25 500 mgll have been reported, and daily sus- pended sediment discharges have been calculated . to be as high as 749500 metric tons per day. An- nual suspended sediment loads can approach 30 million metric tons. METHODS Analysis of Discharge Measurement Data Although discharge of the Green River near the town of Green River has been measured since 1894, only the discharge measurement notes since 1909 are available. Between 1909 and 1912, measurements were made at various loca- Geological Societv of Americll Rnl1",tin n.."..mh..r lQQQ ..,"",10 .~ ... tions. Between 1912 and 1930, measurements were made at a ferry/cableway located approxi- mately 9 Ian downstream from the present cable- way; we resurveyed this site in 1997. After 1930, most measurements were made at the present ca- b1eway, and these measurements provide an un- interrupted record of channel geometry to the present. Each discharge measurement includes water depth and velocity at a minimum of 20 points across the channel, and each measurement point was referenced to markings on the cable that have never changed. Each cross-section mea- surement also includes river stage, channel top- width, mean-section velocity, mean depth, cross- section area, and discharge. We analyzed these data using methods de- scribed by Topping (1997) and Smelser and Schmidt (1998) with some modifications. Mea- surements made during periods with extensive shore ice were inconsistent with the average stage-discharge relation, and these measurements were not analyzed. Minimum streambed eleva- tion was computed by subtracting the largest measured water depth from river stage, and time series of minimum and mean bed elevation were analyzed. The water-stage recorder at the town of Green River is located approximately 1 Ian up- stream from the present cab1eway. Water-surface elevation was surveyed at a variety of discharges near the cableway to establish a local stage-dis- charge relation. The stage-discharge relation at the cableway is the same as the present relation at the water-stage recorder for discharges greater than 600 m3/s (Allred, 1997). Adjustments were made to stage data for discharges less than 600 m3/s to ensure proper analysis of the vertical po- sition of historical cross-section data. Since one focus of this research was to deter- mine changes in channel width, it was necessary to extract appropriate data from the large data set. Measurements made at or just below bankfull stage best describe bankfull channel width (Fig. 2). Inclusion of measurements made at very high dis- charge introduces excessive variability due to large differences in the width of floodplain inundation; measurements made at very low discharge intro- duce variability caused by the topography of chan- nel bars that migrate past the cross section. For purposes of determining the history of changes in bankfull channel width, we therefore analyzed only those discharge measurements made at a dis- charge approximately equal to the modem 2-yr flood of 589 m3/s. Aerial Phot()gl"aph Interpretation and Analysis Stereo aerial photos taken in 1938, 1952, 1962, 1985, and 1993 (Table 1) were analyzed to char-