A Sand Budget for Marble Canyon, Arizona—Implications for Long-Term Monitoring of Sand Storage Change

ecent U.S. Geological Survey Rresearch is providing important insights into how best to monitor changes in the amount of tributary- derived sand stored on the bed of the Colorado River and in eddies in Marble Canyon, Arizona. Before the construction of Glen Canyon Dam and other dams upstream, sandbars in Glen, Marble, and Grand Canyons were replenished each year by sediment- rich floods. Sand input into the Colorado River is crucial to protecting endangered native fish, animals, and plants and cultural and recreational resources along the river in Glen Canyon National Recreation Area and Grand Canyon National Park.

View of the Colorado River in Grand Canyon, Arizona, showing sandbars that form downstream from There is longstanding interest in tributary debris fans, which constrict the flow of the river and cause eddies. The view is looking upstream the condition and trend of river-related at a point approximately 217 miles downstream from Glen Canyon Dam. (Photo by Sam Jansen.) resources in and along the Colorado River in Glen Canyon National Recreation Area and Grand Canyons are maintained by occurred because releases of water from and Grand Canyon National Park, Arizona. replenishment of sand that is supplied by Lake Powell are virtually free of sediment. These resources include endangered native tributaries and redistributed by occasional The tributaries that enter the Colorado River fish, native riparian flora and fauna (plants controlled floods released from Lake downstream from the dam supply only a and animals that live along streams), Powell. The two largest sources of tributary fraction of the pre-dam sand supply (Topping riverside campsites, and many sites of sand are the Paria and Little Colorado and others, 2000), and the capacity of the cultural significance to Native Americans. Rivers, which enter the Colorado River 17 post-dam river to transport that sand greatly The physical underpinnings for many of and 79 miles downstream from the dam, exceeds this limited supply. Normal dam these resources are river-deposited sandbars respectively. The sand delivered by these operations, therefore, tend to erode, rather that occur intermittently along the banks of tributaries is temporarily stored on the than build, sandbars. more than 300 miles of the Colorado River riverbed until the water released during By experimentation, scientists have in Glen, Marble, and Grand Canyons. controlled floods transports it to sandbars learned that controlled floods, if released In the Colorado River, most sandbars along the river’s banks. from the reservoir immediately following form along the shoreline in eddies, which large inputs of sand from tributaries, are areas of recirculating, relatively low Sandbars in Grand Canyon— can build sandbars (Schmidt and Grams, velocity flow. The sandbars are typically Previous Research 2011). These sandbars are built during inundated during floods and exposed controlled floods when sand is carried at times of low stream flow. Before the Decline in the size and abundance of from the riverbed and temporarily completion of Glen Canyon Dam in 1963 sandbars since the pre-Glen Canyon Dam suspended at high concentration in the and the construction of other large dams era has been documented by analysis of old flow. The suspended sand is transported upstream, sandbars were replenished each aerial and ground-level photographs and into eddies where it is then deposited year by sediment-rich floods. Sandbars by topographic surveys that began in the in areas of low stream-flow velocity. are now smaller because Lake Powell, mid-1970s. Scientists have estimated that Sandbars enlarged by this process provide the nearly 200-mile long reservoir formed sandbar area in the upstream 100 miles of larger camping beaches for river-rafting by the dam, traps all upstream sediment Glen, Marble, and Grand Canyons was 25 trips and create backwater habitats used before it reaches the canyons downstream. percent less in 2000 than in average pre-dam by native fish. Newly deposited sandbars The sandbars that remain in Glen, Marble, years (Wright and others, 2005). This decline also provide areas for riparian vegetation

U.S. Department of the Interior Fact Sheet 2013–3074 U.S. Geological Survey August 2013 to grow and are a source of windblown exposed sandbar is merely the highest part is based on repeat measurements of the sand. Windblown sand carried upslope of a much larger sandbar, most of which is underwater topography of the sand deposits from sandbars helps to cover and submerged. Thus, tracking trends in sand using an instrument called a multibeam potentially preserve some of the culturally storage requires tracking both the exposed and echosounder. This topographic method, significant archeological sites in Grand submerged parts of sandbars, as well as the for which uncertainty does not increase Canyon (Draut and Rubin, 2008). total amount of sand stored elsewhere on the for longer monitoring periods, provides a Scientists have also learned that riverbed. Monitoring the submerged parts of relatively precise portrayal of the locations controlled floods may erode sandbars sandbars in the Colorado River has required and amounts of changes in sand deposits. if the concentration of suspended sand the development of specialized methods. Repeat topographic measurements are, during a controlled flood is too low. The Two complementary methods are used however, difficult to implement over long concentration of sand during a flood is to monitor changes in sand deposits for river segments and are made infrequently. directly proportional to the amount of long segments of the Colorado River. The Flux-based measurements are ideally the riverbed covered by sand and the flux-based method for measuring changes suited to monitoring short-term changes in size of that sand. Higher concentrations in sand deposits is based on measurements sand storage, which are essential for making of suspended sand occur when the sand of total sand transport, referred to as “flux,” decisions about dam operations. Topographic is relatively fine and large amounts past measurement stations spaced 30 to 60 measurements, on the other hand, are more of the riverbed are covered by sand. miles apart. The difference in flux between appropriate for monitoring long-term trends These findings are incorporated in the stations represents the amount of sand that in the deposits, which are needed to evaluate current reservoir-release management has been deposited in or eroded from the the cumulative effect of management actions strategy for Glen Canyon Dam, which intervening river segment. Because stations over several years. Grams and others (2013) involves releasing controlled floods— operate continuously, this method provides integrated these two measurement methods administratively referred to as High a continuous record of changes in storage for the 61-mile length of Marble Canyon. Flow Experiments (HFEs)—whenever between the measurement stations. These Because mapping channel topography is time the Paria River has recently delivered continuous measurements can be compared consuming and costly, five short (less than 3 large amounts of sand to the Colorado to dam releases during the same time period mile) reaches were selected for monitoring as River. The magnitude and duration of the to better understand how dam operations a practical alternative to mapping the entire controlled floods is adjusted to transport affect sandbar resources. Disadvantages 61-mile segment. For such an approach to just the amount of sand that has recently of this method are that uncertainty be successful, topographic changes in the been delivered from the Paria River. grows in proportion to the length of the monitoring reaches must be representative of In support of the protocol to monitoring period and that it does not changes throughout the study area. Thus, the implement HFEs, the U.S. Geological provide information about how much each primary purpose of the study was to compare Survey’s Grand Canyon Monitoring and individual sand deposit changed within the two measurement methods and thereby Research Center (GCMRC) monitors the each long river segment. The alternative determine if this sampling strategy could be amount of sand supplied from the Paria method to measure change in sand deposits River and other tributaries and measures 112°W the size and distribution of sand on the Before the construction of Glen P Canyon Dam and other dams a riverbed and in eddies. Data on long-term r i Glen Canyon National upstream, sandbars in Glen, a trends in the amount of sand stored on the R Recreation Area Marble, and Grand Canyons, i v riverbed and in eddies allows scientists e Arizona, were replenished each r to evaluate whether the cumulative effect year by sediment-rich floods. To UTAH of the recently adopted HFE management provide information crucial to 37° strategy is causing a net increase or managing these sandbars, U.S. ARIZONA Glen Canyon Dam Lake Powell decrease in the size and volume of sand Geological Survey scientists are Lees Ferry deposits along the Colorado River. A monitoring changes in sandbars significant and progressive decline in and the amount of sediment sand storage would indicate that sandbars supplied to the river in Marble Canyon. This map shows the created during controlled floods are locations of the five short (less short-lived and that the total volume of than 3 mile) monitoring reaches sandbars is declining. Such a long-term used to measure topographic ColoradoRM River 30 decline in sand storage would likely result changes in sandbars and the in declining future effectiveness of HFEs locations of stream-flow and in maintaining desired sandbar size. sediment-transport measurement stations. The brown-shaded area Recent Findings—How Much is within Grand Canyon National RM 61 Sand is in the River? Park. Glen Canyon extends from Glen Canyon Dam to Lees Ferry; RM 87 Although sandbars that are exposed upper Marble Canyon extends from 36° Lees Ferry to river mile (RM) 30; Little above the water surface are the resources Co lower Marble Canyon extends from lor of management interest, the visible part of ad RM 30 to RM 61; and Grand Canyon o a sandbar represents only 10 to 20 percent R i begins at RM 61. Short monitoring reaches v e Map of the total sandbar. Like icebergs, a small r Streamflow and sediment Area transport measurement stations ARIZONA 0 30 MILES

0 50 KILOMETERS Elevation change < -2 m -1.9– -1.0 -0.9– -0.5 1 -0.4– -0.3 -0.2– -0.1 0 0.1 0.2–0.3 0.4–0.5 2 0.6–1.0 1.0–2.0 >2.0

3

4

5

56 0 200 METERS

This aerial image with superimposed map shows areas of erosion (red) and deposition (blue) between December 2004 and May 2009 for a segment of the short monitoring reach near river mile 56 in Marble Canyon (see map). A multibeam echosounder was used to measure the changes in the underwater topography of sediment deposits. The numbered gray outlines show individual eddies, where changes in sediment storage are entirely sand. Of the three largest eddies in this short river segment, one (2) is almost entirely dominated by erosion and two (1 and 4) are Glen Canyon Dam was completed in 1963, affecting stream flow and sand movement in the dominated by deposition. The Colorado River Colorado River below the dam. These photographs show the river in Grand Canyon, 150 miles channel outside of the eddies includes areas of downstream from the dam. The top image taken in 1952 shows a large sandbar; in the bottom deposition and erosion. (Changes in elevation image taken in 2003 little sand remains. View is downstream from the right (north) bank of the river. shown in meters (m); 1 m=~3.28 feet. <, less (Photos courtesy of Kent Frost, top, and Steve Tharnstrom, bottom; from Webb and others, 2002.) than; >, greater than.) used to reliably track long-term trends in that local changes in the topography of sand few locations in a few short reaches are sand storage. deposits are very large when compared to the unlikely to be representative of the average Each of the short topographic changes in storage calculated from flux-based topographic response over a long segment monitoring reaches was mapped at least measurements. The amount of sand that is of the Colorado River. four different times between 2002 and either deposited or eroded from just a few 2009. Comparison of the maps for different hundred yards of river channel can equal The “Sand Budget” times allowed determination of the amount the total change in sand storage for an entire and spatial pattern of changes in riverbed 30-mile river segment. Tracking sand storage in Grand Canyon and eddy sand deposits. These results were Thus, changes in sand storage that are is like managing a budget or a bank account. compared with sediment-flux measurements measured over long segments of the river “Deposits to the account” occur when sand is spanning the same time periods. channel by the flux-based measurement supplied by tributaries, but these “deposits” The repeated mapping of riverbed program are not the result of uniform are typically only made to the riverbed. Sand topography demonstrates that very large changes evenly distributed throughout the transported further downstream towards changes in individual sand deposits can occur channel. Instead, changes in sand storage Lake Mead reservoir (along the Arizona- in the Colorado River in Marble Canyon. that are measured over long segments result Nevada border) by normal dam operations Adjacent sand deposits often change in from the sum of topographic changes in and controlled floods are withdrawals from different ways—one deposit may erode (1) small areas where there has been large the account. Sandbars built during controlled greatly, but another nearby deposit may erosion and (2) other areas where there floods may be viewed as “transfers” from grow significantly. For example, between has been large deposition. This finding the “riverbed account” to the “eddy sandbar December 2004 and May 2009, some parts highlights the difficulty in identifying short account.” Thus, it is essential that river of the river channel near river mile 56 “representative” topographic monitoring managers are provided information about eroded, whereas other areas accumulated reaches for extending results to longer the amount of sand in the total bank account, sand. Grams and others (2013) also found river segments. Changes measured at a the amount of sand in the eddy sandbar 3.0 50,000 These graphs show the “sand upper Marble Canyon budget” (sand available to build 40,000 sandbars) for upper and lower Marble 2.0 Canyon, as well as Colorado River References Cited water discharged from Glen Canyon 30,000 Draut, A.E., and Rubin, D.M., 2008, The role of Dam, during 2002–2009. The solid eolian sediment in the preservation of archeo- 1.0 brown lines show the sand budget logic sites along the Colorado River corridor in 20,000 based on sediment flux recorded at Grand Canyon National Park, U.S. Geological measurement stations (see map). Survey Professional Paper 1756, 71 p., available 0 The upper solid line is the upper limit 10,000 at http://pubs.usgs.gov/pp/1756/. of the uncertainty range, and the lower line is the lower limit of the Grams, P.E., Topping, D.J., Schmidt, J.C., -1.0 0 uncertainty range. The blue dots with Hazel, J.E., and Kaplinski, M., 2013, Link- 3.0 50,000 “error bars” show the sand budget ing morphodynamic response with sediment lower Marble Canyon based on repeat measurements of mass balance on the Colorado River in Marble 40,000 the topography of sand deposits. 2.0 Canyon—Issues of scale, geomorphic setting, The error bars show the estimated and sampling design: Journal of Geophysi- Sand budget, in million metric tons 30,000 Discharge, in cubic feet per second measurement error. These estimates cal Research, v. 118, no. 2, p. 361–381, doi: have additional uncertainty (not 1.0 10.1002/jgrf.20050, accessed July 29, 2013, shown) associated with the 20,000 at http://onlinelibrary.wiley.com/doi/10.1002/ extrapolation from short monitoring jgrf.20050/abstract. 0 reaches (see map) to the 30-mile river 10,000 segments between measurement Schmidt, J.C., and Grams, P.E., 2011, The high stations. The thick blue lines are flows—Physical science results,in Melis, T.S., -1.0 0 discharge measured at Lees Ferry. ed., Effects of three high-flow experiments on 2002 2003 2004 2005 2006 2007 2008 2009 2010 (1 metric ton =~1.10 U.S. short ton.) the Colorado River ecosystem downstream from Glen Canyon Dam, Arizona: U.S. Geological account, and the amount of sand that has been these long-term trends in storage indicate that Survey Circular 1366, p. 53–91, available at withdrawn from the account and transported to storage in Marble Canyon did not decline http://pubs.usgs.gov/circ/1366/. Lake Mead. between 2002 and 2009. This period was one Topping, D.J., Rubin, D.M., and Vierra, Repeat topographic mapping of the of average to above average tributary sand L.E., 2000, Colorado River sediment trans- riverbed shows where large deposits and inputs and average to below average dam port—1; natural sediment supply limitation withdrawals of sand from the account typically release volumes. Grams and others (2013) and the influence of Glen Canyon Dam: Water occur. Grams and others (2013) showed that conducted their study during a period that was Resources Research, v. 36, no. 2, p. 515–542, doi: 10.1029/1999WR900285, accessed the sand is stored in hundreds of separate favorable to sand accumulation. Periods when July 29, 2013, http://onlinelibrary.wiley.com/ accounts, some larger than others, distributed dam release volumes are greater and tributary doi/10.1029/1999WR900285/abstract. throughout the river channel. Understanding sediment inputs are less frequent will likely this sand budget—where the sand-storage result in less sand accumulation. Grams and Webb, R.H., Melis, T.S., and Valdez, R.A., locations are, the size of these accounts, others (2013) showed that measurements of 2002, Observations of environmental change in Grand Canyon: U.S. Geological Survey Water- and how much these accounts change—is channel change made in short reaches can be Resources Investigations Report 02–4080, 33 p., essential to improving predictions about how used to track changes in deposits and transfers available at http://pubs.usgs.gov/wri/wri024080/. dam operations will affect sand storage and, of sand among the storage locations, but the ultimately, sandbars. results cannot be extrapolated to long segments Wright, S.A., Melis, T.S., Topping, D.J., and The sand budget also showed that of the river, because the size and distribution of Rubin, D.M., 2005, Influence of Glen Canyon Dam operations on downstream sand resources between 2002 and 2009 there were periods of sand storage locations is highly variable. of the Colorado River in Grand Canyon, in depletion of sand from storage and periods of Gloss, S.P., Lovich, J.E., and Melis, T.S., eds., accumulation of sand in storage. Sand storage Next Steps The state of the Colorado River ecosystem in always declined between tributary sediment Grand Canyon: U.S. Geological Survey Circular On the basis of the above findings, input events, whereas tributary sand inputs in 1282, p. 17–31, available at http://pubs.usgs.gov/ GCMRC scientists and cooperators now 2006, 2007, and 2008 were large and resulted circ/1282/. measure topographic changes over long in progressive sand accumulation. Although segments of the Colorado River channel. much of this sand was transported downstream Paul E. Grams Instead of monitoring short reaches, which in the March 2008 controlled flood (that is, a are unlikely to be representative of the entire “withdrawal” from the account), some of this Edited by James W. Hendley II system, scientists now map a much larger sand was transferred to newly enlarged eddy Graphics and layout by Jeanne S. DiLeo proportion of the sand-storage locations in long sandbars or remained on the riverbed. river segments. This monitoring strategy will For more information contact: Management Implications provide a detailed accounting of the changes U.S. Geological Survey in sand deposits for entire 30-mile monitoring Southwest Biological Science Center Because the absolute quantity of sand segments. Scientists will learn much more Grand Canyon Monitoring and Research Center available to build sandbars along the Colorado about the size, distribution, and behavior 2255 N. Gemini Drive River cannot be reliably measured, the only of those sand-storage locations that are of Flagstaff, AZ 86001 option is to measure trends in sand storage— significant management interest, hopefully (928) 556-7094 that is, changes in the size of eddy sandbars enabling development of a more cost-effective http://www.gcmrc.gov/ and changes in the amount of sand on the and spatially representative sandbar monitoring This fact sheet and any updates to it are available online at http://pubs.usgs.gov/fs/2013/3074/ riverbed. Initial results from efforts to monitor program.