Hydrology and Geomorphology of the Santa Maria …repository.azgs.az.gov/sites/default/files/dlio/files/...Page 6 Hydrology and Geomorphology of the Santa Maria and Big Sandy Rivers

Page 1Hydrology and Geomorphology of the Santa Maria and Big Sandy Rivers and Burro Creek, Western Arizonaby Jeanne E. Klawon

Hydrology and Geomorphologyof the Santa Maria and

Big Sandy Rivers and Burro Creek,Western Arizona

By Jeanne E. Klawon

Arizona Geological SurveyOpen-File Report 00-02

March, 2000

Arizona Geological Survey416 W. Congress, Suite #100, Tucson, Arizona 85701

46 p. text

Investigations supported by the Arizona State Land DepartmentAs part of their efforts to gather technical information for stream navigability assessments

Investigations done in cooperation with JE Fuller Hydrology / Geomorphology


EXECUTIVE SUMMARY

This report provides hydrologic and geomor-phologic information to aid in the evaluation ofthe navigability of the Big Sandy River, SantaMaria River, and Burro Creek. These streamsflow through rugged mountainous terrain ofMohave, Yavapai, and La Paz counties inwestern Arizona and join to form the BillWilliams River at what is now Alamo Damand Reservoir. The rivers reflect a diversity ofchannel patterns, and include the mainly wideand braided sandy alluvial channels of the BigSandy River, the narrow bouldery channels inbedrock canyons of Burro Creek, and a mix ofbedrock canyons and wide alluvial channels onthe Santa Maria River.

Although none of these tributaries has anabundant water supply, the relative surface-water flow of each river does vary. For most ofits length, Burro Creek contains water thatflows year round or is relegated to discontinu-ous pools during the dry portions of the year.The Big Sandy River maintains perennial flowalong some portions of its length and is fed byupper watershed perennial streams such asTrout Creek. The Santa Maria is the driest ofall three rivers, with most of its surface-waterflow produced by tributaries in the upperwatershed and in the main channel for 10-15miles near its mouth. Low flow dischargemeasurements of each river are generally less

than 10 cfs. Flood events are dramatic incomparison. Historical peak flow estimates forthese rivers have been estimated at 68,700cubic feet per second (cfs) (2/9/93) for the BigSandy River, 47,400 cfs (2/14/80) for BurroCreek, and 23,100 cfs (3/1/78) for the SantaMaria River. The largest flow events haveoccurred during the winter months whenmeteorological conditions cause a series ofstorms to pass through the region, frequentlygenerating multiple floods within a given year.All three rivers in many cases record the sameflood events, reflecting the regional extent ofthe storms.

Substantial channel changes have oc-curred along the alluvial reaches of these riversin the past 100 years. Most channel changeshave occurred during large floods. Channelchanges typically have involved erosion ofstream terraces and tributary fans aroundoutside bends, removal of vegetation withinthe channel, and in some cases, channel widen-ing. Analysis of channel change spans the 20th

century. Based on historical accounts, dramaticchannel changes occurred around the turn ofthe century; based on interpretation of aerialphotographs of the rivers, dramatic channelchanges also occurred during floods from 1978to 1995.


This report summarizes the geomorphologyand hydrology of the three principal tributarystreams of the Bill Williams River in westernArizona. The Bill Williams River basin drainsmost of west-central Arizona, straddlingMohave, Yavapai and La Paz counties (Figure1). The principal tributaries of the Bill Will-iams are Big Sandy River, which heads east ofKingman, and Santa Maria River, which headsnear Hillside. These tributaries join to form theBill Williams River at what is now AlamoLake. Burro Creek is the largest tributary ofBig Sandy River, entering south of Wikieup.

The rugged terrain of the Bill WilliamsRiver basin has limited human habitation to afew small towns and land development toirrigation agriculture and ranching mainly inthe Big Sandy Valley and along the Bill Will-iams. Although the area was host to manymines which sought gold, silver, copper, andother associated minerals, the rugged terrainand limited water resources in the area im-peded the development of large-scale enter-prises. A notable exception is the Bagdadcopper mine in the Santa Maria River drainagebasin. Thus, for the most part, this area re-mains a remote location in Arizona.

Purpose and Scope

This section provides hydrologic and geologicinformation to aid in the determination ofnavigable versus non-navigable status for themajor tributaries of the Bill Williams River inwestern Arizona. It begins with a brief intro-duction to the geology, climate, and physiogra-phy of the region. Hydrologic information isthe focus of the paragraphs which followintroductory information and includes thehistory of stream gages within the basin andanalysis of their surface water records. Se-lected floods from this century are discussed in

some detail in order to understand the meteo-rology and hydrology of the largest flows inthe gage record. Pre-gage historical documentsand paleoflood records provide additionalinformation on the hydrologic behavior of theBill Williams River basin. Following thehydrologic discussions, channel characteristicsare described as well as channel changes thatmay have occurred within the last century.Because of the extensive length of each river,specific reaches were selected and studiedusing aerial photographs, topographic maps,and general land office surveys from the early1900’s. Appendices include a list of contacts,rating curves for primary stream gages, andphoto documentation of river reaches.

Acknowledgments

Kyle House and Pat Shafroth provided infor-mation and feedback that facilitated prepara-tion of this report. Phil Pearthree reviewed andedited the report. Pearthree and Tim Orr helpedprepare figures for inclusion in this report. PeteCorrao is responsible for final layout of thetext and illustrations.

Geologic Setting

The Bill Williams River basin is located in theBasin and Range and Transition Zone geologicprovinces in west-central Arizona. The Basinand Range province extends from the SnakeRiver Plain in Idaho through southern Arizonaand into Mexico. The Basin and Range prov-ince is characterized by generally north-trending mountain ranges (Spencer andReynolds, 1989) which are separated by basinsformed by normal faulting along mountainfronts. In southern and western Arizona, basinsare deep, well-defined grabens which trendnorth to northeast, and have fairly regular

INTRODUCTION


Figure 1. Location of the Big Sandy and Santa Maria rivers and Burro Creek inwestern Arizona.


spacings of 25 to 50 km (Menges andPearthree, 1989). The Big Sandy Valley is themost prominent basin in the study area and iscomposed of basin fill over 3000 ft in depth(Davidson, 1973). The Transition Zone is therugged, mountainous country between theBasin and Range and the Colorado Plateau. Ithas geologic and physiographic characteristicsthat are transitional between the highly de-formed Basin and Range and the relativelyundeformed, fairly high Colorado Plateau tothe northeast.

The geology of the Bill Williams Riverbasin reflects the complex history of the Basinand Range province, with several periods ofmagmatism and overprinting of compressionaland extensional terranes in the past 80 millionyears. A period of wide-reaching magmatismand crustal shortening associated with theLaramide orogeny occurred in middle to lateCretaceous to early Tertiary period, about 80 to60 million years ago. This same area wasextended in the middle Tertiary between about20 and 10 million years ago, forming majorlow-angle normal faults trending east-north-east-west-southwest. Large amounts of exten-sion occurred in the area south and west of theBig Sandy - Santa Maria confluence, resultingin the exposure of rocks in the Buckskin,Rawhide, and Harquavar mountains that wereburied beneath many miles of rock prior to theperiod of extension. Lesser amounts of middleTertiary extension occurred in the TransitionZone to the northeast. As a result of thisextension, the region experienced a reversal ofsurface drainage direction from north-flowingto south-flowing streams as the extended partof the crust subsided. Relatively modest lateTertiary extension in an east-west to east-southeast-west-northwest direction was accom-modated by higher-angle normal faults(Zoback and others, 1981). The change inextension direction and style in the late Ter-tiary was accompanied by magmatic composi-tion changes from primarily silicic to interme-

diate volcanism to predominantly basalticvolcanism.

Sediments that filled the structural basinsformed by regional extension can be groupedinto upper and lower basin fill, and Quaternarydeposits. Lower basin fill ranges in age frommiddle to late Miocene, was deposited asnormal faulting continued, and has a thicknessof at least 3,000 ft in the Big Sandy Valley.Upper basin fill is as much as 300 ft thick andwas deposited in the waning stages of defor-mation during the Pliocene and early Quater-nary. Pedimentation along mountain fronts wasbegun in this period and continued into themiddle Quaternary along with the deposition ofthin coarse-grained fans. Through the Quater-nary, alluvial fans and terraces have beendeposited, modified, and eroded throughprocesses which are for the most part driven byclimate changes. The oldest and highest allu-vial fans in the Big Sandy Valley occur in theupper watershed of Tule Wash and are about200 ft thick. The bedrock mountains andalluvial basins have continued to be dissectedduring the Pliocene and Quaternary as regionaldrainage systems have continued to downcut.Sediments from coalescing drainages, depos-ited in the floodplains and streams of the studyarea, usually amount to no more than 40 ft inthickness (Menges and Pearthree, 1989;Davidson, 1973).

Physiography

Most of the Bill Williams River basin is in theBasin and Range province and the ruggedTransition Zone along the margin of the highelevation Colorado Plateau (Figure 2). The BigSandy River basin has a total relief of 7297 ftand an area of ~2810 mi2 (House, 1997), whichincludes 687 mi2 (Markman, 1996) of theBurro Creek watershed. The Big Sandy Riverbegins at the confluence of Knight and Troutcreeks in the Pilgrim Wash Quadrangle andflows south through Big Sandy Valley, then


Figure 2. USGS 7.5’ topographic quadrangles covering the study area and active stream gages. These tributaries ofthe Bill Williams River flow south to the confluence of the Big Sandy and Santa Maria rivers, where they combine toform the Bill Williams River.


cuts through the mountainous terrain of Green-wood Peak, Signal Mountain, Arrastra Moun-tain, and Artillery Peak to its mouth, where itconverges with the Santa Maria River. It has amain channel gradient of 0.008 (Garrett andGellenbeck, 1991) above the gaging station 17miles south of Wikieup; other calculatedgradients include 0.011 north of Cane SpringsWash and 0.006 in the southern portion of BigSandy Valley (Davidson, 1973). Burro Creekheads in western Yavapai county between theMohon and Santa Maria Mountains near theLuis Maria Baca Float No. 5, flows southwestthrough Bozarth Mesa and the rugged terrainof the Poachie Range, and joins the Big SandyRiver approximately 12 miles south ofWikieup. Total relief from its source to itsmouth is 5979 ft. Selected stream reacheslocated below the confluence with BoulderCreek and an 11-mile-long stretch that begins15 miles above the confluence with the BigSandy River, have gradients of 0.006(Markman, 1996). The Santa Maria Riverheads in the Santa Maria Mountains near theLuis Maria Baca Float No. 5 and has a water-shed area of 1520 mi2 (House, 1997). It has agradient of 0.009 above the gaging station nearBagdad and a lower gradient of 0.008 (Garrettand Gellenbeck, 1991) upstream from thegaging station near Alamo, which includes agradual alluvial reach downstream from theBagdad gage. From its source, the drainageflows to the southwest past Grayback Moun-tain and Ives Peak and turns to the west tomeet the Big Sandy River near Alamo. Withthe highest elevation in the watershed at 7626ft in the headwaters of Kirkland Creeksubbasin, and the lowest on the order of 1120ft at its confluence with the Big Sandy, theSanta Maria River watershed has a total reliefof 6506 ft.

Climate

The climate of the Bill Williams River basin is

semi-arid to arid. Temperature and precipita-tion values as well as dominant vegetation varysubstantially within the basin, with pinyon-juniper woodlands in the mountain ranges andcacti and riparian species in the interveningvalleys. Average January temperatures rangefrom 35° to 45° F at higher elevations to 45°-50° F at lower elevations. Average July tem-peratures range from 65°-75° F at higherelevations to 80°-90° F at lower elevations inthe watershed (Sellers and Hill, 1974).

Precipitation also varies considerablywithin the Bill Williams River basin. Annualprecipitation is generally between 10 and 15inches for valleys and low elevations, droppingas low as 6 inches near the mouth of BillWilliams River (United States Army, Office ofthe Chief Engineer, 1944), and 15-20 inches inthe mountainous high-elevation areas (Sellersand Hill, 1974).

Precipitation falls mainly in summer andwinter rainy seasons. Summer rains occurprimarily during the afternoons and eveningsin July and August and also in some valleysbetween 6 and 9 am. Summer storms are forthe most part generated by convection, inwhich moisture from the Pacific Ocean and theGulf of Mexico encounters heated mountain-ous terrain, which causes the air to increase intemperature and rise. These unstable air masseslead to high intensity rainstorms of shortduration often accompanied by thunder, light-ning, and strong winds (Sellers and Hill, 1974).

Dissipating tropical storms from theeastern Pacific occasionally bring heavyprecipitation to portions of western Arizonaduring the late summer and early fall. For thesestorms to cross Arizona, the jet stream isrerouted to the northeast and commonly inter-acts with a cutoff low or low pressure trough(House, 1997). The largest floods recorded atthe Santa Maria River gage near Alamo oc-curred in the late summer and early fall duringthe 1939, 1951 and 1964 water years. In hisinvestigations in western Arizona, House


(1997) found that large late summer and earlyfall floods occurred in 1939 and 1951, as wellas 1970 and 1976. It is not possible to know ifthe gage on the Santa Maria River near Alamoalso experienced large floods during 1970 and1976 as it was discontinued in 1966.

In the upper Bill Williams River basin,stream gages have recorded high peak flows inthe late summer-early fall months that presum-ably resulted from dissipating tropical stormprecipitation events. Their magnitudes aresmaller in comparison to winter flood events atthe same gages, however. The largest flood ofthis type was 15,200 cfs recorded at the SantaMaria River gage near Bagdad on September24, 1983; this discharge is among the top tenlargest floods recorded at this gage. Largeflows were also recorded on this date at BurroCreek (11,900 cfs) and Big Sandy River nearWikieup (12,900 cfs).

Although winter rains are generallylighter and longer lasting than summer rains,unusual meteorological phenomena may bringabundant precipitation across the Bill WilliamsRiver basin during the winter months. These

storms result from the westward displacementof the ridge of high pressure in the PacificOcean and the formation of a semi-permanentlow-pressure trough over the western UnitedStates. In this situation, storms tend to movesouth along the west coast instead of throughWashington and Oregon. They then travel eastfrom the coast through California and Arizona.Usually this storm track may produce precipi-tation in Arizona for a couple of days andcontinue to the east; this pattern tends to recurand may produce several precipitation eventswithin a short time interval. However, if thehigh-pressure system in the north becomesespecially well developed, the low pressurecell tends to become isolated from the trackand may stagnate and intensify off the coast.By the time the storm reaches Arizona, it isfully developed and capable of producinglarge-magnitude events. In fact, when the floodrecords of large drainages of western andcentral Arizona are analyzed, it is almostalways the case that the largest floods onrecord have occurred during the winter season(House and Hirschboeck, 1997).

This subsection summarizes various aspects ofthe hydrology of the Big Sandy and SantaMaria rivers and Burro Creek. I briefly discussreaches with perennial flow on each of thesestreams, summarize historical flow records,and discuss in more detail the genesis of a fewof the largest floods that have occurred in thisdrainage system.

Areas of Perennial Flow

The Big Sandy, Santa Maria Rivers, and BurroCreek all have perennial reaches along their

HYDROLOGY

lengths (Figure 3). In this report, a reach isperennial if it usually flows year round; how-ever, some reaches included in the perennialcategory run dry during periods of extremedrought (Brown, et al, 1978). Reaches that aretermed perennial by Brown et al (1978) andFreethey and Anderson (1986) are described ashaving a base flow of less than 10 cfs. In itsheadwaters, the Big Sandy River is supplied byperennial reaches of Willow Creek and theTrout Creek basin. The main stem perennialreaches stretch from Wikieup downstream tothe Burro Creek confluence and from Signal to


Figure 3. Perennial stream reaches in the upper Bill Williams River basin (datafrom Brown, et al., 1978, and Freethey and Anderson, 1986).


Figure 4. Active and discontinued stream gages in the Bill Williams River basin. Discontinuedgages typically have short records from the 1960’s and 1970’s. Exceptions are of the gage onBurro Creek, which records flow through the 1980’s, and the Santa Maria gage near Alamo,which began in 1939 and was discontinued when Alamo Dam was completed and a new gagingstation was established upstream.


Alamo Lake. The Whipple expedition, whichtraveled through this area during 1853 and1854, described the Big Sandy as having awater source which sinks and rises in the sandychannel along the length of the river (Foreman,1941). Perennial flow in the Wikieup reachaveraged 2.52 cfs from 21 miscellaneousmeasurements taken between 1959 and 1964(Davidson, 1973). Burro Creek is perennialalong most of its length with the exception ofthe reach near U.S. Route 93 bridge crossing.Tributaries of Burro Creek that are perennialinclude sections of Francis Creek, BoulderCreek, and Pine Creek. The reaches in thesetributaries are all located within the first fivemiles upstream of the river mouth. The SantaMaria River basin has perennial reaches in the

headwater tributaries of Cottonwood, Smith,and Sycamore creeks. The main stem SantaMaria is perennial approximately 5 milesabove and below Bridle Creek, and 5 to 10miles upstream from Alamo Lake. Portions ofKirkland Creek and Date Creek also flow yearround.

Surface Water Records

Gage History. Stream flow volumes were firstrecorded in 1939 in the Bill Williams Riverbasin, beginning with stream gages on the BillWilliams River and Santa Maria River nearAlamo, Arizona (Figure 4; Table 1). Thesegages evolved as Alamo Dam near Alamo, AZwas proposed and built, and Alamo Lake

Table 1. List of stream gages of the Bill Williams River Basin above Alamo Dam.


developed behind it. The gage near Alamo onthe Santa Maria was discontinued in 1966because it was within the area of Alamo Lake;it was replaced by a new gage located about 10miles upstream (Santa Maria near Bagdadgage). The USGS first began measuringstreamflow on the Big Sandy River and BurroCreek in 1966 and 1980, respectively, but theydiscontinued the Burro Creek station followingthe 1993 water year. Stream gages were alsoinstalled on a number of tributaries to the BigSandy, Santa Maria and Burro Creek; however,these gages were short-lived and mainly recordstream flow from the early 1960’s to the late1970’s. Active stations in the Bill WilliamsRiver basin upstream from Alamo Dam in-clude the Big Sandy River near Wikieup, AZ,Santa Maria River near Bagdad, AZ, and theBill Williams River below Alamo Dam.

Pre-gage Flood Records. Gage recordsgive a picture of the past 60 years, at best;other accounts of pre-gage late 19th century toearly 20th century flooding suggest that thiswas a wet period with many floods, some ofwhich probably exceeded the floods in thegage record. Years that have been noted tohave large floods on the Bill Williams Riverinclude 1884, 1891, 1905, 1906, 1910, 1911,1915, 1916, 1920, 1923, 1927, 1930, 1931,1932, 1937, 1938, and 1939. Although themagnitudes of these most of these floods arenot known, indirect estimates were made forthe largest of these, which include the flood ofFebruary 1891 estimated at 200,000 cfs andJanuary 1916 at 185,000 cfs. This is in contrastto the largest recorded discharge measurementof 92,500 cfs which was made on February 7,1937 at the Bill Williams River near Alamo(United States Army, Office of the ChiefEngineer, 1944). Historic accounts of thisperiod by Irene Cornwall Cofer who lived inthe Big Sandy Valley at the turn of the centuryrelate that prior to the 1880’s, the Big SandyRiver channel was narrow, allowing for wideagricultural fields, orchards, and ranch houses.

Many of these properties were washed awayduring floods in the 1880’s and early 1900’s.Cofer describes her experience of one of thesefloods on New Year’s Eve, 1908:

“The first really big flood that I saw in theSandy River; one that cleaned the trees andbrush from the river bottom and took cotton-wood trees from the banks that were perhaps ahundred years old and tossed them end overend like they were two by fours.” [Cofer1969:76]

Paleoflood Records. Paleoflood hydrol-ogy is an interdisciplinary approach used tostudy historic and prehistoric records of floodsin river systems around the globe and is appli-cable to a wide range of scientific topics andpractical problems. By extending the floodrecord beyond a relatively short historical dataset, geologists are able to extract more repre-sentative samples of the largest floods forflood frequency analysis and to address long-term issues such as aggradation and degrada-tion in the fluvial system, as well as flood-climate linkages for a particular region. Themethod produces results that are very usefuland needed in urban planning, flood controlprocedures, and reservoir operation (Stedingerand Cohn, 1986; Hereford, et al., 1996; Graf, etal., 1991; Patton, 1977; Costa, 1978; Macklin,et al., 1992).

The slackwater deposit-paleostage indica-tor (SWD-PSI ) technique in paleoflood hy-drology allows for indirect discharge measure-ments following a flood by using featureswhich record the height of the flood waters(Baker, 1987; Kochel and Baker, 1982; Patton,et al., 1979). Slackwater deposits are fine-grained deposits, typically composed of siltand sand, which accumulate in backwaterzones where reduced velocities allow fineparticles to fall from suspension. Reaches mostconducive for preserving slackwater depositsinclude those that have fixed channel bound-aries and features which initiate flow separa-tion, such as alcoves, channel constrictions,


minor tributary mouths, or bedrock obstruc-tions. Paleostage indicators include all otherhigh water marks such as flotsam piles andmats, scour lines, non-exceedance indicators,water stains on trees and bedrock, and treescars.

The Bill Williams River basin has notbeen extensively studied from a paleofloodperspective, although a paleoflood project iscurrently underway in this basin (House, oralcommunication, 1998). Ely (1992) studiedselected paleoflood sites along the Bill Will-iams River, Big Sandy River and Burro Creekin order to extract regional information aboutvariations in large-flood frequency in Arizonaand southern Utah within the past 5000 yearsand their relation to long-term meteorologicalpatterns and other paleoclimatic indicators.Ely’s Big Sandy River paleoflood site islocated 9 miles downstream from theconfluence with Burro Creek at 34°27’14” N,113°37’33” W, and is composed of 3 insetstacks of deposits, the highest of which iscovered by mesquite trees. Although a radio-carbon date from this uppermost depositsuggests that the entire package of floods isrelatively young, the lack of flood evidence onthe top surface indicates that the highest extentof deposits had not been overtopped in thehistoric period prior to 1992 and thus thatlarger floods than those measured historicallyhave occurred. Ely’s paleoflood site at BurroCreek near Andy’s Hole (34°32’43” N,113°29’14” W) is similarly young, with 5 or 6flood deposits preserved.

Hydrologic data

Mean daily flow series. The magnitudes ofmean daily flow values fluctuate seasonallyand annually on rivers in the Bill WilliamsRiver basin. The largest flow volumes occurduring the winter (December through March)and early fall (August though September) andarrive in the form of floods as well as sustained

high flow for periods of days. The driestperiods for these rivers occur during earlysummer (June and July) and late fall (Octoberand November), in which flow may cease ordrop to a few cfs.

The three rivers in this study exhibitrecords of daily flow values that differ inmagnitude and seasonal variability accordingto the hydrologic characteristics of each river.The Big Sandy River near Wikieup, supportedby a shallow aquifer in the Big Sandy Valleyalluvium as well as several springs in the area,continues to flow year round. Daily flowvalues are generally less than 10 cfs during thedrier months; wetter months have slightlyhigher daily flow values and sharp flood peaksduring storms. No flow days (i.e., days whichno flow is recorded) have not occurred at thisparticular gage, meaning that this reach of theBig Sandy River is consistently perennial.Burro Creek experienced sustained flow formost of the gage period of 1984-1993, al-though many times flow was less than 1 cfsduring the dry months of the year. Duringsome years, however, especially in the late1980’s and early 1990’s, no flow days wererecorded. These totaled 388 out of 4922 daysin the gage record; in other words, approxi-mately 8% of the total gage record measuredmean daily flow values of 0 cfs. The SantaMaria at the gage near Bagdad has experiencedseasonal variability similar to the Big SandyRiver and Burro Creek, but there is a greaterdifference in magnitude between high flowperiods and low flow periods during the year.For example, 7,112 of 10,393 or 68% of dailyflow values have been no flow days on theSanta Maria River. Thus, the Santa MariaRiver in its upper reaches is more sensitive tofluctuations in precipitation, whereas the othergage locations discussed above have alternateground water sources that sustain flow duringthe drier portions of the year.

The Santa Maria River near Alamo,located near the river’s mouth, had a strikingly


Figure 5. Mean annual flow at the USGS stream gage on Big Sandy River near Wikieup, Arizona.

Figure 6. Mean annual flow at the USGS stream gage on Burro Creek at Old U.S. 93 Bridge.


different record than the gage upstream. Only1.6% of the gage near Alamo record is com-posed of no flow days compared to 68% at thegage near Bagdad. While this gage does recorda completely separate time interval than theupstream gage, the difference in records isalmost certainly due to the change of locationon the river rather than a change in the overallhydrology of the system. The larger drainagearea above the gage near Alamo may contrib-ute to the larger peak discharge values re-corded there. The maximum peak dischargerecorded near Alamo (August 29, 1951) isapproximately 10,000 cfs greater than any peakdischarge recorded more recently at the up-stream gage. The two gage records also differmarkedly in the seasonality of the largestfloods. The largest floods at Alamo (1939-65)occurred during the late summer-early fall,while those floods which dominate the Bagdadrecord (1966-present) occurred during thewinter months. Based on the time of year inwhich they occurred, the late summer-early fallfloods were generated by summer monsoonalstorms or dissipating tropical storms. Hydro-logic data for the Santa Cruz River in southernArizona exhibit a somewhat similar trend(Webb and Betancourt, 1992). From 1930-59,the Santa Cruz flood record is dominated bysummer monsoonal floods, none of whichwere very large, while the record from 1960-86is dominated by fall and winter floods causedby dissipating tropical storms and regionalwinter storms. Webb and Betancourt point to atrend in meteorological phenomena, in whichan increased number of dissipating tropicalstorms are brought into the southwesternUnited States during ENSO (El Nino-SouthernOscillation) conditions (Webb and Betancourt,1992). On the Santa Maria River, only onelarge winter flood occurred between 1939 and1965 (3/14/41). A number of dissipatingtropical storms generated floods during thisinterval, however, including the largest floodof the gage record (8/29/51).

Mean annual flow series. The mean annualflow series, the arithmetic average of meandaily flow values for each year in the period ofrecord, is best used to identify high flow andlow flow years on record for a given gage. TheBig Sandy River near Wikieup experienced itsrecord high flow year in 1993, which includesthe flood of record (68,700 cfs, 2/9/93) and anumber of other large flood peaks during thewinter of 1992-93 (Figure 5). The record lowflow year for the gage was in 1996, when itslargest flow value amounted to 275 cfs on 9/9/96. For Burro Creek, the record high flow yearwas again 1993 with the flood of record esti-mated at 55,300 cfs on February 8, 1993(Figure 6). The record low flow year for BurroCreek was 1987, which had a peak dischargeof 565 cfs on 3/5/87. The Big Sandy gage alsorecorded a low flow year in 1987. On the SantaMaria River, the record low flow year occurredin 1956 at the Santa Maria River near Alamo,in which mean annual discharge measured 2cfs and the peak flow value measured 107 cfs(Figure 7). The record high flow year occurredin 1941, in which multiple flood peaks throughthe winter contributed to a high mean dis-charge for the given water year. The largestpeak of this year measured 20,600 cfs on 3/14/41. The Santa Maria River near Bagdadexperienced comparable high flow years in1979 and 1980, both of which included anumber of flood peaks during the wintermonths (Figure 8). The record low flow yearfor this gage was 1996, in which the gagerecorded no flow during the entire year.

Annual flood series. Annual flood series of theBill Williams tributaries represent the maxi-mum peak flow volume for each year of recordwhich have occurred during the last 50 years ateach USGS gaging station (Figures 9-12).Years of extreme flooding on the rivers in thestudy area, with record flows and multiplelarge floods during winter months include1978, 1980 and 1993, among others. Figure 13


Figure 7. Mean annual flow at the former USGS stream gage on the Santa Maria River nearAlamo, Arizona.

Figure 8. Mean annual flow at the current USGS stream gage on the Santa Maria River nearBagdad, Arizona.


Figure 9. Annual flood series for the USGS stream gage on the Big Sandy River near Wikieup,Arizona.

Figure 10. Annual flood series for the USGS stream gage on Burro Creek at Old U.S. 93 Bridge,Arizona.


Figure 11. Annual flood series for the former USGS stream gage on the Santa Maria River nearAlamo, Arizona. The period of record is 1939-1967.

Figure 12. Annual flood series for the USGS stream gage on the Santa Maria River near Bagdad,Arizona. The period of record is 1966-1995.


Figure 13. Peak discharges during 1978, 1980, and 1993. Gages on these streams record multiple flood peaks andare the result of a series of storms during the winter months.

Figure 14. Precipitation values for the storm of February 27-March 3, 1978.


displays the largest peak flows occurring forthe above years on Burro Creek, Big SandyRiver and Santa Maria River. Most of thesefloods appear to have been felt throughout theBill Williams River basin, since stream gageson all three rivers typically record large flows.For the sake of brevity, only the most extremeevent for each year will be discussed in theparagraphs to follow.

Analysis of Selected Large Floods

Regional meteorological and hydrological datafrom storms that have impacted the Bill Will-iams River basin show similar large-scalemechanisms and basin responses to large-magnitude precipitation events. The largestfloods are generated from frontal and convectivewinter storms occurring in the winter monthsthat carry above-normal amounts of moistureinto the southwest. In this region, precipitationis enhanced by orographic effects from theMogollon Rim, which serves as a barrier tonortheastward-moving moist air. The storms aregenerally characterized by anomaloushydroclimatological conditions, in which ablocking high pressure ridge creates split west-erly flow such that a jet stream track occursfurther south than normal, and allows moistureto be delivered to the Southwest. This trackbecomes strengthened as the trough deepens andmay become stationary, allowing precipitationto fall over the region for a number of consecu-tive days. This jet stream condition may occurmany times throughout the winter to steerstorms over the region. Rapid warming andcooling trends create antecedent conditionsconducive to flooding as subsequent stormsdevelop. Some of the antecedent conditions thathave been present in the studied floods include:high soil moisture content, above averageprecipitation preceding the event, and highwater content in snow (Aldridge and Eychaner,1984; Aldridge and Hales, 1984; Chin, et al.,1991; House and Hirshboeck, 1997).

Although gages in the Bill Williams Riverbasin recorded large peaks for several years ateach gage, I selected three floods that wererecorded as large-magnitude floods at morethan one gage in the basin, suggesting thattheir effects were felt basin-wide.

Storm of February 27-March 4, 1978. Thestorm pattern for 2/27-3/4/78 follows thegeneral scenario outlined above. The winterstorm track was displaced further south than itwould normally occur and included warmingand cooling trends which delivered precipita-tion during January 15-24 and February 6-16and increased soil moisture content. Significantantecedent moisture in the form of aboveaverage rainfall occurred for many gages in theregion. On February 27, a low-pressure troughwas situated across Arizona, and broughtshowers to the Bill Williams River basin;moisture trailing behind a high-pressure ridgecontinued the precipitation trend. On March 1,warm, moist air from the tropics travelingnortheast collided with cold air over Arizonaand the mountain fronts of the transition zone,and produced the greatest precipitation of thestorm event for the gages in western Arizona.Areas with the largest precipitation totals onMarch 1st include the Bagdad (3.08 in) andWikieup (1.91 in) precipitation gages (Figure14). Total precipitation from February 27-March 4, 1978 for gages in this region rangefrom 3.01 in. at Prescott to 6.51 in. at Bagdad(Aldridge and Eychaner, 1984). Large dis-charges were recorded on March 1 at streamgages in the basin; a discharge of 36,500 cfs atthe Big Sandy River near Wikieup was thethird largest in the gage record. A discharge of23,100 cfs recorded at the Santa Maria Rivernear Bagdad was the peak of record. Peakdischarge in Burro Creek is unknown since thegage was not installed until 1980.

Storm of February 13-21, 1980. The storm ofFebruary 13-21, 1980 was characterized by a


series of six storms, which originated off thecoast of California and moved across centralArizona carrying warm moist air from thePacific. This event was preceded by aboveaverage rainfall during the month of Januarydue to the southerly displaced storm track;sizable floods of 19,800 cfs at the Big Sandyand Santa Maria stream gages occurred onJanuary 30,1980. The result was that theavailable water holding capacity of the soilwas decreased by the time the February stormsarrived. High flows were recorded throughoutthe basin and were reflected in the level ofAlamo Lake, which was higher than normaldue to prior flood events in 1978 and 1979(Chin, et al., 1991). The largest flood peakduring the storm of February 13-21 occurredon February 14th at the Burro Creek gage andmeasured 47,400 cfs. The Big Sandy and SantaMaria gages record flood peaks of unknownmagnitude on February 15th and estimated at38,500 and 19,500 cfs, respectively, on the20th. These peaks rank as the second largestdischarges in the gage record of all threetributaries.

Storm of February 7-10, 1993. The generalcirculation pattern during the winter of 1993was characterized by the development of ahigh-pressure area in the eastern North PacificOcean that persisted through the winter. Thispersistent high pressure caused a branching ofthe polar jet stream and forced the associatedPacific storm tracks further north and souththan they would normally occur. This broughtgreater cyclonic storm activity across the stateof Arizona. Sea surface temperatures (SST’s)remained above normal from December 1992through February 1993 such that warm moistair from the eastern Pacific was delivered to

Arizona, increasing rainfall totals during thewinter months. Persistence of the large scalecirculation anomaly and the repeated occur-rence of split westerly flow also acted toincrease precipitation above average levels,create antecedent conditions conducive toflooding, and bring frontal passages throughthe area, initiating alternating cooling andwarming trends. Storms occurred during bothwarm and cool episodes; rainfall was thedominant precipitation type during warmevents while some snowpack accumulated athigher elevations during cool events. Fourmajor storms occurred in Arizona during thewinter of 1993 that were responsible forflooding in Arizona: (1) January 6-9; (2)January 13-19; (3) February 7-10; (4) February18-21 (House and Hirshboeck, 1997). Al-though all four storms produced flood peaks inthe Bill Williams River basin, the storm ofFebruary 7-10 produced the largest peaks forthe 1993 water year and peaks of record forselect gages.

Meteorological conditions for this eventwere such that moisture directed north behind awarm front was quickly followed by a coldfront from the west, a scenario which providedabundant precipitation for the flood event.Precipitation ranged from approximately 60 to80 mm with >80 mm at the highest elevationsin the watershed. Flood peaks were producedon February 8-9, 1993 in the Bill WilliamsRiver basin. Burro Creek experienced its peakof record of 55,300 cfs on February 8th at 2300hrs. Three hours later, the Big Sandy nearWikieup also experienced its peak of recordmeasuring 68,700 cfs on February 9th at 0200hrs. Santa Maria near Bagdad measured a flowof 15,700 cfs at 0300 hrs, which is the fifthlargest discharge on record for this site.


In order to characterize the geomorphology ofthe channels and document any changes inchannel position, I first conducted an aerialphoto search for coverage of the Bill WilliamsRiver basin through the 20th century. Aerialphoto sets that were used included sets from1953-54, 1963, 1972, 1976, 1978, 1980, 1985,1987, 1990, 1992 and 1995. Other sources ofinformation include General Land OfficeSurveys from the early 1900’s and 7.5’ USGStopographic maps (Table 2). Because of thelength of the three rivers and the prevalence ofcanyon reaches that experience negligiblechannel change over short time periods, Iselected specific alluvial reaches to study indetail. Reach selection was based on aerialphoto reconnaissance using 1:60,000 ArmyMap Service (AMS) photos from 1953-54 toidentify alluvial reaches which have the poten-tial for channel change and the availability ofphotos for alluvial reaches. Reaches were thenmapped by aerial photo interpretation, delin-eating active channels, terraces, and olderalluvium, and then comparing mapped reachesfrom different years to identify channelchanges. Field checking and ground photosprovide additional documentation for recentchannel changes.

Channel Characterization

The Big Sandy River, Burro Creek, and SantaMaria River have channel configurations thatrange from narrow steep-walled canyons towide alluvial reaches. Confined reaches incanyons typically contain a single channel andhave little space for lateral channel migration.Alluvial reaches may also maintain a singleprimary flow channel; however, secondarychannels commonly are used during largemagnitude flow events. Some reaches must becharacterized as an intermediate type between

these two end members, as they are confinedwithin canyon walls relatively close to channelmargins but are wide enough such that sedi-mentation occurs along channel margins andsecondary channels can develop. Channelbedload sediments are composed mainly ofcobbles, gravel and sand with some boulders,which are transported mainly during floodevents. Sand, silt and clay size particles makeup the suspended load during flood events.

The Big Sandy River is mainly a wide andbraided alluvial system in Big Sandy Valleywith incision of less than 20 ft into higherterraces and alluvial surfaces when they arepresent. In its lower reaches, the river entersinto a series of alternating confined and allu-vial segments. Confined reaches stretch fromSection 25, T15N R13W, to the north side ofBurro Peak and through Signal Canyon in theSignal Mountain Quadrangle.

In contrast to the Big Sandy, Burro Creekcontains proportionately more confinedreaches rather than alluvial reaches. Alluvialreaches that do exist are narrower and morelimited in extent than those of the Big Sandy.Alluvial reaches may also contain “bottle-necks”, or abrupt contractions and expansionsin channel width. Alluvial reaches include:Happy Hollow Canyon to Salt Creek; BoulderCreek to Six Mile Crossing; U.S. Route 93bridge crossing; and confluence with BigSandy River to ~1 mile above that confluence.

The Santa Maria River in its upper reachesexhibits confined narrow canyons and interme-diate reaches that are composed of alluvialchannels which narrow to bottlenecks in places.The intermediate reaches extend from QuailSpring to sec 25 of T13N R9W and from SantaMaria Ranch to Black Canyon. The lower SantaMaria River, from the gaging station to AlamoLake, is a wide alluvial reach similar to the BigSandy River in Big Sandy Valley.

GEOMORPHOLOGY


Channel Change

The purpose of this section is to document thenature and magnitude of channel changes thathave occurred during the 20th century. For eachreach channels were delineated by aerial photointerpretation and were designated by theprevalence of channel sediment and the ab-sence of thick vegetation. Areas that containedchannels with significant vegetation were notdesignated as the main channel but insteadwere included as channel bars and overflowareas within the floodplain. Some mid-channelvegetated bars that were between channels arenot represented in order to simplify the figures.

Big Sandy River, Burro Creek Confluence toWikieup. The Big Sandy River from Wikieupto Burro Creek confluence is highly varied inits channel character. On the Wikieup 7 ½’quadrangle, the channel is wide and braided,but abruptly narrows as it flows south througha narrow canyon with tight bends. The low

flow channel in straight reaches of the canyonis moderately sinuous and slightly incised intoalluvium; around the bends the channel is wideand shallow and supports wetland vegetation.From its crossing with Signal Canyon road, theBig Sandy River returns to a wide alluvialchannel to the confluence with Burro Creek.The greatest potential for channel change liesin alluvial reaches; for this reason as well asphoto availability, the reach from Wikieup toits narrowing point serves as an illustration ofchannel changes that have occurred on thisriver reach of the Big Sandy. Two subreacheswill be discussed: the Wikieup reach is locatedin the northern half of the Wikieup quadrangle(Figure 15) and the Hopewell Ranch reach islocated in the southern half of the Wikieupquadrangle (Figure 16). Data sets used in thisstudy include General Land Office (GLO)cadastral surveys and aerial photos from 1953,1963, 1978, 1990, and 1995.

Channel changes have been expressedprimarily in the form of changes in vegetation

Table 2. List of aerial photo sources used for this geomorphologic study of the Big Sandy River, Santa Maria Riverand Burro Creek.


Figure 15. Historical channel changes on the Big Sandy River, Wikieup reach.

Figure 16. Historical channel changes on the Big Sandy River, Hopewell Ranch reach.


density and channel width and position, pre-sumably in response to flood flows. GLOsurveys completed in 1912 and 1917 show achannel markedly different from that of 1953and later in the historic record. In general, thechannel delineated by GLO surveyors waswide compared to later channel patterns of1953 and 1963 and had greater sinuosity thanany later channel documented by aerial photos.This is especially true of the Wikieup studyreach. Although GLO survey data is lessreliable than photo documentation, channelpositions seem reasonable by the followingarguments: (1) The greatest bends in thechannel occurred at the intersection of thechannel with section lines, which shouldaccurately depict channel position. (2) Largefloods are known to have occurred in the late1800’s and early 1900’s. Because surveyingfollowed this interval, it seems reasonable thatthe channel was wide and sinuous, reflectingthe influence of large flows on channel pattern.In fact, according to historic accounts, the late19th century to early 20th century was a timeperiod in which many channel changes wereoccurring. Many homesteads that had existedalong the Big Sandy River in these reaches andin northern Big Sandy Valley were washedaway during the 1880’s and later. Landholdingsthat are specifically noted include: the CornwallRanch (see Figure 16), the Boner Ranch, theScott Place, and the town of Greenwood (Cofer,1969). Channel incision of 5-10 ft and channelwidening in the Big Sandy River since the1860’s was also noted by Kam through writtencommunication to Davidson in 1966 (Davidson,1973). In 1953, multiple low-flow channelsexisted within the reach and were generallyconfined to a small portion of the larger chan-nel. Photos from 1963 show an increase invegetation in portions of active channels of1953 and abandonment of a right bank meandernear Wikieup in the Wikieup reach; generally,flow routes were very similar to that of 1953,but occupied a smaller portion of the channel.

The 1978 photos post-date major floodsduring the winter of 1977-1978 and reflectchanges caused by large flows. Although thechannel occupied a similar position to thechannel of 1963 deposition on the Big SandyRiver floodplain obstructed the pre-floodchannel and effectively in many cases, itswidth changed dramatically, and in someplaces diverged to flow around vegetatedchannel bars. In the Wikieup reach, channelwidening generally followed the 1963 channelsnorth of Wikieup and reoccupied and accentu-ated the 1953 meander near Wikieup. South ofWikieup, channel changes occurred in the formof right bank erosion along agricultural fields.Split flow existed downstream of the left bankerosion in which flow occupied the 1963channel and initiated a new channel thatdiverged toward the left bank and utilized aroad oblique to flow to widen its channel evenfurther. The two channels rejoined approxi-mately ½ mile downstream from their split.

The Hopewell Ranch reach also displaysprominent channel changes, the most spectacu-lar of which occurred at the Bronco Creekconfluence and downstream from theconfluence. On August 19, 1971, BroncoCreek experienced a flood estimated at 26,000to 30,000 cfs (House and Pearthree, 1995).This flood caused major channel changes inBronco Creek and deposition of fan material atthe junction of Bronco Creek and Big SandyRiver. The Bronco Creek fan rerouted it towardthe left bank, narrowing the new channel. Thischange in channel position is reflected in theposition of the river channel in 1978 and thedirect abutment of the Bronco Creek fan againstthis new channel. Rerouting toward the leftbank was followed downstream by right bankcutting along the agricultural fields of the bridgeterrace and erosion on the opposite bank. Down-stream of the U.S. Route 93 bridge, right bankcutting occurred, eroding vegetated channel barsand low terraces to bedrock.


Photos from 1990 capture a channel thatwas very similar to that of 1978 in both widthand position. However, the 1990 channel wasoverall more vegetated and one channel hadbecome dominant over another in areas of splitchannel flow. For example, the eastern channelwas dominant whereas the western channelwas semi-vegetated. In the Hopewell Ranchreach, the channel that had eroded agriculturalfields to accommodate large flow volumes washeavily vegetated.

In 1995, the channel was reduced insinuosity and increased in width in theWikieup reach with noticeable erosion on theoutside bends of the main channel, especiallyalong the agricultural fields on the right bankin Wikieup. In the Hopewell Ranch reach, the1995 channel was actually more sinuous thanthat of 1978 or 1990; channel width wasvariable with some reaches being wider andsome narrower than previous channel configu-rations. The most noticeable channel changesoccurred upstream of Bronco Creek, where thechannel widened and migrated toward the rightbank, and in the vicinity of the U.S. Route 93bridge, where bank erosion is evident along theoutside bends and the channel narrows as itflows underneath the highway bridge. Thechannel bends in this reach are nearly at 90degree angles. It is most likely that channeldifferences between 1990 and 1995 aerialphotos are largely a result of the floods of 1991and 1993.

Burro Creek, U.S. 93 Bridge Site. BurroCreek at the U.S. 93 bridge is located in theKaiser Spring 7 ½’ quadrangle in T14N R11Wand is characterized by channel bends withpoint bars and minor terraces in conjunctionwith a narrow canyon reach below the 93bridge. This reach is not perennial but typicallycontains water for a good portion of the year.The channel itself is characterized by pools andriffles typical of the creek as a whole. Twomajor tributaries join Burro Creek within this

reach; Raster Wash enters upstream of thebridge, while Black Canyon enters down-stream. In contrast with the selected reaches onthe other rivers, bedrock control affords lesspotential for channel change. This is especiallytrue in the vicinity of the bridge, where channelposition has changed minimally during the 20th

century. However, around tight bends in thecanyon, channel widening has occurred suchthat point bars and alluvial terraces havedeveloped. In these reaches the channel hasexperienced lateral scour during floods whichhave repeatedly truncated Black Canyon fanand exaggerated meander patterns on BurroCreek upstream of Raster Wash (Reach 1).Reach 1 is the focus of the following para-graphs as it demonstrates the most pronouncedchannel changes that have occurred within thestudy site. Aerial photos used to documentReach 1 were taken in 1953, 1963, 1976, 1980,1990, and 1995. GLO surveys and topographicmaps completed in 1980 also provide informa-tion on channel position.

Reach 1 is characterized by a meanderingchannel with low sinuosity issuing from a leftbend in upper Burro Creek (Figure 17). Twolarge terraces (A; B) joined by a narrow stripare located on right bank on the outside bendof the channel meander while smaller terracesoccupy the left bank. An overflow channel (C)carved into an alcove curves along the rightbank immediately upstream of Raster Wash. In1953, the channel was positioned centrally inReach 1, and skirted the left bank at C. Vegeta-tion was sparse in overbank area C, indicatingthat it must have experienced recent flow. In1963, the channel appeared very similar. By1976, however, lateral erosion had begun onthe right bank of terraces A and B, with contin-ued use of overflow area C. By 1990, thechannel had eroded the right bank to bedrockso that terraces A and B were no longer con-nected. Left bank terraces downstream werealso eroded, though not as dramatically. Thiseffectively routed flow toward the left bank,


Figure 17. Historical channel changes on Burro Creek just upstream of the U.S.Route 93 bridge.


isolating overbank area C and depositingsediment at its opening. Area C also experi-enced bank cutting along the edge adjacent tothe main channel. Low relief bars decreased invegetation density from 1976, while vegetationin area C increased in density. Reach 1 appearsvery similar in photos taken in 1995 withminimal reduction in vegetation density at theupper end of the reach. These changes areminimal given the fact that the 1993 flood wasthe peak of record. Thus, most channel changesoccurred between 1976 and 1990, and mostlikely resulted from the floods of 1978 and1980.

Santa Maria River, U.S. 93 Bridge Site. TheSanta Maria River channel at the U.S. 93bridge is an alluvial reach that issues from amore confined setting in Section 10 T12NR9W and reenters a confined setting down-stream of the U.S. 93 bridge in Section 21T12N R9W. In this reach, the Santa MariaRiver is bordered by bedrock, Pleistoceneterraces greater than 50 ft above the modernchannel, Holocene terraces, and active gravelbars. Changes have occurred in both channelposition and vegetation density on terraces andbars; a few of these features will be used tohighlight areas of change. These include thebridge terrace, located north of and underneaththe U.S. 93 bridge on the right bank and theleft bank terrace in a minor alcove on the leftbank downstream of the bridge (Figure 18).Site A designates the channel upstream of thebridge while site B represents the channeldownstream of the bridge. Photo sets used forthis analysis include 1953, 1963, 1978, 1990,and 1995.

1953: A single low flow channel was presentin area A along the left bank, and splitin area B to form two channels whichreconnected at the abrupt left benddownstream; low relief bars were notvegetated and appear to have experi-

enced a recent flood event, possiblythat of 1951.

1963: Channel configuration was very similarto that of 1953; the low flow channelalong the left bank was vegetated on itsmargins, and additional vegetation hadgrown on bars.

1978: Channel configuration was very similarto 1963 photos.

1990: Channel in area A remained verysimilar to past channel positions;however, vegetation along the low flowchannel in area B had been partlyremoved and a new channel cut moreabruptly from left to right; vegetationcontinued to increase on the bridgeterrace and terrace A. Multiple chan-nels had developed where previouslyonly two could be defined

1995: Channel formed in 1990 was vegetatedand cross cut by a new low flow chan-nel down the center of the river. Thechannel on the left bank in area A hadbecome a minor flow path. Bars hadbecome more vegetated since 1990;however, the main channel was veryclear of vegetation. In area B, split lowflow channels were evident once again.

The Santa Maria River channel upstreamfrom the U.S. 93 Bridge has been relativelystable during the past century, but the down-stream reach has experienced channel migra-tion from left to right and morphologies rang-ing from braided to split flow. Changes thathave occurred are reflected in differences inchannel position from 1978 to 1995. Channelchanges correspond to a time period in whichlarge magnitude floods were recorded on thisdrainage; these floods were likely responsiblefor much of the lateral migration and erosion.

Santa Maria River near Bagdad GagingStation to Alamo Lake. The Santa MariaRiver in its lower reaches is an alluvial channel


Figure 18. Historical channel changes on the Santa Maria River at U.S.Route 93 bridge.


Figure 19. Historical channel changes on the Santa Maria River near Valencia Wash.


Figure 20. Historical channel changes on the Santa Maria River near Date Creek.


comparable in morphology and scale to allu-vial reaches of the upper Big Sandy River.Channel widths range from less than ¼ mile atthe gage near Bagdad to nearly one mile at itswidest in Section 15 T11N R11W. Channelmorphology is best described as a braidedsystem with semi-vegetated bars betweenchannels. This section reviews the generalchannel changes that are evident in this reachand highlights a few areas that show dramaticchanges for this photo set. General channelchanges include decreases in channel widthand increases in sinuosity from 1953 to 1976,and increases in channel width and decreasesin sinuosity from 1987 to 1992 (Shafroth,written communication).

Figures 19 and 20 highlight two areas onthe Santa Maria alluvial reach that have experi-enced changes in channel width and sinuosity.The first area is located in the vicinity ofValencia Wash and demonstrates the trend inchannel width (Figure 19). In the early 1900’s,surveyors recorded the position of the mainchannel as they surveyed section lines for theTownship and Range grid. Along the westsection line of sec 13 T11N R11W, the SantaMaria River was positioned in what is now asemi-vegetated bar. By 1953, channel positionremained relatively stable near the gage buthad shifted toward the left and right (north andsouth) of its previous position along the west-ern border of section 13 (Figure 18). In 1953,the active flow area was wide with multiplechannels transporting flow. Photos from 1979show an overflow of sediments onto cultivatedland on the right bank immediately down-stream of the gage, and abandonment of thesouthernmost channel upstream of ValenciaWash. In general, channels within the braidedsystem were diminished in size and partlyvegetated when compared to those that existedin 1953, while the active channels were less

vegetated. By 1992, the channels upstreamhadwidened to incorporate more of the bottom-land into the primary flow path, but did notreach the extent of the 1953 flow pattern

Near Date Creek, channel position differsmarkedly during the 20th century (Figure 20).In 1917, the main channel as noted by thesurveyor was located along the right bank ofthe Santa Maria River. In 1953, the channel inthe vicinity of Date Creek was composed oftwo channels occupying the left and rightbanks. Major channel changes are evident onthe 1967 topographic base, where the northerlychannel is dominant and cuts to the left banknear the mouth of Date Creek. This pattern isaccentuated in 1979. By 1987, the dominantchannel upstream of Date Creek had shifted toa more central position; continued left bankcutting shifted the channel bend downstream ofDate Creek. This trend continued into 1992and in effect, decreased the sinuosity of thisreach.

In summary, the most prominent channelchanges occurred from 1953 to 1979, whenchannel width decreased near Valencia Washand sinuosity increased near Date Creek, andfrom 1979 to 1987, when the channel experi-enced left bank lateral erosion near Date Creekand channel widening near Valencia Wash.The largest floods in the historic record at theBagdad gage occurred during the winters of1978, 1980, 1991, and 1993 and seem to be themost obvious catalysts for channel change onthe Santa Maria River. Generally, changes onthe Santa Maria River took the form of channelwidening, increased sinuosity, and decreasedvegetation density in response to large floods.Study reaches on the Big Sandy River andBurro Creek show similar records of channelchange, both in the type of change and occur-rence during periods of large floods.


The Big Sandy River, Santa Maria River, andBurro Creek flow through rugged terrain ofsome of the most remote lands in Arizona.Although the three rivers do not have abundantwater resources, they all contain perennialreaches along their lengths. The Big SandyRiver has perennial surface water along someportions of its length and is fed by upperwatershed perennial streams such as TroutCreek. Burro Creek maintains perennial sur-face water along most of its length, whichduring the driest intervals exists as discontinu-ous pools of standing water. The Santa Mariais the driest of all three rivers, with most of itssurface water supplied by tributaries in theupper watershed and in the main channel for10-15 miles near its mouth. Low flow dis-charge measurements of each river are gener-ally less than 10 cfs. Flood events are dramaticin comparison. Peak flow estimates for theserivers are estimated at 68,700 (2/9/93) for the

CONCLUSIONS

Big Sandy River, 53,300 cfs (2/8/93) for BurroCreek, and 23,100 cfs (3/1/78) for the SantaMaria River. The largest storm events resultedin heavy, basin-wide precipitation events thatcaused floods of varying magnitudes at allthree gages.

Substantial channel changes have oc-curred in response to large floods during thehistorical period. Channel changes typicallyhave involved erosion of stream terraces andtributary fans around outside bends, removal ofvegetation within the channel, and channelwidening. Analysis of channel change spansthe 20th century; according to historic accounts,severe channel changes occurred around theturn of the century, when large floods washedaway farmlands and the old mining town ofGreenwood along the Big Sandy River. Themost dramatic channel changes documented byaerial photo interpretation occurred from 1978to 1995.


Aldridge, B.N., 1971, Northwestern Arizona, Floods ofDecember, in southwestern United States, inRostvedt, J.O., and others, Summary of floods in theUnited States during 1966: U.S. Geological SurveyWater-Supply Paper 1870-D, p. D62-D69.

Aldridge, B.N., 1972, Investigation of floods from smalldrainage basins in Arizona, in Proceedings of theTwenty-first Annual Conference on Roads andStreets, held at University of Arizona, April 13-14,1972: Tucson, University of Arizona, ArizonaTransportation and Traffic Institute, p. 107-126.

Aldridge, B.N., 1978, Unusual hydraulic phenomena offlash floods in Arizona, in Peck, E.L, ed., Conferenceon flash floods, Hydrometeorological aspects, LosAngeles, Calif., May 2-3, 1978: American Meteoro-logical Society, p. 117-120.

Aldridge, B.N., and Eychaner, J.H., 1984, Floods ofOctober 1977 in Southern Arizona and March 1978in central Arizona: U.S. Geological Survey WaterSupply Paper 2223, 143 p.

Aldridge, B.N., and Hales, T.A., 1984, Floods ofNovember 1978 to March 1979 in Arizona and West-Central New Mexico: U.S. Geological Survey WaterSupply Paper 2241, 149 p.

Anderson, T.W., 1995, Summary of the Southwestalluvial basins, regional aquifer system analysis,south-central Arizona and parts of adjacent states:U.S. Geological Survey Professional Paper 1406-A,p. A1-A33.

Anderson, T.W., and Johnson, A.I., eds., 1986, Regionalaquifer systems of the United States - southwestalluvial basins of Arizona: Bethesda, Md., AmericanWater Resources Association Monograph Series no.7, 116 p.

Baker, 1987, Paleoflood hydrology and extraordinaryflood events: Journal of Hydrology, v. 96, p. 79-99.

Brown, D.E., Carmody, N.B., and Turner, R.M., 1978,Drainage map of Arizona showing perennial streamsand some important wetlands, 1 sheet, scale1:1,000,000.

REFERENCESCarmody, T., 1980, A critical examination of the

‘largest’ floods in Arizona: A study to advance themethodology of assessing the vulnerability of bridgesto floods for the Arizona Department of Transpor-tation, General Report 1: Tucson, University ofArizona, College of Engineering, EngineeringExperimental Station, 53 p.

Chin, E.H., Aldridge, B.N., and Longfield, R.J., 1991,Floods of February 1980 in southern California andcentral Arizona: U.S. Geological Survey ProfessionalPaper 1494, 126 p.

Clark, R.J., 1979, Hydrogeochemical and streamsediment reconnaissance basic data report for thePrescott NTMS quadrangle, Arizona: U.S. Depart-ment of Energy Report GJBX-122(79), 92 p., 6sheets, 7 microfiche.

Clark, T.C., 1975, The Arizona water plan, a statusreport, in Arizona Watershed Symposium, 19thannual, Phoenix, September 24, 1975, Proceedings:Arizona Water Commission Report no. 7, p. 9-23.

Cofer, Irene Cornwall, 1969, The Lunch Tree: Theo.Gaus Sons, Inc., Brooklyn, NY, 210 p.

Costa, J.E., 1978, Holocene stratigraphy in floodfrequency analysis: Water Resources Research, v. 14,p. 626-642.

Davidson, E.S., 1973, Water-resources appraisal of theBig Sandy area, Mohave County, Arizona: ArizonaWater Commission Bulletin 6, 40 p., 2 sheets, scale1:32,000.

Ely, L.L., 1992, Large floods in the southwestern UnitedStates in relation to late-Holocene climatic variations:Tucson, University of Arizona, Ph.D. dissertation,326 p.

Enzel, Yehouda, Ely, L.L., House, P.K., Baker, V.R.,and Webb, R.H., 1993, Paleoflood evidence for anatural upper bound to flood magnitudes in theColorado River Basin: Water Resources Research, v.29, no. 7, p. 2287-2297.


Evans, L.G., and Haimson, J.S., 1982, SWAB/RASAaquifer parameter study, final report for U.S. Geo-logical Survey contract 14-08-0001-18268: ArizonaDepartment of Water Resources, 21 p.

Foreman, Grant, 1941, A Pathfinder in the Southwest:The Internerary of Lieutenant A.W. Whipple duringhis Explorations for a Railway Route from FortSmith to Los Angeles in the Years 1853-1854:University of Oklahoma Press, Norman, 298 p.

Freethey, G.W., and Anderson, T.W., 1986,Predevelopment hydrologic conditions in the alluvialbasins of Arizona and adjacent parts of Californiaand New Mexico: U.S. Geological Survey Hydro-logic Investigations Atlas HA-664, 3 sheets, scale1:500,000.

Garrett, J.M., and Gellenbeck, D.J., 1991, Basincharacteristics and streamflow statistics in Arizona asof 1989: USGS Water Resources InvestigationsReport 91-4041, 612 p.

Gillespie, J.B., Bentley, C.B., and Kam, W., 1966, Basichydrologic data of the Hualapai,

Graf, J.B., Webb, R.H., and Hereford, Richard, 1991,Relation of sediment load and flood-plain informa-tion to climatic variability, Paria River drainagebasin, Utah and Arizona: Geological Society ofAmerica Bulletin, v. 103, p. 1405-1415.

Grimm, N.B., and Fisher, S.G., 1986, Nitrogen limitationpotential of Arizona streams and rivers: Arizona-Nevada Academy of Science, Journal, v. 21, p. 31-43.

Halpenny, L.C., and others, 1952, Ground water in the GilaRiver basin and adjacent areas, Arizona—a summary:U.S. Geological Survey Open-File Report, 224 p.

Hereford, Richard, Thompson, K.S., Burke, K.J., andFairley, H.C., 1996, Tributary debris fans and the lateHolocene alluvial chronology of the Colorado River,eastern Grand Canyon, Arizona: Geological Societyof America Bulletin, v. 108, no. 1, p. 3-19.

House, P.K., 1993, The Arizona floods of January andFebruary 1993: Arizona Geology [Arizona Geologi-cal Survey], v. 23, no. 2, p. 1, 6-9.

House, P.K., 1997, An unconventional,multidisciplinary approach to evaluationg themagnitude and frequency of flash floods in smalldesert watersheds: Arizona Geological Survey Open-File Report 97-9, 61 p.

House, P.K., and Hirschboeck, K.K., 1997,Hydroclimatological and paleohydrological contextof extreme winter flooding in Arizona, 1993, inLarson, R.A., and Slosson, J.E., eds., Storm-InducedGeologic Hazards: Case Histories from the 1992-1993 Winter in southern California and Arizona:Boulder, Colorado, Geological Society of AmericaReviews in Engineering Geology, v. XI., p. 1-24.

House, P.K., and Pearthree, P.A., 1994, A geomorphologicand hydraulic evaluation of an extraordinary flooddischarge estimate: Bronco Creek, Arizona: ArizonaGeological Survey Open-File Report 94-19, 21 p.

House, P.K., and Pearthree, P.A., 1994, Hydraulic andgeomorphic re-evaluation of an extraordinary flooddischarge estimate: Bronco Creek, Arizona [abs.]:Geological Society of America Abstracts withPrograms, v. 26, no. 7, p. A.235-A.236.

House, P.K., and Pearthree, P.A., 1995, A geomorphologicand hydraulic evaluation of an extraordinary flooddischarge estimate: Bronco Creek, Arizona: WaterResources Research, v. 31, no. 12, p. 3059-3073.

House, P.K., Pearthree, P.A., and Wood, M.L., 1999,Hydrologic and geomorphic characteristics of the BillWilliams River, Arizona: Arizona Geological SurveyOpen-File Report 99-4, 47 p., 2 sheets, scale 1:24,000.

Johnson, M.G., 1972, Placer gold deposits of Arizona:U.S. Geological Survey Bulletin 1355, 103 p., 1 sheet.

Kochel, R.C., and Baker, V.R., 1982, Paleofloodhydrology: Science, v. 215, p. 353-376.

Macklin, M.G., Rumsby, B.T., and Newson, M.D.,1992, Historical floods and vertical accretion of fine-grained alluvium in the Lower Tyne Valley, North-east England, in Billi, P., Hey, R.D., Thorne, C.D.,and Tacconi, P., Dynamics of Gravel BedRivers:New York, John Wiley and Sons Ltd., p. 573-589.

Markman, S.G., 1996, Assessmemt of water resourceconditions in support of instream flow water rights,Burro Creek, Arizona: U.S. Bureau of Land Manage-ment, 25 p.


Melton, M.A., 1965, The geomorphic and paleoclimaticsignificance of alluvial deposits in southern Arizona:Journal of Geology, v. 73, no. 1, p. 1-38.

Menges, C.M., and Pearthree, P.A., 1989, Late Cenezoictectonism in Arizona and its impact on regionallandscape evolution, in Jenney, J.P., and Reynolds,S.J., 1989, Geologic evolution of Arizona: Tucson,Arizona Geological Society Digest 17, p. 649-680.

Patton, P.C., Baker, V.R., and Kochel, R.C., 1979,Slackwater deposits: a geomorphic technique for theinterpretation of fluvial paleohydrology, in Rhodes,D.P., and G.P. Williams, eds., Adjustments of theFluvial System: Dubuque, Iowa, Kendall/Hunt, p.225-253.

Richmond, D.L., and Richardson, M.L., 1974, Generalsoil map and interpretations, Mohave County,Arizona: U.S. Department of Agriculture, SoilConservation Service, 48 p., 1 sheet, scale 1:600,000.

Robertson, F.N., and Garrett, W.B., 1988, Distributionof fluoride in ground water in the alluvial basins ofArizona and adjacent parts of California, Nevada,and New Mexico: U.S. Geological Survey Hydro-logic Investigations Atlas HA-665, 3 sheets, scale1:500,000.

Sacramento, and Big Sandy Valleys, Mohave County,Arizona: Arizona State Land Department WaterResources Report no. 26, 39 p.

Sellers, W.D., and Hill, R.H., 1974, Arizona Climate,1931-1972, Tucson, Arizona, The University ofArizona Press, 616 p.

Spencer, J.E., and Reynolds, S.J., 1989, Middle Tertiarytectonics of Arizona and adjacent areas, in Jenney,J.P., and Reynolds, S.J., 1989, Geologic evolution ofArizona: Tucson, Arizona Geological Society Digest17, p. 649-680.

Stedinger, J.R., and Cohn, T.A., 1986, Flood frequencyanalysis with historical and paleoflood information:Water Resources Research, v. 22, p. 785-793.

Tellier, A.H., 1973, Geothermal waters of Arizona:Tempe, Arizona State University, M.S. thesis, 30 p.

Texas Instruments, Inc., 1982, National uraniumresource evaluation, variable premise evaluationsystem data report, interpretation ofhydrogeochemical and stream sediment reconnais-sance data, Williams quadrangle, Arizona: U.S.Department of Energy Report GJD-93(81), 264 p.

Tucci, Patrick, 1989, Geophysical methods for water-resources studies in southern and central Arizona, inSymposium on the Application of Geophysics toEngineering and Environmental Problems, Proceed-ings: Denver, Society of Engineering and MineralExploration Geophysicists, p. 368-383 [available forinspection at Arizona Geological Survey, 416 W.Congress, Suite 100, Tucson, Ariz.].

United States Army, Office of the Chief Engineer, 1944,Bill Williams River and Tributaries, Arizona,Washington, D.C.: Government Printing Office, 78th

Congress, 2d Session, House Document No. 625.

United States Bureau of Land Management, 1990,Kingman Resource Area, Resource management planand environmental impact statement - Draft: U.S.Bureau of Land Management, 282 p., 8 sheets, scales1:168,960 and 1:320,000 [BLM-AZ-PT-90-012-1600].

United States Bureau of Land Management, 1993,Kingman Resource Area, Proposed resource manage-ment plan and final environmental impact statement:U.S. Bureau of Land Management, 606 p. [BLM/AZ/PL-93/009-4410].

Wagoner, J.L., 1979, Hydrogeochemical and streamsediment reconnaissance basic data report forWilliams NTMS quadrangle, Arizona: U.S. Depart-ment of Energy Report GJBX-71(79), 74 p., 6 sheets,scale 1:250,000, 5 microfiche.

Webb, R.H., and Betancourt, J.L., 1992, Climaticvariability and flood frequency of the Santa CruzRiver, Pima County, Arizona: U.S. GeologicalSurvey Water-Supply Paper 2379, 40 p.

Wendt, G.E., 1972, General soil map, Yavapai County,Arizona; original text and map by G.E. Wendt, 1968-69; text and mapping legend revised and interpreta-tions added April 1972 by M. L. Richardson: U.S.Department of Agriculture, Soil ConservationService, 46 p., 1 sheet, scale 1:500,000.


Wendt, G.E., Winkelaar, P., Wiesner, C.W., Wheeler,L.D., Meurisse, R.T., Leven, A., and Anderson, T.C.,1976, Soil survey of Yavapai County, Arizona,western part: U.S. Department of Agriculture, SoilConservation Service, 121 p., 176 sheets, scales1:31,680 and 1:570,240 [errata: symbols on originalGeneral Soil Map are incorrect].

Wilson, L.G., DeCook, K.J., and Neuman, S.P., 1980,Regional recharge research for southwest alluvialbasins - final report, U.S. Geological Survey contract14-08-0001-18257: Tucson, University of Arizona,Water Resources Research Center, 389 p.

Wolcott, H.N., Skibitzke, H.E., and Halpenny, L.C.,1956, Water resources of Bill Williams River Valleynear Alamo, Arizona, in Contributions to the hydrol-ogy of the United States: U.S. Geological SurveyWater-Supply Paper 1360-D, p. 291-319, 1 sheet,scale 1:64,000.

Wright, J.J., 1971, The occurrence of thermal ground-waterin the Basin and Range Province of Arizona, in Evans,D.D., ed., Hydrology and water resources in Arizonaand the Southwest, v. 1: American Water ResourcesAssociation, Arizona Section, and Arizona Academy ofScience, Hydrology Section, Annual Meeting, Tempe,Ariz., 1971, Proceedings, p. 269-290.


APPENDIX A

Stage-discharge rating curves for USGS stream gages

Figure A1. Composite rating curve for the Big Sandy River nearWikieup, Arizona.

Figure A2. Rating curve for Burro Creek at Old US 93 Bridge, Arizona.rating table 4; start date (time): 9-30-90 (2400)


Figure A3. Rating curve for Santa Maria River near Bagdad, Arizona.rating table 4; start date/time: 9/30/90 (2400)

Figure A4. Rating curve for Francis Creek near Bagdad, Arizona.rating table 5; start date (time): 10-01-92 (0001)


APPENDIX B

Photographs in this appendix are not a comprehensive set that document river morphology butrather provide a glimpse of selected reaches for each river. The Big Sandy River and BurroCreek photo sets also contain comparative photos, the first photo most likely taken in the early1960’s and the second in 1998. A photo index describes the location and characteristics for eachphoto in the following paragraphs.

Photo documentation of river reaches


Figure B1. Big Sandy River at U.S. Route 93 Bridge—photo comparison: looking upstreamLocation: SW ¼ SW ¼ sec 1 T15N R13W

A historical photo (top) taken by Carlos Elmer1, who was active in Mohave County between1948 and 1975, depicts an autumn flood on the Big Sandy. The lower photo taken in 1998 byJ.Klawon illustrates channel morphology of the Big Sandy in low flow conditions. Duringperiods of low flow, the Big Sandy occupies a narrow semi-entrenched channel and is boundedby an extensive floodplain of fine sediments and full-grown trees outside of what is presumablythe high flow channel. Extreme events occupy a greater portion of the floodplain; in the Elmerphoto, floodwaters are seen inundated even well vegetated portions of the floodplain and dis-lodging some of this debris to carry downstream.

1 Carlos Elmer, “Big Sandy River in flood nr. Wikieup.” CP Photographs Collection, Arizona Collection (CP SPC 214A: 4462). Dept. ofArchives and Manuscripts, University Libraries, Arizona State University, Tempe.


Figure B2. Big Sandy River downstream of Trout CreekLocation: NE ¼ sec 35 T18N R13W

Sandy and gravelly wide alluvial channel with low mid-channel terraces infrequently inundated.

Figure B3. Big Sandy River near SignalLocation: NW ¼ SE ¼ sec 25 T15N R13W

Entrenched moderately sinuous channel of fine-grained sediment with abundantvegetation along channel margins and filling the channel in some locations.


Figure B4. Burro Creek high bridge—photo comparison: looking upstreamLocation: NW ¼ SW ¼ sec 19 T14N R11W

The high bridge over Burro Creek on U.S. Route 93 was completed sometime between 1963 and1976. The photograph on the top by Carlos Elmer was most likely that the photo was taken soonafter the bridge’s completion; The high flow event shown in this photo narrows the probablydates to August or September of 1963, 64 or 71. The photo on the right was taken in the fall of1998 by J. Klawon. Comparison between the two photos demonstrates the differences andsimilarities between high flow and low flow channel patterns. Floodwater occupies 2 channels inthe foreground of Elmer’s photo, with the channel along the right bank as the main channel andthe channel along the left bank as an overflow channel with a noninundated bar separating thetwo. Channel configuration in 1998 is very similar with the low flow channel hugging the rightbank and a vegetated channel bar on the left. Much of the existing vegetation along the left bankis inundated by the floodwaters of the Elmer photo; generally there is less vegetation in thechannel than that of 1998 (note the absence of species on the downstream gravel bar on the leftbank). The early photo may have been taken in the waning stages of this particular flood, shownby the evidence of recent flow on the gravel bar to the far right on the photo.

2 Carlos Elmer, “Low Bridge at Burro Creek.” CP Photographs Collection (CP SPC 214A: 2958). Dept. of Archives and Manuscripts,University Libraries, Arizona State University, Tempe.


Figure B5. Burro Creek low bridge—photo comparison: looking downstreamLocation: NW ¼ SW ¼ sec 19 T14N R11W

The low bridge at Burro Creek was in use as the primary traffic route over Burro Creek for U.S.Route 93 until the completion of the high bridge. After this, the low bridge continued to bemaintained and in 1980, a USGS stream gage was installed on its downstream side. Note that thisgage can be seen in the 1998 (bottom) photo but not in Elmer’s (top) photo2. Possible dates forthis photo range from 1948 to 1975; the truck on the bridge appears to be a make and model fromthe early 1960’s (Ferguson, personal communication). Although channel position in both photosis very similar, vegetation density had decreased from the Elmer’s photo to 1998. Major differ-ences in vegetation occurred in the main channel downstream of the low bridge, and on the leftbank low terrace upstream of the bridge.

3 Carlos Elmer, “Low Bridge at Burro Creek.” CP Photographs Collection (CP SPC 214A: 2958). Dept. of Archives and Manuscripts,University Libraries, Arizona State University, Tempe.


Figure B6. Burro Creek near Big Sandy River confluenceLocation: sec 19 T12N R9W

Sandy and gravelly wide alluvial channel with low fine-grained to gravelly Holocene terraces(center of photo) and higher Pleistocene terraces (banks on sides of photo).

Figure B7. Santa Maria River near U.S. Route 93 bridgeLocation: NW ¼ NE ¼ sec 22 T12N R9W

This photo was taken looking west-northwest (upstream) from a high Pleistocene terrace on theleft bank. The river bends from right to left on the photo around the low terrace of Santa MariaRanch and turns sharply to the southwest off the bottom of the photo. In this reach, the riverchannel is composed mainly of sand and gravel with vegetation along channel margins and lowterraces. The low flow channel occupied the outside bend of the river in this photo.


Figure B8. Santa Maria River near Bagdad, AZLocation: SE ¼ SE ¼ sec 12 T11N R11W

The Santa Maria River channel issues from the upstream reach of intermediate width to a widealluvial braided channel of mainly sands and gravels. The gaging station for the Santa Maria issituated on the bedrock protrusion to the left of the photo. Although it is not evident in thisparticular photograph, most portions of the alluvial reach from the gage to Alamo Lake containrelatively stable vegetated channel bars.

Documents

Hydrology and Geomorphology of the Santa Maria …repository.azgs.az.gov/sites/default/files/dlio/files/...Page 6 Hydrology and Geomorphology of the Santa Maria and Big Sandy Rivers