Volume 14 NO.2 Eorrh Sciences and Mineral P-esources in Arizona Summer 1984

"

by H. Wesley Peirceand Peter L. Kresan

::lJoinft Q1jhat

crJomes @/Vaturatly

"

1983

Figure 1. Raging waters of the brimful Santa Cruz River. Looking upstream to the south from SI. Mary'sBridge. Photo taken on October 2,1983 by Peter Kresan.

INTRODUCTION

Excerpts from The Arizona Daily Star,Tucson:

Sept. 10, 1887:About 2 o'clock yesterday morning, it startedto rain hard and poured unceasingly untilafter daylight, flooding many parts of the cityand causing great loss to railroad east andwest of Tucson.... Mr. Hancock's apiarywas two feet under water.... Mr. Wetmoretold a Star man yesterday that there was 9.5feet of water in the river and that trees andother articles were floating with the cu rrent ata very brisk rate....

Dec. 24, 1965:FLOOD PERIL CONTINUES AS SEWERSWASH OUT; STATE ASSISTANCE SOUGHT.... FLOWING WELLS AREA STUNNED BYWILD RILLITO. The roiled, brown waters ofthe flooding Rillito Creek tore into two trailerparks in the Flowing Wells area yesterday,demolishing two trailers. Residents bitterlytermed it a disaster and scorned publicofficials for apathy about their plights....

Dec. 31, 1965:RUNOFF CRISIS REPEATS ITSELF. Rainand rapidly melting snow in the Catalinasswelled the Rillito River again yesterday....

Oct. 4, 1983:ONLY TWO IN MARANA HAD FLOOD IN­SURANCE. Only two national flood insur­ance policies were issued in the Marana areabefore flooding inundated the whole area,because town officials "didn't believe it floodsthere," a flood insurance official said....

RIVERS' CURVES FIGHT CITY'S STRAIGHTLINES.... Where the rains had collided withroads, houses, and power lines, the floodripped, swallowed, and snapped....

Excerpts from The Arizona Daily Star,Tucson:

Dec. 23, 1914:WORST FLOOD FOR GENERATIONS....LOSS OF SEVERAL LIVES UP THE VAL­LEY.... BELOW MARANA AND CORTARO,TRACK OF MAIN LINE INUNDATED FORABOUT 4 FEET: 25 MILES OF TRACKWASHED OUT.... TWO PEOPLE BELIEVEDDROWNED AT SAHUARITA; 25 PEOPLEMAROONED ON HOUSETOPS AND WIND­MILLS....

Excerpts from The Tucson Citizen:

Oct. 3, 1983:FLOODS RAM TUCSON.... ROARINGRIVERS EAT AWAY BRIDGES, HOMES....MARANA IS SUBMERGED; RESIDENTSEVACUATED.... HOMES, LIFE POSSES­SIONS SWALLOWED BY SANTA CRUZ....4,000 ARIZONANS EVACUATED IN FACEOF MASSIVE FLOODS....

Oct. 17, 1983:THE FLOOD OF '83 - A SPECIAL REPORT:

THE BIG ONE. This was the flood we'll re­member. This was the flood our children andgrandchildren will be told about time andagain as we warn of the awful power of thearea's normally dry rivers. At least 10,000Arizonans were at least temporarily homelesswhen flood dangers forced evacuation of

Page 2

entire communities. Other areas were cut offfor days as the rivers toppled bridges andblocked roads....

CLIFTON'S BEST PREPARATIONS FAILED.Clifton knows floods.... Most of the city's4200 residents were evacuated. Over 600h~mes and 86 of Clifton's 126 businesseswere damaged severely....

LOSSES TOTAL HUNDREDS OF MIL­LIONS....

If experience is a great teacher, thenrepetitious experience should be d.oublyeffective as an educator. Teaching ISineffective, however, if the pupils aren'tpaying attention. In October 1983, na­ture taught many Arizonans a lessonthat they will not soon forget. Certainly,the natural events that occurred duringthat time inspired many questions, andquestions inevitably must precede an­swers. Local reaction varied from one oftragedy among those who were directlyaffected to one of glee among those whoenjoyed watching nature "do its thing,"even at the expense of humankind. En­gineers learned a great deal and. arealready applying their newly acquiredexperiences and insights.

The dynamics of the hydrologic eventcan be analyzed in detail and are prob­ably among the easiest aspects of theevent to discuss. There would be noconcern about the event itself, however,were there not direct social, economic,and political impacts and implications.Where there is an interface with humanactivity, some natural earth proce~ses

can be both hazardous and damaging.In the desert country of southern Ari­zona, processes associated with waterrunoff are the dominant natural hazard.

The events of October 1983 providethe incentive for this brief and basicreview of the nature of the runoff hazardin this desert region. Although all of theexamples are from the Tucson area, th.eprinciples involved are generally appli­cable to other desert regions.

DYNAMICS OF DESERT RIVERS

The network of natural drainagewaysin the desert country of southern Ari­zona is exceedingly intricate. The inte­grated network is a part of the larg.er Gilaand Colorado River systems, which arenaturally designed to carry surfacewaters toward the Sea of Cortez. Al­though most of the network occupiesvalleys and foothills, the headwaters arein the higher reaches of adjacent moun­tain ranges. Many of these ranges are amile or so higher than the desert valleysand are, therefore, subjected to muchhigher precipitation rates. The excessprecipitation in the mountains i~ con­veyed to the valleys, where drainagesare naturally enlarged to accommodatethe total flow.

Oureau of Geology and Mineral Technology

Within an integrated drainage net­work, the size of any particular flow orrunoff event is proportional to the areareceiving precipitation. Only at tim~s ofregional rainfall is it possible to activateall of the existing drainages. Such wasthe case in October 1983. Regionally,the land surface had been well-wettedby previous rains; then, in 2 days: aidedby tropical storm Octavo, about 6lnchesof rain fell. More rain fell in the moun­tains, swelling waterways even further.

Those who witnessed one or more ofthe major drainages in action were re­minded of the frightening power of rush­ing, roily water (Figure 1). A flow.rate of25 000 cu bic-feet-per-second (estl matedfo; Rillito Creek) is about equivalent toan 800-ton mass moving past a givenpoint each second. (An 800-ton massweighs more than two 747 Jumbo Jets,which weigh 775,000 pounds each.)

A basic law of physics states that anymass, once in motion, will continue in astraight line until acted upon by anoutside force. What happens when amass of moving water is "asked" to flowaround a bend in a channel? The onlyway the moving water can be made toturn is if the outside bank exerts enoughforce to redi rect the flow. If the banks arerelatively weak, as they tend to be insouthern Arizona (Figures 2a-d), therewill be a compromise: the river will con­tinually "chew" at the bank in its effort toflow in a straight line, but will eventuallyturn in response to the resistance thatthe wasting bank will offer. This "chew­ing" causes banks at curv~s, and thusthe curves themselves, to migrate down­stream. The amount of land removed is afunction of bank strength; radius of cur­vature; rate, amount, and duration offlow; etc.

There are, therefore, two measure­ments used to describe the extent ofbank alteration: (1) the amount ofstraightening in the direction of riverflow; and (2) the distance betwe~n oldand new bank, measured perpendicularto the direction of flow. For the largehistorical runoff events, these measure­ments ranged from near zero to about1,500 feet, and from near zero to about600 feet, respectively, for a single bank.In other words, an area as large as10 acres is known to have been trans­posed from riverbank to river bottom.Losses of up to 5 acres occurred atseveral sites along the Rillito lastOctober.

The vulnerability of banks to destruc­tion is also a function of geometricposition at any given time. Like a cueball, rapidly flowing water literallybounces from one side of a stream to theother, wherever there are curves in thechannel. Unless they are adequatelystabilized, these curves will not remainsteadfast for long.

Summer 1984

Flood vs. Flow Event

It is conceptually important to dis­tinguish between flood and flow. eventsin a desert region. Much confUSion hasarisen because of a lack of appreciationfor the contrasti ng processes involved inthese two types of runoff. A flood occurswhen discharge exceeds the capacity ofan active channel to contain the flow. Inother words, a true flood refers to dis­tinct overbank flow, called flood flow. Ifthere is no flooding, the runoff event issimply a flow event. Flooding may lo­cally occur, but elsewhere along thesame drainage, runoff may be totallycontained within well-defined banks.Flooding is an unusual flow condition,whereas confined flow is the norm.

Most of the damage to humankindwithin the Tucson metropolitan regionhas been done under nonflood condi­tions by the collapse and erosion of riverbanks, especially on the outside of me­ander bends.

If nonflood runoff alters banks enoughto undercut "flood-protected" buildings,regulations that require constructi~n

above a certain elevation on a floodplainwill not spare buildings from disaster.Many of the more dramatic picturestaken along Rillito Creek, Tanque VerdeWash, Pantano Wash, and the SantaCruz River, on or after October 2, 19.83,were related to nonflood bank-cuttingand bank collapse (Figures 2a-d). Evenso, adequate setback regulations. havebeen slow in coming. In recent times,each new experience with severe non­flood runoff damage has led to morestringent setback regulations, especial­ly in areas where there is inadequatebank protection. Because of the Oct~­

ber 1983 experience, Pima County engi­neers now consider "inadequate" anybank protection that is not the relativelynew soil-cement type. At the presenttime SOD-foot setbacks are requiredwhe;e banks are not protected by soilcement. A land user, however, can re­quest a variance if the reque~t is a?e­qU'ately supported by englneenngstudies. A SOD-foot setback might seemlarge, but at selected times and placeson Rillito Creek, bank erosion from asingle runoff event has exceeded thisamount.

Actual flooding did take place wherechannel capacity was not able to containrunoff. The Marana area was the mostdramatic example. Marana is down-

Editor's Note: Related articles on desert-runoff haz­ards and flood-plain management have appeared In

the following issues of Fieldnotes: Vol. 2, No.3(Sept. 1972); Vol. 5, NO.1 (March 1975); Vol. 10,NO.4 (Dec. 1980); and Vol. 11, NO.1 (March 1981).These issues are available from the Bureau for $2.00($1.00 covers postage and handling; $1.00 coversreproduction costs for the March 1975 Issue, whichis out-of-pri nt).

Volume 14 NO.2 FieldnOies

NON FLOOD BANK-CUTTING AND BANK COLLAPSE

------ ---------

Page 3

Figure 2a. Severe bank-cutting along the Santa Cruz River near 1-19and San Xavier Road. Looking northwest.Bridge segment nearest viewer coliapsed when support washed out. Bridge in distance did not fail. Bankretreated from west end of bridge to present position. Distance between bank and midstream end of bridge ismeasure of amount of bank erosion that occurred. Photo taken on October 9,1983 by Peter Kresan.

Figure 2b. Bank-cutting along north bank of RillitoCreek at N. 1st Avenue. Looking downstream. Photoby Peter Kres'an.

Figure 2c. Bank-cutting on outside of bend along Riliito Creek. Looking down- Figure 2d. Bank erosion along north bank of Riliito Creek. Looking downstream.stream near Prince and Country Club Roads. Photo by Tad Nichols. Photo by Ken Matesich.

stream from the confluence of RillitoCreek and Canada del Oro with theSanta Cruz River. Water spread out later­ally over a distance of 4 or 5 miles,causing a true flood. The channelsthrough the city of Tucson, on the otherhand, are deeply entrenched and barelymanaged to contain the October runoffwithin their banks. Nevertheless, this"saving grace" did not prevent the tur­bulent waters from damaging bridges,roads, buildings, vehicles, utility lines,crops, livestock, certain bank-protectiondevices, etc. (Figures 3a-d).

Because almost every drainageway inArizona was activated by the rainfallfrom the large storm system, runoff ef-

fects were widespread. Small washesscoured their banks and bottoms, oftenfinding things of man to damage(Figure 4).

When the Water Is Gone

After the last vestiges of runoff haveseeped into the sand, vertical channelbanks remain. That banks can migratehundreds of feet during flow events istestimony to their lack of resistance.Banks fail because of undercutting andcollapse, and this tendency does notdisappear when the water does. Col­lapse of banks on the verge of failurecould be triggered by any destabilizing

mechanism. Vibrations of any sort, load­ing at the top by even one person, andundercutting by cave-making young­sters could cause a bank to collapse,with potentially tragic results (Figure 5).

There are many miles of banks alongmajor drainages that course through theTucson metropolitan area. Some of thesebanks are more than twice as high as twoaverage-sized adults. Most were modi­fied during October 1983 and left invarious states of instability.

Consequently, when the water is gone,there is still reason for concern aboutdry drainages. Although they are in thei rnormal state, dry drainageways continueto be hazardous to the unaware.

Page 4 Oureou of Geology ond Minerol Technology

MISCELLANEOUS DAMAGE

Summer 1984

Figure 3a. Ina Road undercut by water from adjacent Santa Cruz River. Looking Figure 3b. Wipeout of bridge and utility tower at Sunset Road crossing ofnortheast. Photo by Ken Matesich. Santa Cruz River. Looking southeast. Photo taken on October 2,1983 by Steve

Reynolds.

Figure 3c. Same area as Figure 2d. Looking upstream at water well that was on left side of bank before erosion. Photo by Ken Matesich.

WHAT CAN BE LEARNED?

The events of October 1983 reinforcedthe belief that the worst regional runoffevents tend to associate with the largetropical systems that invade the Stateduring the fall months. Because thesetropical systems can be repetitious, theycan set the stage for large-scale runoffby first saturating the ground.

A general survey conducted by theauthors revealed how fortunate manyresidents were that the next scheduledtropical storm failed to materialize insouthern Arizona. Many buildings andobjects, more numerous than those thatwere toppled, were poised for under­mining when the flows of early Octoberabated. Since then, the southern part ofthe State has been in a dry spell. Thisrespite is buying time forthe community

Figure 3d. Wipeout of northern approach to DodgeBoulevard Bridge over Rillito Creek. Looking south.Photo by Peter Kresan.

Volume 14 NO.2

to complete various repairs and addsome protection prior to the anticipatedsummer rainy season.

Local residents might expect a futurerash of aggravating maintenance workwhere utilities were buried along andbeneath foothill washes several yearsago. In some cases, various lines, burieduntil 1983, were uncovered by wash­bottom scour and broken at least threetimes in the latter half of the year(Figure 4). Many more lines are nowcloser to the surface because of erosionabove them.

Considerable experience was gainedabout various methods of protectingbanks. Almost anything will protect abank, as long as the protective device isnot tested too severely. A recently de­veloped technique, which involves theuse of soil cement, received its baptismin October. Except for minor problems,the technique tested well (Figures 6aand 6b). On the other hand, some of themore classic protective measures failedduring the big October test (Figures 7aand 7b). For regulatory purposes, PimaCounty now recognizes only one type ofbank protection: soil cement. Severalsoil-cement projects, funded by Federalmonies, are underway at the placesdeemed to be most critical.

The October runoff event demon­strated how difficult it is to protect worksof man that encroach upon major drain­ages. The largest drainages, such asPantano Wash, Rillito Creek, and theSanta Cruz River, reached man-madestructures that had been built manyyears ago. The runoff event was large,powerful, and persistent enough tocause hundreds of feet of lateral bankmigration in several places. The areaswhere the soft banks would be cut awaywere predictable (Figure 8 with inset);the size and power of the runoff event,however, were not anticipated.

Because the channels, banks, and ad­jacent flood plains along major drain­ages are usually privately owned, it hasnot been possible to treat these system­atically. A shopping-center owner canafford to invest more heavily in protec­tion than can an average home or trailer­park owner. The result is "piecemeal­ing," a condition that a raging flow ofwater will test in search of a weakness.Bank protection devices necessarily endat property boundaries, a situation thatleaves the devices especially vulnerableat their points of termination. This is alsotrue of soil cement. Water can erode theefficacy of any protective device if it getsbehind the upstream end or overtops thestructure (Figure 9). Selective applica­tion of soil cement is itself a form of"piecemealing" that will leave unpro­tected banks free to migrate (Figure 9).How this migration will eventually affectthe protected parts remains to be seen.

Fleldnores

SOME REMAINING QUESTIONS

Desert drainage systems are complex,interwoven, dynamic, and vital, charac­teristics that combine to test engineer­ing and management skills. On the onehand, there is a demand to stabilizebanks, especially around houses, busi­nesses, and bridges that carry dailytraffic. On the other hand, major drain­ages playa vital role in recharging theonly indigenous water supply in south­ern Arizona: ground water. Replenish­ment of ground water depends on themaintenance of the sand "sponge" thatusually occurs along drainage bottoms.What would cause the removal of this"sponge"? What would save it?

In the ideal solution to this two-sidedproblem, viable bank protection wouldbe added and the necessary conditionsfor effective ground-water rechargewould be maintained. Realization of thisplan requires a basic understanding ofthe dynamics of the system, appropriateengineering techniques, and adequatefinancing. Appreciation and understand­ing of regional drainage dynamics iscritical to the management of majordrainages. Research into the cause­and-effect relationships within this drain­age system should be encouraged andsupported.

Proper management of drainagewaysinvolves several questions: What com­binations of circumstances would causechannel-bottom scouring (removal ofthe important sand-gravel "sponge") orsand-gravel accumulation? How doesurbanization of the desert floor affectthese processes? Certainly the pavingand development of square-mile-after­square-mile prevents transport of nor­mal sediment loads to the major drain­ages. This leads to clear-water runoff,which, in turn, encourages scour (sedi­ment transport) within the main drain­ages. If the banks of these drainageswere totally protected, the most immedi­ate sediment source would be the loosebottom materials that must be main­tained to aid ground-water replenish­ment. Structural modifications would berequired to prevent large drainages fromscouring and to promote ground-waterrecharge. The enhancement of rechargeshould be a continuing goal of research.

Other questions concern the quanti­tative infl uence of urbanization on desertrunoff. How does urbanization - paving,smoothing, packing, channeling, vegeta­tive removal, etc. - affect runoff amountsand rates? Is the natural drainage sys­tem, at least near urban centers, beingasked to carry a larger burden than it

Figure 5. East bank of Pantano Wash collapsingafter flow ceased. Scene depicts instability of banksand attendant hazard. Photo by H. Wesley Peirce.

Page 5

Figure 4. Sewer line. buried for 11 years, uncoveredby erosion along bottom of Finger Rock Wash infoothills of Santa Catalina Mountains. Smaller pipeis natural gas line, also exposed by erosion. Lookingupstream toward the northwest. Photo by H. WesleyPeirce.

would otherwise? How does this in­crease in urbanization, over time, affectdrainage predictability and planning forthe future? How can urbanizatiofl of theTucson Basin be planned to minimizethe impact? Because of this evolvingfactor, how reliable are past studies andthe regulations based upon them?

THE FUTURE

Runoff in the Southwest desert is anatural process that is vital to life ingeneral, but injurious in specific casesof encroachment. That the process willcontinue is assured. Because the fre­quency and severity of future events areunpredictable, it behooves citizens tolook to themselves for protection by

Page 6

--------~---~------

Oureau of Geology and Mineral Technology

SOIL-CEMENT BANK PROTECTION

Summer 1984

Figure 6a. Same area as Figure 1 after passage of major flow. Right bank with railing is undamaged soilcement. Photo by Peter Kresan.

Figure 6b. Pantano Wash bank undergoing soii­cement process. Photo by Ken Matesich.

exercising judgment about things thatthey can directly control. Most adultshave some say about where they chooseto live. If one is aware of the generaldesert-water hazard, there should be noexcuse for placing oneself in a grosslyvulnerable situation.

There will be future damage to exist­ing man-made structures that have beenrendered vulnerable by virtue of theirlocation and inadequate or nonexistentbank protection. On the other hand,because of the experiences of October1983, the security of many bridges andassociated features will be enhanced bybetter bank protection.

Building will continue near the majordrainages where banks are judged to be

adequately protected by soil cement.Although great faith is being placed inthis form of bank protection, it remainsto be seen whether nature, over time, willbe able to significantly undo even theseman-made attempts to control the nat­ural flow of water toward the sea.

CONCLUSION

Damaging runoff in the deserts ofsouthern Arizona is the rule rather thanthe exception. The region continues togrow in population and urbanization. Itis only logical, therefore, for one toassume that damaging runoffs will oc­cur in the future.

The consequences of large-scale run­offs range from minor harassments totragic destruction. The "floods" of Oc­tober 1983 resulted from pervasive tropi­cal systems that affected much ofArizona. Although this natural event mayhave been the most costly ever inflictedon Arizona, it demonstrated what ispossible. This message alone is invalu­able; "forewarned is forearmed." Morerespect is already being given to theimportant drainages.

Because of the applicability of thelaws of physics and geometry, there isno real mystery as to what a flowingmass of water will attempt to do andwhere it will do it. What are not predict­able are the size and frequency of runoff

Figure 7a. Post in foreground marks position of bank-protection device prior toOctober 1983. South bank of Rillito Creek near N. 1st Avenue. Photo by PeterKresan.

Figure 7b. Rock-and-wire-mesh bank-protection device breached and over­topped in October 1983. Looking north along Santa Cruz River from bridge at W.Grant Road. Photo by H. Wesley Peirce.

FAILED BANK PROTECTION

Volume 14 NO.2 Page 7

Figure 8. Bank-cutting at outside of bend along west bank of Pantano Wash, south of Golf Links Road. Phototaken on October 9, 1983 by Peter Kresan.

Inset: Same area as main photo, as seen in 1972. Dashed line indicates trend of future bank-cutting, aspredicted in Fieldnotes in September 1972. Note erosion along predicted trend. Also note expansion ofdevelopment toward wash. Photo by H. Wesley Peirce.

PREDICTED BANK EROSION

events for which a community shouldprepare itself. What constitutes prepar­edness? How much is enough? Howmuch are citizens willing to spend on theuncertain future? One thing does seemcertain, however: because of their ex­perience with the October 1983 runoff,Arizonans will be willing to spend morethan they otherwise would have. Often­times people have to see to believe.Crying wolf too often tends to lower acitizen's level of concern; seeing a wolf,on the other hand, heightens his or herawareness. Seei ng the "wolf" of Octobergenerated enough support that PimaCounty voters approved a bond sale toredesign and repair the many highwaysand bridges that were damaged. In­cluded will be bank protection mecha­nisms that will better withstand highflows, if they are properly designed andconstructed.

As more channel control is sought toarrest bank collapse and migration, im­portant questions will arise about themaintenance of stream-bottom stability,the potential for increased bank erosion

along unprotected stretches, and theincreased flood potential downstreamfrom highly channelized sections. Majordrainages are vital ecological factors:they are linked to the ground-watersupply upon which much of southernArizona depends. A raging torrent ofwater may appear unfriendly and inneed of control; however, some of thistorrent, if given the chance, will seepunderground and help to restore thelevel of the water table. The trick tomanagement of drainages is to exertcontrol where necessary, but to encour­age and maintain maximum recharge.

High banks continue to be unstablelong after they have returned to theirnormal state of dryness. For wayfarersalong the drainages, caution is thewatchword, whether the drainages arewet or dry. ~

Figure 9. Bank collapse after Rillito Creek gotbehind upstream end of soil-cement protectrvedevice and undermined buildings. Flow is frombottom to top. There was no bank protection forbuildings in lower left position. Near intersection ofPrince and Country Club Roads. Photo taken onOctober 9, 1983 by Peter Kresan.