GEOLOGY OF THE ROSWEll ARTESIAN BASIN,NEW MEXICO, AND ITS RelATION TO THE'. HONDO RESERVOIR

IIIIIIIIII

Technil=al Report

STATE OF NEW MEXICOState Engineer Office

S.E.RevnoldsState Engineer

ROBERT T. BEAN

.1949

and

Number 9

,.'

EFFECT ON ARTESIAN AQUIFER OF STORAGEOF flOOD WAlfR IN HONDO RESERVOIR

CHAR lES V. THEIS

1951~• '0. ,r

/

Prepared in cooperation with theG.eol.ogical Survey and Bureau of Reclamation

. . . Unitec;:l Slates Department of the Interior

STATE ENGINEERLIBRARY 0

EFFECT ON ARTESIAN AQUIFER OF STORAGE

OF FLOOD WATER IN HONDO RESERVOIR

GEOLOGY OF THE ROSWELL ARTESIAN BASIN,

NEW MEXICO, AND ITS RELATION TO THEHONDO RESERVOIR

IIIIIIIIIIIIIIII(

[

l

TECHNICAL REPORT

STATE OF NEW MEXICOSTATE ENGINEER OFFICE

S. E. ReynoldsState Engineer

By

Robert T. Bean

u. S. Geological Survey

1949

and

By

Charles Y. Theis

U. S. Geologicol Survey

1951

Prepared in Cooperation with the

Geologicol Survey and Bureau of Reclamation

United Stotes Deportment of the Interior

NUMBER 9

I;1

CONTENTS

EFFECT ON ARTESIAN AQUIFER OF STORAGE OF FLOOD WATER IN HONDO RESERVOIR

GEOLOGY OF THE ROSWELL ARTESIAN BASIN, NEW MEXICO, AND ITS RELATION TO THE

HONDO RESERVOIR

IIIIIIIIIIIIII

r

I

Abstract .Introduction .

Description of Area .Previous Work .Purpose of the Study .History of the Hondo Reservoir .

Physiography .General Surface Features .Erosion Surfaces .Bottomless Lakes .Further Solution Effects .Physiography of the Hondo Reservoir

Stratigraphy .Yeso Formation .San Andres Formation .

General Characteristics .Glorieta(?) Sandstone Member .San Andres Formation (Excludif!g Glorieta(?) Sandstone

Member) .Chalk Bluff Formation .Goat Seep Limestone .Terrace Deposits .

Structure .General Attitude of the Rocks ..Faulted Anticlines .

General Features . . . . . . . . . . . . .Sixmile Hill Structure .Border Hills Structure .Y-O Structure .

Other Structures . .Bluewater AnticFne and Syncline to East .Black Hills Anticline .Dunken Dome .Structure along'the Rio Penasco .Minor Faults .

Structures Due to Slumping or Collapse .Domes in Gypsiferous Strata .Effect of Structure on Ground Water .

Ground Water .Intake Area .

Area Involved .Differences in Intake Capacity .Amount of Recharge of the Artesian Aquifer .Variations in Water Level .Perched Water .....

The Artesian Aquifer. . . . . . . . . . . . . . . . . .. . .General Occurrence. . . . . . . . . . . . . . . . . .. . .Hydrologic Anomalies along Cottonwood Creek. . .

Effect of Use of the Hondo Reservoir .Protection of Roswell .Recharge of Artesian Aquifer .Rise of Artesian Head .Effect on the Pecos River .

References Cited .

Text .

References Cited .

i

Page

111222

555577

88999

915151515151616161818191919191919202020

202020212425262626262727272830

31

33

36

ILLUSTRATIONS

Plate Following page

1. Aerial Photograph of Hondo Reservoir Area. 8

2. Aerial Photograph of Area ImmediatelyEasr-Northeast of Hondo Reservoirand North of Rio Hondo •• • • • . . • • 8

3. Aerial Photograph of Rio Hondo at Bend about3 Miles Southwest of Roswell ••..• . • . . 8

Figure Page

1. Mal? of Part of Southeastern New MexicoShowing Total Area that Contributesto Artesian Reservoir ...•....... 3

2. West-Central Chaves County and SoutheasternLincoln County Showing PrincipalStructures in Eastern Intake Area ofRoswell Artesian Basin and Wells inVicinity of the Hondo Reservoir. . . . 6

ii

II

['

!

[

IIII~

IIIIII

GEOLOGY OF THE ROSWELL ARTESIAN BASIN,

NEW MEXICO, AND ITS RELATION TO THEHONDO RESERVOIR

GEOLOGY OF THE ROSWELL ARTESIAN BASIN,

NEW MEXICO, AND ITS RELATION TO THE

HONDO RESERVOIR

IIIIIIIIII

ByRobert T.

ABSTRACT

In the Roswell Basin in southeastern New Mexicoartesian water is produced from cavernous zones inthe carbonate rocks of the San Andres formation andthe lower part of the Chalk Bluff formation, both ofPermian age. The Hondo Reservoir, 9 miles wesr­southwest of Roswell, was completed by the U. S.Bureau of Reclamation in 1907, to store waters ofthe Rio Hondo for irrigation. The project was notsuccessful, as the impounded water escaped rapidlythrough holes in the gypsum and limestone of theSan Andres formation constituting its floor. Of27 ,000 acre~feet that entered the reservoir between1908 and 1913, only 1,100 acre-feet was drawn Ollt

for use, the remainder escaping through the floorof rhe reservoir. Since 1939, plans have beendrawn up by the State Engineer and by Federalagencies to utilize the reservoir to protect Roswellfrom floods. It has also been suggested that waterfrom the Pecos River might be diverted into under­ground storage through the reservoir.

Sinkholes in the Roswell Basin are largelyclustered in areas where gypsum occurs in the bed­rock. Collapse of strata is due to solution ofunderlying rock commonly containing gypsum.Domes occur in gypsiferous strata near Salt Creek.The Bottomless Lakes, sinkhole lakes in the es­carpment on the east side of the Pecos, are be­lieved to have developed in north-south hinge-linefractures opened when the westernmost beds in theescarpment collapsed. Collapse was due to solu­tion and removal of gypsiferous rock by artesianwater which now fills the lakes.

The Hondo Reservoir was built in a natural sinkwhich was deve loped in gypsiferous limestoneswhere folded and faulted in the Sixmile Hill struc­ture. Development of the sink was due in large partto solution by floodwaters from the Rio Hondo.

A section of 770 feet of the San Andres formationalong the Rio Penasco is described in this reporr.In the eastern part of the intake area of the RoswellArtesian Basin the San Andres is composed almostentirely of carbonate rocks from the Rio Penasconearly to the Rio Hondo, north of which gypsumis also present.

1

Bean

The principal structures in the eastern part of theintake area are three long, narrow faulted anti­clines. The Sixmile Hill structure, more than 60miles long, is the most variable of these, for inplaces it is a simple fault and in others probablyan unfaulted anticline.

The intake capacity of the San Andres formationin the eastern part of the intake area is classifiedas moderate to very high from Salt Creek to the RioHondo, and in general as moderate to low from theRio Hondo to the Rio Penasco. The Rio Hondoappears to lose about 19,400 acre-feet per year byseepage into the artesian aquifer, and the RioPenasco about 8,700 acre-feet per year. Theseamounts are respectively about 8 percent and 4 per­cent of the total recharge of the artesian reservoir.

As precipitation and seepage losses in the east­ern part of the intake area seem inadequate toaccount for the recharge of the artesian aquifer,particularly in dry years, significant amounts ofrecharge are probably contributed by precipitationon a total intake area of about 7,000 square miles.

The sharp decrease in artesian head to the eastalong Cottonwood Creek is probably due to heavyleakage from the artesian to the shallow aquiferthere, the principal supporting evidence being thehigh water level in the shallow aquifer.

If the Hondo Reservoir should be used again,nearly all water impounded there would enter theartesian aquifer. If this did not result in increasedwithdrawal from wells, practically all this waterwould eventually reach the Pecos River, but ifground-water use were increased, the amount reach­ing the Pecos would be less.

INTRODUCTION

DESCRIPTION OF AREA

The Roswell Basin in southeastern New Mexicois one of the most important areas of artesian­water production in the United States. In it, atpresent (947), about 56,000 acres of land is irri­gated with artesian water, and about 45,000 acresmore is irrigated with shallow ground water, mostof which was originally derived from the artesian

aquifer. The area of artesian wells is elongated ina north~south direction paralle 1 to the Pecos Riverand is about 65 miles long and from 7 to 12 mileswide.

Recharge of the artesian reservoir is effected byprecipitation on an area to the west, the total areathat contributes to recharge probably being about7,000 square miles (fig. 1). A part of the precipita­tion on this area percolates directly to the watert;ble where it begins to move slowly toward theartesian reservoir, and a part flows in surfacestreams for some distance before being added tothe ground-water supply by influent seepage. Theprincipal aquifer of the artesian reservoir is theSan Andres formation, in which the water occurs incavernous zones in limestone and dolomite. How­ever, in the southern part of the basin, according toMorgan (1941, p. 781), artesian water is producedprincipally from cavernous zones in limestones inthe lower part of the Chalk Bluff formation overlyingthe San Andres formation.

The Hondo Reservoir lies 9 miles west-southwestof Roswell and about 18 miles west of the PecosRiver. It was built in the first decade of the presentcentury as a storage reservoir to utilize the watersof the Rio Hondo for irrigation. It was not a suc­cess, however, as its floor proved to be very leaky,and water would not remain in it longer than a fewdays after floods. In more recent years plans havebeen put forward by State and Federal agencies tous e the reservoir for flood control in order to pro­tect the city of Roswell. It has also been sug­gested that water from the Pecos River might bediverted into the reservoir, from whence it wouldgo into underground storage, thus eliminating evapo­ration and transpiration. As yet (1947) none ofthese plans has been put into effect.

PREVIOUS WORK

Several studies dealing either entirely or in partwith the ground-water geology of the Roswell Basinhave been made. The earliest paper of note was areconnaissance study of the Roswell Basin byFisher, which appeared in 1906. Another briefpaper by Renick (1926), on the geology and ground­water resources of the Rio Penasco area, deservesmention.

By far the mOst complete study of the ground­water resources of the Roswell Basin was made byFiedler and Nye, whose excellent work was pub­lished in 1933 as Water-Supply Paper 639. Athorough study of the shallow-water resources ofthe basin by Morgan appeared in 1939, and in thesame year Theis described the origin of both baseflow and additional flow in Major Johnson Springs

2

at the south end of the basip-. Morgan (1941) wroteanother paper appearing 2 years later in which hestated that the "Dog Canyon limestone" was theprincipal artesian aquifer in the southern part of thebasin. The name "Dog Canyon" has since beenabandoned and replaced by the Goat Seep limestone(King, 1942). It is approximately equivalent to andgrades into the lower parr of the Chalk Bluff forma­tion, the lower unnamed member and the Queensandstone member. Further contributions to theliterature on the ground-water geology of the Ros­well Basin by Theis, /l.forgan, and others (NationalResources Planning Board, 1942a, p. 27-75) ap­peared in 1942 as parr of the study of the entirePecos drainage basin.

PURPOSE OF THE STUDY

The purpose of the present study is to determinethe hydrologic effects of the proposed use of theHondo Reservoir either for flood control or as acatchment basin for underground storage of PecosRiver waters. In order to understand all geologicfactors involved, a geologic study was made of theartesian area and of the outcrop area of the artesianaquifer farther west.

The present study was made at the request of theUnited States Bureau of Reclamation and with fundsmade available by the Bureau.

HISTORY OF THE HONDO RESERVOIR

The Hondo Reservoir was completed in 1907 bythe U. S. Bureau of Reclamation, then called theReclamation Service. It was designed as an off­channel storage reservoir, the impounded water tobe used for irrigation. It was not a success, how­ever, as water diverted into it always escaped veryrapidly through sinkholes in limestone and gypsumin its floor. Many attempts were made to utilize thereservoir from 1907 to 1915, but without success.The following account, taken from the officialreporr of the Reclamation Service,l describes oneof these attempts:

On June 11th (1913) the big flood came and water ranin the river from then until July 5th, and about 900 acreswere irrigated. Some 2,000 acre~feet of water enteredthe reservoir. For two or three days some of this couldhave been drawn off through the outlet gates, but therewas plenty of water in the river. When the rivet dried up,rhe water in the re servoir had fallen so that none couldbe drawn out. All irrigation, both in April and in Juneand July, was direcr from the river. None was taken ourof the re servoir.

IV. S. Reclamation Service, n.d., Project History from Incep­tion of Project, Feb. 24, 1904, to Dec. 31, 1915, Hondo Project,N. Mex.: typewritten rept. in files of U. S. Bur. of Reclam. inAlbuquerque, N. Mex., v. 2, p. 115.

IIIIIIIIIIIIIIII

II

Boundary of orca contributing to recharge of'\.......- .............. arlCliian fnlicrvoir (dashed where approximate).

FIGURE 1. Mop of part of southeastern New Mexico showing total area that contributes to artesian reservoir.

3

After early failures of the reservoir, efforts weremade to plug the principal sinkholes, but thesewere unsuccessful. The method, according to

Foster,2 was to explode dynamite in the holes andthen backfill and tamp them with a "good material."These holes would reopen, however, either at thesame location or very close by, when water wasagain let into the reservoir. The worst leaksdeveloped in the northern part of the reservoir, andon November 15, 1912, it was reported that 75 holeswere present in "gyp rock" throughout the floor.Most of these were from 4 to 10 feet in diameter andfrom 4 to 8 feet deep, but six holes were about 18feet in diameter and from 10 to 12 feet deep. Nofurther attempts to plug the leaks were madeafter 1911.

The following table shows the insignificantproportion of water diverted into the reservoir thatwas actually made available for irrigation. 3 Noother water was released from the reservoir orspilled.

Inflow and Outflow at Hondo Reservoir

Year Inflow Total Outflow(acre-feet) (acre-feet)

1908, May 1 to Dec. 31 3,000 3001909 ............ 500 01910 ............ 2,000 2001911 ...........• 11,600 5001912 ............ 9,000 1001913 .. ........... ...... . 1,760 0

Total 27,860 1,100

No records of inflow after 1913 are available, andafter April 1, 1917, the Bureau of Reclamationturned the project over to the Hondo Water UsersAssociation, which has not attempted since to storewater in the reservoir but only to irrigate directlyfrom flood waters of the river (McClure, 1939a,p. 72).

The flood that inundated Rosweli in May andJune 1937 helped awaken interest in utilization ofthe reservoir for flood control, and an investigationalong that line was made under the direction of theState Engineer. The report of this investigation(McClure, 1939a) recommended that two roliedearth-fill dams be constructed, one across the RioHondo and one across its tributary Rocky Arroyo.The dam on Rocky Arroyo would be locared in thecentral part of sec. 16, T. 12 S., R. 22 E., justdownstream from the "breakover area," an area of

2Ibid•• p. 82.

3fbid•• p. 98-99.

4

low elevation through which a part of flood flow ofthe Hondo moves naturally into the Rocky Arroyodrainage. Storage for a maximum of 11,335 acre­feet would be provided for in the Rocky ArroyoReservoir, and excesses over this amount would bediverted into the Rio Hondo and the Hondo Reser­voir. A combination diversion and storage darn ofrolled earth-fill would be built across the Hondofrom the approximate center of the south line ofsec. 26, T. 11 S., R. 22 E., to a bedrock spur in theNE~ of the same section. The Hondo Reservoirunder this plan would have a capacity of 43,800acre-feet with the water surface 5 feet below thetop of the 20-foot dam.

Two other plans for protecting Roswell fromfloods were put forward by the Federal Governmentduring the course of the Pecos River Joint Investi­gation (National Resources Planning Board, 1942b,p. 158-161). The first called for diversion of floodflows by levees south of Roswell into a draw empty­ing into the Pecos River southeast of the city. Thesecond called for use of the Hondo Reservoir and areservoir in Rocky Arroyo similar to that proposedin the State Engineer's plan. A rolled earth-filldam would be constructed across Rocky Arroyo inthe same place provided for in the State plan, andthe effective flood-control storage of the reservoirso formed would be modified by removing thenorthwest-trending part of the old dike in the easthalf of sec. 26, T. 11 S., R. 22 E., and tying theremaining portion of the dike to the bedrock spurnear the center of the south line of sec. 26. Thetops of existing dikes would be raised by filling onthe downstream side, and the capacity of the reser­voir would then be about 54,900 acre-feet. Emer­gency spills from this reservoir would go across anatural saddle to the north into the drainage ofBerrendo Creek, which passes north of Roswell anddoes not flood the city. The total capacity of thereservoir system thus would be about 64,200 acre­feet. This would have amply controlled the largestflood on record, that of September 1941, whichwould have required about 47,900 acre-feet ofstorage capac ity.

It should be noted that, of the two plans for floodprotection of Roswell, only the first plan is con­sidered to be economically iustified in the NationalResources Planning Board report (l942b, p. 163­164). Any future consideration of this plan, how­ever, would probably have to be modified, as theRoswell Army Air Base is now located across partof the area through which the diversion channelwould pass.

As yet (1947), no plan for flood control or ~1,,,

Rio Hondo has been put into effect.

?-:'---$

•

I

II'

II

PHYSIOGRAPHY

GENERAL SURFACE FEATURES

The region covered by this report lies in theeastern part of the long slope from the Sacramentol\,iountains and the Sierra Blanca to the PecosRiver. Relief and elevation decrease from west toeast. In the west, relief is moderately high, asyouthful valleys have been Cut into the limestoneuplands. These valleys decrease in depth to theeast, and terrace deposits of low relief appear10 to 28 miles west of the Pecos River and extendto the river.

The most prominent physiographic expressions ofstructure are ridges or lines of hills along theBorder Hills, Sixmile Hill, Y~O, and Bluewater anti­clines (fig. 2). All but the last of these wheretypically developed are narrow faulted anticlinestr~I1dingin a northeasterly direction. The Blue­~ater anticline (see p. 19) is much broader, trendsmore nearly due north, and is marked topographi­cally by a rather abrupt increase in elevation alongits eastern limb. The Border Hills is the bestdeveloped, physiographically, of the narrow anti­clines, as it is marked by a nearly continuous ridge.The Hills along the Y-O structure are less promi­nent, and the Sixmile Hill structure is expressed inits central part by valleys and saddles rather thanhills, for there it is not anticlinal but a fault of lowdisplacement. Other faults in the area are alsovery well expressed by alined valleys separated bysaddles. The anticlinal lines of hills are breachedby prominent water gaps where crossed by mainstreams.

Study of the physiographic expression of struc­ture here shows that the present drainage patterndeveloped after the structural movements occurred,or at least after they began. Main streams followvalleys along the principal structures in someplaces, which would not be possible if the drainagewere antecedent or superposed. The Rio Felix, forexample, follows the Y-O structure for 5 miles inT. 15 S., Rs. 21 and 22 E., and the Rio Hondofollows the Sixmile Hill structure for 3 miles justsouthwest of the Hondo Reservoir. In addition,where streams cross linear anticlinal hills, theyoften do so at a point of particular structural weak­ness. For instance, the Rio Hondo crosses BorderHills where the southern pan of the structure isbent or offset to the west and a hinge point in thedirection of the vertical displacement of the faultexi~ts (p. 18). Such a point must have been lowtopographically, and through this point the RioHondo was established. A similar situation exists6%iniles to the southwest, where a tributary of theRio Felix crosses the Border Hills structure at a

point where the structure again bends sharply tothe west.

EROSION SURFACES

The highest erosion surface in the area of thisreport was called the Sacramento Plain by Nye(Fiedler and Nye, 1933, p. 14). It is preserved inthe interstream areas on the east slope of theSacramento Mountains and probably was once con­tinuous with the High Plains east of the Pecos.About 200 feet or more below this surface is theDiamond A Plain, named by Nye (Idem, p. 14).t...forgan (National Resources Planning Board, 1942a,p. 35) states that this surface was developed con­temporaneously with the r-.fescalero pediment eastof the Pecos. Old terrace remnants, called flgravel­capped mesas" by Nye (Fiedler and Nye, 1933,p. 13), probably were coextensive with the DiamondA Plain. Nye (Idem, p. 10) named the surfacesbelow the Diamond A Plain in descending order theBlackdom, Orchard Park, and Lakewood terraces.Most of the irrigated land of the Roswell ArtesianBasin is on the Orchard Park terrace. The Lake­wood terrace forms a narrow belt on either side ofthe Pecos River and is underlain by alluvium.

BOTTOMLESS LAKES

Southeast of Roswell a series of sinkhole lakesknown as the Bottomless Lakes have been de­veloped in the escarpment of the gypsiferous ChalkBluff formation on the east side of the Pecos River.Water levels in these lakes are higher than the riverand the water table in the adjacent flood plain andstand approximately at the level of the piezometricsurface of the artesian aquifer. The level of thelakes evidently fluctuates with the artesian head,as observations in 1947 showed a drop in the levelof Cottonwood Lake during the irrigation season.The lake was overflowing its rim when observed onFebruary 7 and March 21, but was about 1 footbelow the rim on July 23. The lakes are thusbelieved to be pools fed through underground chan­nels by water from the artesian aquifer.

The westernmost beds in the escarpment in thearea where the lakes occur have slumped and nowdip toward the river, and numerous fractures paral­lel to the river have formed along the hinge linesbetween the nearly flat lying beds to the east andthe slumped strata. Sinkholes of all sizes havedeveloped in the hingewline fractures. The processof formation of the Bottomless Lakes thus evidentlybegan with discharge of water from the artesianaquifer at the base of the escarpment. This dis­charge may well have been beneath the surface, thewater passing directly into the sands and gravelsof the flood plain. The water dissolved out some

5

I

I

I

I

I

I

I

I

III

II

II

------- - -f ----

I

COlj!~TYE DDT

:~--rm-

i

EXPLANATtON

.!lRil or o..,tWI,,~. SlfCb-cbll' 11.01 foullotd

Alis 01 ~1'~CIiIH-, plluit-Iy tovll('rj Irl pl'(l~M

Si'm~al indicalinog flQt·h,jtlO beds

Well me.,.ll,/nd ChJf1MO 'hid)'

e

rI.S

>­r ~

17 z.~

o

"

T

r••

r7

•

r >­,0 0

• 0

~cale Miles

FIGURE 2. West-central Chaves County and $outheostern Lincoln County showing princ:i po I structures ineastern intake oreo of Roswe 11 Artesian Basin and we lis in vic:in ity of the Hondo Reservoir.

I...6

IIIII

••••••I

•••IIII

of the gypsum beds in the Chalk Bluff fotmationnear the points of discharge, causing the overlyingbeds nearest the river to collapse and openinghinge~line fractures. Rainfall and surface runofffrom the immediate vicinity formed sinkholes in thefractures, and the first of these sinks to form con­tinued (0 enlarge until their throats tapped thechannels leading from the artesian aquifer to thedisc~arge points. Artesian water then rose in thesinkholes and the Bottomless Lakes were formed.

FURTHER SOLUTION EFFECTS

Large numbers of sinkholes occur in many placesirlthisregion. They are nOt distributed at randombut are present in clusters or groups I and mostcommonly are an indication of gypsiferous bedrockbeneath the surface. They are most numerous inthe northern part of the area, where thick zones Con­taining gypsum occur in the San Andres formation.One area of many sinks occurs in Sixmile Hill justnorth of the Hondo Reservoir and extends into theHoar of the reservoir itself ( see below). The sinkson Sixmile Hill show a tendency toward alinementas they are generally developed along joints trend~

ing N. 42° E., parallel to the Sixmile Hill fault.They occur most commonly on high ground. In theravines they have been largely destroyed by ero~

sian. Farther north in the Salt Creek drainage areasinks are very numerous; for example, in the north­ern part of T. 9 S., R. 22 E.

Sinks are much less common in the southern partof the area studied, where outcrops of gypsiferousstrata were not found. The sinks here have devel­oped almost entirely in undissected areas, un­hindered by erosion. The small area of undissectedupland largely accounts for their scarcity. Thesinks in limestone are partly filled with debris inalmost every case, although those in gypsiferousstrata to the north most commonly have open throats.There are a few broad sinks, apparently formed innongypsiferous strata, along Highway 83 near theRio Penasco. A large sink of this type occurssouth of the Penasco near Dunken. An area con­taining a moderate number of sinks lies southwestof the Flying H Ranch, mainly in the southeasternquarter of T. 15 S., R. 18 E., and these sinks maybe due in part to solution of gypsiferous stratabelow the surface. One of the sinks here is aquarter of a mi Ie long and contained a lake whenobserved in July 1947. A smaller sinkhole arealies in the northeastern part of T. 14 S., R. 19 E.

South of Salt Creek, 1 to 3 miles west of Highway285, lies an area abounding in gypsiferous rock, inwhich occur not only sinks bur arcuate elevatedareaS as well. It is similar to a region south of

Carlsbad near ~Jalaga Bend in the Pecos River,though the structures in general are not as welldeveloped here as near Malaga Bend. In the domesbeds of limestone dip away from the center, andcircumjacent Hrace-track" valleys are developedin gypsum.

\Vest of the western boundary of sec. 1, T. 9 S.,R. 22 E., a zone of gypsum apparently underliesabout 10 to 30 feet of limestone. Here deep sinkshave been developed in the limestone, owing pri­marily to the solution and removal of the underlyinggypsum. At other places in the Salt Creek region,for example in the southeastern part of T. 8 S.,R. 21 E., extensive flats have been developed ongypsum. On these flats are scattered rounded hillsof the gypsiferous rock, nearly all being capped bybeds of limestone or dolomite.

PHYSIOGRAPHY OF THE HONDO RESERVOIR

The Hondo Reservoir occupies a large naturalsink on the north side of the Rio Hondo 9 mileswest-southwest of Roswell. The lowest point of thesink is about 2 miles from the river and about 15feet lower than the Hood plain between the sink andthe river. Sixmile Hill extends in a northeasterlydirecdon from the reservoir sink, and the axis of theSixmile Hill structure, which here is a complexfolded and faulted anticline (see p. 16, 17), ex­tends directly through the reservoir. Sixmile Hillis covered by a great many smaller sinkholes, asseen on the accompanying aerial photographs (pIs.1, 2, and 3) which show the Hondo Reservoir andarea east ther.eof toward Roswell. It will be notedthat the small sinks extend into the northeasternpart of the reservoir itself, and undoubtedly otheropenings are present beneath the silt cover at thecenter of the reservoir.

At the reservoir the Rio Hondo has cut a channel10 alluvium and terrace material the surface ofwhich is 30 feet or more below the level of theupland. The reservoir sink lies directly ahead of anortheastward-flowing segment of the Hondo and isseparated from it only by the flood plain of the river.

The location and development of the reservoirsink apparently have been due to three factors: thepresence of the Sixmile Hill structure, the largeproportion of gypsum in the bedrock, and the loca~

rion with respect to the Rio Hondo. The SixmileHill fault beneath the Hondo Reservoir probably iscomposed of two strands. To the northeast a valleyhas been developed along the fault, and this valleysplits just north of the reservoir, indicating that thefault probably splits also. In fact, a segment of thevalley apparently developed along the easternstn'lnd of the fault can be seen on the floor of the

7

reservoir about three-quarters of a mile northeast ofthe old reservoir head gate. Development of thereservoir sink probably began by the formation ofvalleys tributary to the Rio Hondo along the twofault strands and perhaps along other lines ofweakness in the complex Sixmile Hill structure.Between and around the valleys, sinkholes wereformed by solution of the gypsiferous rocks. Afterthe tributary valleys had become sufficiently welldeveloped, flood waters from the Hondo began toback up in them and to accelerate greatly theprocess of solution as they sank into sink holesand fractures in the rocks. As this process con­tinued the valleys deepened and broadened, umileventually they dissolved away practically all rockbetween them and coalesced into one large sink,the reservoir sink. The reservoir sink then con­tinued to act as a natural basin for receiving floodwaters of the Hondo until construction of the pres­ent dikes, as shown by the position of the sink andthe alluvial silt covering of its floor.

The nature of the material underlying the HondoRe servoir is shown by the following logs of the

Log 01 Hole No. 10

Depth

~,laterial (leet)

From To

Soil 0 1.0Hard limestone 1.0 12.8Broken limesrone 12.8 29.7Clay and broken rock 29.7 31.7Clay and gypsum 31.7 33.6Gypsum 33.6 55.6;{ed Clay 55.6 56.3Gypsum 56.3 59.3Clay 59.3 74.0Gypsum 74.0 78.4

Log 01 Hole No. 11

Clay 0 11.1Brok en limestone 11. 1 22.0Clay :2.0 25.0Cavity 25.0 30.0broken rock, cavities 30.0 64.4Gypsum 64.4 70.2Clay 70.2 71.9Cavity 71.9 73.4Loose rock 73.4 76.8Gypsum 76.8 79.8Clay 79.8 80.2Limestone 80.2 88.4Gypsum 88.4 91.8

8

deepest diamond-drill holes put down by the UnitedStates Reclamation Service in 1903. 4

STRATIGRAPHY

The rocks of the Roswell Artesian Basin arePermian and Quaternary in age. Rocks not involvedin the transmission or confinement of water in thebasin will not be described in this report, as manyexcellent published works on the stratigraphy ofsoutheastern New Mexico are available.

YESO FORMATION

The Yeso formation is the oldest Permian forma­tion in southeastern New Mexico, with the excep­tion of the Abo sandstone, on which it rests. TheYeso is composed of limestone, sandstone, shale,anhydrite, and some salt. The red-bed type oflithology is common. According to Morgan,S theYeso is about 1,000 feet thick in the west face ofthe Sacramento Mountains, but it thickens to theeast, and logs of wells show it to be over 2,000feet thick about half way down the long east slopeof the mountains. The Yeso is the imperviousformation underlying the artesian aquifer, the SanAndres formation, in the Roswell Basin. Nye(Fiedler and Nye, 1933, p. 70-76) assigned thename "Nogal" to these beds, but that name hasbeen dropped in favor of the older Yeso.

Mor,gan states that the depth to the tOp of theYes 0 formation near the Pecos River range s fromabout 1,900 feet near Rosweli to about 2,300 feetnear Dayton. He says that haUte, anhydrite, andred beds in the Yeso are progressively replacedsouthward in the Roswell Basin by limestone, andat Dayton the formation contains only a few thinbeds of clastics with the limestone.

Good outcrops of the upper part of the Ye so canbe seen along Highway 70 west of Riverside in thecanyon of the Rio Hondo. Sandstones and shalesirregularly colored red and yellow are prominent,and many thick zones of limestone are also present.Beds assigned to the uppermost part of the Yeso bymost geologists acquainted with this region areexposed in the Border HiBs structure where it iscrossed by the Rio Hondo. Here white to light­yellow sandstone, commonly identified as theGlorieta sandstone member at the base of the San

4McClure, T. M., 1939, Report On Lower Rio Hondo FloodControl lnvestigatio,l: typewritten rept. in files of State Engi­neer Office in Santa Fe, N. Mex., p. 115-116.

5Morgan, A. M., 1942, Report on the Geology of the PecosValley, N. Mex.: ms. rept. in files of U. S. Geol. Survey inAlbuquerque, N. Mex.

Pl.ATE 1. Acrial photograph of Hondo Rcscrvoir oreO.(10.24·46 1: 10 SCS 1:31,680 DDO.12.169)

I

IIII

II

III~('

o 112

SCALE IN MILES

1111

III

PL.ATE 2. Aerial photograph of area immediately ead.northeast of Hondo Reservoir ond north of Rio Hondo.(6.10.46 10:37 SCS 1:31,680 DDO.3.157)

of

SCALE IN MILES

PLATE 3. A<>rial photograph of Rio Hondo at bend about 3 mil"s southw<>st of Roswell.(6.10.46 10:35 $CS 1:31,680 000·3·151)

I... .•~.;,

I

I

I

I

o 1/2

SCALE IN MILES

••I

••II

•IIIIIIIIIII

Andres formation, occurs above limestone differinglittle from the limestone of the overlying SanAndres. On the south bank of the Rio Penasco, atthe Runyan Ranch in sec. 29, T. 16 S., R. 17 E.,the lowest beds of the Yeso exposed are red,yellow, and light~gray thin-bedded argillaceoussandstone and thio- to medium-bedded gray dolo­mitic' limestone. Overlying these limestones aretwo medium-grained white to yellow-brown calcare­ous sandstones similar to the Glorieta and sepa­rated from the Glorieta sandstone member by 109feet of limestone and dolomitic limestone. Thissequence of sandstone, limestone, and dolomiticlimestone probably represents the uppermost partof the Yeso. (See section, p. 13.)

Notable thicknesses of halite are found in wellspenetrating the Yeso formation north of Roswell,according to Morgan, but salt never appears in theoutcrop area of the formation. It has undoubtedlybeen removed by solution, Morgan states, down todepths to which ground water can circulate freely.Morgan also reports that deep water found in theYeso is nearly always highly mineralized and unfitfor use, but in the outcrop areas of the formationbodies of perched water are used for stock anddomestic purposes.

SAN ANDRES FDRMATIDN

GENERAL CHARACTERiSTICS

The San Andres formation consists mainly of athick succession of lim-=stones, dolomitic lime­stones, and dolomites, some of which to the northare replaced by anhydrite or gypsum. Its base ismarked by a calcareous sandstone, the Glorietasandstone member. A few thin beds of calcareousshale and calcareous siltstone, commonly yellow,occur mainly in the lower part of the formation.East of the Pecos River wells encounter halite inthe San Andres. An erosional unconformity sepa­rates the San Andres from the overlying Chalk Bluffformation. The San Andres is roughly the equiva­lent of the Picacho limestone of Nye (Fiedler andNye, 1933, p. 55).

The San Andres formation averages about 1,000feet in thickness in the Roswell Basin. The sec­tion exposed along the Rio Penasco is 770 feetthick, as measured by the writer, but the uppermostpart of the formation is missing. The San Andresis exposed over most of the area from the crest ofthe Sacramento cuesta to within 10 to 30 miles ofthe Pecos River, where it dips beneath youngerbeds. The depth to the top of the San Andres at thePecos increases southward from about 400 feet eastof Roswell to about 1,200 feet near Lakewood.

The San Andres formation is the principal arte-

sian aquifer in the Roswell Basin. Water occursunder pressure in porous and cavernous Zonesmainly in the carbonate rocks of the upper halfof the formation.

GLORIETA(?) SANDSTONE MEMBER

The sandstone member at the base of the SanAndres is named the Glorieta from its type localityat Glorieta l\lesa, where it forms the cap rock. Thename nHondo sandstone," which has sometimesbeen used in the Roswell Basin, refers to the samerock unit. The Glorieta is typically a white toyellow medium-grained, well-sorted calcareoussandstone. Morgan states that its thickness in­creases to the north at the expense of the overlyinglimestone beds, from a thickness of about 12 feetin the southwestern part of the basin to about 300feet at Glorieta Mesa. The sandstone commonlyconsidered by geologists acquainted with this re­gion to be the Glorieta is exposed in the Rio Hondowater gap through Border Hills. Cross bedding iswell developed in its upper part here. Lower zonescontain many solution cavities, mostly less than aquarter of an inch in diameter, giving the sandstonea pock-marked appearance. The sandstone heregrades downward through about 8 feet of calcareoussand and sandy limestone into silty limestonecommonly referred to as the Yeso formation. Theupper boundary of the Glorieta(?) is much sharper,the upper 6 inches being fine-grained, very thin­bedded sandstone stained by limonite.

A similar sandstone occurs at the Hondo watergap 40 feet below the Glorieta(?), and two sand­stones separated by 109 feet of limestone arepresent at the Runyan Ranch on the Rio Penasco(see YESO FORMATION). It is probable thatseveral sandstones of similar lithologic type occurnear the contact of the San Andres and Yeso forma­tions, and at any given place it may well be impos­sible to prove which, if ;ny, is continuous with thesand capping Glorieta ~lesa.

In the western part of the Roswell Basin the mainwater body apparently extends through the Glori­eta(?) sandstone member and the overlying porouscarbonate rocks with little regard for their differentcharacter. Morgan reports that farther east, in theartesian area and its vicinity, water in the Glorietais trapped between overlying beds of the SanAndres, which are here impervious, and beds of theYeso below I and as a result it is almost invari­ably salty.

SAN ANDRES FORMATION(EXCLUDING GLORIETA(?) SANDSTONE MEMBER)

In most of the Roswell Basin the part of the SanAndres formation above the Glorieta(?) sandstone

9

member consists of alternate beds of limestone anddolomite. Although there are some exceptions, themore calcareous limestone is commonly coarser­grained and thicker-bedded than the dolomiticlimestone and dolomite. In the upper part of theformation it is often lighter-colored also, but thickzones of dark, bituminous limestone occur in thelower part of the San Andres. The typical dolomiteis very dense and semi-lithographic, and occurs inmoderately thin beds having a blocky appearance.Where fresh it is very dark, but the insoluble sub­stances it contains weather out and coat theweathered surface, giving it a much lighter- color.The limestone and dolomite ordinarily contain ahigh percentage of insoluble matter, both silt andclay being represented. Some of the limestone isoolitic and much of it is crystalline. The limestonecommonly weathers to a brown or tan, and the sur­face may appear fluted. Nodular and irregularmasses of chert occur in both the limestone and thedolomite but are more common in the limestone.Stylolites are common in the limestone, particularlyin the lower part of the formation. They are be­lieved to be due to solution under pressure.

Beds of gypsum also occur at the surface in theSan Andres formation in the northern part of theRoswell Basin. Where encountered in the subsur­face, the corresponding rock is anhydrite. NearSalt Creek, hills composed of over 80 percent gyp­sum occur in several places, the interbedded rockbeing dolomitic limestone. Most of the gypsum iswhite or gray. Banding in various shades of grayis well deve loped parallel to the bedding in someplaces, and in others reddish argillaceous or siltymaterial occurs along such lamination planes.Gypsum is much more soluble in water than mostrocks, and in areas underlain by gypsiferous rockssolution effects such as sinks, enlarged joints, andsmall caverns are very common. The zones of gyp:­sum pinch out to the south as the gypsum is re­placed by carbonate rocks. Halite also occurs inthe subsurface in the San Andres. According to

Morgan 6 the western limit of the halite lies a shortdistance east of the Pecos River, and it does notextend south of Acme.

The carbonate rocks of the San Andres formationinclude certain zones containing a great many smallholes about the size of a pinhead, or slightlylarger, which are interconnected to some extent.These rocks are often described as (lworm-eaten."Such zones occur in limestone, dolomite, and

61\1organ, A. 1\.01., 1942, Report on the Geology 0/ the PecosValley, N. t,·lex.: illS. rept. in files of U. S. Geol. Survey inAlbuquerque, N. Mex.

10

dolomitic limestone. They are often very limitedvertically, and horizontally as well. They are muchmore common in the upper part of the formation thanin the lower, although they may occur anywhere init. Nye (Fiedler and Nye, 1933, p. 65) has pointedout that these cavities are due to solution andremoval of anhydrite which was originally pre­~ipitated as crystals or nodules with the carbonaterock. In some places the original anhydrite, nowaltered to gypsum, can be found fHling the cavitiesin unweathered rock. In other places, secondarycalcite has partly or completely filled the holes.

Larger cavities, irregular in shape, also occur inthe San Andres formation. Typical examples can befound along Highway 70 on the wall of Rio HondoCanyon east of Riverside. Here beds of limestonecontain numerous irregular cavities from a fractionof an inch to more than an inch in diameter andlined with secondary calcite. They are not freelyinterconnected, but the calcite linings indicatethat there has been slow percolation of groundwater through them. A large amount of highlyporous rock having a reticulate pattern occurs inBerrendo Creek where it is crossed by the SixmileHill Structure. The porosity has been developed ina dense dark-gray thin-bedded dolomitic limestonecrossed by numerous joints approximately at rightangles to the bedding. The reticulate pattern isformed by secondary calcite which fills the beddingplanes and joints, the dolomitic limestone betweenhaving been partly or completely removed. Theprocess apparently begins by deposition of thecalcite. Then solution of the central block ofdolomitic limestone begins, as many exposuresshow rounded blocks of dolomitic limestone sur­rounded by a rectangle of calcite. In some placesthe central block is missing, its removal probablybeing due to mechanical weathering and erosion inthe stream bed.

Further evidence of movement of ground waterthrough the San Andres can be found in manyplaces, particularly in the walls of Rio HondoCanyon and to a lesser extent in the canyons of theRio Felix and Rio Penasco. Old solution channelsand small caverns in such places illustrate the typeof opening through which water moves in the partsof the formation now buried. Cavernous zones aremost readily developed where the rock has beenfractured and broken, either by diastrophic forces,as in Sixmile Hill, or by solution of underlyingbeds followed by collapse, as is very common fromthe Rio Hondo northward, east of Border Hills.Cavernous zones are by no means limited to suchareas, however. For example, in the cut bank onthe south side of the Rio Penasco, near the bound-

1+,<~.-­"

Partial Section on East Limb of Y-O Structure JustWest of Y-O Crossing-Continued[

rI,

IIIIIIIIIIIIIII

ary between sees. 20 and 21, T. 17 S., R. 20 E.,massive dolomitic limestone in beds as much as 8feet thick changes sharply upstream into irregularthin beds honeycombed with solution cavities asmuch as several inches in diameter and containingmuch secondary calcite. This undoubtedly wasonce a main underground channel which has sincebeen exposed to view by erosion. Water is carriedin the artesian basin in cavernous zones suchas this.

A section of the San Andres formation along theRio Penasco was measured in connection '\\'lth thisstudy and is described below. The section is com­plet~, aside from a few minor covered intervals,from the base of the formation to the highest bedsexposed along the river west of sec. 15, T. 17 S.,R. 20 E., which is about 9 miles west of Hope.East of this point the canyon walls are too low andthe exposures too poor to be followed continuously.The method employed was to measure a sectionwhere exposures were good, and then follow a keybed in that section up the canyon, changing bedswhere necessary, until another good exposure ofdifferent zones in the formation was reached,whereupon it was measured. Among other things itwas discovered thut the beds stratigraphicallyhighest in the formation do not occur at the extremeeastern end of the traverse, us was expected, butat the Y-O structure, 5 miles farther west. Furtherdata on structure along the Penasco are given onpage 19.

Composite Section of the San Andres FonnationAlong the Rio Penasco

Upper part of Sun Andres Formation

Partial Section on East Limb of Y-O StructureJUSt \l.'est of Y-O Crossing

FeetUppermost beds absent.Limestone, coarsely crystalline 2Dolomite, dark gray, dense. . . . . . . . IDolomitic limestone, finely crystalline 6Limestone, highly calcareous, crystalline,

medium-coarse, thick-bedded .. . . . . . 7Limestone, argillaceous, thin-bedded, non-

resistant, characterized by persistentyellow stain; l'yellow bed" .

Limestone, granular, all but upper footthick-bedded . . . . . . . . . . . . • . . . 6

Dolomitic limestone, some limestone, bedsless than 1 foot thick . . . . . . . . . . 8

Limestone, granular. . . . . . . . . . . . . 1Dolomitic limestone, dark gray, dense,

beds less than 1 foot thick. . . . . . . . . . . . . 8Limestone, granular 3Limestone and dolomitic limestone, alter-

nating beds averaging I foot in thickness 9

Dolomitic limestone, and some dolomite,medium-bedded; includes one thinchert zone .

Limestone, upf,er 6 inches containing verymany chert nodules; forms a massivecliff in places .

Dolomitic limestone and dolomite, dark grayLimestone, granular, containing some

chert masses .Dolomitic limestone, dark gray, dense,

blocky, moderately thin-bedded; weathersto light gray. Contrasts markedly withthe limestone below .

Limestone, brown to gray, granular;resistant, forming one massive bed insome places but not all. Containssome masses of chert .

Limestone, brown to gray, upper footthin-bedded, nonresistant, andweathering light gray .

Continuing Partial Section at Bridge inSW~~, T. 17 S., R. 20 E.

Limestone and some dolomitic limestone,dark gray, granular to dense to finelycrystalline, beds averaging 1 foot inthickness, bedding planes irregular;some elongated masses of chert;slightly "worm-eaten': in places .

Dolomitic limestone, thinly laminatedLimestone, light gray, greatly 'lworm-eaten"Dolomitic limestone, some limestone and

dolomite, medium-dark gray, dense tofinely granular, some slightly "worm­eaten." Bedding as much as ly!feet thi ck .

Limestone, shaly, brittle, fissile. At lowercontact, persistent siliceous tabularmasses containing a great many fossilfragments, principally of brachiopods

Limestone and dolomitic limestone, threethick beds. Some siliceous zones

Dolomitic limestone, buff, finely crystalline;forms massive cliff; weathers to veryirregular, hackly surface .

Limestone, shaly, nonresistant .Dolomitic limestone, gray; surface

hackly, bedding planes irregular .

Continuing Partial Section 2 Miles Downstreamfrom Scharbauer Ranch

Dolomitic limestone, very argillaceous,beds averaging 2 inches in thickness

Oolomitic limestone, a little dolomite andchert, mostly medium-dark gray, veryargillaceous; beds mostly less than6 inches thick .

Dolomitic limestone, medium-dark gray,very argillaceous; thicker-bedded thanzone above .

Feet

8

94

2

2

6

4

2021

18

2

6

14

2

II

20

11

Continuing Partial Section from 0.3 Mile to 1.8 MilesDownstream from Clements Ranch-Continued

Composite Sectz'on 01 the San Andres FormationAlong the Rio Penasco-Continued

Upper part of San Andres Formation:-Continued

Continuing Partial Section 2 Miles Downstream fromScharbauer Ranch-Continued

Feel

Dolomite, den se, dark . . . . . . . . . . . . . . . . . 1Limestone, medium-grained. . . . . . . . Y;;Dolomite, dense, dark .. . . . . . . . . . Y;;Dolomitic limestone, medium-dark gray,

very argillaceous 4

Continuing Partial Section from 0.3 ~lile to 1.8 MilesDownstream from Clements Ranch

Limestone, oolitic, somewhat dolomitic 5Dolomitic limestone, near dolomite, dark

gray; fractures contain calcite;resistant, forming cliff . . . . . . . . . . 19

Dolomite, fissile, argillaceous zones atbase and top and I-foot fossiliferousbed between . . . . . . . . . . . . . . . 3

Dolomitic limestone, some very neardolomite, dark, fine-grained; bedsaverage 8 inches in thickness;bedding planes irregular . . . . . . . . . . . . . . 9

Dolomite, dense, dark . . . . . . . . . . . . 1Limestone, fine- to medium-grained, dark

to light gray; some of the fine isdolomitic . . . . . . .. . . . . . . . . . . . 10

Dolomite, dark gray, dense, beds as muchas 1 foot thick. . . . . . . . . . . . . . . . 21,2

Limy material, argillaceous, incoherent,yellow . . . . . . . . . . . . . . . . . . . . }2

Limestone, SOme dolomitic, medium tofinely granular, argillaceous, beddingplanes irregular . . . . . . . . . . . . . . 5~';

Dolomitic limestone near dolomite, dense,dark, weathering to light gray, argillaceous 2

Limestone, medium to finely granular,argillaceous; contains bands of silica neartop; bedding planes irregular; resistant,forming prominent outcrops .. . . . . . . . . . 6

Cover. . . . . . . . . . . . . . . . . . . . . . . . 3Limestone, dark to light, granular, mostly

very argillaceous; beds average 1 footin thickness; a few beds somewhat dolomitic. 14

Limestone, oolitic, light gray, thick-bedded, forms cliff. . . . . . . . . . . . . . . 6

Dolomite, dense, blue-gray; very irregularfracture; forms cliff with limestone above 2

Dolomitic limestone near dolomite, fine tovery fine-grained, dark blue-gray; bedsaverage about 1 foot in thickness;siliceous masses near center. . . . . . . 8

Cover. . . . . . . . . . . . . . . . . . . . . . . . 4Limestone, becoming dolomitic toward top;

medium-grained to dense. . . . . . . . . 7Shale, yellow, limy, argillaceous, thinly

bu't poorly bedded. . . . . . . . . . . . . . 'l2Limestone and dolomitic limestone, alternating 5Limestone, with chert nodules. . . . . . . . 1Dolomite and dolomitic limestone, mostly

dense, ·dark gray, but some granular,blue-gray and light gray. . . . . . . . . . . 13

12

Dolomite, very dense .Limestone, somewhat dolomiticLimestone, granular, blue-gray,

brownish-gray, and light gray .Dolomitic limestone, medium-light gray;

contains patchy llworm-eaten" zones, thesebeing more highly calcareous; forms cliff

Total, upper part of San Andres

Lower part of San Andres Formation:

Limestone, blue-gray and brov.'11, granularto dense; contains calcite bodies whichmay have been fossils; one massive bedin mOSt places, forming prominent cliffwith beds above •...............

Limestone, medium tQ finely granular, blue-gray,argillaceous; mostly in beds 8 inches to2 feet thick, but several thin-beddedshaly zones occur .

Limestone, dark brownish-gray to light gray,medium coarsely granular, highly argil­laceous; very irregularly bedded thoughmuch is thin-bedded and shaly .

Limestone, argillaceous, blue-gtay, granularto dense; granular layers contain fossils;in beds 8 inches to n~ feet thick sepa­rated by thin-bedded shaly zan es as muchas 1 foot thick; nonresistant .

Limestone, argillaceous, blue-gray, granular,somewhar fossiliferous; in tWO massivebeds; contains stylolite seams .

Limestone, argillaceous, blue-gray, medium tofinely granular; beds 8 inches to 2 feetthick, some separated by thin-bedded limyshale 1 to 2 inches thick; fossiliferous,particularly near base, containingbrachiopods and gastropods .

Limestone, argillaceous, blue-gray, dense;one massive bed having very irregularfracture .

Limestone, argillaceous, dark gray, dense,bedding and fracture irregular; contains thinlaminations of differing resistance whichweather into relief; a few fossils andfossil fragments .

Limestone, argillaceous, fine- to medium­grained, medium- to thin-bedded,fossiliferous .

Limestone, argillaceous, blue-gray, fine­grained, beds averaging 8 inches in thick­ness; upper foot very fossiliferous

Limestone, blue-gray, medium-grained,medium-bedded; contains many fossilsand fossil fragments, particularly horncorals in upper 4 feet; stylolitecontact at base .

Limestone, very fine-grained, medium-bedded;contains few fossil fragments .....

Limestone, blue-gray, poorly bedded andirregularly fractured; contains styoliteseams, fossils, and fossil fragments .

Feet

12

16

20360~~

6

21

4

13

8

6

4

6

. 7

j

8

10

6

Yeso(?) Fonnalion-Continued

II~

I,

,,\: ',.,.

CompOS£le Section of tbe San Andres FonnationAlong the Rio Penasco-Continued

Lower part of San Andres Formation:-Continued

Feet

Limestone, argillaceous, gray, medium-bedded;contains some fossil fragments. . . . . . . . . . 5

Sandstone, argillaceous and silty, red, yellow,and light gray, thin-bedded; and medium-bedded gray dolomitic limestone .

Total Yeso(?) formation exposed

Feet

15

144

Percentages of Various Lithologic Types in theSan Andres Formation along the Rio Penasco

Other differences between the two divisions havealready been mentioned. They include the commondark, bituminous limestone in the lower part; thethicker bedding in the lower part; the more numer-

The thickness of the San Andres formation meas~

ured in the section, including the sandstone at itsbase, which geologists ordinarily identify as theGlorieta sandstone member, is 770 feet. Below theGlorieta(?), 144 feet probably belonging to the Yesoformation was described. The San Andres fallsnaturally into upper and lower divisions, the di­vision between them being at the base of a persist­ent zone of resistant dolomitic limestone 409 feetabove the base of the formation. This zone occursalong both sides of the Penasco west of the junc­tion of Highways 83 and 24 and can also be foundfor a few miles east of that junction. It is extremelyunlikely that this zone, at least with a well-definedbase, could be found and identified at more remotepoints in the area, but the general separation of theSan Andres into upper and lower divisions is ofvalue because of differences which are persistentover wide areas.

The greatest difference between the upper andlower divisions of the San Andres formation alongthe Penasco is the predominance of dolomitic lime~

stone in the upper part as against the overwhelmingpredominance of more calcareous limestone in thelower. The following table gives the percentage ofeach rock type occurrinJ; in the two divisions.The measured thickness of the upper division was360~ feet; lower division above the Glorieta sand­stone, 394 feet.

II

••••••I

•••

Continuing Partial Section Across Riverfrom Runyan Ran ch

Dolomitic limestone, argillaceous, medium­dark gray weathering to brownish, fine­grain ed, very thick- bedded, slightlyuworm-eaten"j solution pits common onsurface; forms cliff. . . . . . . . . . . . . . 23

Limestone, argillaceous, medium-dark gray,fine-grained; beds about 1 foot thickexcept thicker-bedded upper 5 feet;layer of fusilinids (?) 3 feet from top 19

Cover 21Limestone, argillaceous, mostly dark gray,

bituminous, medium- to coarse-grained;bedding averages 8 inches in thickness . 15

Limestone, mostly dark gray, bituminous,medium- to coarse-grained, slightly argil­laceous in places; bedding ranges fromabout 1}~ feet in thickness to very massive;irregularly fractured; contains some goodbrachiopods and coral specimens, butfew gastropods . . . . . . . . 191

Cover 4Dolomitic limestone and dolomite, some

limestone toward top; fine-grained toden se, medium- to thin-bedded ..... 12

Glorieta(?) sandstone member, calcareous,white to yellow-brown, medium-grained. 15

Total, lower part of San Andres 409

Yeso(?) Formalion

Limestone, some dolomitic, medium-dark gray,medium- to fine-grained; some cover ... 15

Dolomitic limestone, argillaceous, medium-dark gray, dense; contains many verticaljoints and irregular bedding. . . . . . 8

Limestone, dark gray, medium-grained,medium- to thin-bedded; containsfossil fragments 16

Limestone, slightly dolomitic, argillaceous,medium-dark gray, dense; beds averaging4 inches in thickness . . . . . . . . . . . . . 6

Limestone, argillaceous, medium-dark gray,finely crystalline, massivej containssome fossil fragments 12

Limestone, dark gray, medium-bedded, denseto medium-grained; full of fossil fragments 19

<::Over .......•... . . . . . . . . . 3Dolomitic limestone, very silty,

greenish-gray, massive. . . . 14l)olomitic limestone, argillaceous and silty,

medium-dark graYj beds less than1 foot thick •.................. 16

Sandstone, silty, calcareous, yellow to gray 20

Rock Types

LimestoneDolomitic limestoneDolomiteShale and chertCover

Upper Division

3655

7

Less than2

Lower Di vision(excluding the

Glorieta (?)sandstone

member)

858I

Less than6

II

13

ous fossils in the lower part; and the more common"worm-eaten II rock in the upper part. In additionto these, the intake capacity of the upper divisionalong the Penasco is greater than that of the lower.This is due to thinner bedding, the presence of asomewhat larger number of joints, and the largerproportion of POtous and "worm-eaten" rock in theupper part. The difference is not gteat, the ca­pacity of the upper division being clas sed here asmoderate, and the lower as moderate to low.

The ttaverse along the Penasco showed that ingeneral there is much lateral variation in the rocksof the San Andres formation. Zones distinctiveenough to be followed for a mile or more are rare.Resistant zones often become nonresistant wirhin ashort distance, and in places very thick-beddedzones become thin-bedded within a few feet (p. 11)."Worm-eaten" roc k may lose its porosity in a shortdistance. For example, a typical sequence ofcarbonate rocks over 120 feer thick, nearly allhighly "worm-eaten," forms the north wall of acanyon tributary to the Rio Penasco just south ofHighway 83, 3Yz miles west of the Highway 13j unction. One mile farther east, where this tribu­tary empties into the Penasco, this same sectioncontains practically no "worm-eaten" rock. Thedolomitic limestone zone in the lower division ofthe San Andres, 277 feet above its base, forms aprominent cliff near the top of the canyon wallopposite the Runyan Ranch in sec. 29, T. 16 S., R.17 E. A little more than 2 miles downstream on thesame side the zone is less prominent but is stillcomposed of dolomitic limestone. Across thecanyon from this point, however, it has apparentlygraded into more calcareous limestone, for no traceof dolomitic limestone can be found in the promi­nent zone in the same stratigraphic position.

Beds of distinctive lithology can often be usedfor correlation over shorr distances. Thin beds ofargillaceous limestone or limy material, commonlyyellow, have been found especially valuable in thisregard. Dolomite beds and beds of oolitic lime­stone have also proved usefuL In some placesfossiliferous zones have been very helpful overdistances as great as a mile.

Secondary calcite not only occurs in large open­ings such as solution channels, joints, and faultsalong the Rio Penasco, bur also fills innumerablesmall fractures and pores in apparently solid rock.It is very common, in testing dolomite or dolomiticlimestone near dolomite with a 10-percent solutionof hydrochloric acid, to discover that effervescenceon most of a fresh surface is either lacking or veryslow, but that in minute filled fractures or poresit is very brisk.

14

A large proportion of the nodules and irregularmasses of chert in the San Andres formation aremade up of siliceous fossil fragments, mainlybrachiopods, gastropods, and corals, cemented bysilica. Banded chert is much rarer. A sample ofbanded chert was found in float along the tributaryentering the Rio Penasco just west of the Y-Ostructure about half a mile above its mouth.

No major lithologic changes in the San Andresformation occur north of the Rio Penasco until thevicinity of Rocky Arroyo. Near the eastern marginof the limestone uplands there, numerous sinkholesbegin to appear, indicating almost certainly thatgypsum is present some distance below the surface.Farther north the gypsum itself can be found in thewalls of some of the sinks on Sixmile Hill, and athin zone of gypsum crops out on the south side ofHighway 70, 3.4 miles west of the crest of SixmileHill. This zone is very near the top of the forma­tion, a~ an outlier of red beds of the Chalk Bluffformation occurs about a mile to the northeastwhere it can be seen easily from the highway. Thegypsum does not extend as far southeast as sec.9, T.12 S., R. 23 E., as no gypsum was encounteredthere in the driliing of the F. L. Sherman well. Irev idently also either pinches out or has beeneroded off farther west, as the H. L. Woods wellin sec. 18, T. 11 S., R. 22 E. encountered nogypsum. The available evidence indicates that thegypsum in Sixmile Hill and vicinity is not continu­ous with "the widespread gypsum of the Salt Creekarea. The latter has not been found at the surfacesouth of the second row of sections in T. 9 S.,R. 22 E. Morgan 7 repqrts that in general the SanAndres becomes more gypsiferous northward towardVaughn.

A few thin yellow-brown thin-bedded shales andsiltstones are present in the lower division of theSan Andres west of Roswell. These beds are veryirregular in shape and tend to pinch out within ashort distance. A bed 7 feet thick occurs in a roadcut on Highway 70 near the top of Rio HondoCanyon east of Riverside. Limestone is the mostcommon rock in the lower division in this vicinity,as it is along the Rio Penasco.

The beds of the San Andres along the Rio Hondoand northward have slumped in a great many placesbecause of removal of material below by solution.The resultant fracturing of the rock is an importantfactor in making the intake capacity of this areamuch higher than that farther south.

;Morgan, A. M., 1942, Report on the Geology of the PecosValley, N. Alex.: ms. rcpt. in files of U. S. Geol. Survey inAlbuquerque, N. Mex.

•

IJ

II

••••••••••••l

•••'.<'.'.',... ~"

I

••

CHALK BLUFF FORMATION

The Chalk Bluff formation overlies the SanAndres formation and is separated from it by anerosional unconformity. The Chalk Bluff is theequivalent of the Whitehorse group in the subsur­face farther east. The Chalk Bluff in this generalregion has been subdivided into the followingmembers, from oldest to youngest: an unnamedlower member, the Queen sandstone member, theSeven Rivers gypsiferous member, and the ThreeTwins member. Where these members can be difw

ferentiated in the ouecrop areas, the lower memberisrypically high in dolomite, the Queen in sand­srone, and the Seven Rivers in gypsum. The ThreeTwins member does not occur in the Roswell Basin.In the northern part of the Roswell Basin the lowermember and Queen sandstone member make up allthe Chalk Bluff formation present, but in the south­ern part the uppermost beds are of the Seven Riversgypsiferous member.

To the south the various units of the Chalk Bluffgrade into other formations in the reef zone border­ing the Delaware Basin and in the basin itself.Geologists are not in agreement on these correla­tions, and the only one of concern in this paper isthe gradathm southward of the lowest beds of theChalk Bluff into limestone in the lower part ofthe Whitehorse group. According to Morgan (1941,p. 781) this takes place in the subsurface in theartesian area, and he states that the ttDog Canyon 1t

(Goat Seep) limestone replaces progressively higherbeds of the Chalk Bluff southward, beginning withthe base near Lake Arthur. The Chalk Bluff isroughly equivalent to the Pecos formation of Nye(Fiedler and Nye, 1933, p. 44), bur as srared byMorgan (1941, p. 781) the ftDog Canyon" was in­cluded in Nye's Picacho limestone underlying thePecos formation.

In the vicinity of Roswell the Queen sandstonemember and the unnamed lower member of the ChalkBluff are composed mainly of sandstone and shaleof the red-bed type and anhydrite or gypsum, butthey also contain some thin beds of limestone.East of the Pecos, and also north of Roswell, theChalk Bluff contains halite. Excellent exposuresof the Chalk Bluff occur along the east side of thePecos near Roswell. A large area of outcrop westof the Pecos is present in the upper part of theCottonwood Creek and Walnut Draw drainages. TheChalk Bluff formation forms the confining bedsabove the artesian aquifer from the vicinity ofRoswell southward. North of Roswell the ChalkBluff is largely missing in the artesian area, andclay beds in the valley fill act as the confin~ing beds.

Water occurs in the Chalk Bluff formation in andnear the large area of outcrop in Tps. 15 and 16 S.It is rather highly mineralized but is used fordomestic purposes in some places. East of thePecos water in the Chalk Bluff is commonly verysalty, according to Morgan, and often stock will notdrink it. In the artesian basin water occurs underhead in some places in beds of sand and limestonein the lower part of the Chalk Bluff.

GOAT SEEP LIMESTONE

The lower beds of the Chalk Bluff formationgrade into the Goat Seep limestone in the southernpart of the ~rtesian basin, as noted above. Morganstates that the lCDog Canyon" (Goat Seep) in thesubsurface between Lake Arthur and Lakewoodconsists of dolomitic limestone, anhydritic andsandy limestone, and sandstone. He says that itthickens at the expense of the lowest beds in theChalk Bluff, from about 60 feet near Lake Arthur toabout 700 feet at Lakewood. The principal artesianaquifers in the southern part of the basin occur inthe Goat Seep. They are especially well developedin its upper part in anhydritic limestone beds whichinterfinger with sandstone and anhydrite beds of theQueen-Grayburg division of the Chalk Bluff extend­ing from the north.

TERRACE OEPOSITS

Terrace deposits of Quaternary age overlie thePermian formation west of the Pecos River in a belt12 to 25 miles wide. Their thickness ranges from afeather edge to 350 feet, the thickest depositsoccurring along a line approximately parallel to theriver and about 4 miles west of it. The depositsconsist of clay, sand, and grave I, some of whichare firmly cemented into shale, sandstone, andconglomerate. They were laid down by shiftingstreams flowing from the west to the Pecos River,and consequently the various lithologic types arevery irregular in distribution. Nye (Fiedler andNye, 1933, p. 35-38) refers ro rhese beds as rhequartzose conglomerate. The corresponding bedson the east side of the river were named the Gatunaformation by Lang (Robinson and Lang, 1939, p.84-85). The beds of the quartzose conglomerateand the Gatuna formation in most places dip atvarious angles owing to collapse following solutionand removal of underlying beds of gypsum and salt.

The shallow water of the Roswell Basin occursin the terrace deposits.

STRUCTURE

GE;,ERAL ATTITUOE OF THE ROCKS

The rocks of the Roswell Basin in general dipgently a little south of east on the west side of the

15

Permian Basin, which underlies a large part ofwestern Texas and eastern New ?\.fexico as well assmaller parts of Oklahoma and Kansas. On thewestern rim of the basin lie the Sacramento Moun­tains and the Sierra Blanca. The former is a sedi­mentary cuesta capped by limestone of the SanAndres formation, which extends down its easternslope toward the Pecos River. The latter is acrystalline complex composed mainly of igneousrocks on whose eastern flanks r-..fesozoic sedimen­tary rocks occur, but within 15 miles of whosesummit the limestone appears and shortly assumesa gentle dip to the southeast. Northeast of theSierra Blanca the igneous complex of the Capitan~fountains has been pushed up through the sedi­mentary rocks along an east-west axis.

The average southeastward dip of the rocks isabout 50 to 60 feet to the mile, which is greaterthan the slope of the land surface. Progressivelyyounger beds are thus generally exposed at thesurface to the south and east. There is consider­able variation, however, in the regional dip, and inmany places broad undulations of the strata resultin outCtOpS of younger beds at the surface west ofthe outcrops of older beds. One of these is northof Highway 70 about 2~ miles west of Sixmile Hill,where a small outlier of red beds of the Chalk Bluffformation lies 4 miles or more west of the mainbody of the formation. The most striking structuralfeatures are the faulted anticlines which extend formany miles approximately along the regional strike.These ate the Border Hills, Sixmile Hill, and Y-Ostructures. Other anticlines, synclines, domes, andminOt faults are ptesent, as will be described onthe following pages.

FAUL TED ANTICLINES

GENERAL FEATURES

The Sixmile Hill, Border Hills, and Y-0 structuresare faulted anticlines where typically developed,although their structure varies somewhat from placeto place. The Sixmile Hill structure is a simplefault in places, an anticline containing secondaryfolds and probably a fault or faults in others, andpossibly a simple anticline in still others. Some­what less variation occurs in the Border Hills andY-0 structures.

The similarities of the three faulted anticlinesare notable. They all trend in a northeasterlydirection. All are very narrow, being less than aquarter of a mile wide in places, and are remarkablylong. All are broken by a high-angle fault of rela­tively low vertical displacement which is reversedfrom place to place along the structure, exceptpossibly in the case of the Y-O fault, where re-

16

versal of displacement has not been proved. Nye(Fiedler and Nye, 1933, p. 78) estimates that thethrow does not exceed 11200 to 300 feet" along anyof the faults, and the present writer is doubtful ifthe vertical displacement is anywhere greater thanhis lower figure. It is possible that the horizontalcomponent of movement along the faults is of thesame order of magnitude as the vertical component.

SIXMILE HILL STRUCTURE

The Sixmile Hill structure is best known west ofRoswell, where it is expressed as a hill whosecrest is between 5 anq 6 miles from the center ofthe town. Here the structure is a broad anticlinebroken by a fault a short distance east of the crest.The structure is at least 60 miles long, and it canbe followed southwest from the point where itcrosses Highway 285 about 10 miles north of Ros­well until it intersects the syncline just east of theBluewater anticline about 2 miles north of Dunken.Its strike along most of its extent is about N. 40°E., but the trend changes in several places north ofthe Hondo Reservoir so that it averages aboutN. 31° E. northwest of Roswell. Two faults havinga more easterly trend branch off the main structurebetween Rocky Arroyo and the Rio Felix (fig. 2).

The characteristics of the main structure changein several places. Where it crosses South BerrendoCreek two narrow anticlines are exposed. A faultof very small displacement breaks the crest of oneof these, but this may not be the major fault. About2 miles farther south a scarp on the east side ofSixmile Hill evidently lies along the fault, and fromjust north of Highway 70 nearly to the HondoReservoir the fault lies along a nearly straightvalley broken in places by low saddles. Gooddirect evidence for the existence of the fault wasfound in drilling a well for oil in the bottom of thisvalley, a mile and a quarter south of the highway.Nye (Fiedler and Nye, 1933, p. 79) reports rhat twoholes were drilled a few feet apart, the secondbeing drilled because the bit was deflected, prob­ably along a fault plane, in the first. The samething happened in the second hole. The Glorieta(?)sandstone member of the San Andres formation wasencountered at two different depths in the twoholes, and on the basis of this and other informationNye states that the throw on the fault is "probablyless than 50 feet.» One of the holes is now awater well and was measured in connection withthis study.

The anticlinal nature of the Sixmile Hill structurein this vicinity can be determined by inspection ofthe cuts along Highway 70, but it is obscured bymany variable dips. Some of these are due to col­lapse, bUt others are probably expressions of

••

IIII

"f

--11II~~~~

I

IIIIIIIIIIIIlIIILII

narrower folds on the main anticline. The principalfault does not show in the road cuts, but severalsmaller faults showing displacements up to 10 feetcan be seen. Most of these are probably due toslumping or collapse rather than to the diastrophicforces that formed the main structure.

Two wellwdefined joint systems occur in theSixmile Hill area and undoubtedly extend beneathtbe Hondo Reservoir. One of these parallels thestructure and the other trends about N. 65° W., thusintersecting the first set at an angle of abollt100 degrees.

The intake capacity of the limestone of the SanAndres in Sixmile Hill is very high because of thenumerous sinkholes and the fractured, broken rockcaused both by slumping and by deep-seatedearth forces.

The fault valley in Sixmile Hill bifurcate s abrupt­ly within a mile of the northeast edge of the HondoReservoir, and this probably indicates a bifurcationof the fault. Southwest of the reservoir the rivervalley lies along the structure for 3 miles, and ashort distance to the east shallow valleys and lowsaddles mark another parallel structural line. Thelatter almost certainly lies along a fault, and it maybe that a parallel fault underlies the river, the twobeing continuous with the two branches of thefault just north of the reservoir. The structureimmediately south of the reservoir shows the samecomplexities it does to the north-secondary folds,evidences of collapse, sinkholes (though not asmany), and well~developed joints.

The condition underlying the Hondo Reservoircan be inferred from the condition of the SixmileHill structure just north and south of it, remember­ing that the structure passes directly through thecentral part of the reservoir (see pI. 1). The Six­mile Hill fault underneath the reservoir probably iscomposed of two subparallel strands less than amile apart. The strata around and between thestrands have been folded and fractured further dur~

ing formation of the Sixmile Hill structure, and twosets of joints have been developed. Small sink­holes, which are most numerous in the northeasternpart of the reservoir, show that solution has beenactive at the surface, and solution at greater depthshas resulted in much irregular slumping and col~

lapse of the strata. It was in this fractured andbroken rock that the reservoir sink was formedunder natural conditions by flood waters of the RioHondo before the coming of man (see p. 7), and inthis silt~covered sink the Hondo Reservoir wasbuilt.

Five miles southwest of the reservoir along thetrend of the Sixmile Hill structure, several narrow

flexures and mwor fractures in a zone 250 yardswide appear along the banks of Rocky Arroyo. Noclear indication of a major fault is shown. Continu­ing southwest, the structure decreases in intensity,but it can be followed easily on aerial photographs,as valleys have developed along the fault almostcontinuously, low saddles separating the valleyheads.

Faults branch off the main structure in sec. 1,T. 13 S., R. 21 E., and approximately on the linebetween sees. 7 and 18, T. 15 S., R. 20 E. Thebranch faults have a more easterly strike than themain structure and can be followed southwestwardby their physiographic expression to the easternlimb of the Bluewater anticline, beyond which theydo not show clearly on aerial photographs. Themore northerly fault apparently lies along the axisof an anticline in part of its extent. The magnitudeof the disturbance is not great, however, and thethrow at the fault probably nowhere exceeds 20feet. The vertical throw on the more southerlyfault is only 6 feet just south of the Rio Felix, thewest side being upthrown, and no anticlinal struc­ture is present.

The anticlinal nature of the main Sixmile Hillstructure disappears entirely before it reaches theRio Felix, and the throw of the fault at the Felixis only 4~ feet, the east side being upthrown. Be­tween the Felix and the Rio Penasco the structureincreases in intensity again, and 1 mile north ofHighway 83 the width of the disturbance is about250 yards. The structure here is a complex anti­cline, and beds can be found dipping at a greatmany angles, some nearly vertical, on the limbs ofseveral sharp flexures across the structure. A fewfaults of low displacement are probably presentalso. The structure is of the same nature here asit is in the vicinity of the Hondo Reservoir, exceptthat sinkholes and collapsed areas are missing.The structure is known to continue to the southwestuntil it intersects the syncline east of the Blue­water anticline 2 miles north of Dunken and a shortdistance west of Highway 24.

The Sixmile Hill structure is thus in most placesa broad anticline containing secondary folds andbroken by one or more faults of relatively low dis­placement near the crest. In Sixmile Hill the westlimb of the fault is upthrown, but at the Rio Felixthe east side is upthrown and it is entirely possiblethat there are other reversals along its extent. Nofolding is present at the Felix. In many places theexistence of a fault cannot be proved, but thephysiographic evidence indicates that it is probablycontinuously present or nearly so, except north ofSouth Berrendo Creek, where evidence of a fault islacking and the structure may be simply an anticline.

17

BORDER HILLS STRUCTURE

The Border Hills structure is expressed as aridge of that name which Highway 70 crosses about25 miles west of Roswell. The structure is clearlya narrow faulted anticline throughout most of itsextent, although the fault does not show distinctlyin road cuts at the highway and may die out to thenorth. The structure is well exposed in the wallsof Rio Hondo Canyon between 2 and 3 miles southof Highway 70. There it is clearly a narrow anti~

cline, broken at the crest by a nearly vertical fault.Very surprisingly, the vertical component of thefault reverses itself across the river. North of theriver the east limb is upthrown and on the souththe west is upthrown. This observation was firstmade by Nye (Fiedler and Nye, 1933, p. 79) withsome question, but it was checked by the presentwriter who corroborates Nye's tentative findings.Two excellent sketches of the structure here areincluded in Nye's work (Fiedler and Nye, 1933,fig. 11). There is a pronounced horizontal bend oroffset in the fault where it crosses the river, thesouthern part of the structure being displaced tothe west.

The Glorieta(?) sandstone member of the SanAndres formation is exposed in many places nearthe axis of the structure, and on the north side ofthe Hondo another sandstone which is probably inthe Yeso formation appears below it. These mark­ers are very useful in determining the upthrownside of the fault, but the writer was unable to findthem opposite each other in the same locality todetermine the amount of displacement. Dips oneither side of the anticline are as steep as 50° butare mostly about half that amount. The base of theeastern limb particularly is very sharp, and massivehorizontal beds assume a dip of more than 20° in avery few feet. A well-developed joint systemparallels the structure.

The Border Hills structure is not a typical thrust~

faulted anticline. Instead it evidently was formedby forces that pinched the rocks into a very narrowfold which broke nearly vertically along the axis,further relief from the stresses being gained byrelative displacement of the two limbs. Displace~

ment was partly vertical but may have included acomparable horizontal component as well. Thetotal throw along this fault was not great, andprobably did not exceed 300 feet anywhere alongthe structure. The zone of crushing and breccia­tion along the axis of the anticline is very wide;however, in the saddle 300 yards south of the RioHondo, for example, its width is 200 feet.

The trend of the Border Hills structure variesconsiderably, but it averages about N. 32° E. Its

18

northeastward limit as mapped in this report is insec. 5, T. 10 S., R. 21 E., where available aerialphotograph coverage ends, but it undoubtedly ex~

tends into the Salt Creek area, for a monoclinedirectly in line with the structure was found in thebanks of Salt Creek in sec. 22, T. 8 S., R. 22 E.

South of the Rio Hondo, valleys parallel thestructure along much of its extent. These commonlylie a little west of the fault, which is often markedby less conspicuous saddles. Six and a half milessouthwest of the point where it crosses the RioHondo, the Border Hills strUCture is again bent tothe west, and again the bending takes place at awater gap, here occupied by a tributary to the RioFelix. On the south side of this water gap thestructure bifurcates. The east branch is mainly afault, and it dies out in about 3~ miles. The westbranch continues as a faulted anticline into T. 14S., R. 18 E., and Merritt (1920, p. 55) states thatthe structure continues southwestward as far astl a point below Elk" in T. 16 S., R. 16 E.

Y·O STRUCTURE

The Y-0 structure is a northeast-trending narrowanticline faulted along the axis. It crosses the RioPenasco in three places within 2 miles west of theY~O Crossing, from which it takes its name. Itsdirection varies only a few degrees from an averageof N. 41° E. in the area mapped, and it extends atleast from T. 18 S., R. 18 E. to the terrace depositsin T. 14 S., R. 23 E. The structure is markedtopographically by a discontinuous line of hillsalong much of its extent, but the Rio Felix alsofollows it for about 5 miles, swinging back andforth across it.

The Y~O structure is well exposed in the northwall of the canyon of the Rio Penasco about 0.6mile west of the Y~O Crossing. The anticlinalstructure is well developed here, and the crushedand brecciated fault zone can be seen near theriver. The writer was unable to determine theamount and direction of throw at this point, but 0.7mile southwest, where the structure again crossesthe Penasco, the vertical displacement is only 10feet, the east side being upthrown. The throw isprobably greater at the exposure to the northeast.Dips on the flanks of the anticline along thePenasco are generally less than 10°. The structuredecreases in intensity south of the southern bound~

ary of T. 17 S., but it continues, according toRenick (1926, pi. 1), to sec. 31, T. 18 S., R. 18 E.To the north, it is well exposed along the Rio Felixin several places. In the eastern part of T. 15 S.,R. 22 E., the eastern limb of the anticline is welldeveloped, but dips on the western limb are verylow. The YwO structure cannot be followed north~

•••••••JIIIII

II

II

IIIIIIIII

eastward across the terraces west of the PecosRiver, but Nyc (Fiedler and Nye, 1933, p. 80-81)presents evidence from wells that it may extend asfar north as T. 10 S., R. 26 E., in the latitudeof Roswell.

OTH ER STRUCTURES

BLUEWATER ANTICLINE AND SYNCLINE TO EAST

The Bluewater anticline is a broad, complex foldtrending a few degrees east of north whose easternlimb underlies the sharp weseward increase inelevation and relief extending along a line throughthe junction of Highways 83 and 24. Its axis islargely west of the area mapped in detail in thepresent report, and it is not shown in figure 2. JUSt

east of the anticline lies a narrow syncline (seefig. 2), which has some characteristics in commonwith the faulted anticlines previously described.Where exposed south of the Rio Penasco, a zone ofsecondary folds and r:ontorted strata 100 yards wideoccurs along its axis. A fault may be present, butif so its displacement is small, as there is a strati­graphie break of only 60 feet between beds onopposite sides of the central disturbed zone.

BLACK HILLS ANTICLINE

Renick 0926, p. 124) describes an anticline ly­ing south of the Rio Penasco whose axis trendsabout N. 60° E. and extends from sec. 36, T. 17 S.,R. 19 E., Hat least as far north as" sec. 14, T. 17S., R. 20 E. The structure is known as the BlackHills anticline, and it has been drilled for oil with­out success. The !\'fagnolia Petroleum Co. wasengaged in remapping the structure during thesummer of 1947, with the object of locating itmore accurately.

DUNKEN DOME

An anticline whose axis extends approximatelynorth-south along the east boundary of sees. 30 and31, T. 17 S., R. 18 E., was mapped by Renick(1926, p. 124) and called the Dunken dome. Ananticline crossing the Rio Penasco in the W~ sec.6 of the same township was found during the pres­ent investigation and is believed to be a northwardextension of the Dunken dome. Dips on the flanksof the anticline here are 5° or less.

STRUCTURE ALONG THE RIO PENASCO

Along most of the Rio Penasco between Highway24 and sec. 14, T. 17 S., R. 20 E., the dip of theSan Andres is less than 1°. In the eastern 7 or 8miles of this course, east of the Y-O Crossing, thedip of the rocks is actually less than the gradientof the river, and as a result the highest exposedbeds of the formation occur on either side of theY-O fold. Farther west the dip is somewhat steeper,

though still generally less than 1°, and older bedsappear farther upstream. About 7~~ miles in an airline west of the westernmost crossing of the Y-Ostructure a series of eastward-dipping monoclinescontaining dips of as much as 10° appears, thebeds between dipping very slightly to the east asbefore. The Dunken(?) anticline, a syncline, andthe Bluewater anticline west of these structureshave already been described.

Vertical joint systems trending N. 3So E. andN. 26° W., the latter the better developed, areprominent between the Scharbauer and the ClementsRanches, and peculiar joints striking N. SO W. anddipping 30° to 45° occur for a mile or more abouthalfway between these ranches. These last jointsare unusual in that they are limited in verticalextent to certain beds or zones which range from5 to 50 feet in thickness. The well-developedjointing in this vicinity increases slightly theintake capacity of the San Andres formation.

Slumping due to solution and removal of under­lying beds does not occur along the section of thePenasco described here. In this respect it differsgreatly from the Rio Hondo below Border Hills, andthis difference is an important factor in causing theinfluent seepage of the Penasco to be much lessthan that of the Hondo. (See p. 21-24.)

MINOR FAULTS

A very unusual series of minor faults occursbetween the Felix and the Penasco in and adjacentto sec. 36, T. 15 S., R. IS E. The faults all strikeabout N. 7° W., and occur in an area about % milewide and 2 miles long. The individual faults aremostly very short, the longest extending about %mile and the shortest only about 200 feet. Aneast-west line drawn at random across the areawould encounter from 2 to 12 of the structures.

The displacement along the faults is very small.In places the faults are little more than joints, butin other places the measurable throw may be asmuch as 5 feet. Strangely enough, in some placesthe faults occur along the axes of anticlines verysimilar, except for their minute scale, to the BorderHills and Y-O structures, even to the developmentin some places of tiny fault valleys along theircrests. Dips on the flanks of anticlines of thistype may reach 35° even thou,gh the whole structureis only 20 feet across.

These structures occur just south of an area ofnumerous sinks, and they may have resulted fromstresses set up by solution and removal of rockunderlying the sinkhole area.

Other minor faults occur in the limestone uplands.Their length may be as great as several miles, but

19

their displacement is not great, probably nowhereexceeding 10 feet. They strike in various direc­tions, some nearly at right angles to the trend of theprincipal faulted anticlines. Valleys have beendeveloped along the minor faults in most places,saddles occurring where the faults cross ridges.

STRUCTURES DUE TO SLUMPING OR COLLAPSE

In certain parts of the Roswell Basin strata arecommonly found fractured, bent, and tilted at vari­ous angles as a result of slumping after removal ofunderlying material by solution. These collapsestructures are most common from the Rio Hondonorthward. They Can be seen, along with otherevidences of solution such as small caverns andsolution channels, in the walls of Rio Hondo Can­yon for many miles below Border Hills. They areabundant in Sixmile Hill, occur along Highway 70nearly to Border Hills, and are very numerous inmuch of the Salt Creek area. Farther somh col­lapse structures are much more rare. Even in thecliffs south of Rocky Arroyo in T. 12 S., R. 22 E.,only a few miles south of the Rio Hondo, the strataof the San Andre s are re lative1y undisturbed exceptwhere crossed by the Sixmile Hill structure. Nocollapse structures can be seen in the walls of RioFelix Canyon along the Flying H Ranch road eastof R. 17 E. The few folds which do occur wereprobably caused by tectonic forces. Between theFelix and the Rio Penasco, within a few miles eastof the eastern limb of the Bluewater anticline,slumped beds and sinks occur. Collapse structuresare not present along the Rio Penasco from T. 17S., R. 20 E. to the Runyan Ranch 10 T. 16 S.,R. 17 E.

All the areas of slumped beds JUSt describedprobably were originally underlain by anhydrite orgypsum interbedded with limestone. Removal ofsome of the easily soluble anhydrite or gypsumalong with some limestone resulted in collapse ofthe overlying beds. In the Salt Creek and SixmileHill areas the gypsum probably was removed fromthe San Andres formation, but farther west, alongaQd to the north of the Rio Hondo, it probably wasdissolved from the underlying Yeso. South ofRocky Arroyo collapse structures are very rare, asthe gypsum has pinched out and been replaced bylimestone. The sinks and collapse structUres be­tween the Rio Felix and the Rio Penasco may havebeen developed on gypsiferous strata near themiddle of the San Andres.

The areas containing collapse structures areparticularly important because they are areas oflarge ground-water recharge. Streams flowingacross these areas lose a high proportion of theirflow by influent seepage into the rocks below; for

20

exall1ple, the flow of the Rio Hondo passing BorderHills usually disappears entirely before the HondoReservoir is reached.

DOMES IN GYPSIFEROUS STRATA

Both domes and sinks occur in strata of inter­bedded gypsum and limestone south of Salt Creekin and near sec. 6, T. 9 S., R. 24 E. The structurehere is very complex, and partly developed domesare much more common than complete structures.One such partial dome, elliptical in shape, iscrossed by a westward-trending secondary roadabout 1.2 miles west of Highway 285 and 1.4 milessouth of Salt Creek. The minor axis of this partialelliptical dome is about 100 yards long. Theorigin of domal structures in gypsiferous rocks hasnot been determined, but it has been suggested thatthey may be due to the expansion that accompanieshydration of anhydrite to gypsum.

EFFECT OF STRUCTURE ON GROUND WATER

The most important structural feature affectingground water in the Roswell Basin is the gentlesoutheastward dip of the rocks, which causes theimpervious strata of the Chalk Bluff formation tocap the San Andres formation west of the PecosRi ver and thus produce artesian conditions. Theeffects of the local structures described on the pre­ceding pages are not great. The principal effect ofa fault or fold is to increase the intake capacity byincreasing the amount of fracturing of the rock inthe vicinity. The capacity is further increased bythe development of valleys along the faults, so thatrunoff from the surrounding hills flows over thefractured rock of high intake capacity in the bottomof the valley. Solution in fault valleys was prob~

ably an important factor in the formation of theHondo Reservoir sink, as described on page 7.

GROUHD WATER

INTAKE AREA

AREA INVOLVED

The ultimate source of all the artesian water ofthe Roswell Basin is precipitation that falls on thearea west and northwest of the artesian area. Thetotal area that is believed to contribute to rechargeto some extent is about 7,000 square miles. Of thisamount, about 1,200 square miles was called then principal intake area" by Fiedler and Nye (1933,p. 238). This is the area just west of the upperconfining beds of the artesian reservoir in whichthe water table lies in the San Andres formation.An area of about 2,800 square miles west of thetlprincipal intake areao is the catchment basingiving rise to surface streams which cross that

•'.•

•••••••

IIIIIII

•••II

••LIII

area and contribute to its ground·water supply byseepage into the underlying rocks. It is probablethat a significant amount of water also moves under­ground into the ltprincipal intake area ll from thewest. In addition, precipitation and influent seep·age in an area of about 3,000 square miles under­lain by the San Andres formation northwest ofRoswell also probably contributes to the artesianbasin, as noted by Morgan (National ResourcesPlanning Board, 1942a, p. 45). This area is rough­ly triangular in shape and extends from a baseappc"ximately along Salt Creek and the CapitanMountains on the south to an apex near Vaughn onthe north. Its southern part is drained by SaltCreek and its tributaries, but drainage of the north~

ern part is entirely underground.

Several considerations indicate that a largerproportion of recharge for the artesian aquiferprobably comes from areas outside the ((principalintake area" than has previously been believed.Fiedler (Fiedler and Nye, 1933, p. 246-247) showsthat 29 percent of the average precipitation on thettprincipal intake area" would have to reach theartesian reservoir to recharge it completely, but hestates that nit is probable, however, that less than25 percent of the precipitation on the principalintake area plus the run~off from the tributarystreams that flow across that area percolates intothe artesian reservoir and is available for the useof the artesian wells." The present writer is inagreement with this statement, and suggests thatthe amount may be (tconsiderably" less than25 percent.

Recharge for the artesian reservoir, probably insignificant amount, comes in part from the regionwest of the ltprincipal intake area." Precipitationhere is greater than in the I!principal intake area."The main water table lies at considerable depth inthe Yeso formation in much of the area, but perchedwater bodies exist in many places and furnish waterfor many wells. The principal movement of theperched water, and also to some extent the deepwater, is undoubtedly down the dip of the stratatoward the east. Some of the perched water probablynever drops below the base of the San Andresformation as it moves slowly eastward, and thus itfinally contributes to the reservoir supplying theartesian basin. Water deep in the Yeso must seepthrough or find breaks in beds of shale beforeentering the main water body to the east, butsignificant amounts of it must do so.

Movement of ground water from the area northwestof Roswell to the artesian reservoir has beenpointed out by Morgan (National Resources Plan­ning Board, 1942a, p. 45, fig. 5). As the PecosRiver does not gain in flow between Fort Sumner

and Acme, it is believed that ground water enteringthe San Andres formation west of that segment ofthe river must eventually be discharged in theartesian basin. It is probable that the amount ofwater contributed by this area of about 3,000 squaremiles is significantly large.

DIFFERENCES IN INTAKE CAPACITY

During the present study particular attention waspaid to the differences in recharge characteristicsin the eastern part of the intake area. The exist­ence of these differences is shown well by thecomparative behavior of the Rio Hondo and the RioPenasco as they cross the area. The Rio Hondonormally has a strong flow in the vic inity of Picachoand Riverside, but after it passes Border Hills itsflow begins to decrease notably, even before theirrigation season, and it ordinarily disappears com­pletely before reaching the Hondo Reservoir. TheRio Penasco, on the contrary, continues to flowunder ordinary weather conditions at least as fareast as the Hope Diversion Dam in sec. 14, T. 17S., R. 20 E., after diminishing in flow above thispoint much less than the Rio Hondo. The differ­ence in the behavior of these streams is due todifferences in the characteristics of the rocks overwhich they flow. Furthermore, the character of therocks underlying the stream bed in any given local­ity is undoubtedly similar to that shown in thecanyon walls at the same locality. Study of therocks in these canyon walls and in other places,plus other considerations, points toward the follow­ing features as the principal characteristics ofareas of high recharge capacity: collapse strucR

tures, sinkholes, and numerous small caverns andsolution channels. Characteristics contributing ina smaller way to high recharge are joints and otherfractures, thin bedding, and "worm-eaten" limestone.

The San Andres formation shows greater evidenceof a higher recharge capacity in the walls of RioHondo Canyon below Border Hills than along anyother stream observed. The San Andres is bestexposed above the former Diamond A Ranch, but theunderlying rock generally increases in rechargecapacity along the river until past the Hondo Reser­voir. The recharge capacity at the reservoir is veryhigh. An excellent exposure in a cliff in the SEXsec. 27, T. 11 S., R. 20 E., shows irregularlyslumped bedding and many old solution channe Islined with secondary calcite. About a mile west ofthis point an old sink has been Cut into by a north­ward bend of the river and is now well exposed incross section. The beds on both sides of the sink'sthroat dip downward owing to collapse after removalof underlying material by solution. Scattered out­crops along Highway 70 north of the Rio Hondo

21

indicate that the recharge capacity of the SanAndres there is also high but in many places in theSalt Creek area the recharge capacity is probablyhigher. Sinkholes and collapse structures are vetynumerous in the northern part of T. 9 S., R. 23 E.,and the southern part of T. 3 S., R. 23 E. In someparts of T. 8 S., R. 22 E., along the banks of SaltCreek, the recharge capacity is as high as alongthe Rio Hondo Canyon but in others somewhat less.In general the recharge capacity of the San Andresis moderate to very high east of the Border Hillsstructure from the Rio Hondo northward to SaltCreek. This includes all of Fiedler's llprincipalintake area" north of the Rio Hondo.

South of the Rio Hondo the rock characteristicscommon to areas of high recharge disappear quiteabruptly. The strata on the south side of RockyArroyo are largely undisturbed except where crossedby the Sixmile Hill structure, and the recharge ca­pacity to the south is less than along the Rio Hondoand about equal to that along the Rio Felix in thenorthwest part of T. 15 S., R. 22 E., except whereexcessive fracturing at the Y~O structure has in­creased the recharge capacity. Along the Flying HRanch road east of sec. 31, T. 15 S., R. 20 E., theSan Andres formation in the banks of the Rio Felixis undisturbed and flat lying, and the rechargecapacity is less than to the east. Farther west verygentle folds and a few faults of low displacementappear in places in the canyon walls, and moreevidences of solution are present. The rechargecapacity increases only slightly, however, to about3 miles west of Flying H headquarters. West of thispoint to the junction of the road to Picacho moredeformation occurs and the recharge capacity in­creases greatly. Intake capacity is also high in asmall area in the northeastern part of T. 14 S.,R. 19 E., owing to the presence there of sinks.

Recharge capacity along the Rio Penasco in Rs.17 to 20 E. is moderate to low as compared withthat along the Rio Hondo. Collapse structures donot occur in the canyon walls, and solution effectsare not numerous, although some are present (p. 10).A few sinks occur in the vicinity and are mostnumerous east of the Bluewater antic line betweenthe Penasco and the Felix, where the intake ca­pacity locally is high. The strata of the upperdivision of the San Andres between the ScharbauerRanch in sec. 18, T. 17 S., R. 19 E., and sec. 14,T. 17 S., R. 20 E., dip less than 1° and are moder­ately thin-bedded, moderate ly fractured, and containa notable amount of Itworm-eaten" limestone.Evidences of solution are more common than theyare upstream, and the intake capacity can beclassed as moderate. Farther west, lower, thicker

22

beds of the formation are exposed. The effect ofthe thicker bedding is to decrease the rechargecapacity, but this is offset in part by prominentjoints which occur in T. 17 S., R. 18 E. (see p. 19),and by fracturing connected with the increased fold­ing and faulting in this township, and also fartherwest. The intake capacity of this segment ismoderate to low, about equal to that along thestretch from Rs. 17 to 20 E.

Study of aerial photographs and reconnaissanceon the ground indicate that there are no other siz­able areas showing characteristics of high intakecapacity between Rocky Arroyo and the Rio Pen­asco. The recharge capacity in that part of Fied­ler's ltprincipal intake area" between Rocky Arroyoand the Rio Penasco is therefore moderate to low,except for the few very small areas of high capacitynoted above.

Comparisons of seepage losses along the RioHondo and Rio Penasco illustrate well the differ"ences in intake capacity along various segments ofthe rivers. The figures in the table llDischarge atF our Gaging Stations on the Rio Hondo in 1908R

09 1 " on the following page, are taken from measure­ments made during the Hondo Hydrographic Surveyin 1908 and 1909 (New Mexico State Engineer, 1925,p. 156-159). These are the only figures availablewhich show flow at as many as four different sta­tions on the Rio Hondo.

The losses during the winter months between thestations listed in the table are due almost entirelyto influent seepage, and are thus of most signifi­cance for this study. During the growing seasondiversions for irrigation are considerable, andevaporation and transpiration direct from the riverare greater than during the winter. There are noadditions from tributaries here except during periodsof very heavy precipitation, and, as the records atRoswell and the Hondo Reservoir show subnormalprecipitation for both 1908 and 1909, it is veryunlikely that any additions of this kind are repre­sented in the table.

The figures for December 1908 to February 1909,inclusive, show that an average of 117 acre-feet permonth disappeared in the 13 miles between Picachoand Border Ranch, where the river crosses theBorder Hills structure. In this stretch the Hondoflows over the highest beds in the Yeso formationand undeformed, thick-bedded limestone in thelower part of the San Andres. In the next 12 milesbetween Border Ranch and Diamond A Ranch anaverage of 816 acre-feet per month disappeared asthe river flowed over beds of high intake capacity.In the 12~~ miles between the Diamond A Ranch andthe Hondo Reservoir, an average of 1,097 acre-feet

I

••••II

••,•JIIJII

Discharge at Two Gaging Stations on theRio Penasco, April 1927

The loss in 16 miles between the two stationswas thus 212 acre~feet in 1 month, which is pro­portionately of the same order of magnitude as theloss on the 12-mile sttetch of the Rio Hondo above

Loss in a 16-mile segment of the Rio Penascoabove the Y-0 Crossing in April 1927, is shown inthe table below (New I\texico State Engineer, 1928,p. 98-101). Nearly complete data are availablefor only one month of simultaneous operation of thestations at the ends of this segment, and even sothe discharge for five of the last six days of themonth at the lower station is estimated. There areno diversions for irrigation of any magnitude alongthis part of the Penasco. Additions from tributariesoccur only during periods of very heavy precipita­tion , and it is unlikely that any of the flow at thelower station represents such additions.

IIIIIIIII

per month was lost. Intake capacity of the SanAndres here is classified as high to very high. Theeffect of high capacity is probably increased by theface that here the river flows in part on alluvium,some of the water seeping first into the alluvium andthen into the bedrock below.

High seepage losses on the Hondo were alsoshown in the investigations by W. W. Follett in1913-14 (Fiedler and Nye, 1933, p. 240) and in theBooito Hydrographic Survey by the State Engineerin 1930-31 (~lcClure, 1939h, p. 110-111). TheBonito survey showed a loss of 39 second-feetbetween Border Ranch and the intake canal to thereservoir when the flow at the ranch was 70 second­feet. This amounts to about 2,320 acre-feet permonth, which is about 400 acre-feet greater thanthe seepage shown during the Hondo survey in 1908and 1909. The difference is undoubtedly due to

measurement during the heavy flow of 70 second­feet, for it has been shown that seepage lossesincrease with increased discharge. The Bonitosurvey also found that a loss of about 20 second­feet occurs in the first 4 miles below the HondoReservoir. Follett's measurements show a stillgreater loss of 45.3 second~feet or 2,715 acre-feetper month between the l\loncano gage, which is be­tween Picacho and Riverside, and the reservoir.

Location

2,000 feet below road to present ClementsRanch, sec. 6, T. 17 S., R. 18 E.

200 yards above Y-O Crossing, sec. 1,T. 17 S., R. 19 Eo

Acre-Feet

2,770

2,558

IIILII

I

Discharge at Four Gaging Stations on the I?io Hondo in 1908-09

Picacho Border Ranch Diamond A Ranch Hondo Reservoir'"

t\lonth (WY; sec. 15, (S\V~4 sec. 30, (NE~/.i sec. 20, (Sec. 26.T. ] 1 S., T. 11 5., T. 11 S., T. 11 5.,R. ]8 E.) 1{. 20 E.) R. 21 E.) R. 22 E.)

ACTf'- r;eet Acre- Feet Acre-Feet Acre-Feet1908

l\!ay 16-31 .. 2,073 ],410 1,ll9 460J line. · .. 1,146 795 ]3 0July ... · . 4,908 3.537 2. ]88 1,116AUi<ust · . S,() 16 3.744 2,834 1,995September 3,082 2,542 2,OB4 873October .. 2,4% 1,578 648 ll2November .. 2,263 1,895 1,603 153December · . 2,546 2,377 1,740 380

1909

J anuMy . .. 2, 3~) 2,202 1,}1(! 37February 1,628 1,57B 653 0March 817 617 143 0April. ],035 439 13 °May · . 092 343 0 0June. · . 97/j 273 100 uJuly · . · . 2,745 1.404 917 511

"'Total of discharge .le g,lge on in] et canal from Rio Hondo Reservoir and discharge of gage belowheali of inlet canal to llondo Eeservoir.

23

Border Hills. The intake capacity of both segmentsis classified here as low to moderate.

Renick (1926, p. 135) lists stream-flow measure­ments made at several points along the Rio Penascoby Teeter in 1924 I(before the irrigation seasonstarted." 1feasurements in and near the nprincipalintake area" are given in second~feet as follows:Laramore Ranch, 36.7 second-feet; Y-0 Crossing,29.2 second-feet; Hope Diversion Dam, 24.0 second~

feet. In order to compare these with data in thetables above, the writer has converted them intodischarge in acre-feet per month, as shown in thefollowing table.

Discharge along the Uio Penasco Before IrrigationSeason in 1924, in Acre-Peet per Month

Oi stance Along

Station LocationStream from DischargeLast Station

(miles) (acre-feetJ

Laramore Ranch Sec. 26, T. 16 S.,R. 17 E. - 2,184

1'-0 Crossing Sec. 1, T. 17 S.,1\. 19 E. 19 1,738

Hope Diversion Sec. 14, T. 17 S.,Dom R. 20 E. 7 1,428

These figures show a loss of 446 acre-feet permonth in the 19 miles above the Y-0 Crossing and310 acre~feet in the 7 miles below it. These lossesare somewhat higher than might be expected whencompared with other data. However, the greaterloss above the Y-0 Crossing may be due to con­siderable seepage where the river crosses thesyncline east of the Bluewater anticline, which hereis highly fractured. Between the Y-O Crossing andthe Hope Diversion Dam the intake capacity hasbeen classed as moderate, and the loss on thePenasco is notably less than that along an equalsegment of the Rio Hondo having high-intake capac­ity, between Border Ranch and the former DiamondA Ranch.

AMOUNT OF RECHARGE OF THE ARTESIAN AQUIFER

Records of seepage loss from the Rio Hondo andRio Penasco make it possible to estimate theamount of recharge to the artesian aquifer con­tributed by each of these streams. Each gainswater in its upper reaches, and the first continuousloss from the Hondo occurs between Picacho andBorder Ranch. The river loses, in round numbers,120 acre-feet per month in this segment and 820acre-feet per month between Border Ranch and theformer Diamond A Ranch, making the total lossabove Diamond A Ranch about 940 acre-feet permonth, or 11,280 acre-feet per year. The percent­age of this loss due ro evaporation can be estimated

24

from the approximate area of the surface of theri ver and the rate of evaporation from a small sur­face in the general area, which is about 8 feet peryear (National Resources Planning Board, 1942a,p. 19-21). The yearly evaporation loss thus com­puted is about 400 acre~feet, which makes theground-water increment above Diamond A Ranchabout 10,880 acre-feet per year. Below the DiamondA Ranch, s.e:veral complicating factors make anyestimate considerably rougher. All the flow of theriver usually disappears before the eastern limit ofthe intake area is reached, but a small amountof this loss represents diversion for irrigation, andat times of heavy discharge large amounts of waterflow beyond the intake area. In addition, someinfluent seepage probably moves eastward asperched warer and never reaches the artesian aqui­fer. In order to estimate loss of flow in a recentyear of average precipitation, records of flow at theDiamond A Ranch during 1944, a year of approxi­mately average precipitarion on the drainage basinof the Hondo, can be examined as shown in thetable below (D. S. Geol. Sutvey, 1946, p. 282;1947b, p.285).

It will be noted that in no month did the dischargeexceed the 1,097 acre~feet per month estimated,from records of the 1908-09 survey, to disappearabove the Hondo Reservoir, plus the 20 second-feet,or 1,190 acre-feet per month, estimated, by the1930-31 survey, to disappear in the first 4 milesbelow it.

Discharge at Diamond A Gaging Station, 1944

Month Acre-Feet Month Acre-Feet

January 1,760 July 930February 1,480 August 1.360March 904 September 1,240April 284 October 696}.jay 75 November 1,220June 17 December 1,390

Total for year 11,356 acre-feet

From a study of these figures, plus the dailyrecords of discharge in second-feet, it is estimatedthat about 75 percent of the discharge at theDiamond A Station, or about 8,517 acre-feet, enteredthe artesian reservoir in 1944 below that station.The total increment from seepage from the Hondobelow Picacho was then approximately 19,400 acre­feet, which can be taken as the contribution of theHondo to the artesian reservoir during a year ofnormal precipitation. This amount is 8.3 percent ofthe probable normal recharge of the artesian basin,which is 235,000 acre-feet, as estimated by Fiedlerand Nye 0933, p. 252).

IIItII

I.'..~

IJJJ

[I

IIIIIII11II

Sherman \icll. S\\.!4N\V~4 sec. 9, T. 12 S., H. 23 E.

25

Brown v/ell. NE:'4SE~~j sec. 27, T. 11 S., R. 23 Ie.:.

*'.'.inJmill pumping.

"'~Sonle work done on well since last r""lding.

212.39212.64212.84213.25213.45

\iar. 3 *211.3~ \!ny 21

\Iar. 10 21l..J6 June 2'.Jar. 25 211. 57 June 16Apr. II 211. 62 June 30Apr. 21 211. 76 July 14\lay 5

Jan. 22 256041 l>.:ay 14 **257.49;"lar. II 255.98 '\lay 19 257.86Mar. 14 256.18 June 2 257.82l',idr. 25 256.81 June 16 258.37Apr. Il 257.45 June 30 258.88Apr. 22 257.37 July 14 259.42~lay 6 257.92 Sept. 9 260.82

I\Jar. 3 Iti4.75 t-.lay 19 148.37:\1ar. 12 * 145.25 June 2 148.28'\!ar . 24 ... 146JJ4 June 16 149.49Apr. 10 147.95 June 30 150.17!>.~ay 5 148.91 July 14 151.22

Water WaterDate Level Date Level

Feb. 26 310.87 May 19 312.661\lar. II 310.98 June 2 312.63~,jdr. 25 311.58 June 16 '313.41Apr. 11 312.20 June 30 313.61Apr. 22 312.45 July 14 314.05May 6 312.82

may be due to changes in atmospheric pressure.The Sherman Well and Bloom Well No.3 seem toshow such variations.

The effect of rainfall is shown in the periodbetween ifay 5-6 and J\Jay 19-21, when levels insix of the wells rose. Further rains between ~1ay

21 and June 2 caused the levels in three of thesewells to continue to rise, and the fall in the otherthree was not great. The variable effect of rainfallindicates that it benefits certain portions of theaquifer more quickly than others, probably becausethey are nearer the underground points where down­ward-percolating water reaches the water table.

l\leasuremenlS 0/ Ij/ater Level(In feet below land-surface datum, 1947)

\Voods \Vell No. 2. SW~4N\\'~ sec. 2, T. 11 5., R. 22 E.

Woods Well No. 1. N\Vl4S\V~4 sec. 1, T. 11 S., R. 22 E.

(i>.leasurements made in this well be,ginning in 1945are published in the annual water-levelreports of the U. S. Geological Survey.)

VARIATIONS IN WATER LEVEL

\Vater levels in nine wells, all in the intake areaand within 5 miles of the Hondo Reservoir, weremeasured every 2 weeks from late February to themiddle of July, 1947, in connection with the presentstudy. All were equipped with windmills and a few\vith supplememary gasoline engines. i\lost wereused mainly for stock, but a few for domesticsupply. A record of the measurements is given inthe table below. During the period of measurement,the warer level showed a continuous drop, with afew exceptions. This decline during the summer isnormal and is due to the spread of tbe regional coneof depression which results from use of tvater forirri,gation from the artesian wells to the easc. Therate of fall at the beginning of rhe period of meas­urements was generally much less than at rhe end;for example, Woods Well No. 2 dropped 0.11 footbetween February 26 and :-"Iarch 11, and 0.58 footbetween June 30 and Juiy 14. Levels in many ofthe wells dropped quite regularly except whenaffected by precipitation and overflow of the RioHondo, but in other wells there arc variations that

In veurs of normal precIpItation the Rio Penascoflows' continuously as far as the Hope DiversionDam, where all except flood flows are divertedduring the irrigation season. Channel losses abovethe diversion dam can be roughly estimated from therecords made by Teeter in 1924, although thesefigures show larger losses per mile than records fora shorter segment of the river in the State Engl'neer's report for 1926 and 1927. (See p. 23.)Teeter's records, when converted to acre-feet permonth, show a loss of 756 acre-feet between theLaramore Ranch, where the Penasco first begins tolose water continually, and the Hope DiversionDam. This amounts to 9,077 acre-feet a year, ofwhich about 400 acre-feet is probably evaporationloss~ Thus, on the basis of figures which may be alittle large, but ignoring possible losses from floodand winter flows below the Hope Diversion Dam, thecontribution of the Rio Penasco to the artesianreservoir can be roughly estimated at about 8,700acre-feet per year, or abom 3.7 percent of the totalrecharge of the artesian reservoir.

No discharge measurements are available for theRio Felix, Salt Creek, or any of the other screamscrossing the intake area. Such measurements wouldbe valuable, but it should be noted that no otherstreams contain perennial flow as far east as do thePenasco and the Hondo, and rhus no others con­tribute as much seepage, with the possible excep­tion of Salt Creek and its tributaries. Nevertheless,the contribution by influent seepage of streamsorher than tbe Hondo and the Penasco when takenall together undoubtedly is significantly large.

i ;........ d

••I.

••

Measurements 01 Water Level-Continued

Bloom Well No. L NWy';SWY.; sec. 5, T. 12 S., R. 23 E.

(Measurements made in this well beginning in 1941are published in the annual water-level reports of

the U. S. Geological Survey, where it is calledthe "J. Herbst WelL")

Jan. 22 *236.47 May 19 237.40Feb. 28 *236.06 June 2 237.51Mar. 14 235.63 June 16 237.85Mar. 24 *236.57 June 30 238.16Apr. 10 236.51 July 14 238.60Apr. 21 *236.71 Sept. 9 240.09May 5 237.26

Blnom Well No.2. NEY.;NEY.\ sec. 6, T. 125., R. 23 E.

Feb. 25 246.62 May 19 248.51Mar. 10 246.82 June 2 248.59Mar. 24 247.08 June 16 ~248.81

Apr. 10 *246.85 June 30 '249.20Apr. 21 247.94 July 14 249.15May 5 248.36

Bloom Well No.3. NEY4SWX? sec. 12, T.12 S., R.22 E.

Mar. 10 291.04 May 21 292.40Mar. 25 291.66 June 2 292.56Apr. 10 291.46 June 16 293.42Apr. 21 291.77 June 30 293.02May 5 **293.70 July 14 293.98

Hondo WelL SWY.;SWy'\ sec. 25, T. 11 S., R. 22 E.

Mar. 14 316.48 May 19 318.22Mar. 24 316.97 June 2 *318.31Apr. 10 *318.31 June 16 318.77Apr. 21 317.94 June 30 319.17May 5 318.47 July 14 319.60

Sixmile Hill WelL NEY.;NEY.i sec. 8, T. 11 5., R. 23 E.

Mar. 3 262.87 May 19 264.82Mar. 11 263.49 June 2 264.83Mar. 25 263.52 June 16 265.52Apr. 11 264.50 June 30 265.94Apr. 22 264.77 July 14 266.48May 6 265.10

'Windmill pumping."Some work done on well since last reading.

The Rio Hondo was out of its banks over a con­siderable area southeast of the Hondo Reservoir onJune 24, but the water levels in the wells that seemmost likely to be affected were not as high on June30 as was expected. All declined during the periodof the flood. The Brown Well shows the greatest

26

effect of recharge from the flood, as it dropped only0.68 foot berween June 16 and June 30 as against1.21 feet during the 2-week period before and LOSfeet the period after. From its location one wouldexpect the Sherman Well to show an effect from theflood, but the water level declined 0.41 foot duringthe flood period and only 0.20 foot each during theperiod before and the period after. High atmos·pheric pressure on June 30 may have caused thelarge decline for the period ending on that day.

PERCHED WATER

Perched water bodies exist in many places in theintake area, particularly in its western part. Two,which are probably fed by influent seepage ftom theRio Penasco, were struck in the Taylor Well aquarter of a mile south of the river at the crossingof Highway 24. The upper aquifer, 148 feet from thesurface, went dry aftet 2 years of domestic use,and the well was deepened to the second at 255feet. The depth to the main water table here isabout 600 feet.

THE ARTESIAN AQUIFER

GENERAL OCCURRENCE

Water in the artesian reservoir exists under pres­sure in cavernous zones in the San Andres formationand· the lower part of the Chalk Bluff formation.These zones are extremely limited in extent bothhorizontally and vertically; hence it cannot be pre­dicted at what elevation a well will strike waterat any given location. The cavernous zones atenumerous, however, and the writer knows of no wellthat failed to strike at least one "water rock" in anarea of known artesian production.

North of Lake Arthur, all artesian productioncomes from the San Andres, except for minoramounts from aquifers in the overlying Chalk Bluffformation. The aquifers in the San Andres occuralmost entirely in the upper half of the formation.South of Lake Arthur, Goat Seep ("Dog Canyon")limestone begins to interfinger with the correspond­ing membets of the Chalk Bluff formation, beginningwith its base and increasing in thickness to thesouth. Artesian production comes from the GoatSeep in an increasing amount southward, althoughwater continues to occur in the San Andres.

F Or a complete description of the artesian aqui­fer, the reader is referred to reports by Fiedler andNye (1933, p. 140-156), and Morgan (1941).

HYDROLOGIC ANOMALIES ALONG COTTONWOOD CREEK

Hydrologic conditions along Cartonwood Creekare unusual in that the area of artesian flow thereextends farther west than in any other part of theRoswell Basin, and a marked decrease in head

I

•I

•••••

I.11IIIIIIIII

II

LII

occurs from west to east. Areas of large yield byartesian flow occur along all major .?rainage lines,Cottonwood Greek being included even though itspresent drainage area is not a (lmajor" one. Nye(Fiedler and Nye, 1933, p. 183) shows that largevie Ids of this kind probably are dl.le to very active~irculation of ground water beneath the drainages.The writer believes that the high head in the upperpart of the Cottonwood drainage is probably duesi;mply to an underground connection with a part ofthe intake area where the water level is at suffi­ciem height to cause such a head. The decreasein head to the east, as mapped by Fiedler and Nye0933, pI. 30), occurs along one piezometric breakcunning approximately north-somh through sec. 5,T. 16 S., R. 25 E., and a second break and steepgradient about a mile wide running about north­south through section 2 of the same townshipand range.

A studv of the piezometric surface in the Cotton­wood are"a was made during the present investiga­tion, using rhe figures of Fiedler and Nye (1933,p. 325-332) on the head of wells, either whendrilled or as measured by them in 1926. The headof the wells was computed as of 1911, using thefigures given by Fiedler and Nye (1933, p. 203,204) for the yearly decrease in head for the areainvolved, and adding or subtracting as necessary.This method was believed to be of particular value,as information from all the wells in the area couldbe used instead of only the few whose heads weremeasured in 1926. It was found that the sharpreductions in head eastward could be explained bysteep gradients in the piezometric surface as wellas by sharp breaks, and it is believed that steepgradients are the more accurate representation. Thesteepest gradient on the upper Cottonwood wasfound to lie in the eastern part of sec. 1, T. 16 S.,R. 24 E., and the western parr of sec. 6, T. 16 s.,R. 25 E. This probably corresponds to the westernpiezometric break of Fiedler and Nye, though itscenter lies about a mile and a half west of theirbreak. The artesian head here, computed as of1911, drops about 24 feet in a mile. A much greaterdrop occurs in the steep gradient farther east, cen­tering in sec. 2, T. 165., R. 25 E., where Fiedlerand Nye showed their eastern break and steepgradient. The drop from the average head of sec.3. T. 165., R. 25 E., to the average head of sec. 6,T. 16 S.', R. 26 E., a distance of 3 miles, is about11 0 feet.

The writer be lieves that the sharp reductions inhead in the Cottonwood Creek area are due to heavyunderground leakage from the artesian aquifer incathe sballow aquifer in the zones \vhere the piezo­metric gradients are very steep. This tbeory

adequately explains the sbarp gradients and thelow head east of them, and it is supported byadditional evidence, chief of which is the highwater table in the shallow aquifer along CottonwoodCreek. The high elevation of the water table isshown best by the fact that the creek bed intersectsthe water table, and flow from ground-water seepagebegins farther west than in any other stream channelin the Roswell Basin. Flow begins, in fact, in thevicinity of the point near the eastern boundary ofR. 24 E. where heavy leakage into the shallowaquifer is believed to occur. Flow in Walnut Draw,which approximately parallels Cottonwood Creek2 or 3 miles to the north, also begins farther westthan in most streams, its beginning point beingapproximately straight north of the eastern zone ofheavy leakage ioro the shallow aquifer. Furtherevidence of strong shallow-water recharge is therelatively high elevation of the water table in theupper Cottonwood, as mapped by Morgan (1939,pi. 2), who shows thar the slope 0-[ the water tablefrom the eastern boundary of R. 24 E. to the PecosRiver is notably greater than the slope in an equaldistance anywhere else in the shallow-water basin.

EFFECT OF USE OF THE HONDO RESERVOIR

PROTECTION OF ROSWELL

The most obvious beneficial effect of use of theHondo Reservoir for flood control would be the pro­tection of Roswell from the disastrous floods whichhave periodically hit the city. The flood of Sep"tember 1941, for example, caused damages totaling$ 558,644 in Roswell and vicinity. The annualdamages from floods here have averaged $ 55,888(National Resources Planning Board, 1942b, p. 152),and these damages would be eliminated with theutilization of the reservoir for flood control.

RECHARGE OF ARTESIAN AQUIFER

Nearly all the flood waters impounded in theHondo Reservoir when it was in use escaped throughsinkholes and solution channe Is in its floor into theSan Andres formation beneath. According to theofficial Reclamation Service report,ll following aflood of 2,000 acre-feet the water level in thereservoir was so low thar after !ltwo or three days"no water could be drawn out. Of 27,860 acre-feetdiverted into the reservoir between 1908 and 1913,only 1,100 acre-feet was utilized for irrigation,which was the intended purpose of the project.

Use of the I-Iondo Reservoir for flood control thuswould result in nearly all the impounded water

8U. S. Reclnmation Service, n.d., ProJect Hislory from [nce!)­Uon 0/ Project, Feb. 24, 1904, to Dec. 31, 1915, /londo Project,N. ,\lex.: typewritten rept. in files of U. S. Bur. of Recl'llll. inAlbuquerque, N. Mex., v. 2, p. 11'5.

27

entering the San Andres to replenish the artesian­water supply. A small amount of warer would belost by evaporation, and a somewhat larger propor­tion would not percolate into the limestone butwould seep out of the reservoir in alluvium andterrace material, in part remaining in the uncon­solidated material to form a perched water body onthe surface of the bedrock. That this would occuris known from reports that very shallow wells interrace material east of the reservoir producedwater when the reservoir was in operation. Theamount of evaporation, of course, would depend onthe length of time water remained in the reservoirbefore going into underground storage, For largefloods, this time can only be estimated, as themaximum intake for anyone yeat during operationof the reservoir was onlv 11,600 acre-feet, ( Seetable, p.4.) If 40,000 acre-feet of water shouldenter the reservoir and should take 30 days to seepout during one of the summer months, about 920acre-feet, or 2.3 percent, would be lost by evapora­tion, using 0.8 foot of evaporation per month as abasis for calculation (National Resources PlanningBoard, 1942a, p. 20). Diversion of as large anamount as 40,000 acre-feet into the re servoir wouldbe very exceptional, as control of the heaviest floodon record, that of September 1941, would have re­quired only 47,900 acre-feet of storage. Evapora­tion from smaUer amounts of storage would be lessbecause the surface area would be les s and thewater would go inco underground storage in ashorter time.

It is believed that the percentage of water fromthe reservoir that would go into storage in terracedeposits would be small, as all available accountsof leakage describe the water as running Out throughopenings in lime stooe aod gypsum in the floor andthus entering the San Andres formatioo, It is there­fore estimated that less than 10 percent of thewater entering the reservoir would escape ioto aodremain in unconsolidated material. This perchedwater would move slowly east toward the PecosRi ver. It would move over impervious beds of theChalk Bluff formation which occur a few miles fromthe reservoir, and it would continue eastward toaugment the supply of shallow water in the artesianarea and eventually, if not pUl11ped froll1 wells, to

discharge inco the Pecos Ri ver.

As evaporation from the Hondo Reservoir wouldamount to less than 2 percent of the \vater enteringit under normal conditions, 98 percent of the wate,rwould enter the two underground aquifers, theshallow and the artesian. The writer estimatesthat about 8 percent of the original amount wouldenter the shallow aquifer, and 90 percenc the arte­sian aquifer.

28

If the reservoir should be used for flood control,the tate at which water would be lost by seepagewould naturally decrease as more and more siltentered the reservoir. This effect probably wouldbe small at first, for after eight previous years ofuse, during which repeated efforts were made to

plug the leaks, F oster9 reported that "leakagethrough the reservoir bottom was apparently as badin the season of 1915 as it was in 1907-08."Nevertheless, in the course of many years, leakagemight be greatly reduced by natural silting of theoutlets.

If the reservoir should be used as a catchmentbasin through which water from the Pecos Riverwould be directed into underground storage, siltingwould be negligible, as the water of the Pecoscontains less silt than the I-lando, and the reservoirprobably could be used for this purpose indefinitely,

RISE OF ARTESIAN HEAD

An estimate of the amount the artesian headwould be raised by introduction of specified amountsof water inco the Hondo Reservoir can be made asfollows: the method is to determine as accuratelyas possible the amount the he ad has been raised orlowered in the past after years of excessive or de­ficient precipitation and, by estimating the depar­ture from normal recharge during those ye ars, tocompute the number of acre-feet of water requiredto bring about each foot of rise or fall in head.Partial records from three recorders on artesianwells in the Roswell Basin are available for theperiod 1926 to 1940 (National Resources PlanningBoard, 1942a, p, 46-47), and records of six arte­sian wells and five or six wells near the easternedge of the intake area are available for the period1941 to 1946 (U. S. Geo!. Survey, 1943, 1944, 1945,1947a, 1949), Records of annual precipitation inthe Pecos River Basin up to 1939 are inc luded inthe Pecos River Joint Investigation (NationalResources Planning Goard, 1942a, r. 6-14); lacerre cotds are available in repofts of the U. S. WeatherBureau.

The only trustworthy correlations of prec ipitationand change in head are those made for the years1941-46, for which records of intake-area wells areavailable. Inasmuch as a rise or decline of thewater level near the edge of the intake area shouldresult in nearly an equivalent fluctuation of theartesian head in the valley, these records havebeen used primarily for the correlations becauseheavy pumping from the artesian aquifer does notaffect them as strongly as it does the water levelsin the artesian weils, It is as sumed hete that the

9[bid, p. 82.

IIIIIIIIII

I

recharge of the artesian aquifer after a year ofnormal precipitation is 235,000 acre-feet, as esti~

mated by Fiedler (Fiedler and Nye, 1933, p. 252).It is also assumed that the deviation from the nor­mal in recharge is directly proportional to thedeviation from the normal of the effective precipita­tion, effective precipitation being determined byweighting the records so that precipitation on theeastern part of the total intake area is given twicethe effect of precipitation farther west and north.It is further assumed that the chan,ge in water levelfrom one] anuary to the next is caused entirely bythe excess or deficiency of recharge during th~

intervening calendar year. Errors are of courseintroduced by these assumptions, but it is believedthat the errors lie within the limits of accuracy ofthe estimates made which, because of the greatnumber of complicating factors involved, must ofnecessity be rather rough.

The table below shows the effective precipitationin the years from 1941 to 1946, in percent of normal,tabulated with the average change in level by thefollowing January, of the artesian and the intake­area wells. In the last column is shown the numberof acre~feet of recharge above or below normal thathas been computed to cause a rise or fall of 1 footof head, on the basis of the intake-area records.This quantity is not given for 1944, as the precipi~

tation during that year was practically normal. Thefigures for 1945 suggest that the decline in head foryears of greatly deficient precipitation may not beas great as would be expected. The rise in headof the artesian wells between 1946 and 1947 wasdue to rains late in 1946, which greatly reduced the

lrngation draft on the reservoir and thus allowedmore complete recovery of the artesian head. Arough check was made on the figures in the lastcolumn of the table by comparing the departure fromnormal recharge with the change in head of artesianwells for the years between 1926 and 1940. It wasfound that the computed variation in the amount ofrecharge per foot of change is much greater thanthat shown in the table because of the much greaterinfluence of pumping, but that the average is aboutthe same.

It is, therefore, estimated that the artesian headrises about 1 foot for each 30,000 acre-feet ofadditional recharge. The rise is proportional to theamount of recharge so long as the effective porosityof the aquifer is constant through the interval ofrise or fall, and such is believed to be essentiallythe case.

This, of course, represents the rise that occursbecause of increased recharge throughout the intakearea. It is to be expected that rapid addition ofwater in a comparatively small area such as theHondo Reservoir would result in a much more rapidand greater local rise than would be indicated bythis average. The amount of rise and the durationof the higher stage of water level would depend uponthe rapidity of leakage of water from the reservoir,the transmissibility and porosity of the aquifer, andthe effect of the large springs at Roswell in drain­ing off the excess watet.

All these factors are unknown. The hydrauliccharacteristics of the limestone aquifer do not lendthemselves to determination by the methods used

LII

Ue/ation oj Excess or Deficiency oj Recharge to Change in Waler Level oflnlake~i\rea Wells and Artesian Wells, 1941-46

Average Change in Level of Wellsfrom January of Year Considered Amount of l\eclurge

Effective to Following January Excess or Causing ChJ.nge ofYear (+ denotes rise; -denotes fnll) Deficiency 1 Foot in Level

Precipitation of Recharge of Intake-AreaIntake-Area Artesian Wells

\<.'ells \\'ells(flereen! of normal) (feet! (feet) (acre-feel) (acre-jeet)

PH 1 205 +10.67 +14.5 +247,000 23,0001942 114 + 1.93 - 1.0 + 33,000 17,00019,j J 7~ - 1,35 - 1.7 - 50,000 37,0001944 ;'7 ,+ .48 - ,6 - -1945 5(, ,- 1.81 - 4,3 -103,000 57,0001946 N6 ,,- .93 + J.3 - 33,000 35,000

'"Includes measurement of one well in !\larch 1945. No January reading avaihtble.

'"'"Includes measurement of one well in November 1946. January 1947 reading shows abnormal fal! of 3.0 feer from

November reading.

29

for sandy aquifers. Preliminary work indicates thatbecause of the very high transmissibility of theaquifer the rise in head in the artesian basin willprobably be less than 10 feet at any time in theevent that 50,000 acre-feet of water should bestored in the reservoir. The effects of leakage ..fromHondo Reservoir on the artesian aquifer are con~

sidered in a supplementary report by CharlesV. Theis.

EFFECT ON THE PECOS RIVER

The opinion is held by some that all water im~

pounded in the Hondo Reservoir and escaping fromit by seepage would be forever lost to water usersdownstream on the Pecos River. Such is not neces­sarily the case. Addition of water to the artesianaquifer causes the head to rise, as described above,

30

and increased pressure on the confining beds Causesan increased amount of leakage upward into theshallow aquifer. Water is discharged continuallyfrom the shallow aquifer into the Pecos River, andan increase in amount of shallow water would causean increase in discharge into the Pecos. Use ofthe reservoir would further increase the amOunt ofshallow water by the small proportion which seepsout of the reservoir and moves eastward in terracematerial. Thus, if the amount of water pumped fromthe shallow and artesian aquifers were not increasedbecause of-the availability of additional water con~

tributed to the aquifers from the Hondo Reservoir,practically all such additional water would eventu~

ally be discharged into the Pecos River. Anyadditional pumping, however, would result in reduc­tion of this discharge.

IIIIII

III..IIIII

'IIIIIIIIII

•••••III

••IIIIIIIII

REFERENCES CITED

Fiedler, A. G., and Nye, S. S., 1933, Geology and Ground·Water Resources of the Roswell Artesian Basin,N. Alex.: U. S. Geol. Survey Water-Supply Paper 639.

Fisher, C. A., 1906, Preliminary Report on the Geologyand Underground Waters of the Roswell ArtesianBasin: U. S. Geo!. Survey Water-Supply Paper 158.

King, P. B., 1942, Pennian of Western Texas and South·eastern New Afexico: Am. Assoc. Petroleum Geolo­gists Bull., vol. 26, no. 4, p. 535-763.

McClure, T. M" 1939a, Abridged Report on Lower RioHondo Flood Control Investigall'on: New MexicoState Engineer 14th and 15th Bienn. Reprs., p.61-105.

1939b, Effect oj Lower Rio Hondo----=pc:Z-o-o-:d-C::o-n-'-r-o! Project on Downstream Flow: New

Mexico State Engineer 14th and 15th Bienn. Repts.,p. 107-114.

Merritt, J. W., 1920, Structures of Western Chaves County,N. Mex.: Am. Assoc. Petroleum Geologists Bull.,vol. 4, p. 5.5-57.

Morgan, A. M., 1939, Geology and Shallow-Water Re*sources of the Roswell Artesian Basin, N. Alex.:New Mexico State Engineer 12th and 13th Bienn.Rep,s., p. 115-249.

1941, Depth of Active Solution by-.,,---;-;;;--,---Ground Waters in the Pecos Valley, N. Alex.: Am.Geophys. Union Trans., p. 779-783.

t-.Iorgan, A. M., and ~ayre, A. N., 1942, n Geology," inNational Resources Planning Board, The Pecos RiverJoint Investigation, Reports of the ParticipatingAgencies.

National Resources Planning Board, 1942a, The PecosRiver Joint Investigation, Reports of the Participat­ing Agencies.

1942b, Regional----;P"Z-a-n-n7in-g-P,;-a-r',--,X.,----oY"h;-e-P;,--e-,c-,-o-,-s"R""iver J0 in tInvest iga-

tion, Summary, Analyses, and Findings.

New Mexico State Engineer, 1925, Surface Water Supplyof New Mexico, 1888-1925.

__-;-:;----:::--:-_-;--:-;-__1928, Surface Water Supp 1yof New Mexico for the years 1926 and 1927.

Renick, B. C., 1926, Geology and Ground-Water Re­

sources of the Draznagf' Basin of the Rio Penascoabove Hope, N. Alex.: New 1iexico State Engineer7th Bienn. Rept., p. 123-127.

Robinson, T. W., and Lang, W. B., 1939, Geology andGround-Water Conditions of the Pecos Valley in theVicinity of Laguna Grande de la Sal, N. Alex.: Newt-.Iexico State Engineer 12th and 13th Bienn. Repts.,p. 77-100.

Theis, C. V., 1939, Origin of Water in Major JohnsonSprings, near Carlsbad, N. Alex.: New Mexico StateEngineer 12th and 13th Bienn. Repts., p. 251-252.

Theis, C. V., Morgan, A. M., Sayre, A. N., Hale, W. E.,and Loeltz, O. J., 1942, II Geology and Ground Water,"in National Resources Planning Board, The PecosRiver Joint Investl'gation, Reports of the Participat­ing Agencies, p. 27-75.

Theis, C. V., Morgan, A. M., Hale, W. E., and Loeltz,O. J., 1942, H Ground~Water Hydrology of Areas in thePecos Valley, N. l'Ilex.," in National Resources Plan­ning Board, The Pecos River Joint Investigation,Reports of the Participating Agencies, p.41-51.

U. S. Geo!. Survey, 1946, Surface Water Supply of theUnited States, 1944, pt. 8: U. S. Geol. Survey Water~

Supply Paper 1008.

1943, Water Levels and Artesian--'P;-r-e-s-s-u-re-Ci"-n-O""b-'s-:Cervation Wells in the United Slates in

1941, pt.6: U.S. Geol. Survey Water-Supply Paper 941.

1944, Water Levels and Artesian----;:P,.-r-e-ss-ur-e---:in-CO;;-;-b-se-rvation Wells in the United States in

1942, pt. 6: U. S. Geo1. Survey Water-Supply Paper949.

_---=:'--__-:-_;:::-__1945, Water Levels and ArtesianPressure in Observation Wells in the United States in1943, pt. 6: U. S. Geol. Survey Water-Supply Paper991.

1947, Water Levels and Artesian----=P=-r-e-s-s-u-re-i:-n---;:O"b-s-e-rvation Wells in th e Uni ted State s in

1944, pt. 6: U. S. Geol. Survey WaterMSupply Paper1021.

1947, Surface Water Supply of the----:U"'"n-',.-'-e-:d,.-S::'-:-a-:-'-e-s.---:l"9'"'45, pt. 8: U. S. Geol. Survey Water­

Supply Paper 1038.

1949, Water Levels and Artesian----:P=-r-e-s-s-u-re-i,.-n-O=b-s-e-rvation Wells in the United States in

1945, pt. 6: U. S. Geol. Sutvey Water-Supply Paper1028.

31

EFFECT ON ARTESIAN AQUIFER OF STORAGE

OF FLOOD WATER IN HONDO RESERVOIR

ISTORAGE

RESERVOIR

EFFECT ON ARTESIAN AQUIFER OF

OF FLOOD WATER IN HONDOIII

ByCharles V. Theis

IIIIIIII

Prediction of the effect of storage of flood watersin Hondo Reservoir upon the artesian aquifer of theRoswell Basin is made difficult by several factors.The size of the maximum flood is unknown. Theeffect of continual use of the reservoir upon thecaverns and passageways permitting leakage isunknown. The permeability and porosity of thelimestone of the San Andres formation in the vicini­tV of the reservoir are indicated by the history ofthe reservoir to be large but are unknown quanti­tatively. Because of the factors that apparentlywill remain unknown until large quantities of waterare placed in the reservoir, it appears that only theorder of magnitude of the effect produced by im­pounding can be ascertained. It is believed, how­ever, thar the order of magnitude has considerablesignificance.

The largest flood of record in the Hondo River,that of September 1941, would have required astorage capacity of 47,900 acre-feet (NationalResources Planning Board , 1942, p. 159). In mostyears little or no water would enter the reservoir.As the flood of 1941 coincided with an annualprecipitation that far exceeded that in any otheryear of a 75-year period of record , it is probablethat the storage needed would exceed 50,000 acre­feet rarely, if ever. The following computations ofthe effect of a flood on the artesian aquifer arebased upon such a figure. It is probable that somuch water would never go underground in thereservoir - that in case of a large flood a consider­able part of the water would be re leased throughsurface channels after the flood peak had passed.The period during which the water would go under­ground is assumed for computation as 25 days.Although there is necessarily considerable conjec­ture as to the rate at which Water could seep under­ground in the reservoir, the old records of operationof the reservoir show (Bean, p. 4) that when morethan a few thousand acre-feet of water entered thereservoir it was possible to release some water at alater time through the gates. A rate of leakage

in excess of that assumed would notaffect the rise in artesian head in the irri­

area, if the total quantity of leakage remainedsame.

The factors that would control the flow of waterin the aquifer are the transmissibility and storagecoefficients of the aquifer. For the comparativelylarge openings in the artesian aquifer which wereformed by solution, the storage coefficient is equalto the porosity of the aquifer. Both these factorsare difficult to determine for the artesian aquifer,because the methods for determining the coefficientsby pumping tests that are applicable to uniformlyporous aquifers, such as sands or sandstones, can­not be used with an aquifer composed of cavernouslimestone.

The coefficient of transmissibility in the notthernpart of the artesian basin is shown, by the veryflat gradient of the piezometric surface in that area,to be very large. The water level in a well at theHondo Dam in the SW\iSW\i sec. 25, T. 11 S., R.22 E. was at an altitude of 3,578 feet in rvfarch1947, and that in a well at the Bloom Ranch in theNW14SW14 sec. 5, T. 12 S., R. 23 E. was at an alti­tude of about 3,575 feet. It appears that the waterlevels in these wells were somewhat higher in1927 , when a piezometric map of the aquifer wasprepared by Fiedler (Fiedler and Nye, 1933, pI. 30),for the depth to water in the Bloom Well (idem,

p. 358, well 1684) was given as 230 feet at thattime as against 236 feet as measured in l\Iarch 1947.The altitude of the water levels in these two wellstherefore probably can be taken as 3,583 and 3,580in 1927. Fiedler's piezometric map of 1927 showsthe 3,565-foot contour to lie 10 miles east of theserespective wells, indicating a gradient of about 1.5feet per mile. In the south end of the basin thegradient is at present apparently considerablyhigher but it is doubtful thar it was much higherbefore drilling of artesian wells and the consequentgreat drop in artesian head in tde south part of thebasin.

The quantity of water passing into the aquifer inthe northern part of the artesian basin is also difii­cult to determine. Fiedler and Nye (1933, p. 252)estimated the recharge of the artesian basin as235,000 acre-feet annually, of which about 210cubic feet per second or 155,000 acre-feet per )rearwas formerly discharged at the large springs near

33

Roswell. J\Jr. Bean concludes that the northern partof the basin presents a much better opportunity forrecharge than the southern part. The estimated re­charge of 235,000 acre-feet, divided by a length ofoutcrop of about 75 miles, represents an averagerecharge per mile of about 3,100 acre-feet. Con­sidering the greater opportunity for recharge in thenorthern section, it would appear that somethinglike 5,000 acre-feet per year per mile probablypercolates eastward in this section into the arte­sian area. This quantity is equivalent to 4.5 milliongallons a day per mile, and at a hydraulic gradientof 1.5 feet per mile it represents a transmissibilityof the aquifer of 3,000,000 gallons a day per foot.This figure for transmissibility is very high butapparently must be at least of the right order ofmagnitude.

An estimate of the porosity may be made fromBean's data. He estimates (p. 29) that the watertable rises 1 foot for each 30,000 acre-feet of re~

charge. This amount of water over the 1.200 squaremiles of ({principal" intake area (p. 20) indicatesa porosity of about 4 percent. It is believed againthat, because of the greater amount of gypsum inthe limestone of the northern part of the basin, theporosity here is somewhat larger than this averagefigure. The porosity here is assumed to be 5 per­cent for the purposes of computation, but it may besomewhat higher.

If it is assumed that water would enter the SanAndres formation at a rapid rate by leakage from theHondo Reservoir, a cone of water-table elevationwould be built up beneath the reservoir, which inthese permeable beds would spread rapidly. Waterwould be stored in the limestone as the water tablerose. The cone would presumably spread freely to

the north, south, and west, and to the east to theline of intersection of the water table with the over­lying Chalk Bluff formation which forms the confin­ing bed for the artesian aquifer. From this lineeastward the head of the water would rise butprobably no water would go into storage. The conepresumably would spread to this point fairly uni­formly in an aquifer having a coefficient of trans­missibility of about 3,000,000 gallons a day perfoot and a coefficient of storage of about 5 percent.After part of the pressure cone entered the confinedpart of the aquifer, the rise in pressure wouldproceed at a more rapid rate than it would in anentirely unconfined aquifer, if no other compensat­ing factors required consideration. However, withthe rise in pressure it is probable that the springsat Roswell would again start flowing as they didafter the heavy rains in 1941. The increased dis­charge would relieve the hydraulic pressure in parr.It seems probable, therefore, that in this area the

34

effects of increase of head because of a rapidleakage of water from Hondo Reservoir probablywould not exceed the head that would be caused bysuch a leakage into an aquifer without partialconfinement and \vithout nearby natural outlets.

The effects of a recharge of 2.000 acre-feet aday or 450

1000 gallons a minute into an aquifer

with the assumed characteristics can be computedaccording to methods previously published (Theis,1935).

The water in the reservoir probably would leakthrough many openings in the floor. For computa­tion it is assumed that the leakage would takeplace over an area of 1 square mile through nineholes - at the corners, the centers of the sides, andat the center of the square mile. Such an arrange­ment would simulate the unknown actual condition.Computations are given for distances of 2,500 feetand 25,000 feet from the nearest side of the squarearray of wells. It appears that the line of intersec­tion of the water table with the overlying confiningbed is at a distance of about 5 miles, or approxi­mately 25,000 feet, from the Hondo Reservoir. Itis believed that the rise in head at that distance\vould represent the maximum rise of water level inthe artesian basin. Computations are given in thefollowing table for an aquifer having what are be­lieved to be the most probable hydraulic charac­teristics: a transmissibility of 3,000,000 gallons aday per foot and a porosity of 5 percent; for one ofthe same transmissibility but a porosity of 10 per­centi and for twO aquifers having a transmissibilityof 1,000,000, one having a porosity of 5 percent andthe other, 10 percent. It is believed that theserepresent the greatest possible range of values forthe San Andres formation in the vicinity of theHondo Reservoir.

It appears, because of the exceptionally hightransmissibility of the limestone of the San Andresformation in the vicinity of the Hondo Reservoir,thac the effect of the addition of a large quantity ofwater to the aquifer at the reservoir would bequickly spread over a large area and that waterlevels would rise over a large area in the outcroparea of the San Andres formation to the extent of afew feet. For a short time, and in a small area inthe immediate vicinity of the reservoir, the waterlevel would rise perhaps 100 feet. In most of thisarea the depth to water exceeds 200 feet. Theincreased head would be large ly dissipated beforethe confined part of the aquifer was reached.

The increased head within the irrigated areaprobably would be so small as not to cause anyincrease in well discharge. If the flood occurredin the summer and as a result of rainfall that did

IIIIIIIJ

IIII

IIII

IIIIIII

Expected Increase in Head, in Feet, in Roswell Artesian Basin, Caused by Leakage 0/50,000 Acre-Feet ofWater lrom Hondo Reservoir in 25 Days. FOUT Combinations of Assumed Coeificients of Transmissibility

and Storage were Used in Order to CoveT Possible Range of These Coelficients

Coefficient of1,000,000Tran smissibility 3,000,000

(gpd/ft!

Coefficient of0.10Storage 0.05 0.10 0.05

Distance from 2,500 25,000 2,500 25,000 2,500 25,000 2,500 25,000Reservoir (feet)

Days Since Beginningof Leakage

25 54 6 42 2 108 2 77 1

50 12 7 11 4 33 7 31 2

75 7 5 7 4 22 8 19 3100 5 4 5 3 14 9 14 4200 2 2 2 2 7 5 7 4

300 2 2 2 1 4 4 4 3400 1 1 1 1 3 3 3 3

IIIII

not include the irrigated area, the head in the irri~

gated area would be low and the effect would be toraise water levels somewhat in the wells. Theeffects on the flow into the river probably would notbe felt until winter when, with the seasonal rise inwater level, the flow from the springs on the lowerBerrendo would increase. If the flood occurred inthe winter the effect would be to increase the flowfrom the natural outlets of the springs. If it oc­curred in conjunction with a widespread storm, andtherefore when irrigation from wells was suspended,the head in the irrigated area would rise rapidlybecause of both increased recharge and cessationof discharge from wells, and the large springs wouldstart flowing, as they did in 1941. If the PecosRiver were in flood, the effect of impounding theflood water of the Hondo in the reservoir (.in effectdi verting a part of the flood underground) would beto furnish additional regulation of the flood and topostpone part of the flood runoff.

It is possible that the effects of leakage from thereservoir and the consequent increase in head mightbe localized somewhat by more or less continuouscavernous passageways running from the reservoirsite. The effect of such passageways, insofar asthey exist, would be to cause a relatively largeincrease in artesian head in their vicinity and acorrespondingly smaller increase elsewhere. In­asmuch as such passageways were developed byground water tributary to the large springs at Ros­we 11, it is probable that they would tend to divertthe water rapidly from the reservoir site to thesprings, cause thereby a relatively rapid bleeding

off of the increased head, and on the whole dimin~ish the effect of the increased head from thatcomputed.

The increased artesian head probably wouldcause some increase of leakage from the artesianaquifer into the valley fill (shallow~water aquifer).Some of this water might be diverted by pumps nearthe western edge of the irrigated area and neverreach the river. However, as the amount of leakagethrough the confining bed is only a small part of thetotal discharge of the artesian aquifer and as theincreased head would be only a small part ofthe total head that causes the leakage, the propor~

tion of the water that leaked inco the alluviumprobably would be small. In the areas near theriver any increased leakage from the artesian aqui h

fer probably would find its way through the alluviumto the river through drains and by direct inflow.

The effect computed is for a flood on the Hondothat would be very rare. The much smaller amountsof water that would be diverted into the reservoirevery few years apparently would cause no sig­nificant increases in artesian head. It appears,therefore, that the diversion of water undergrounqthrough the Hondo Reservoir would not furnishencouragement for increased use of artesian waterfor irrigation. If the use of water from wells werenot increased, the effect of diversion of waterunderground would probably be beneficial to theflow of water in the Pecos River, as it would helpto regulate the flow and, at least in large measure.would avoid the loss of water by percolation intothe alluvium that now occurs when the Hondo inflood spreads widely over the alluvium.

35

REFERENCES CITED

Fiedler, A. G., and Nye, S. S., 1933, Geology and Groun.d~

Water Resources 01 the Roswell Artesian Basin, N.Alex.: U. S. Geo!. Survey Water~Supply Paper 639.

36

National Resources Planning Board, 1942, Regional Plan­ning Part X - The Pecos River joint Investigation,Summary. Analyses, and Findings.

Theis, C. V., 1935, The Relation between the Loweringof the Piezometric Surface and the Rate and Durat£onof Discharge 0/ a Well Using Ground·1Vater Storage:Am. Geophys. Union Trans., p. 519-524.