University of South Florida Scholar Commons School of Geosciences Faculty and Staff School of Geosciences Publications
8-1992 Evidence of Regional Structural Controls on Vent Distribution: Springerville Volcanic Field, Arizona Charles B. Connor Florida International University, [email protected]
Christopher Condit University of Massachusetts
Larry S. Crumpler Brown University
Jayne C. Aubele Brown University
Follow this and additional works at: https://scholarcommons.usf.edu/geo_facpub Part of the Earth Sciences Commons
Scholar Commons Citation Connor, Charles B.; Condit, Christopher; Crumpler, Larry S.; and Aubele, Jayne C., "Evidence of Regional Structural Controls on Vent Distribution: Springerville Volcanic Field, Arizona" (1992). School of Geosciences Faculty and Staff Publications. 1655. https://scholarcommons.usf.edu/geo_facpub/1655
This Article is brought to you for free and open access by the School of Geosciences at Scholar Commons. It has been accepted for inclusion in School of Geosciences Faculty and Staff ubP lications by an authorized administrator of Scholar Commons. For more information, please contact [email protected]. JOURNAL OF GEOPHYSICAL RESEARCH,VOL. 97, NO. B9, PAGES 12,349-12,359,AUGUST 10, 1992
Evidenceof RegionalStructural Controls on Vent Distribution: SpringervilleVolcanic Field, Arizona
CHARLESB. CONNOR
Departmentof Geology,Florida International University, Miami
CH]H•OP• D. CONDrr
Departmentof Geologyand Geography, University of Massachusetts,Amherst
LARRY S. CRUMPLERAND JAYNE C. AUBELE
Departmentof GeologicalSciences, Brown University, Providence, Rhode Island
Quantitativeanalysis of the geographicdistribution of ventsand comparison with regional structural, petrologle,and vent age data provide insight into the processes governing the emplacement of vents in theSpringerville volcanic field, Arizona. A totalof 409vents in theSpringerville volcanic field (SVF) havea meandistance to nearestneighbor vents of 955 m, a muchcloser spacing than is commonin some platform-typevolcanic fields. Basedupon a clusteranalysis search radius parameter of 4500m, these ventscomprise seven geographic clusters, with only five outlyingvents occurring in the entirefield. Cindercone clusters in the westemportion of the field are significantlyolder than clusters in the eastern portionof the field (p valueof <0.001),and there is a tendencyfor cluster age to decreaseto the east. This is particularlyevident when mean cluster ages are calculated for tholelite,alkaline olivine basalt, and evolvedalkaline rock types independently. Application of thetwo-point azimuth and Hough transform methodsdemonstrates that regional cinder cone alignments transect these clusters. The most prominent of thesealignments trend ENE in theeastern portion of thefield and WNW in thewestern portion of thefield, creatingan overall arcuate pattern that is subparallelto the trend of theMogollon Rim andthe Colorado Plateau/TransitionZone boundary. These observations suggest that vents(and clusters) migrated from westto eastin responseto platemotion, but the generalpattern of ventmigration was complicated by regionalstructures, which enhanced the volume and duration of magmatismin some areas. The fractures or faultsimplied by vent alignments indicate that Shmin is orientedradial to theColorado Plateau in theSVF. Preferredvent alignment orientations may be relatedto extensionresulting from plateau uplift, and to a muchsmaller degree from a minorBasin and Range imprint. While regionalin extent,the implied structuresappear to differsignificantly from some of thosein severalother plateau-marginal fields in that theycannot be relatedto majorreactivated Precambrian structures. Our vent alignment data differ from thoseseen by otherworkers in theZuni-Bandera and Mount Taylor fields, suggesting the stress field for the SVF is differentfrom otherfields in the proposedJemez lineament. The stressfield impliedby vent alignmentdata, combined with structuraldata, suggests that the southwesterntectonic boundary of the ColoradoPlateau of Brumbaugh(1987) should be extendedsoutheastward to include the SVF at theplateau's southernboundary.
1. INTRODUCTION 1979], clusteranalysis [Connor, 1987, 1990], the Hough 1.1. Intent transform[Wadge and Cross,1988], and the two-pointazimuth method[Lutz, 1986;Wadge and Cross,1988; Zhangand Lutz, The apparentcorrelation between regional structures and 1989]. The results of our vent distribution analysis are ventalignments in manycinder cone fields has suggested that comparedwith structuraldata collectedin the field. This ventsare locatedat the top of vertical fracturesalong which comparisonis usedto showthe relationshipbetween observed magmaascends [Kear, 1964; Nakamura,1977; Settle, 1979]. patternsin cinder cone distributionand regional crustal Ventalignments and parallel dikes [Delaney et al., 1986]have structure and to comment on the implied stress field alsobeen used by many workers[e.g., Zobackand Zoback, surroundingthe SVF. The analysisserves to clarify the 1980, 1989; Zoback, 1989; Aldrich and Laughlin, 1984] as relationshipbetween SVF volcanismand the neotectonic regionalstress orientation indicators. One of the fundamental structureswhich dominate the geology of the region. In weaknessesimplicit in usingthese kinds of data is that there addition, these data are examined for clues regarding the has beena lack of rigor in definingvent alignmentswithin mechanismsgoverning the timing andlocation of cindercone volcanic fields, where vent density often makes alignment emplacementwithin the field. recognitiondifficult. In thispaper we useseveral quantitative methodsto searchfor vent clustersand alignmentsin the late 1.2. Background:Regional Setting Tertiary-QuaternarySpringerville volcanic field (SVF), located on the southernmargin of the ColoradoPlateau. These Cindercone volcanismhas been widespreadon the southern methods include univariate statistics [Porter, 1972; Settle, part of the ColoradoPlateau physiographic province during thelate Tertiary and throughout much of theQuaternary. With the exceptionof the Hopi Buttes,all of this volcanismhas Copyfight1992 by the AmericanGeophysical Union. been concentrated within seven volcanic fields located near the Papernumber 92JB00929. marginsof theplateau. The Springervillevolcanic field is the 0148-0227/92/9ZIB-00929 $05.00 southernmostof thesefields (Figure 1); its southernmostflows
12,349 12,350 CONNORET AL.: CONTROL•ON VENT DISTRIBUTION, SPRINOERVIU.E VOLCANIC FIELD
114ø 109ø 103ø
•ii:'"'"•WesternGrandCanyon ' - Jemez• •iiiii• i ,_, """•iii:•:':"::::'"':'::::.-....-'•lI ••' i!•SanFranciscø' Ho...... Colorado IPlateau Mount Taylor ':,.•,_ ':'••'.:.:..:.':••i•i•::::'':::''' '::•:':':• '-• !
I i•'• r-.•,.'"'""--"""'"'"'""'"'• .....•"' '•:•:•• " •"" • If iel d. field,•:'.'..'"•i•t ' / v /. . - I •...... ' • I • •' "• '• .•jljt•Flagstaff I \ - ':•..'?f •. / ' } _ •iiiiii'.:::•,,::.•u-.-o- I '.'•? • J #Albuquerque • \••' • • ••::'"•::::•c[""• Hackberry..•.:•, •__••'•:':'•'':• •., •!, • , High '. •[•• ' ",•':':•::'"":":•:iii:'":':"::•--I •' \ Plains ', "'-' X..._ __ )), ! I(white Uøun I • • Zone Grande • • I Rift BasinandRange '• "----I.i.':•;•/• • .... x, • .... • • 32ø ___1 , km
Fig. 1. The Springervillevolcanic field is locatedalong the southernmargin of the ColoradoPlateau, in centraleast Arizona.Physiographic provinces of the areadelineated. The Moõollon Rim is a topographicescarpment which defines the boundarybetween the TransitionZone and the Colorado?lateau, south of the SVF. Shadedareas indicate volcanic fields lessthan 5 m.y.old, outlined areas indicate volcanic fields 5 to 16m.y. old [fromWolfe et al., 1983].The Jemcz lineament [Aldrichand Laughlin,1984] is thoughtto be comprisedof volcanicfields and associated structures extending from the Jemezvolcanic Held to thesouthwest, through the MountTaylor, Zuni-Bandcra, and Springerville fields.
straddle the Mogollon Rim, which forms the physiographic 1.3. Background: SpringervilleVolcanic Field boundary between the Colorado Plateau and the Transition Zone. Geophysical models across the Transition Zone The SVF comprises409 vents distributedover an area of betweenthe ColoradoPlateau and Basin and Range(Figure 1) approximately3000 km 2, locatedjust north of the9- to7- show a significant change in depth to the mantle from m.y.-old Mount Baldy (White Mountains) trachyte shield approximately40 km in the interior of the plateau,to 22 km volcano (Figure 2; Nealey [1989]). The vent structuresare in the Basin and Range [Brumbaugh, 1987; Warren, 1969]. dominantlycinder conesbut include spattercones, two shield Near the SVF, this changestarts near the northernboundary of volcanoes,four fissurevents, and five maar craters. Although the TransitionZone (the Mogollon Rim) about25 lcmWSW of thereare someolder lava flows [Conditet al., 1992; Cooperet the SVF, where the seismicrefraction study of Warren [1969] al., 1990; Condit, 1984], the vents in this study were active suggestsa depth to the Moho of about 40 kin. South of this between2.1 and 0.3 m.y. ago and eruptedapproximately 300 area the TransitionZone is further characterizedby a change km3 ofbasaltic lavas of dominantlyalkalic affinities (alkali- from flat lying relatively undisturbedPaleozoic rocks to a olivine basalt(=47 vol %) and hawaiite (=28 vol %) [Condit et setting dominated by Basin and Range faults and increased al., 1989]). Tholeiite (=24 vol %) and a limited number of seismicity [Brumbaugh, 1987; Keller et al., 1979; Thompson more evolved alkalic rocks (mugeariteand benmoreite,<1 vol and Zoback, 1979]. The SVF itself lies within an aseismic %) are alsopresent but includeonly a few vents. In contrastto zone [Brumbaugh, 1987]. Although the physiographic othervolcanic fields along the margin of the ColoradoPlateau, boundaryof the ColoradoPlateau is well defined and abrupt, suchas the San Franciscofield [Wolfe et al., 1987a, b] and the the tectonic boundary of the plateau on its southernmost Mount Taylor field [Crumpier, 1982], large silicic centersare margin is indistinct. not present in the SVF, and no high-silica flow units have Aldrich and Laughlin [1984] have suggestedthe SVF lies at been identified [Condit et al., 1989; Ulrich et al., 1989]. the southwestern end of the Jemez lineament. The Jemez Detailed geologicalmapping [Condit et al., 1992] together lineamenthypothesized by these workers is a broad (=50 lcm with petrologic and stratigraphicinvestigations in the SVF wide), N52øE trending, tectonicallyactive zone, characterized have led to several additional observations that are relevant to by volcanic fields and N25øE striking en echelon faults, the origin and history of volcanism in this area. First, the which, along with an inferredPrecambrian province boundary, locus of active volcanismhas shifted throughtime, and lava delineatesthe southeasternmargin of the plateau (Figure 1). composition has evolved similarly in all locations Clearly, interaction of the differing tectonic provinces of accompanyingthis shift. Early lava flows in the SVF are sheet plateauand Basin and Range,and possiblythe uniquestructural flows of tholeiitic composition and are widespread; the featuresof the Jemezlineament, may complicatethe structural locationsof the few unburiedvents for theseflows suggestno settingof the SVF. The relative importanceof each has not geographic preference in their distribution [Condit et al., been fully assessed. Therefore a primary goal of our 1989]. Following the effusion of tholeiite, alkali-olivine investigationhas been to evaluate the relationshipbetween basaltbecame the dominanteruptive product. Early formed patternsin cinder cone distributionand tectonicfeatures of the alkali-olivine basalt vents tend to be concentrated in the region. western portion of the field, and through time, eruptions CONNORE'q' AL.: CONTROLSON VENT DISTRIBUTION,SPRINGERVILLE VOLCANIC FIELD 12,351
migrated from west to east, coincident with the direction of expressionfrom 7 1/2-min topographicmaps and airphotos. motion of the North American Plate, a pattern similar to that These41 vents are found in an unmapped =200-km 2 area in the suggestedby Tanaka et al. [1986] for contemporaneousrocks southcentral part of the field. Within the mappedarea, field of the San Francisco volcanic field. Condit et al. [1989] relationssuggest that 40 cinder conesare not associatedwith speculatedthat the SVF may be locatedover a thermalanomaly lava flows, being either barren cones or cones with buried fixed with respect to the sublithosphericmantle. Evolved flows. Some 266 vents are clearly associatedwith single lava alkaline rocks (hawaiite, mugearite, and benmoreite), which flow units. Of the remaining 102 vents, 60 are associatedwith are generally younger than other lavas, show a similar shift 30 flows (2 vents/flow), 21 with 7 flow units (3/unit), 8 with 2 from westto east throughtime. flow units (4/unit), and one flow unit has 5 vents. A secondobservation is that mappedregional faults do not Compositionsare known for 240 vents; most analyseswere transect the SVF. Unlike the San Francisco and Western Grand obtained from associated flows near vent locations. A Canyonfields, no major basement-controlledregional faults complete listing of vent coordinates is available from the or flexures have been identified in the sedimentaryrocks authorson request. surrounding the SVF. Some vents in the San Francisco The distance to nearest neighbor vent in the SVF is [Tanaka et al., 1986; Ulrich and Nealey, 1976], Mormon lognormallydistributed with >99% confidence.The geometric [Ulrich et al., 1989; Holm et al., 1989], Zuni-Bandera[Kelley mean distance to nearest neighbor vent is 955 m. Settle and Clinton, 1960; Aldrich and Laughlin, 1984], and Mount [1979] characterized cinder cone distribution in several cinder Taylor [Kelley and Clinton, 1960; Crumpler, 1980a, b, 1982] cone fields by reporting the nearest neighbor distribution in volcanicfields are associatedwith, or overlie, prominentdip- terms of quartiles. Using this approach,Settle found that slip faults. For example,four major volcanic centerswithin platform-type fields, those consistingsolely of independent the San Franciscofield line up alongregional fault systems monogeneticvents, have a greater vent spacing than volcanic- and many cinder cones are located over fractures and are type fields, which consist of vents parasitic to a larger elongateparallel to the fault system(e.g., vent 6735 [Ulrich polygenetic volcano. Comparing the SVF to two other and Bailey, 1987], vents 6604 and 6609 [Wolfe et al., 1987a], platform-type fields, it is clear that the SVF has much more vents 2418 and 3132 [Newhall et al., 1987], and vent 80 closelyspaced vents than the other platform-typecinder cone [Moore and Wolfe, 1976]). Numerous vents in the Zuni- fields cited (Table 1). As this is true at each quartile, the Banderafield showsimilar relationships to mappedfault zones observation is true of the distribution as a whole and is not [Aldrich and Laughlin, 1984; Kelley and Clinton, 1960]. In dependenton the presenceor lack of outlyingcinder cones or a the Mormon volcanic field, vents align along north and few very closelyspaced vents. The SVF has medianand upper northweststriking structures. Within the Mount Taylor field, (75%) quartiledistances which are shorterthan two of the three which is located on the Acoma Sag, vents tend to form volcanic-type fields cited (Mauna Kea and Kilimanjaro). A distinctivefissure patterns aligned parallel to or along NNE map of the density distribution of cinder cones in the SVF oriented structures or to be clustered in linear trends with (Figure 3) showsthat vents are most densely concentratedin orientationssimilar to basementstructural trends [Crumpler, the southcentral portion of the field. Contoursshowing high 1980a, b, 1982]. ventconcentrations (greater than 14 vents/177km 2) are 2. METHODS elongatein a WNW orientation(Figure 3).
2.1. Distributionof Cinder Cones 2.2. Cluster Analysis The 409 vents used in this analysis(Figure 2) include 368 Univariate descriptive statistics cannot fully characterize mappedvents and 41 vents delineatedby their topographic vent distribution because they cannot describe spatial
3800 _ [--ISNOWFLAKE 109ø30' /',• I 34•30'
3780
3760
3 4ø ß
•10 ø
580 600 620 640 660 Fig. 2. Map of the locationof all (409) ventsin the Springervillevolcanic field (triangles).Mount Baldy, a Tertiary trachyticshield volcano (large solid triangle), is locatedabout 20 km southof the SVF. The townsof Springerville,Show Low, andSnowflake are shown as open boxes. Universal Transverse Mercator CUTM) coordinates are given at themargin of the map,labelled at 20-kinintervals (10-kin tick spacing). 12,352 CONNORET AL.: CONTROLSON VENT DISTRIBUTION, SPRINGERVILLE VOLCANIC FIELD
TABLE 1. Near-Neighbor Distributionsfor Some Volcanic Fields densitycontour map will changeif thesesearch parameters are changed.The validity of using one size search area or grid Distanceto Near-NeighborVent by Quartile, m Number spacingover another is difficult to assess.Cluster analysis 25% 50% 75% of cones avoidsthe interpolationassociated with contouringaltogether and provides a systematicapproach to dealing with search Platform-TypeFields areas.Here, a uniform kernel densityfusion cluster analysis is Springerville field 710 979 1407 409 used[Wong, 1982; Wong and Lane, 1983; Wong and Schaak, Nunivak Island 1047 1547 2955 83 Michoac•n, Mexico 988 1580 2732 1016 1982; Sarle, 1985]. The applicationof this techniqueto Volcanic-TypeFields cindercone distributionproblems was describedby Connor Mauna Kea 569 813 1524 168 [1987, 1990]. Mount Etna 425 780 1373 87 Briefly, this cluster analysis provides a means of Mount Kilimanjaro 679 1152 1879 205 recognizingmodes in vent distributionin a quantitativeway. The method is considered robust because even if two clusters Distributionof distances(in meters)to nearest-neighborvent by overlap slightly, they will still be recognizedas two distinct quartile for various platform-type and volcanic-typecinder cone fields clusters. A circle of radius r is drawn about each vent within [Settle,1979; Connor, 1987; this study]. SVF ventsare moreclosely the field. The number of cones to fall within this circle is spacedthan vents inthe Nunivak Island and Michoac•n platform-type f(Xi), for the ith vent. Ultimately, ann x n matrixis calculated fields and, at some quartiles, than vents in the Mauna Kea and Mount K'fiimanjarovolcanic-type fields. forthe n numberofvents in thefield. For each element in the matrix,
variation in distribution. The goal of cluster analysis is to identify natural vent clusters within the SVF. This is 1] , d(Xi,X) 3800 r'"] 1ø9ø3ø' SNOWFLAKEff'--"X• 34ø30' SHOWLOW o-, 4- 2 3780 I ERVILLE 3760 34ø 110ø MT. BALDY i i i i i •1 i i ,i 580 600 620 640 660 Fig. 3. Vent densitydistribution contour map basedon a contouringgrid spacingof 10 km and searchradius of 7.5 km abouteach grid point. The grid pointvalues were contouredusing a minimumcurvature contouring algorithm. Contour intervalis2 vents/172km2. Notethat the highest concentration ofcinder cones inthe SVF is found in the south central portionof the field. In this area, contourlines are elongatein a WNW orientation.UTM coordinatesare given at the margin, as in Figure 2. CONNORETAt,.: CONTROLS ONVENT DISTRIBUTION, SPRINGERVILLE VOLCANIC FIELD 12,353 adverse effect on the results unless vent density is low and producedby some underlying geological mechanism, such as clustersare poorly developed.The single linkage clustering anisotropyin the stressfield. processcontinues repetitively until a stablenumber of clusters The basis of the two-point method is a Monte Carlo is found. A map is then made, showing the distributionof simulation. Any two vents in the volcanic field lie along a individualvents by clustermembership. Of course,changing line. As a farststep in the analysis,lines are drawn from each the searchradius, r, will changethe numberof clusterson this vent through all other vents in the field. The azimuth of each map. Therefore the analysis is repeatedusing many search of these lines is measured, resulting in a total of n(n-1)/2 radii. If clusteringis a significantfeature of vent distribution, azimuths, where n is the number of vents. The number of two- then the number of clusters, and more particularly the point alignmentsof a given orientation will depend on both distributionof clusters,will not change or will change only the presenceof vent alignmentsand the shapeof the boundary slightly with changingsearch radius. of the volcanic field, since vent pairs tend to align in a For SVF vents, the change in the number of vent clusters preferredorientation if the entire field is elongate. The Monte with changing search radius is illustrated in Figure 4. At Carlo simulationis made to correctfor the effect of field shape. searchradii greaterthan 6000 m, a single cluster exists with In this case,we followed Lutz's [1986] example and chosethe four outliers. Between6000 and 5500 m, severallarge clusters vents at the margin of each cluster as the vertices of a form, and at search radii less than approximately 4200 m, polygonal area for the Monte Carlo simulation. For each theselarge clustersbegin to pull apartin an apparentlyrandom simulation, n points were randomly plotted within this area, manner, and the number of clusters with decreasingsearch each point representingthe location of a vent. The azimuth radiusbegins to increasemuch more rapidly (Figure 4). from each point to every other point within the polygon was A total of 12 maps was produced, illustrating the found, and the cumulative frequencieswere comparedwith the distributionof vents by clustermembership at different search observedvent azimuth frequency distribution at 10ø intervals. radii. Sevenclusters persist over a range of searchradii (see If actual vents tend to align preferentially in a given Figure 5 caption) with only slight changesin membership. orientation, this will emerge through comparison with the The distribution of these clusters is illustrated in Figure 5, Monte Carlo results,using a t test [Lutz, 1986]. usinga searchradius of 4500 m. Theseclusters have between The two-point azimuth method was appliedon a clusterby 27 and 101 cinder cones each. In addition to these seven clusterbasis to the seven largest clustersin the field (Figure clusters, five cones are classified as outliers because they 5). The outcome of the analysis indicates that significant either do not cluster with other cinder cones at all at this search anisotropyexists in the distribution of cinder cones within radiusor they form a clusterof only three cones(Figure 5). A each cluster. Significant azimuthal orientations are found total of four vents appearto be misassignedat a searchradius within each cluster at the 95% confidencelevel (Table 2). The of 4500 m as a result of low vent densities in some areas, such azimuthal directions identified as statistically significant are asbetween clusters I and2 (Figure5). not identical in all clusters, but they appear to vary in a consistentmanner. Significant azimuthal directions are WNW 2.3. Two-Point Azimuth Analysis oriented in clusters 2 and 3 (Figure 5), coinciding with the elongate area of highest vent density (Figure 3). Significant The two-pointazimuth method, developed by Lutz [1986] azimuthal directions are ENE oriented in clusters 5 and 6 in the and first applied to volcanodistribution problems by Wadge easternpart of the field and in clusters1 and 3 in the western and Cross [1988], providesa statisticalmeans of identifying and central part of the field. Clusters 4 and 7, in the central preferredorientations and/or anisotropyin vent distribution. part of the field, have significantN-S azimuthaldirections, and Within a given area, somevents will align even if thesevents significantNE orientationsare identifiedin cluster4. are distributedaccording to a uniform random process. For example,vents emplaced in an isotropicstress field may form an alignment by chance. Such an alignment could be 2.4. The Hough Transform misinterpretedas an indicationof a preferredhorizontal stress orientation.The two-point azimuth method helps distinguish The two-point azimuthmethod does not provide information between alignments formed by random chance and those on the actual location of alignments; it simply indicates the orientation of significant anisotropy. Wadge and Cross [1988] applied a computerenhancement technique, the Hough 80 transform, to cinder cone distributions in Michoac•n, Mexico, to determinethe actual locationsof alignments. We apply it in conjunctionwith the two-point azimuth method to identify 60- alignmentsin the SVF, again on a clusterby clusterbasis. Each vent in a cluster lies along an infinite number of lines, each line having a unique azimuth. These lines can be representedin polar coordinatesas curves. A point on one of these curves has the coordinates p and O, where p is the shortest(normal) distancefrom that arbitrarypoint to the line, 20- and 0 is the angle of that normal from zero [Wadge and Cross, 1988]. In this case, we chose the mean cluster centroid as the , center of the coordinate system. If, for example, four vents i i i i 2000 4000 6000 8000 within a cluster align exactly, the four sinusoidal curves associatedwith these vents in p, 0parameter space will Search radius, r (m) intersectat a single p, 0 coordinate,which in turn yields the Fig. 4. Changein the numberof clusterswith changingsearch radius orientation and position of the line, relative to the cluster about each vent. The clusteringmethod used is a modification of centroid. Wong's[1982] densityfusion method [Sarle, 1985]. Solid circlesare In practice,discrete A0 and Ap are used. Here, we use Zi0 = plottedto indicatethe numberof clustersfound by the analysisat a specificsearch radius. Note the rapidincrease in the numberof clusters 2ø andAp = 400 m, the samevalues used by Wadge and Cross at searchradii less than approximately4200 m. [1988] and Connor [1990], taken to represent reasonable 12,354 CONNORET AL.: CONTROLSON VENT DISTRIBUTION,SPRINGERVILLE VOLCANIC FIELD 110'00 109'45 109'30 34'30 i i i 34'30 Cluster1 Cluster3 Cluster4 Cluster5 C,.str. C,str, Sho Faults-markdownthrownon side Flexures N 0 5 km . . oo ½ o¸ Springerville!-! 0' ße •e D D 34'30 110;00 109;45 109"30 Fig. 5. Ventsplotted by clustermembership. Vents belonging to the samecluster are plotted using the samesymbol, centeredon the ventlocation. This map illustrates the resultsof a clusteranalysis using a searchradius of 4500m, whichis wellrepresentative of the large clusters in thefield. Othersearch radii produce different cluster configurations. The clusters shownare stableover the followingsearch radii, plus or minusa few vents: cluster1, 4200 to 5500 m; cluster2, 4200 to 5500 m; cluster3, 4200 to 5500 m; cluster4, 2900 to 4500 m; duster5, 3700 to 5000 m; cluster6, 3500 to 4500 m; cluster7, 2700to 4500m. Vent alignmentsidentified using the Houghtransform are indicated by solidlines and are labelledA-L. Thesealignments were identified on a clusterby clusterbasis; nonetheless, they often meet or lean be extrapolatedacross cluster boundaries, forming arcuate trends in thesouthern part of thefield. fracture zone sizes. Experimentationshowed that the number In five cases,the Hough transformidentifies alignments and orientation of alignmentsdid not vary significantly by which consist of six or more aligned vents and do not changingthese parameters.The Hough transformis sensitive correspondto significantorientations found using the two- to the number of cinder cones used in the analysis and the point azimuthmethod (Table 2; Figure 5). Two of theseare shapeof the cinder cone cluster. The more cinder cones in a foundin the NW part of the field within an elongatecluster cluster, the more likely it is that several cinder cones will (cluster 1). The two-point azimuth method minimizes the align. If a clusteris elongate,alignments will likely be found effect of cluster shape, and as a result any trend in this in the directionof elongation. Consequently,we were careful orientationis difficult to identify with the two-point method. to comparethe resultsof the Hough transformwith the results Another such alignment is located in cluster 2 and has an of the two-point azimuth method, which takes cluster shape azimuth of 020ø . This alignment consistsof nine cones and into account. transects the entire cluster. Within each cluster, one or two alignmentsthat consistof numerousvents were identified using the Hough transform. 3. STRUCTURAL FEATURES AND VENT Most of these alignments have orientationssimilar to those DISTRIBUTION PATrERNS recognizedas significantusing the two-point azimuth method at the 95% confidencelevel. These alignmentsconsist of six Lavasof theSVF capa thicksequence of sedimentaryrocks and usually seven or more vents, usually in proportionto the of Permianto Cretaceousage that dip to the NNE at --•-0.5ø numberof cindercones in the entire cluster(Figure 5; Table 2). [Condit et al., 1989] and form the MogollonSlope, These alignments were identified within each cluster southernmost tectonic division of the COlorado Plateau independently. Nonetheless, the alignments often meet, [Kelleyand Clinton,1960]. This sedimentarysequence is nearly meet, or can be extrapolatedacross cluster boundaries. nearly flat lying and is essentiallyundeformed. The entire This is particularly true for alignments in the central and MogollonSlope area around the SVF lacksthe north striking southernportion of the field. Vent alignmentsC, E, and I normal faults characteristicof the westernmargin of the create a nearly continuousalignment with an arcuateshape plateau[Wernicke and Axen, 1988], the northeasttrending which spans the southern half of the field and which is structures characteristic of the southwesternplateau approximately65 km in length. This alignmentdivides in the [Shoemakeret al., 1978;Tanaka et al., 1986;Holm and Cloud, central portion of the field, within cluster 3, east of which a 1990],or the pronouncednorth and northwest striking faults secondarcuate alignment forms (alignmentsF, t3, and H of of the Mormon volcanic field [Holm et al., 1989]. In the Figure 5). sedimentaryexposures to the north of the SVF, with the CONNORET AL.: CONTROLSON VENT DISTRIBUTION,SPRINOERVII.LE VOLCANIC FIELD 12,355 TABLE 2. Summaryof the AlignmentAnalysis Cluster and Two-Point HoughTransform No. of Vents Azimuth Direction Azimuth No. Vents Length Vents/km 1 27 070-080 .... - 295 (A) 6 16.3 0.37 - 298(B) 6 19.0 0.32 2 101 280-290 291 (C) 9 19.8 0.45 - 022 (D) 9 21.2 0.42 3 84 07 0-080 071 (E) 9 10.0 0.90 *300-310 307 (F) 9 10.4 O.87 4 38 020-030 - _ - - 040-050 042 (L) 6 11.4 0.53 350-360 356 (K) 6 12.2 0.49 - 331 (J) 6 17.0 0.35 5 66 060-080 061 (I) 9 2O.7 0.43 6 29 050-060 055 (H) 7 9.2 0.76 070-080 - - - - 7 56 000-010 090 (G) 8 11.3 0.71 ClUstersare numbered asin Figure5. Theresults of thetwo-point azimuth analysis are reported for each cluster.All orientations,calculated at 10ø intervals,found to be significantat the 95% confidencelevel are given (except the asterisk,for which the confidenceinterval is 90%). Specific alignmentsare identified using the Hough transform;these are identified by letter as in Figure 5. The azimuth, numberof vents, alignmentlength measured from first to lastvent in the alignment,and number of vents/kilometerare given for each alignment. There is generalagreement between the resultsof the Hough transformand the two-pointazimuth method. In casesin which the two-pointazimuth orientation does not correspondto the azimuth of an alignment identified using the Hough transform, or vise versa, a dash is entered in the correspondingcolumn(s). *Confidence level of 90% for this interval. exceptionof three normal faults of minor displacementand Figure 5) and locally faulted and tilted basalticflow units, most less than 15-km lengths, the few structuralelements include found in the north and northeasternpart of the field. In the gentle anticlines and synclines [Wilson et al., 1960, 1969; westernand centralpart of the field, many of thesefeatures are Kelley and Clinton, 1960], most with WNW to NW trends, aligned WNW to NW, parallel and subparallel to vent some of which extend for distances over 20 km. Several of alignments(Figure 5). Structuralfeatures of other orientations these trends continue into the field, expressedas local normal are present as well, indicating the local complexitiesof the faults that gradeinto flexures[Crumpier et al., 1989; Condit et stressfield acting on the area concomitantto and following al., 1992]. That theseflexures postdate flows has been amply the emplacementof cinder cones. In order of prominence, demonstrated[Aubele et al., 1986]. thesestructures trend NW, ENE, NE, andnorth [Crutr•ler et al., These flexures form topographicescarpments and are the 1989]. most prevalent structuralfeatures of the field. Starting from Four widely separated eruptions appear to have been the southcenter of the field (centerof cluster3, Figure 5), there controlled by fissures, as suggested by their elongate is a decreasein elevationto the northeast,with a total drop of pyroclasticdeposits [Condit et al., 1992]; in addition,a single approximately1 km. This drop occursas a seriesof three dike is found within the field. These features tend to confu-m discrete topographic steps, expressed as arcuate northeast the interpretationthat cinder cone alignmentsare related to facing escarpments(Figure 5). From southwestto northeast, structural trends. Two of these fissure-controlled vent the topographicrelief decreasesfrom approximately 250 m structures are located on and elongated parallel to vent acrossthe first step to 100 m acrossthe second,and to 80 m alignments.Elongate vents form the central part of alignment acrossthe northeaststep. All of these steps have variable I (Figure 5). To the north of this alignment,two cindercones strikes reflecting their arcuate traces: WNW in the are connected by a fissure ridge of spatter and form the northwesternand central portions of the field changing to northernend of alignment K (Figure 5). Located roughly NNW in the northeastern portion of the field. The betweenthese alignments,two vents erupted along a 200-m- topographicallylowest area in the region occursimmediately long fissureoriented 080 ø. In additionto thesefissure vents, a NE of the field. Approximately half to three fourths of this single dike about 1.1 km long has been mapped. Located in dropin elevationcan be accountedfor by the 0.5øNNE regional the south central part of the field, the dike trends 058 ø, dip and by the accumulationof flows which stack up to the subparallelto cinder cone -alignmentsE, H, and I. southwesttoward the center of the field; the rest appearsto result from downwarpingto the northeast. The northeastmost 4. DISCUSSION of these flexures correspondsto a vent alignment (alignment J, Figure 5). Reasonsfor the emplacementof platform-type cinder cone The overall structural fabric within the field is further defined fields instead of single, polygenetic centers are not entirely by two other passive folds (between alignments B and C, clear but may relate to the relative rates of magma production. 12,356 CONNORET AL.: CONTROLSON VENT DISTRIBUTION,SPRINGERVILLE VOLCANIC FIELD Cinder cone fields may form in areas having relatively low classified rocks in the SVF as alkali olivine basalts (AOB), magma supply rates [Fedotov, 1981; Hasenaka and evolved alkaline rocks (EAR, including hawaiite, mugearite, Carmichael, 1985]. Given a low magma supplyrate, conduits and benmoreite),and tholeiite. Applying their classification, are not maintainedin the crust and individual magma batches we find a systematicchange in the agesof the three rock types ascendvaried pathwaysto the surface,rather than ascendingin from clusterto cluster. Within individual clusters,the average the same conduit repeatedly [Fedotov, 1981]. The SVF age of AOB rocks is greaterthan the averageage of EAR rocks certainly does have a low magma supply rate comparedwith (Table 3), perhapsa result of longer residencetime neededfor many individual volcanic centers. The SVF producedlavas at magmatic differentiation. Between clusters, however, this anaverage rate of 1.5x 10-4 km3/yrbetween 2.0 and 0.3 m.y. relationshipdoes not hold true. For instance,the averageage ago [Condit et al., 1989]; peak lava productionrates between of EAR rocks in cluster 1 is significantly greater than that of 2.0 and1.0 m.y reached 2.8 x 10-4 km3/yr.This is 1 to 2 AOB rocks in cluster 3. Comparisonof similar rock types ordersof magnitudeless than the output rate of most basaltic betweenclusters suggests a systematicvariation in the timing centers listed by Crisp [1984]. The Mount Baldy of eruptionof magmasof differing compositions.Clusters 1 stratovolcanojust to the south of the SVF, with an estimated and 2, located in the western portion of the field (Figure 5), containsignificantly older rocks of a given type than clusters volumeof 280 km 3 [Merrilland Pewe, 1976], appears tohave locatedfurther to the east (p value of <0.001). This is true for formed in about 0.5 m.y. [Nealey, 1989], with a production tholeiite,AOB, and EAR rocks. In general,there is a decrease rateof 5.6 x 10-4 km3/yr,about twice that of thepeak in the ages of rocks from west to east, among all petrologic productionof the SVF. The outputrate of basalticlavas in the types, consistentwith the fixed-source model proposedby San Francisco Peak volcanic field has been similar to that of Condit et al. [1989], in which eruptive activity migratedfrom the SVF, with a maximum long-term eruption rate (between west to east in responseto the westwardmotion of the North 0.73and 0.! m.y.)of about3.0 x 10-4 km3/yr;silicic peak American Plate relative to a fixed source [Condit et al., 1989]. outputfor this field was about 2.0 x 10-4 km3/yrbetween 1.0 Volcanism in the SVF reflects this motion by shifting from and 0.25 m.y. if adjusted to reflect only extrusive rocks one cluster to another through time, rather than by the (without correcting for porosity) [Tanaka et al., 1986]. continuousmigration of vent loci through time. This is Similar rates are suggestedfor the Mount Taylor volcanicfield analogousin some respects to Hawaiian hot spot volcanism byCrumpler [1990] at between1.1 and 1.6 x 10-4 km3/yr [Jackson et al., 1972], where individual shields are built between3.5 and 1.5 m.y., althoughPerry et al. [1990] give through time in response to the continuous motion of the Pacific plate with respectto a fixed source. Also analogousto peakproduction rates as low as 4.0 x 10-5 km3/yr for a central Hawaiian volcanism, activity does not cease in one cluster coneof the samefield. Low magmasupply rate and cindercone simply becauseit has begun in another. volcanismare coincidentin the SVF and perhapsin many of The picture in the SVF appearsto be more complex than in the plateau-margin cinder cone fields. This may indicate a its Hawaiian counterpart in several respects, not surprising causal relationship. given the differing lithospheric regimes. Comparatively Univariate descriptive statistics and cluster analysis have young EAR rocks are found in cluster3, in the southcentral demonstrated that vents are closely spaced in the SVF portionof the field (Figure 5) where cindercone densitiesare comparedwith some other active volcanicfields and occurin greatest(Figure 3). Althoughthe EAR rocksin cluster3 are on clusters, rather than having regular or uniform random averageyounger than those found further east, the variancein distribution. In a field of 409 vents, only five outlying vents the ages of rocks in this cluster is larger than that of other occur. These observationssuggest that throughtime there is a clusters(Table 3). In fact, someof the oldestEAR rocksfound tendency for successivemelting events to occur near one in the field occur within this cluster. Similarly, cluster 2, another,rather than randomlyover the entire area of the field. which containson averagecomparatively old AOB and EAR This result is consistent with the idea that cinder cone clusters rocks, also contains some of the youngest. Apparently the owe their originprimarily to low ratesof magmageneration in eruptionof magmas is influencedby additionalfactors, some localized areas, rather than to the dispersal of magmas by discussedbelow, complicatingthe generalwest to eastpattern crustal structures. of vent migration. Using K/At, paleomagnetic,and stratigraphicdata gathered Several authorshave suggestedthat a coincidencebetween by Conditet al. [ 1989], we have comparedthe agesof rocksof areas of magma generation and faulting or fracturing of the differing petrologiesbetween clusters. Condit et al. [1989] crust is necessary for the emplacement of platform-type TABLE 3. Mean Agesof Rock Typesin DifferentClusters Cluster Tholeiite AOB EAR N M V N M V N M V 1 2 1.56 0.03 7 1.42 0.15 4 1.35 0.03 2 5 1.67 0.01 36 1.38 0.14 18 1.30 0.18 3 0 - - 16 1.17 0.05 19 1.05 0.26 4 4 1.11 0.08 11 1.29 0.02 17 1.21 0.04 5 1 1.07 - 27 1.20 0.12 22 1.19 0.05 6 0 - - 11 1.20 0.06 6 1.10 0.01 7 0 - - 18 1.14 0.02 14 1.11 0.07 WhereM is meanage (m.y.), V is variance,and N is numberof samples.Cluster numbers are thosegiven in Figure5. Clusters1 and 2, in the westemSVF, have the oldestmean vent agesfor tholeiite,alkaline- olivine basalts(AOB), and evolvedalkaline rock (EAR) compositions.In generalthere is a decreasein meanvent agefrom westto east. However,there are someexceptions. In cluster3, for example,eruptions of EAR typeshave increased over time, giving this cluster a youngmean age and a highvariance. Age data from Condit et al. [1989]. CONNORET At..: CONTROLSON VENT DISTRIBUTION,SPRINGERVILI.E VOLCANIC FIEt.D 12,357 volcanicfields. Without this coincidence,magma will likely Given the tectonicsetting in a transitionallocation at the edge form intmsives [Kear, 1964; Settle, 1979; Fedotov, 1981]. of the plateau,just north of the Basin and Range, all three of Patternsin vent migrationin the SVF are somewhatsimilar to theseShrni n stressorientations (NE, NW, and E-W) were likely those observed in the San Francisco volcanic field, where presentduring eruption of the cindercones of the SVF. known and mappedregional structures exert influenceson vent The limited number of vent alignments(D and possiblyL; migration patterns, in addition to a general west to east Table 2; Figure 5) which fall near the 025ø orientationfound migrationin responseto plate motions [Tanaka et al., 1986]. by Aldrich and Laughlin [1984] to be the characteristicof the Most vents in the SVF, however, are not part of the Jemezlineament suggests that the SVF, if indeed part of this alignmentswe haveidentified. Local structuresmay play a role lineament, is at best on the distal end and was little influenced in the emplacementof thesevents, but regional structuresare by the stressfields that controlled the orientation of cinder not a controlling factor. Mapped structures and vent conesand dikes in other fields along this proposedlineament. alignmentswhich do occurin the SVF are subtle,and thereis As Brumbaugh [1987] points out, a clear distinctionshould no evidence that they are related to major reactivated be made between tectonic and physiographic provinces: a Precambrianstructures, as alignmentsare in the San Francisco tectonicboundary (or province) shouldbe defined by changes field. The structuresand vent alignmentsthat are presentin the in tectonic elements,some of which include structuralstyle, SVF are thereforemore likely to be directlyrelated to Plateau stress orientations, volcanism, heat flow, seismicity, and uplift andBasin and Rangeextensional tectonics, rather than changesin crustalthickness. Using thesecriteria, Brumbaugh preexistingstructures. [1987, Figure 8] has suggestedthat the southwesternboundary The quantitativemethods used in this studyhave enabledus of the ColoradoPlateau extends south from 112øW longitudein to accuratelymap subtle vent alignmentsin the SVF in a a concave east arc through the San Francisco and Mormon reproduciblemanner. These vent alignmentsare similar in Mountain volcanic fields, suggesting both fields (and the natureto regionalvolcanic alignments defined by Kear [1964], Westem Grand Canyon field) are tectonicallypart of the Basin in that they are regionalin extent(some are over 25 km long) and RangeProvince. His proposedboundary stops in the area and vents comprising the alignment are not necessarily of the Mormon Mountain volcanic field. We suggestthat, aligned,but are nearly aligned. All vent alignmentswe have given the geophysicalcharacter (depth to Moho of -.•-40km, identifiedmeet the highestquality ranking (A) usedby Zoback aseismic)and the minor imprint of a Basin and Range tectonic and Zoback [1989] to characterizehow accuratelya particular signature(as shownby thesealignment patterns), the SVF has datapoint records the tectonicstress field. The WNW-trending characteristics more common to the Colorado Plateau, and zone of high vent density (Figure 3), azimuthal directions Bmmbaugh'sproposed tectonic boundary should be extended found through the application of the two-point azimuth southeastward,including the SVF on the tectonicboundary of method(Table 2), and alignmentsA, B, C, and F (Figure 5) the Colorado Plateau. indicate that vent emplacementhas been influenced by NE extensionin the westernpart of the field. A changein the 5. CONCLUSIONS orientationof Shmin from dominantlyNE in the westernpart of the field to NW in the easternpart of the field is indicatedby A total of 404 out of 409 vents mapped in the SVF form vent alignmentsE, I, H, and L. The overall arcuatetrend of sevenclusters, using a clustersearch radius of 4500 m. These thesealignments mimics the trendsof the physiographicand vent clusters have significantly different ages, especially tectonicboundaries of the plateau. when differentiatedby petrologictype. The observedgeneral Based on the analysisof local structuralfeatures and the decrease in cluster age from west to east supports the patternsof faulting and local folding of basalts,significant previouslyproposed hypothesis that volcanism in the SVF structurallineaments in the SVF are oriented NW, ENE, NE, and migratesowing to the motion of the North American Plate north (in order of prominence). The NW orientation of relative to a fixed mantle source, with activity waxing and regional topographicsteps and linear deformationzones, waning cluster by cluster in a partially overlappingmanner. interpretedas complexfault zones,may be relatedto the WNW However, the distribution of clusters and some inconsistencies trendingvent alignments(Figure 5). The orientationsof folds in the age data suggestthat this vent migration pattern is and what Crumpler et al. [1989] interpretedas small pull apart complicatedby additionalfactors, including regional structure basins(e.g., 3 km NE of alignmentK, Figure 5) along the as inferredfrom vent alignmentsthat transectparts of the SVF. deformationzones suggesta componentof left-lateral strike- Most prominentalignments trend WNW in the westernSVF slip and that the topographicsteps are not a result of simple and ENE in the eastern SVF. These different trends intersect in normal faulting [Crumpier et al., 1989]. These observations the southcentral portion of the field, where magmatismwas are consistentwith NE extensionin this area, perpendicularto most enduring and cinder cone density is greatest. Other the trendof WNW orientedvent alignments(Figures 3 and5). orientations,particularly those observedin local structures The structuralpattern deducedfrom the analysisof vent and some cinder cone alignments, indicate that local distributionand the mappingof faults and flexuresis the result variabilityin the stressfield complicatedthe patternsof cinder of tensionalstresses that have been presentduring the Plio- coneemplacement considerably. The vent alignmentsindicate Quaternary. Stressorientations in the SVF have, on average, the presence of fractures or faults, along which magma beenradial to the ColoradoPlateau during the formationof the ascendedmore readily than elsewhere. The fact that most are field. Models of plateau uplift have demonstrated that subparallelto regional physiographicfeatures, such as the tensionalstresses in the SVF shouldbe radial to the plateau if Mogollon Rim, suggests that the overall arcuate pattern the field restson or near the tectonicboundary of the plateau observed in cinder cone alignments is a reflection of the [Thompsonand Zoback, 1979; Brumbaugh, 1987]. Some structuralmargin of the Colorado Plateau. This supportsthe structures,such as pull apart basins, are consistentwith conclusionof Zoback and Zoback [1989] that stressfields near clockwiserotation of crustal blocks in this region [Crumpier tectonic boundaries reflect structural transitions. The fractures et al., 1989] and of the plateauas a whole [Bryan and Gordon, or faults implied by vent alignmentswithin the SVF may be 1986, 1990; Steiner, 1986], also suggestingthat the field lies related to extensionassociated with deformationof the plateau on or near the plateau's tectonic boundary. The stress margin, and to a lesser degree to a minor Basin and Range orientationof N-S trendingalignments may be a reflectionof a imprint. While the implied structuresare regional in extent, small componentof the dominantlyE-W tensionthat is the they appearto differ significantlyfrom thosein otherplateau- hallmarkof the Basin and RangeProvince [Zoback, 1989]. marginalfields in that they cannotclearly be related to major 12,358 CONNORET AL.: CONTROLSON VENT DISTRIBUTION,SPRINGERVILLE VOLCANIC FIELD reactivatedPrecambrian structures, which are lacking around Hartigan,J. A., Clustering Algorithms, 456 pp., John Wiley, New the SVF. Our vent alignment data differ significantly from York, 1975. thoseseen by other workersin the Zuni-Banderaand Mount Hasenaka, T., and I. S. E. Carmichael, The cinder cones of the Taylor fields, suggestingthat if responseto a commonstress Michoac•in-Guanajuato, central Mexico, their age, volume, distribution, and magma discharge rate, J. Volcanol. Geotherm. field is a major criterionfor inclusionin the proposedJemez Res., 25, 105-204, 1985. lineament,the SVF is not a part of this feature. Finally, we Holm, R. F., and R. A. Cloud, Regional significanceof recurrent suggest that the southwesterntectonic boundary of the faulting and intracanyonvolcanism at Oak Creek Canyon, southern Colorado Plateau of Brumbaugh [1987] be extended ColoradoPlateau, Arizona, Geology, 18, 1014-1017, 1990. southeastwardto include the SVF on its southernboundary. Holm, R. F., L. D. Nealey,F. M. Conway,and G. E. Ulrich, First-day field trip: Mormon volcanic field, in IAVCEI volume, Field Acknowledgments.We appreciatethe commentsof R. E. Stoiber, Excursionsto Volcanic Terranes in the Western United States, whoreviewed an earlydraft, and discussions with D. U. Wiseand M. L. (IAVCEI vol.), editedby C. Chapinand J. Zidek, Mem. N.M. Bur. Williams. Careful reviews of the manuscriptby Ken Tanaka and Ed Mines Miner. Resour.,46, 41 pp., 1989. Wolfe were extremelyhelpful. The SoutheastRegional Data Center Jackson,E. D., E. A. Silver, and G. B. Dalrymple,Hawaiian-Emperor providedcomputer facilities for this project. chainand its relationto Cenozoiccircumpacffic tectonics, Geol. Soc. Am. Bull., 83, 601-617, 1972. Kear, D., Volcanicalignments north and westof New Zealand'scentral volcanicregion, N. Z. J. Geol. Geophys.,7, 24-44, 1964. Keller, G. R., L. W. Braile, and P. Morgan, Crustal structure, Aldrich, M. J., and A. W. Laughlin, A model for the tectonic geophysicalmodels and contemporarytectonism of the Colorado developmentof the southeasternColorado Plateau boundary,J. Plateau, Tectonophysics,61, 131-147, 1979. Geophys.Res., 89, 10,207-10,218, 1984. Kelley, V. C. and N.J. Clinton, Fracture systemsand tectonic Aubele, J. C., L. S. Crumpier,and M. Shaffiqullah,K-At agesof Late elementsof the Colorado Plateau, Univ. N.M. Publ. Geol., 6, 104 Cenozoicrocks of the central and easternparts of the Springerville pp., 1960. volcanic field, east-centralArizona, lsochron West,46, 3-5, 1986. Le Maitre, R. W., NumericalPetrology, Statistical Interpretation of Baker, M. C. W., Volcanoes,spacing, fractures, and the thicknessof GeochemicalData, Dev. in Petrol., vol. 8, 281 pp., Elsevier,New the lithosphere- A discussion,Earth Planet. Sci. Lett., 23, 161- York, 1982. 162, 1974. Lutz, T. M., An analysisof the orientationsof large-scalecrustal Brumbaugh,D. S., A tectonic boundaryfor the southernColorado structures:A statistical approach based on areal distributionsof Plateau,Tectonophysics, 136, 125-136, 1987. pointlike features,J. Geophys.Res., 91, 421-434, 1986. Bryan, P., and R. G. Gordon,Rotation of the ColoradoPlateau: An Merrill, R. K. and T. L. Pewe, Late Cenozoicgeology of the White analysisof palcomagneticdata, Tectonics,5, 661-667, 1986. Mountains,Arizona, Spec.Pap. 1, 65 pp., Ariz. Bur. of Geol. and Bryan,P., and R. G. Gordon,Rotation of the ColoradoPlateau: An Miner. Technol., Tuson, Ariz., 1976. updatedanalysis of palcomagneticpoles, Geophys.Res. Lett., 17, Moore, R. B. and E. W. Wolfe, Geologic map of the easternSan 1501-1504, 1990. Franciscovolcanic field, Arizona, scale 1:50,000, U.S. Geol. Surv. Condit,C. D., Field, petrologicand petrochemicaldata for the western Misc. Invest. Ser. Map, MI-953, 1976. Springerville-Show Low volcanic field, east-central Arizona Moore, R. B. and E. W. Wolfe, Geologicmap of the eastpart of the (abstract),Eos Trans. AGU, 64, 689, 1984. San Francisco volcanic field, north-central Arizona, scale Condit,C. D., L. S. Cmmpler,J. C. Aubele, andW. E. Elston,Patterns 1:50,000,U.S. Geol. Surv.Misc. Field Stud.Map, MF-1960, 1987. of volcanismalong the southernmargin of the ColoradoPlateau: Nakamura, K., Volcanoes as possible indicators of tectonic stress The SpringervilleField, J. Geophys.Res., 94, 7975-7986, 1989. orientation-Principlesand proposal,J. Volcanol. Geotherm.Res., Condit,C. D., L. S. Cramplet,J. C. Aubele,J. L. Cooper,L. L. Brown, 2, 1-6, 1977. and J. Castro, Thematic geologic maps of the Springerville Nealey, L. D., Geology and petrologyof the Mount Baldy trachytic volcanic field, east-central Arizona, scale 1:100,000, U.S. Geol. volcanic complex, White Mountains volcanic field, Apache and Sur. Misc. Invest. Ser. Map, in press,1992. Navajo counties,Arizona, Ph.D. dissertation,320 pp., Univ. of N. Connor,C. B., Cluster analysisand two-dimensionalFourier analysis M., Albuquerque,N.M., 1989. of cinder cone distribution: Central Mexico and SE Guatemala Newhall,C. G., G. E. Ulrich, and E. W. Wolfe, Geologicmap of the (abstract),Eos Trans.AGU, 68, 1526, 1987. southwestpart of the San Francisco volcanic field, north-central Connor,C. B., Cinder cone clusteringin the TransMexicanVolcanic Arizona, scale 1:50,000, U.S. Geol. Surv. Misc. Field Stud.Map, Belt: Implicationsfor structuraland petrologicmodels, J. Geophys. MF-1958, 1987. Res., 95, 19,395-19,405, 1990. Perry, F. V., W. S. Baldridge, D. J. DePaolo, and M. Shafiqullah, Cooper,J. L., J. L. Aronson,C. D. Condit,and W. K. Hart, New K-Ar Evolutionof a magmaticsystem during continental extension: the agesof lavasfrom the ColoradoPlateau-Basin and Rangetransition Mount Taylor volcanicfield, New Mexico, J. Geol. Geophys.,95, zone, east-centralArizona, lsochron West, 55, 28-31, 1990. 19,327-19,348, 1990. Crisp, J. A., Rates of magma emplacementand volcanic output,J. Porter,S.C., Distribution,morphology, and sizefrequency of cinder Volcanol. Geotherm. Res., 20, 177-221, 1984. cones on Mauna Kea volcano, Hawaii, Geol. Soc. Am. Bull., 92, Crumpier,L. S., Alkali basalt throughtrachyte suite and volcanism, 448-456, 1972. Mesa Chivato, Mount Taylor Volcanic field, New Mexico, part I, Sarle,W. S., The clusterprocedure, in SAS User's Guide: Statistics Geol. Soc. Am. Bull., 91,253-255, 1980a. (Version5), 956 pp., SAS InstituteInc., Cary, N. C., 1985. Cmmpler, L. S., Alkali basaltthrough trachyte suite and volcanism, Settle,M., The structureand emplacementof cindercone fields, Am. J. Mesa Chivato, Mount Taylor volcanicfield, New Mexico, part II, Sci., 279, 1089-1107, 1979. Geol. Soc. Am. Bull., 91, 1293-1313, 1980b. Shoemaker,E. M., R. I. Squires,and M. J. Abrams,Bright Angel and Crumpier,L. S., Volcanismin the Mount Taylor region,Field Conf. Mesa Butte fault systemsin northernArizona, CenozoicTectonics Guide& N.M. Geol. Soc., 33, 291-298, 1982. andRegional Geophysics of the WesternCordillera, edited by R.B. Cmmpler, L. S., The Mount Taylor volcanic field, in Volcanoes of Smith, and G. P. Easton, Mem. Geol. Soc. Am., 152, 341-368, North America, editedby C. A. Wood and J. Kienle, pp. 300-302, 1978. CambridgeUniversity Press, New York, 1990. Steiner, M. B., Rotation of the Colorado Plateau, Tectonics, 5, 649- Crumpier, L. S., J. C. Aubele, and C. D. Condit, Influence of 660, 1986. Quatemarytectonic deformation on volcanismin the Springerville Tanaka, K. L., E. M. Shoemaker, G. E. Ulrich, and E. W. Wolfe, volcanic field, Colorado Plateau, USA (IAVCEI abstract),Bull. N. Migration of volcanism in the San Francisco volcanic field, M. Bur. Mines Miner. Resour.,131, 64, 1989. Arizona, Geol. Soc. Am. Bull., 97, 129-141, 1986. Delaney,P. T., D. D. Pollard, J. I. Ziony, and E. H. McKee, Field Thompson,G. A., and M. A. Zoback, Regional geophysicsof the relations between dikes and joints: Emplacementprocesses and ColoradoPlateau, Tectonophysics, 61, 149-181, 1979. palcostressanalysis, J. Geophys.Res., 91, 4920-4938, 1986. Ulrich, G. E., and N. G. Barley,Geologic map of the SP Mountainpart Fedotov,S. A., Magma ratesin feedingconduits of differentvolcanic of the San Francisco volcanic field, north-central Arizona, scale centers,J. Volcanol. Geotherm. Res., 9, 379-394, 1981. 1:50,000, U.S. Geol. Surv. Misc. Field Stud.Map, MF-1956, 1987. CONNORET AL.: CONTROL5ON VENT DISTRIBUTION,SPRINGERVILLE VOLCANIC FIELD 12,359 Ulrich, G. E. and L. D. Nealey, Structuraland eruptive patternsin the Wong, M. A., and T. Lane, The kth nearest neighbor clustering San Francisco volcanic field, northern Arizona (abstract), Geol. procedure,jr. R. Star. Soc.,Set. B, 45, 362-368., 1983. Soc.Am. Abstr. Programs,8, 641-642, 1976. Wong, M. A., and C. Schaak, Using the kth nearest neighbor Ulrich, G. E., C. D. Condit, K. J. Wenrich, E. W. Wolfe, R. F. Holm, clusteringprocedure to determinethe number of subpopulations, L. D. Nealey, F. M. Conway, J. C. Aubele, and L. S. Cmmpler, Am. Star. Assoc.1982 Proc. Star. Cornput.Sect., 40-48, 1982. Excursion 5A: Miocene to Holocene volcanism and tectonism of the Zhang, D., and T. Lutz, Structuralcontrol of igneouscomplexes and southern Colorado Plateau, Arizona, Mere. N.M. Bur. Mines. kimberlites:A new statisticalmethod, Tectonophysics,159,137- Miner. Resour., 46, 41 pp., 1989. 148, 1989. Wadge, G., and A. Cross,Quantitative methods for detectingaligned Zoback, M. L, State of stressand modern deformation of the northern points: An application to volcanic vents in the Michoacin- Basin and Range Province, jr. Geophys. Res., 94, 7105-7128, Guanajuatovolcanic field, Mexico, Geology,16, 815-818, 1988. 1989. Warren, D. H., A seismic-refractionsurvey of crustal structurein Zoback, M. L., and M.D. Zoback, State of stressin the conterminous central Arizona, Geol. Soc. Am. Bull., 80, 257-282, 1969. United States,jr. Geophys.Res., 85, 6133-6156, 1980. Wemicke, B., and G. J. Axen, On the role of isostasyin the evolution Zoback, M. L., and M.D. Zoback, Tectonic stress field of the of normal fault systems,Geology, 16, 848-851, 1988. continental United States, Geophysical Framework of the Wilson E. D., R. T. Moore, and R. T. O'Haire, Geologicmap of Navajo Continental United States, edited by L. C. Pakiser and W.D. and Apache counties, Arizona, scale 1:375000, Ariz. Bur. of Mooney, Mere. Geol. Soc. Am., 172, 523-539, 1989. Mines, Tucson, Ariz., 1960. Wilson E. D., R. T. Moore, and J. R. Cooper, Geologic map of Arizona, scale 1:500,000, Adz. Bur. of Mines and U.S. Geol. Surv., J. C. Aubele and L. S. Crumpier, Department of Geological Reston, Va., 1969. Sciences, Brown University,Providence, RI 02912. Wolfe, E. W., G. E. Ulrich, and R. B. Moore, San Franciscovolcanic C. D. Condit,Department of Geologyand Geography,University of field, Arizona, Volcano News, 13, 1-3, 1983. Massachusetts,Amherst, MA 01003. Wolfe, E. W., G. E. Ulrich, R. F. Holm, R. B. Moore, and C. G. C. B. Connor, Departmentof Geology, Florida International Newhall, Geologicmap of the centralpart of the San Francisco University,Miami, Fl 33199. volcanic field, north-central Arizona, scale 1:50,000, U.S. Geol. Surv.Misc. Field Stud.Map, MF-1959, 1987a. Wolfe, E. W., G. E. Ulrich, and C. G. NewhaH,Geologic map of the northwestpart of the San Franciscovolcanic field, north-central Arizona, scale 1:50,000, U.S. Geol. Surv. Misc. Field Stud. Map, MF-1957, 1987b. (Received June 18, 1991; Wong, M. A., A hybridclustering method for identifyinghigh density revised April 13, 1992; clusters,jr. Am. Star. Assoc., 77, 841-847, 1982. acceptedApril 23, 1992.)