t~:- ••. ,-,.,._ ..... .-' ..

I'ub!Jc ;:vt:c k concvurs fin:u;dc:r clu Conscil N;tfional des Itcchcrchcs d'Ilalic

ORGANE DE

Tome XXIX (

B. V. , :·, ., n·~.r ...

•' :·· { .J '~ ... ~ .... ~

L I::~:~;/. ·, f r·_, ~

I i,J: ,,.(;'/ '

PriHIC<1 by S1htlli.1~1E'<iO Tll'ilGll.\nco Fl;,\)>;CI:~co GLI~:-<1Xt & FH;u

Via Cistclna dell'Olio · J~apoli, It;lly 1965

I \IIIII 1\1\\1\1\11\\\1 \\Ill 1\\\ \Ill 14625

I I

The Bandelier Tuff: A Study of Ash-flow Eruption Cycles from Zoned fvfagma Chambers ~' 1

R. L. SMlTH and R. A. BAILEY

U. S. Geological Sun·cy, W~1shington, D. C.

Introduction

The Bandelier Tuff (H. T. U. S:\liTH, 1938; GHIGGS, 1964; Ross, Sl\UTH, B..\! LEY, 1961 ), Jemez Mountains of no1·thern New Mexico, U.S.A. (fig. 1 ), is a Pleistocene rl1yolitic ash-now formation consisting of two major members. In this preliminary paper the two major members arc desigm1tccl informally as the upper member of the Bm1-dclicr and the lower member of the Brmdclicr. Formal subdivision is planned for a later p8per. Each member is the product of a giant c::d­dcra-forming eruption: the eruption of tl1e lower member of the Ban­delier approximately 1.4 million years ago, was followed by subsid­ence in its sow·ce area to form the Toledo Caldera; the upper mem­ber of the Bandelier was erupted approximately 1.0 million years ago, and formed the Valles Caldera (SJ\tJTli, BAlLEY and Ross, 1961; K/Ar ages from G. B. DALRY:'>1PLE, personal communication). The upper and ]ower members of the Bandelier thus form two cycles of ash-now eruption ancl caldera form::~tion that together arc the culmination of a long history of bnsaltic and anclesitic to quartz-latitic and rhyolitic volcanism in the Jemez Mountains.

Because of the simibrity between the upper and lower members of the Bandelier, both in physical nnd chemical properties, nnd be­cause sp;~cc docs not permit mon.! than a prvliminary account, de­tails of the upper Bandelier only nrc described here. The following pages are ::1 brief summary of the distribution, stratigraphy, and pre­liminary mineralogy and clll'misti·y o[ the upper member of the Ban-

* Puhlil'ation ;Hllhori,,·d lw !he Dirt•ctnr, U. S. G,·olorkal Suney. 'P;~pe•· n·;"\ :11 th.: IAV lnicrnariun:d S~·,uposiunl on Vuk;mulogy (New Zcabnd).

sl'it'lllific !<t'ssion of Nvv. 25, 1965.

. '

~ -~ "-·~. c>. ... ·- ..... ' • pt- ... ... A<........ • ..... . \, ~ ... r~":"""" .. ," "':'--................ -.: .... ~....,..~...,....~\~-to' ........ ~ .. ~ .. .., .... 'P': ... !f...,. ..... ~~ ........ ... , •• , •c."'-~~ ... ~.-- ..... ""'_,. .. _... ____ ~-~·-~

I I l } i ~

I i j i . ;

.l

·I j l '

~

. ' ~ ..... ......_ -~---, ....,. •. -. ........... _-.;...L~ ........ -..,..._··.~ •_u..-•~·"".-....--~---v ... ~-..-~ .... ~~~'--·........,...,."'JU........,..,.;.:I-~~-...~-·-.A'f-~ .. ~~ /,.'"<.~ • . ,. --- ... ------~ .... -- ·~--...-__ .. __ ,.,..___..._, ____ .. __ .. _..._ ___ ---- -~-----·- ·-· ------ ~ -· ... ..· .. ......... -..... ~ -~~~~

.J

-84-

ddicr. The lower member of thl.! Bandelier will be referred to for significant comparison.

r-------- -------, ; I

I • I I

I ; i j

i I

I

NEW

i r----• I

L_ _ _J

o SANTA f£

MEXICO

. . t i •

I •

I •

I i I

l ________ _j

FIG. 1 - Index map of New Mexico showing location of Jemez Mountains volcanic field.

Distribution

The Bandelier Tuff forms broad plateaus that slope radially from the Jemez Mountains volcanic highland. The platcnus are dissected by numerous deep canyons radial to the Jemez Mountains, and these canyons provide many hundreds of miles of exposure of the Ban­delier (Fig. 2). Figure 3 shows the physiography o[ the !cmcz volcanic highland, the Valles Caldera, and the flanking plateaus of Bandelier Tuff. The total Jemez volcanic fidd has dimensions of about 60 miles north-south and about 40 miles cast-west.

R. L. S~tllll ;m Cyct.

r'" ,. ....... ~· .· .

. . . .. ,.,.... ........ .. ::.:., . . . ..... ~ .

~;..::· ....

r:~_.:-. l )~ l>' t. •: .. p.~ .. . · . .. ..

.;,...; ,_ .

fiG. 2 • Acrinl the c (Naci the c band!

FIG. 3 • Ph< Cal SOt

~~~~_,~tif¥..-,RF•~-.'f.4:o;¥,1~C~·'·'~~~~JIII~;,;..-::,-~-..:,.;;;:r,-~~-~.;;.o..:$-;:~·;.;~-.~ .. ~- ._.,_..,, •. :..,_.,._....-";-:""~~-~~ ... ~

red to for

1 •

t i •

' •

I t

I • I I

1 I

.J

\•olcanic field.

dially from :'! dissected , and these f thc Ban­ez Yolcanic ·Bandelier ut 60 miles

"

Pt. I

R. L. Sllnnr and R. A. RliU:Y - Tire /Jall(/dicr Tuff: A Study of Aslt·flow Emptiuu Cycles /l'om Zu11cd Ma~:mu Clwmlu:rs.

r ····.k:-.':'---· ... - .. ·:··· :;.· .. _'7· ... , : ......... .. . . . ....... . _.,

I r f.' ~-

... ~ .

.·... .. -

, . .;'.! .-..... . > ...... ~

...•. ,. .... , .... ~ .. ,,-.,,t• ... ;~:·--~·.·.( .. ; ...,~,~(·~ .. -~~~~-- .. ~,,~·· ........... .•• •1'+&-o." •

.. . · .. ~:}~:.I::'_::: 2 ::~ -;~~<~.;, :~- . ;;;- ~;-(;.,• . " ' • . i"

0 ,,

;:~; .. -~- ·::·: ·;~Kii:~t.~;.,~ ... ~-.... ~.;,-~),-·'".· ..,~ ... ·fol.~~'".·,. ... :.·~r.;:~ .. -~··.};·;~:o.'· .. · .:rii"'"'-..;.:to:.. ..... ..,.,..;;,..:;.:.,.y;..:.;.<!Z,~::..; fiG. 2 - Aerial \'icw of Banddicr Tuff cliffs on the west side of San Diego Canyon and

the dissected surface of the Jemez Plateau, ·west side of the Jemez Mtns. (Nacimiento Mount<Jins in distance). The lower 1\\0 dark and light bands in the clif[ constitute the lo\\'er member of Bandelier; uppermost dark and light bands constitute the upper member of Bandelier.

FiG. 3 • Photo:;r:1ph of a rdid mudd or the kmc7. ~1ountnins nr,·n showing the V:~lk·s C:1hlo:ra (center) and flanking plale:ms of l3anddi,·r Tuff. View ns seen from the southeast.

..

-86-

The present outcrop area of thc Bandelier ash-now sheets is about 500 squ::H:e miles. The upper member of the Bnnddier formerly covered about 830 square miles (Fig. 4} and had a volume of nbout

16t••s· I .oc···

'

0 • 40 «llf ~-<1 'L..,l_,._,.~....,, IJ.-..-~I...J' o 1 ••tu

_.,. ...

FJG. 4- Distribution of Brmt.ldicr Tuff :wound the Valles Caldera (topographic rim, solid black line; ring fracture, dashed line) and Toledo Caldcrn (topo·•raphic rim, <bsh-dot line). Outcrop of lowc1· m..:mber of the I3;~mldicr, solid" bl;tck. Maximum extent of upper member of the Bandelier, dott~:d.

50 cubic miles. The lower member of the Bandelier had npproximntdy the same distribution and volume, but pre-upper Bandelier erosion has reduced its distribution, and burial by the upper member o( the Banc.iclicr restricts its present outcrop (fig. 4).

..........

In the intt the lower men incised by rad

F

emplaced 01

and pre-BCtn variable thi( size the inn bution. Loc thickness b1

(

• sheets is ~r fonncrly c of about

_ ..,. ...

.. ~ .. _j

ogr::~phic rim, (lopogrnphic

·, solid black.

>roximatcly ier erosion 1bcr of the

,

..

I

-87-

In the interval between the upper and lower Bandelier eruptions, the )ower member of the Bandelier was extensively cro<.kd and deeply incised by radial canyons. The upper member of the J3anc.lclier was

_,. I

D 0-200 FEET

D 200-.000

0 40o-~oo

0 c;oo-ooo

r::l >800

0 •

1""1 r 'I

0

F1c. 5 • Isopach map of upper member of Bandelier Tuff.

I

_.,. _

10 ... I

lo IIIL[I

emplaced on the much dissected topography of the lower member and pre-Bandelier rocks. Simplified isopachs (Fig. 5) show the highly variable tl1ickncss of the upper Bnndclier ash-flow sheet and empha­size the inOuencc of topogr::~phy in con1rolling thickness and distri­bution. Locally it is more than 800 feet thick, the grcutcst known thickness being in the Toledo Caldera basin. ,

-88-

The ash-flow eruptions of both the upper nnd lower members of the Bandelier were pn:ctxl~d by pumkc falls and these deposits arc nearly everywhere present bent.:ath the ash-flow sheets.

Stratigraphy

The present study is based on observation nnd measurement of numerous detailed stratigraphic sections. All sections were hand leveled and all observable features recorded. The diagram of Figure 6 illustrntes some of the features noted and serves to illustr::tte the basis for construction of Figures 7, 8, ll and 12. Degree of welding was determined by porosity estimates every 5-10 feet. Kind and intensity of groundmass crystallization or absence of it were noted. Partic­ularly significant were pumice block accumulations, crystal concen­trations, and other features that mark partings between individual ash flows. « Fossil'> fumarolic pipes were found to be valuable indi­cators of flow tops. Composition and size of lithic fragments; size and mineralogy of phenocrysts; size and texture of pumice blocks; color both on fresh and weathered surfaces; and joint, weathering, and erosion characteristics were recorded. All these features con­sidered collectively provide valuable means of correlation over the entire Bandelier sheet. Typically no two sections are alike, and no single section is representative of the entire sheet. Sections were cor­related by visual field relations, by best fit of observed features, and by contouring of lines of equal porosity (degree of welding).

Reversals in porosity trends in vertical sections, common in the Bandelier, indicate different rates of welding and reveal variations in emplacement temperatures of individual ash flows or groups of ash flows. In Figure 7 four simplified sections of the upper member of the Bandelier that span about 8 miles are joined by lines of equal porosity. The uppermost stratigraphic layers have been removed by erosion, and the principal field key to correlation is the position of the zone of maximum welding. Locally this zone is composed of at lc~st six successive ash flows and can be traced for 12 miles in con­tinuous outcrop. Figure 8, n greatly simplified cross section, spans a distance of 15 miles radially from the VaiiL!s Caldera and is con­structed from 5 measured sections (4 of which arc shown in Fig. 7) and inter-section inspection. The porosity contours, besides aiding in correlation of measured. sections, s}ww how wddin~ dccrc;:~sq away

R. L. S!~IITH nntl Cydes

POROSITY

(WELDI!~G)

40~;..:..·

401------­~o~,..,..,.. 20 : ... 10

J 20l ... '30 •,. '.·

40'~===1 45~

~ B,....,.,,...-.---1 ~o~,.,.,~

30iL~~ 40 :."'_._ ··'.·

45

FIG. 6 • A S!ll\1( show in cryst;tll ings ar

!T members !Sc deposits ls.

;urcment of were hand of Figure 6

1te the basis t,.•elding was nd intensity )ted. Part ic­stal concen-1 indiddual 1luable indi­ments: size 1ice blocks; weathering,

!atures con­:m a\'er the ike, and no ns were cor­eaturcs, and tg) . . 1mon in the :ariations in oups of ash member of

1es of equal remoYed by position of

1posed of at llilcs in con­ion, sp3ns a and is con­·n in rig. 7) ics aiding in n:ascs away

R. L. Swnr nne! R. A. D\IU!Y - The /Jwul<!lier Tuff: A Study of Ash-flow Empticm C>•cles /rum Zu11cd /Hagma Clwm/Jers.

POROSITY (WELD!I·!G)

401--_:....o..~

401---'--~

40 ~.---.;.... 45

40f------~g _-_-:-~ 10

10 20 30~'---'--Y

401-=::=::1 451-

· ...... :-.·. 40 45·1---~

···.

CRYSTALLI· ZATION

PARTINGS e.

FUM.t1ROLES

\'f f' I,\

r·wtt

.. ---- ...

•

JOINTIIJG FEET I I

IIIII 1111111

I I l\ I \I

500

400

l 300

I~ !111

I I ! l 11!:, d

'

200

100

0

FJc. 6 ·A sample stratigrnphic l>l'Ction of the Uf>fK'r memht•r of the Bnn<klk•r Tuff. showing r~·conled t:hangt•S in dc'!)J"l'C of welding (percent porusit~·, shndt:d), crystallization. (\':1Por r>h~1sc. dashnl; devitrifk;llion, dotted) position of part· ings und fum<lrulc~. nnd spacing uf joints.

rl

.• t

1 l ~

j 1 ~ + J ·~ ) :i 1 ';.-<

~·

·~ ,1

' ..

;j ' ~ . ',

• 1

., ·: ! l I

·( ., 'j ; '!

{

l

\ l } i • I j •

-.......... -~ ...... ~ ~ ........ """"' .......... , .. · .. ~. :·~- ·-·. -~·-

~~

45------

J-1

····~ ZONES OF WELDING

30-40 20•30 10-20 J•2 ~

> 45 "to POROSITY

40-45

.... ": . •• ~·f·· .::: .• ~ .. ,;'.J .; 10 ~ l..:

' ' ' ' , . ... -- ir;· r-1 ·.--·-:· .. ,.~)~

~~~~~;;~~~ ,, ,, ... ; .. ·r·' .. , :t.~:-~~\t.t~

J•4

ZONES OF CRYSTALLIZATION,

~ O(VITRIFICATION

~VAPOR· PHASE CRYSTALLIZATION

1------l PARTINGS

~ BASAL PUMICE FALL·

FEET

300

200

100

0

FIG. 7 • Correlation diagram of stratigraphic sections of upper member of the Bandelier Tuff in San Diego Canyon, Jemez Pia· teau, showing projected porosity contours and the relation of zones of crystallization to zones of welding.

It

H ' r. -i !•

~ ~ . ~ (':'

d. I .· I

r- ' I C/'1 :: l ....

l"J::

11 ~tl i>:::l :,.':'C..

.. -)l'j 2' ::::> [t t;:o:l : l - :: ~;;; -.-: . ~I

. I •

~;_j

!t :::. ~

l")t;;l ;:;-!:>

~~ ~~ ~ ~· ~ .,

!f ::;! :::::: .. ;l:..

~ j.

.:; jf

0 lf~ -;l:.. "' ;:;-

f -::: ~ \':1 ., k .§ :.? - . •

s· ::: - -

- --------------'f -~ ----·- - t

I II W\ l \

t-4 \ M \ ~\1-1 I ' •

\ \\ \~. \\ \ \V~ ,.._

\\ \ \. "l'l

~ '

/~~

l" r (/1 ~

f

t f

I J J i 1 ~

f ' ! i

j

! ~ i I i ~ i 4

l

-1 1 1

i l

I l 1 J I

I, t .!

R. L. S~I!Tfl and R. A. D.\tu;y The lJcmiiL'licr Til/[: A Slltcly of Asli.flow Enrplion Cycles from Zoned Ma~ma Cluwrbr:rs.

;-

;. i.

-.

I

> 45 % POR0SI1"Y 40•40 30-40 ~0-30.

10· 20 < 10

ZONES OF

ZONES OF CRYSTALLIZATION

OEVITRIFICATION 8 ' ' VAP0/1· PHASE CRYSTAL· LlZATION

VERTICAL EXAGGERATION 20 X

0

I 0

-·---

I 2

5 KM I 1 3 MILES

TRUE SCALE

FIG. 8 • Cro:;s s.:c:tion sho\\'ing zon<•s of welding and crystallization in lh.: upp.:l' mcml,cr of D<tnddit•r Tuff in Sa Diego C:tnyon, Jcmc-7. Plalt.:~tu. Roman numentls n:fer to slmtigraphio.: subunirs shown in Fiuurc 9.

J

o-

-91

from the source area even though the total cooling unit thickcns. This Jatcral decrease in welding is related to heat loss during em­placement of the ash flows. The heat loss toward the distal end of th~ sheet is also confirmed by the decrease in thickness of the zone ~f crystallization rcbtivc to the thickness of the cooling unit. TJ1is change in. zone thickness is related to the total heat budget of the cooling unit rather than to individual ash flows as shown_by the fact that the lower zonc-bouncbry between crystalline and noncrystalline tuff steadily rises in the cooling unit toward its distal end and crosses partings between individual ash flo\vs.

To understand the stratigraphy of either the upper or lower members of the Bandelier it is necessary to visualize the eruption and emplacement of the ash-flow materials from a volcanic highland onto a much dissected lovdand. The early ash flows were confined to the valleys and basins, successive flows filled the valleys, and the later ash flows merged to form continuous sheets from which scat­tered mountain peaks protrude.

In general the thickest sections are most likely. to reveal the earliest ash flows, although the location of lowland areas and the assymctrical distribution of successive flO\vs from the highland source vents combine to complicate the flow sequences and create gaps in any given section where certain flow units or their correlatives are missing.

Because of many factors, it is impossible or impractical to cor­relate individual ash !lows over long distances Consequently other means were devised to frame a useful stratigraphy.

Correlation of nearly 100 ddailed stratigraphic sections has per­mitted recognition of fi~tigr_<mbi~subunil~-}Yith.l.!L~~r Ef~mb.cJ:-.nLthe...JiaudclieL.l.E!~- These subunits are made up of groups of ash flows and are iden!ifi<:9.J?YJ.hs.~~~nt­temperature characteristics and. by their .. m,inerglpgy. The subunits are not c\'ci.Y~T1al~ply·'-·dcfined~b~Ttl;cy represent str<~tigraphic zones that can be correlated over the entire outcrop arcf!. The pumice l Pu--rl.A-(C.

fall layer is present at the base of the upper Bandelier except where J --;u,-\ . ·: .,l •·

the underlying topography was too steep for lodgment. Subunits I, II, and III contain prominent phenocrysts of quartz

and sodic sanidinc <~nd are recognizable in the field as a progrcssiyc temperature sequence. Subunit I is composed of ash flows th:1t nm:ly T'" LJ ... :.J.. show welding or groundmass crysl<lllization even where overbin by several hundred feet of section. This subunit occurs in lower parts

..

,\

-92-

of all major valleys filled by the upper member of the Bandelier. Subunit II is composed of ash flows that typically show partial weld­ing and vapor-phase crystallization, bliLI19cl:ens£ \Y9J9ing. $J,t_b_tmiUII contains a dcn,sely )ycJdc_d apd deyit_rifi~d zone over a large P.QX1ipn of the outcrop area, although in its distal portions, where it is thin, or where it has chilled against buried topographic highs, it physically resembles subunits I or II. Subunit V, the topmost unit, is charac­terized b)' large zoned pl~;nocrj;sts of anorthoclase rather than of sodic s;nfdine. -- . .

VOLUME Ml, SUB-UNIT

~:.~fF:~:;-.~:::::}·:=~:::~·::·:::·::-:~::.?~:::::::·,.;ii;_:;:<::;:£~:;·~~:.:-;.:~~:::·:;:~::-.;:.::. :: _::(=:~:-:-::;::::::·:~:·£:::~~-~~:~-~~·:;:;;.!~~-~:·~:·;~:~:~:~:·q~:i:·;~::;r:~-::::::-::~.;~s::

. 15 • • • • • • • • • • V .(Anorthoclasa) • • • • • · : ·_. :;_ .:;_ ·~ ~ ~. _:_ _:_, ~ :.:_ ~ ~ + -7 ·_; ~ ._; _:_.· .. 4 • • • • · IV· (Transitional) • • • , • . •

~~~~~~~~~~~~~7~~-.. • 15 •• •• •• • • · • • 'm (Na Sanidine) • • • •

FIG. 9 • Schematic stratigraphic diagram of the upper member of the B<~ndclicr Tuff.

Subunit IV, designated th,t::; ~Jn:tnsition unit_,,, js most critical. It is a zone of mixing where anorthoclase and hypersthene-bearing pumice and sanidine and fayalitc-bearing pumice occur togcth_cr. It is also a zone of abrupt upward decrease in modal quartz. Subunit IV is not equally distributed and in some areas is not present or is not recognized. Four of the subunits are represcntt:d in the exposure shown in Figure 10. Subunit V has been removed by erosion at this locality.

Topographic restriction of subunits I and II, and the wide dis­tribution cf subunit III nre well illustrated by the cross section of Figure 11. As a consequence of diff crentinl comJX1Ction of subunits II and III and the limited distribution of subunit IV, subunit V rests directly on subunit III at the eastern edge of the cross section. To the west, ash flows of subunit IV occur in nornnd succession between subunits III and V (Fig. 11).

The increasing temperature sequence of subunits I-III is well illustrated by Figure 1 I. The lowermost portions of the cooling unit

R. L. S:.tllll nne Cycl<'

r.-- . I

v"~~ .......... ...... -

:. _ .... . ...

FIG. 10 • A 551: Jcmc abO\·

:! Bandelier. 1artial weld-Subunit III

1rgc portion ;c it is thin, it physically l, is charac­.han of sodic

Bandelier Tuff.

1ost critical. bene-bearing together. It

. Subunit IV :nt or is not he exposure :>sion at this

he wide dis­;s section of - subunits II unit V rests ; section. To sion bct\\'ecn

1-III is well cooling unit

-93-

Pt.. V

R. L. SMITII and R. A. B.11U:Y - Tile !Jwu!dicr Tuff: A S111dy of Ash-flow EmJl[iun Cycles from Zonal M(lgllla Clwmbers.

. r . '.!- .·.:"1 . :.) ..

:;

FIG. JO - A 550-foot cliff in upper member of the Bandelier Tuff, C;mon de Ia Can:1da, Jemez. Plateau. Dark band in middle of cliH is subunit III; subunit IV is above, and I nnd ll below.

•. •.

-94-

are nonwcldcd and noncrystalline and all sections show an upward increase in degree of welding (as shown by d<.:crcasc in porosity) into subunit Ill. This sequence is the reverse of that in a simple cooling unit C!11placcd (lt uniform tcmp(!lilJU.J"_f,? (S!vllTII, 1960). Re­versals in porosity (welding) vertically in subunit II show the .Y~r­iability ofemplaccn1ent tcmi?crature in successive flows.

Figure 12 shows a cross section that is closely representative of the total upper Bandelier sheet. All five subunits are present. Com­plications induced by underlying topography arc shown by the abrupt appearance in subunit III of a thick zone of dense welding, and the absence of the same zone upslope as a result of the chilling effects of concealed topographic highs on the upslope portion of the subunit. The cross section of Figure 12 also confirms the emplacement tem­perature gradient previously discussed and illustrated by Figure 11.

Ncar the top of subunit III and in subunits IV and V the recog-nizable ash flmYs are thinner than those lower in the cooffi\gunit, ' 1

· · '

and thep;rtlngs -b~t'~~~~e .. ash n~~;~-.--~~~~-~9.[~-~ffi§tT~~l.-i~nes--o£ dens-e welding, separated by steep porosity gradients (Fig. 12) and by local vitric zones along partings, indicate cooling breaks of greater\ frequency and longer duration than observed in the lower part of subunit III and below. The thin zones of dense welding, occurring as they do near the top of the cooling unit, also indicate a continuation (with some reversals) of the trend toward higher temperature of sue-·) fJ .... "- ·

ccssive ash flows shown more clearly by subunits I-III.

Preliminary Mineralogy

The mineralogical studies of the upper member of the Bandelier arc incomplete, yet sufficiently advanced to illustrate significant mag­matic trends. Figure 13 shows the general nature of the information stored in a composite section similar to that depicted by the western portion of the cross section of Figure 12. Alkali feldspar ~:ompositions {determined from a-parameters, derived by computer from X-ray cHtta for heat-treated phenocrysts) are plotted against sample position in the section. The sanidinc phenocrysts show a slight but significant trend toward a more sadie composition from the air-fall pumice bnse, through subunits I-IV. Thc1·c is then a marked compositional change to strongly zoned anorthoclase in subunit V.

R. L. S:.m1 (

> ..

-IOOOH

-~ --~

- '

-o

A -Fw. l

I· l l !

·• /" i

R. L. S:\11111 and R. A. lJ,Iti.EY - Tltc llmufrlia Tuff: A Study of A.\IJ.flow Eru]Jiioll Cycles from Zo11ed Ma~:ma Cfu:mhcrs.

VERTICAL EXAGGCRATION 20 X

~ 45 '%

40•45

30-40 20-30 tO· 20

< 10

0 I

0

POROSITY

ZONES OF WELDING

I 2

B .

.

S KM I I 3 MILES

DEVITRIFICATION

VAPOR- PHASE CRYSTAL· lllATION

PL. \':

v

-o ________ _,

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

ZONES OF CRYSTALLIZATION

A B

iii~ J!)

TRUE SCALE

FIG. 11 • Eas.t·wc•st cross section of upper member of Bandelier Tuff, Jemez Platc:tu, showing zones of welding nnd crys. tallirA'll ion.

·)

' .

I ..

R. L. S~lllll <tntl R. A. nw.EY - T1u~ Dam/dicr T11{{: A Sllld)' of Asfl.[low Eruption Cycles from Zuncd Magma ClwmlN:rs.

' ' ' "' ' ' " ' '

VAPOR-PHASE

VERTICAL EXAGGERATIOii 20 X

0· I 0 ' 2

5 KM

I 3 MILES

> 45 POROS!T Y (%)

40·45 30-40 20-30 to- 20 <10

1"1.. VII

1000 f([T-

!>00-

o-

CRY ST ALLIZ AT ION

B

A ~:n.~:w>t-T'.~e=~~"J:M'!«SL*&tf.!..YJ¢.J-aee:;1;estw!~'++i¢i!"'St'••---------

F1C. 12 • Cross $L'ction of upper member of the Bandelier Tulf, Frijoles C:myon, Pajarito PlatL·au, showing zones of wdd· in.t: nnd cryst;llliz:llion.

.2 . I!, ...

. . ---~-- ... - 1; ... ~ .... ::..:...:--:- __ ~ · ·, ··•~:~.-~ .... ~ ......... .,..,,_..,"""'""'' ..,e . .._._...._._._,,...e.,..,,..,~i•""'•w·.:....· ~A;..~~.w·.-." \' 'l. "'~..,.; . .; . ._,"",;..g..;.:;,.,'""·"· ,.;.· ........ 1/'',,. •••·,.~~~~

-95-

In subunit IV both anorthoclase and sanidinc occur in bulk welded tuff, but they ha\'e not been found together in individual pum­ice blocks. Both the anorthoclase and the sanidinc in subunit IV are variable in composition (fig. 13), yet the sanidinc is generally

800

700

600

500

400

300

200

100

0

Ill • c Ill J: .., ..

r-:- C; i 0 •

Ill 1-:;:; .,

>. ., 11.

c. >. ' • J: ·~ • I , ! oo \ I I

<t! v

0 I ' I

I l f

i ¢

! ol '

I I

; 0

I 0

I

l 0

45 40 35 30 25 % OR IN ALKALI FELDSPAR

0 10

0

I \

20

Anorthoclase Unit

30 40 % PHENOCRYSTS

v

IV

Ill

II

FIG. 13 - Modal percent phenocrysts and composition of alknli fcldspnr phenocrysts in an 800-foot section of upper member of Bandelier Turr. Open circles, sodic sanidine (monoclinic); dosed circks, anorthoclase (triclinic).

more sodic than in subunit III. The evidence from the fcldspnrs in­dicates that subunit IV is a zone of mechanical mixing transitional between subunits III and V and suggests that some significant event occm-red either in the magma chmnbr.:r, in the conduits, or during the course of the eruption.

The distribution of fa_yalitc and hypersthene offers additional in­formation. raynlite is restricted to sanidinc-bcaring pumice. H~·pcr­sthcnc, on the other h;:md, although ubiquitous in :mort bocbse pum· icc, also occurs instead of fayalitc in sanidinc-bcaring pumice, in some outcrops of subunit IV (and possibly III). Thus the changc

...

-96-

from fayalitc to hypersthene is independent of the change from san· idine to anorthoclase and provides critical evidence for continuity of crystallization, suggesting that the entire upper Bamldicr was erupted from a single magma chamber.

Modal phenocryst content also shows significant change. The modal data (calculated porosity free) in Figure 13 were determined from single thin sections of bulk rock, hence are not rigorously rcp­rescntntivc. Nevertheless the changes are sufficiently large to be significant. Total phenocrysts increase from about 5 percent in the air-fall pumice to nearly 35 percent near the top of subunit III, where a marked decrease to about 20 percent occurs. The total-phenocryst curve is paralleled by the quartz curve, whereas the feldspar· shows relatively steady increase with only minor variutions. The sudden decrease in total phenocrysts at the top of subunit III is probably the result of preferential resorption of quartz, although both quartz and feldspar show resorption textures. This resorption may be most plausibly explained as clue to the sudden release of pressure following remm·al of more than 20 cubic miles of overlying magma erupted as subunits I-III. Abundant field evidence indicates changes in the erup­tive pattern beginning near the top of subunit III. Partings between ash flows become more distinct; wind-blown sand layers are more frequent and thicker, and fossil fumaroles arc better deYcloped. All these criteria indicate slightly longer time intervals between eruptive pulses.

In addition to these field features, to mineral mixing, and to phe­nocryst resorption at this level, an unusually high percentage of coarse lithic inclusions occurs at the top of subunit III. These in­clusions may be the result of structural disruption during the early stages of collapse of the Valles Caldera.

The foregoing relationships suggest the following sequence of events:

1) Rapid eruption of some 20 cubic miles of magma with con­comitant lowering of tot::~l pressure in the magma chamber.

2) Preliminary sagging or partial foundering of the cauldron block and wedging of the conduits, bringing the erupt ion to a tem­porary halt. The culminating ash flows of this phase (subunit III) carry abundant lithic inclusions derived from the conduit walls.

3) Disturbance of equilibrium due to lowered pressure causing resorption of quartz and feldspar in the magm;1 chamber.

I I I

l i

l

4) Cantin ing dc\·clopmcl mamles.

S) Rcopc miles of magm

6) Final thus forming

The 50 ct equivalent to of the subunit and im-crted magma cham four chemica '

data shown i spars from t

but the tren minations. T' client invo!Yi percent crysl

As migh chemical ch; ical analyse Bandelier s} III to about and II have real as is sl there is a p decrease in (fig. 15, or shows a si1

Additic the distrib data) arc n of bulk rc well an ar 3- to 4-fol

1

:from san-continuity

dclicr was

lange. The determined >rous!y rep-arge to be :::ent in the t III, where :phenocryst spar· shows l'he sudden is probably )Oth quartz. my be most ·e following erupted as

in the crup-1gs between ·s are more \'eloped. All :en erupth·e

and to phe-rcentage of I. These in-1g the early

sequence of

m with con-ber. he cauldron •n to a tcm-~ubunit Ill) walls.

mre causing er.

j

J I

I I

-97-

4) Continued sporadic eruptions with sho1·t time breaks, allow­ing development of local disconformities, snndy partings, nnd fu­maroles.

5) Reopening of conduits and eruption of remaining lO+ cubic miles of magma (subunit V).

6) Final colbpse of the cauldron block into the magma chamber, thus forming the Valles Caldera.

Preliminary Chemistry

The 50 cubic miles of upper member of the Bandelier Tuff are equivalent to 35 cubic miles of magma. In Figure 14 total volumes of the subunits (Fig. 9) arc rcclaculated to volume percent of magma and inverted in sequence to show their relative positions in the magma chamber and their proportions as erupted. The data from four chemically analyzed feldspars agree remarkably well with X-ray data shown in Figure 13. The chemical differences between the feld­spars from the pumice fall and from subunits II and III are small, but the trend is clearly confirmed by the CaO, Ba, and Sr deter­minations. The feldspars reveal a small but significant chemical gra­dient involving 35 cubic miles of silicate magma that is less than 25 percent crystallized.

As might be expected, the foregoing mineralogical changes reflect chemical changes \\'ithin the upper member of the Bandelier. Chem­ical analyses of bulk rock samples of the upper member of the Bandelier show that SiO! decreases from about 77 percent in subunit III to about 72 percent in subunit V. (Chemical analyses of subunits I and II have not yet been made). Other differences, though small, are real as is shown by the trends in Figure 15. From subunit III to V there is a progressive decrease in total alkalies and a complimentary decrease in FeO relative to MgO. The lower member of the Bandelier (Fig. 15, open circles) has similar bulk chemical characteristics and shows a similar trend.

Additional chemical data arc shown in Figure 16. The data on the distribution of U, Th, and Nb (GOTTFIHED, et a/., unpublished data) are most complete and the curves arc based on a large number of bulk rock analyses averaged from several mensured sections as well an nnalyscs from selected samples. All three clements show a 3- to 4-fold decrease from the first to last erupted material. It is

7

-98-

important to note tha·l these clements arc not pari icularly sensitive to cooling history dcvitrification and vapor-phase crystallization or to the minor diagenetic alterations that arc recognized in the Banddicr .

% MAGMA

0 Pumice

fall

10 - I

20 - II

30

40 -

1!1

50

60

IV 70 -

80 -

v 90 -

100

.------ FELDSPAR-------,% PHENOCRYSTS

<;OR %CAO %BA %SR Q F TOTAL

43 .18 .oo&7 .0010 5 ... ---- ---------- ------ ------ ---·-- --. . -- - --------- ---------- ------ -- ... -- ----- --- --- -.-

42.7 .23 .0076 .0011 2 12 14 I I I I I I I I I

10 15 25 -- ---- --------- ----- ---- - ... -- --- --. ---

<41.6 .35 .026 .0047 10 15 25 I I I

41 I I f

.I I I I .. I I I

!:! .. I I I 111 c I I I » '0 -;:; I I Ill I lL. c Ill I I I 0::: I 111 x I I (f) I

----~--l------J~-I I I I I I I I I

15 20 35 ----- ----- ---- --- -- ---

---+--l------ 5 15 20

----- ----- ----- --- --- ---.. .. c "' Gl Ill 32.6 .82 .078 .015 5 15 20 r: u ... 0 .. r: ... 4) t: 0. 0 >. c

I 1 FJC. 14 - Summary of mincr<~logical data from suhtmils of the upper member of the

B::mtldicr Tuff, showing the n.:lath·c position of the subunits in the m:Jgma chamber nnd thdr relative magmatic voluml!s.

Although some scatter of values may be attributed to these processes, they seem not to significantly a!Tecl the major trends. The Bnndclicr, however, is very young, and ulder ash flows may show greater mod­ification resulting from prolonged diagenetic alteration.

Data for LbO, Rb20, Pb, F, Cl, and TiOt shown in Figure 16 are

! !

FIC.

L Kz.O

15- (K,O solid as F

P•m"1 fall

I

\1

m

lV

v

Pumice fall

11

Ill

IV

v

Fw. 16- Dis Sh(

ly scnsi1ivc ~ation or to ! Danc.ld icr.

::NOCRYSTS

TOTAL

5

2 14 I I I I I I 5 25

5 25 I I I I I I I I I I I I I I I I I I I I I I I I ~0 35

5 20

5 20

ncmhcr of the in the mngmn

e processes, ~ D:~nddicr,

rcatcr mod-

gurc 16 nrc

i • I I

I I ' i i

FIG.

99 FeO

~

'20

15- (K,O + Na,O} · FcO · MgO diagram for the Bandeli<"r Tuff. Upper member, solid triangles; lower member, open circles. FcO "' FcO + Fe,O, recalculated as FcO.

Pumice fall

I

11

Ill

lV v

Pumice f•U

11

Ill

IV

v " /

•"

•Pb I I I I I ,•

/

.002 .006 WT%

Nb

120 PPM

.u,o I I I I I ./ • / I

/ ./

.01

I • I

.02

•" .ct I I I I I I I I I I .......

,... I ,......- I

,."' I ..... . .1

/

•Rb20 /

/ /

/ /

.03

•TiO• I \ \ \ \ • \

.1

.04 WT%

\ \

.2 W"l"%

Ftc. 16 - Distribution of minor dements in the uppa member of thL' Bmuldk•r Tuff, shown in s•·qu.:n<:c as in the m;~gma chamh.:r.

I(P.,ll!i •• 4'!"'i . .-(\J4n; .. ;_ts.t."!;1''<$4.A'f\41(.,., .. il,4·

-100-

from 3 selected glass analyses. Although insunicicnt for· detailed discussion, they show trends consistent ·with the U, Th, anLl Nb data.

Some Preliminary Conclusions

1) The upper Bandelier ash flows were erupted in a remark­ably systematic way from a single magma chamber containing de­monstrable physical and chemical gradients.

2) At least the first 30 cubic miles of ash flows record a sys­tematic increase in emplacement temperature. The remaining 20 cubic miles probably continue the trend but documentation is more difficult.

3) The emplacement temperatures probably ranged from about 550"C to over 800"C. These approximate temperatures \Vill be discussed more fully in a later paper but are based on welding and crystalli­zation experiments (SMITH, FRIED:\IA!'-1, and L001G, 1958, and unpub­lished), phase relations, and other factors.

4) The empl<~cement temperature range seems best explained by the cooling effect of mixing with air in a vertical eruption column before the <~sh flows formed. The height of the column, and hence the amount of mixing may be related to a changing vapor pressure in the magma. The evidence implies that a falling vapor pressure is conducive to greater heat conservation, hence to successively higher emplacement temperatures during the course of the eruption. A higher vapor pressure at the top of the magma chamber is also im­plied, and whereas this condition is in agreement with some of our data, it is at \'ariance with other data. More information is needed.

5) The bulk compositions of the upper member of the Bandelier show the first erupted material to be most salic and the last erupted most mafic. In a schematic w::1.y the evidence indicates a magma chamber (Fig. 17) with the most pot<~ssic feldspar crystallizing at the top, more sodic feldspar below; fayalite crystallizing at the top, hy­persthene below; and U, Th, Nb, Pb, Li, Rb, F, and Cl concentrated upward in the liquid, Ba and Ti concentrated downward.

6) These gradients arc perhaps to be expected in the magma chmnbcr, but most significant is the fact that they can be recon­structed by means of detailed stratigraphic, miucralogical. and chem­ical studks of the ash-flow sheet.

The lower Bandelier ash-llow sheet, though Jess perf c<.~tly pre-

SCr\'Cd, sho sheet. All C'

about 400,0 and th::~t cc cycles.

The sy others like more rdim: chambers.

Ftc. 17 · Schc in t! was

Recen indicates : us to limi1 silicic volt RATTE anc LIP?\IAX, c and UI (1 and EWAR

We\~

butions tc arations ;

..

r detailed :i Nb data.

a remark­laining de-

:ord a sys­Jaining 20 m is more

'rom about ::discussed d crystalli­.nd unpub-

explained on column and hence pressure in pressure is i'ely higher ruption. A is also im­>mc of our is needed. ~Bandelier

1st erupted a magma

zing at the he top, hy­mcentratcd

ihe magma he n.'con­

<Jnd chem-

{cctly pre-

-101-

served, shows many physical and chemical parallels with the upper sheet. All evidence indicates that the two sheets, separated in time by about 400,000 years, had comparable histories and common origins; and thaL considered together they broaden our concepts of volcanic cycles.

The systematic eruptions of the Bandelier ash-flow sheets, and others like them, arc remarkable, and careful studies should lead to more refined knowledge of the processes actually at work in magma chambers.

U,Rb,F,CI Fayalite K Feldspar

Ba,Ti Hypersthene Na Feldspar

FJG. 17 - Schematic diagram showing gradations in composition of crvstals and liquid in the magma chamber from which the uppt!r member of the Bandelier Tuff was derived.

Recent work published or in progress by a number of workers indicates similarities among these silicic magmas that may permit us to limit the number of hypotheses for the origin and e\·olution of silicic volcanic rocks. Of the current work we should note thnt of RATTI? and STEVE:\ (J 964) in the Snn Juan Mountains of Colorado; LIPi.1AN, CillUSTI,\:>:SEN, and O'Co~:\OR (in press) in Nc\'ada; ARAII.IAKI

and Ur (1965), O:xo (1965, and persona] communication), in Japan; and EWART (1965) in New Zealand.

Aclmowlcdgcments

We wish to thank the following indh·iduals for spcciflc contri­butions to this study: Mrs. Kathryn DICKSO:\ for careful mineral sep­arations und X-ray measurements; D. B. S·t:E\\'Al~T for guidance in

- 102

our feldspar studies; Laura RE.JCllE:X for feldspar analyses; antl David GonFHJED and his associates for U, Th, and Nb data. R. G. SCHMIDT and B. C. HE,\RN, Jr., we thank for critical revic\-vs of the manuscript.

References

ARAMM<I, S., and Ul, T., 1965 - Aira a11d Ata Calderas i11 Soutl:ern Ky11slm, Japat!, and tire Related Pyroclascic Flow Deposits. Program IAV, Intcrnnt. Symposium on Vol­canolog~·. New Zealnnd, 1965 (abstract). Also this volume.

EWART, A., 1965 - Mi11cra!ogy alld Petrogenesis of tile Wlwkamam !g11imbritcs in tile Maraetai Area of tile Taupo Volcanic Zo11e, New Zealand. New Zealand Jour. Geol. and Gcophys. Second Special Issue, Vol. S, No. 4, p. 611·677.

GRJCGS, R. L., 1964 - Geology and Gror111d Water Resources of tile Los Alamos Area, Nell' Mexico. U. S. Gcol. Survey Water-Supply Paper, 1753, 107 p,

LJP).iAN', P., CHRISTIANS!:!-:, R. L., and o· Cox:-.:OR, J. T. (in press) - A COIII[JOSitionally zoned ash-flow slll!el ill sowheastem Nevada. U. S. Gcol. Survey Prof. Paper 524-F.

Oxo. Koji, 1965 - Geology of tire eastem pari of Aso Caldera, central Ky11slw, soutlnvest Japa11. (Jap. with English abstract), Jour. Ceo!. Soc. Japan, V. 71, pp. 541-553.

R~TTt. J. C .. and STEHX, T. A., 1964 - Magmatic differentiatioll in a volcanic sequmce related /o the Creede Caldera, Colorado. Art. 131 in U. S. Ceo!. Survey Prof. Paper, 475-D, p. D49-DS3.

Ross, C. S., S:--urn, R. L .• and BAILEY, R. A .. !961 - Outline of rl~e zeology of tlze Jemez Mountains, New Mexico. Guidebook New Mexico Geol. Soc. Twelfth Field Conf., Albuquerque, N. M., 1961, p. 138·143.

S.\t!TII, II. T. U., 1938 • Tertiary geology of rlte Abiquiu quadrangle, New Mexico. Jour. Geology, V. 46, p. 933-965.

SM!Til, R. L., !960 - Zones mtd z.otzal variations irl weldetl ash flows. U. S. Geol. Survey Prof. Paper 354-F, p. 149-159.

S:-.llTII, R. L., B.\ILEY, R. A., and Ross, C. S., 1961 - Structural evolution of the Valles Caldt•ra, New A-texico, cmd its bearing on the elllp!acemcnt of ring dikes. U. S. Grol. Sun·cy Prof. Pnper, 424-D, p. 145-149.

S:o.um. R. L.. FRit:nl\IAX, I., and Lo:-.:G, W. D., 1958- Welded tuffs, Experiment 1 (abstract): Am. Ceophy. Union. Trans., V. 39, No. 3, p. 532-533.

Discussion

A. EwART: We are p~rticularly interested in the results of your study of the Ycrticnl variations within the J3amlelier Tuff. Recently, similar variations were found ami described from the Whakamaru Ignimbrite, Taupo Volcanic Zone. Thcse includctl an upw<~rd ~..·nrkhment in sodium, upward increase in total pht.'IH.>CI-yst content, quartz resorption, and upward dccn.:asc in modal pbg.iudast•/qw.lrtz mtios. These were also interpreted as of primary (pre~ eruption) origin.

I should like to ask wltcthcr there arc conccntt·;\lions of ll'nticules within

tho! ba~al porr minl!ra logical

R. L. S~tt

you found to !cntkks in tl· volcanic gbss of gbss may ob\'ious chanf rcspcct to so real cnrichm< l ween the gl<:l css might h~

but for the f; K. YAGI:

subunit? What mir

posits? R. L. Sr-n

served- that but the aver; to the next, : progressive i1

The prod' of direct crys balite and al ticalion arc which have shards.

s; and David . G. St:I!M1DT

manuscript.

shu, Japan, aml ;posium on Vol·

imbrites in t11e land Jour. Gcol.

s Alamos Area,

compositionally

of. Paper 524-F. tshu, SO!tllrll'est pp. 541-533. Jeanie s~:quence ·cy Prof. Paper,

:y of the Jemez ·111 Field Conf.,

1• }.fexico. Jour.

S. Gcol. Sun·cy

•t of tlzr? l' al/cs

1g dikes. U. S.

nt J {abstract):

,·our study of br varbtions mpo VL'lcanic d incn-;1:::c in ase in mod;~!

:wimary (pre·

tkllks within

-103-

the bnsal portions of the \';lrious now units, nnd if so, do these show different mineralogical nnd chemical characteristics from the enclosing rock?

R. L. S:..um: I think you arc referring specifically to vitric Jcnticlc:s that you found to be mot·e potassic than their mntrix. We ha\'e not found such !cnlides in the B:mddicr TufT. However, spe<1king on the general probll!ms of volcanic glasscs, we do lind in the Unircd States, that the diagenetic hydr;~tion of glass may be accompanied or followed by chemical changes. The most obvious change is loss of sodium causing rclath·e enrichment of potassium with respect to sodium. In many hydrated glasses, however, I have found .::tlso a real enrichment of potassium that I relate to low temperature reactions be­tween the glass and ground or surface waters. I would suggest that this proc­ess might have altered the original chemistry of the Whakanntru lcnticlcs, but for the fact that you have also found mineralogical differences.

K. YAGI: Is there any difference in the degree of welding within each subunit?

What minerals are formed as the products of devitrification in these de· posits? ···

· R. L. Sll.nTH: Within each subunit diiTerences in degree of welding are ob­served- that is, zones of greater welding alternate with zones of lesser welding; but the average porosity of the subunits decreases upward from one subunit to the next, suggesting a general upward increase in degree of welding and a progressive increase in emplacement temperature of the ash flows.

The products of dcvitrification in the Bandelier- specifically, the products of direct crystallization of the glass in shards and pumice- are inno~riobly cristo· halite and alkali feldspar. In the vapor phase zone these products of deYitri· llcation arc accompanied also by tridymite, alkali feldspnr, and iron oxides which have crystallized from vapors in vesicles and in interstices between shards.

.. ·~ ......... ~"!"'