JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 98, NO. E8, PAGES 15,033-15,047, AUGUST 25, 1993

Erosion of Ejecta at Meteor Crater, Arizona

JOHN A. GRANT AND PETER H. SCHULTZ

Brown University, Providence, Rhode Island

New methods for estimating erosion at Meteor Crater, Arizona, indicate that continuous ejecta deposits beyond

1/4-1/2 crater radii from the rim (0.25R-0.5R) have been lowered less than 1 m on the average. This conclusion is

based on the results of two approaches: coarsening ofunweathered ejectainto surface lag deposits and calculation of the sediment budget within a drainage basin on the ejecta. Preserved ejecta morphologies beneath thin alluvium

revealed by ground-penetrating radar provide qualitative support for the derived estimates. Although slightly great- er erosion (2-3 m) of less resistant ejecta locally has occurred, such deposits were limited in extent, particularly

beyond 0.25R-0.5R from the present rim. Subtle but preserved primary ejecta features (e.g., distal ejecta lobes and blocks) further support our estimate of minimal erosion of ejecta since the crater formed --50,000 years ago. Uncon- solidated deposits formed during other sudden extreme events (e.g., landslides) exhibit similarly low erosion over

the same time frame; the common factor is the presence of large fragments or large fragments in a matrix of finer

debris. At Meteor Crater, fluvial and eolian processes remove surrounding fines leaving behind a surface lag of coarse-grained ejecta fragments that armor surfaces and slow vertical lowering.

INTRODUCTION

Because impact craters on the Earth, Mars, and Venus are instan- taneously created with common initial morphologies, their degrada- tion can be used to infer the intensity and style of erosional processes through time. Here we develop geological methods for estimating average vertical erosion over most of the continuous ejecta around one of the largest and youngest terrestrial craters, Meteor Crater, Ari- zona. These methods assess the differential transport of ejecta frag- ments and the sediment budget within a local drainage basin on the ejecta blanket to describe erosion of the continuous ejecta located be- yond 1/4-1/2 crater radius (>0.25R-0.5R) from the present rim. By contrast, prior studies concerning erosion of other constructional landforms (lava flows) emphasized analyses of changing surface re- lief, amount of local eolian deposition, and soil maturity [e.g., Wells et al., 1985, 1987; Dohrenwend et al., 1986; McFadden et al., 1986,

1989]. Our methods can be applied to the study of other impact cra- ters that are in a variety of geologic, climatic, and planetary settings in order to assess quantitatively the amount and styles of degrada- tion.

Meteor Crater (350 1 ' 30" N; 11101' 15" W) was formed .--50,000 years ago [Sutton, 1985; Nishiizurni et al., 1991; Zreda et al., 1991] on the southern Colorado Plateau, 2.5 km east of Canyon Diablo. The crater is nearly 1200 m across, 180 m deep, and has a 30- to 60-m- high raised-rim (Figure 1). Previous studies of Meteor Crater pro- vided fundamental data regarding the mechanics of crater formation and ejecta emplacement [e.g., Barringer, 1905, 1910, 1914; Tilgh- man, 1905; Nininger, 1956; Shoemaker, 1960; Shoemaker and Kief- fer, 1974; Roddy et al., 1975; Roddy, 1978]. Conclusions regarding the degradational state of the ejecta, however, vary. Several workers [Shoemaker and Kieffer, 1974; Nishiizurni et al., 1991 ] infer consid- erable erosion of ejecta based mostly on examination of steep near- rim areas. Others [Phillips et al., 1991 ] estimated only minor denu- dation of ejecta after examining large ejecta blocks farther from the rim to about •-1.0R. Such disparate estimates for erosion, however, are not necessarily contradictory but simply reflect rates and styles

Copyright 1993 by the American Geophysical Union.

Paper number 93JE01580. 0148-0227/93/93JE-01580505.00

that depend on slope and lithology [Grant and Schultz, 1990]. The following systematic assessment of reworked, transported, and re- deposited ejecta tests the hypothesis that local slopes and ejecta li- thology has controlled erosion from the steep inner rim area to the lower relief outer-continuous ejecta.

PHYSICAL SETTING

Flat-lying sedimentary rocks generally comprise the low-relief plains surrounding Meteor Crater (Figure 1) [Moore et al., 1960]. Fragments ejected from the fine-grained Permian Coconino sand- stone mark the deepest rocks excavated by the impact [Shoemaker, 1960]. The Coconino is conformably overlain by 2.7 m of the Per- mian Toroweap sandstone and 80 m of the mostly limestone/dolo- mite Permian Kaibab Formation [Shoemaker, 1960]. Approximately 9 m of the lithologically diverse Triassic Moenkopi Formation un- conformably caps the sequence (Figure 1) [Shoemaker, 1960]. More detailed descriptions of these rocks are found in McKee [ 1934, 1938, 1945, 1954], Shoemaker and Kieffer [ 1974], Roddy et al. [ 1975], and Roddy [ 1978]. Estimated volumes of Coconino/Toroweap, Kaibab, and Moenkopi excavated during crater formation vary [Roddy et al., 1975; Croft, 1980] (Table 1). For brevity, fragments from the Coco- nino/Toroweap, Kaibab, and Moenkopi are referred to as Coconino, Kaibab, and Moenkopi ejecta, respectively.

A continuous ejecta blanket derived from all excavated litholo- gies surrounds the crater out to ~ 1 km from the rim crest and was capped by a thin layer of fallback ejecta (Figure 1) [Shoemaker, 1960,1987; Shoemaker and Kieffer, 1974; Roddy et al., 1975]. Ejecta near the rim rapidly thin outward following a steep power-law decay [Garvin et al., 1989] with radial to subradial troughs and ridges simi- lar to the hummocky ejecta around simple pristine craters on the Moon and Mars [Schultz, 1976; Mutch et al., 1976; Carr, 1981 ]. Ex- posed outer-continuous ejecta can extend from -0.5R-2.0R and are mostly comprised of Kaibab. Primary or reworked Coconino ejecta are rare beyond -0.5R, even in long-term depositional traps.

Soils on the ejecta possess a carbonate Bk horizon between stage II and stage IV of development [Gile et al., 1966; Macbette, 1985]. Such a range of features (e.g., carbonate laminae 0.1-1.0 cm thick, thick coatings on clasts that sometimes coalesce into massive and cemented accumulations) probably reflects the carbonate-rich na- ture of the parent ejecta and complicates age determination. Most al- luvium at the crater, however, is mantled by soils correlating with the

15,033

15,o34 GRANT AND SCHULTZ: EROSION OF EJECTA AT METEOR CRATER, ARIZONA

Kaibab

Moenkopi

Canyon

Museum

C hav-ez Pass R o a d '•

ß '-A p proxima te

Ejecta Limit

ß •. •' "-•::•;;t':•.... .......... -T- Ranch .....

.... .:...:....:..::.......'::.... ,..

Fig. 1. Air photograph of the 1200 m diameter Meteor Crater and surrounding region in north central Arizona (approximate location (350 1' 30" N' 111 ø 1' 15" W). The crater is located 2.5 km east of Canyon Diablo. The approximate limit of the continuous ejecta (dotted line) and contacts (dashed lines) between the Permian Kaibab Formation (light) and the Triassic Moenkopi Formation (dark and immediately around the crater) are indicated.

late Pleistocene Jeddito Formation and the Holocene Tsegi and Naha Formations in the Hopi Buttes region -80 km to the northeast [Shoe- maker, 1960; Shoemaker and Kieffer, 1974].

Radial gullies incise near-rim and upper flank portions of the continuous ejecta and deposit thin alluvial fans and diffuse drainages beyond -0.6R [Shoemaker, 1960; Shoemaker and Kieffer, 1974; Roddy et al., 1975] that locally bury outer-continuous ejecta (Fig- ures 2 and 3). Confined fluvial activity is limited to erosion of gully floors and deposition in small, distal fans. On a broader scale, Can- yon Diablo and circumcrater drainages are separated by a low divide -1.5 km west of the crater (Figure 2).

The presently arid climate at Meteor Crater is characterized by strong (>30 m/s), southwesterly winds [Green and Sellers, 1964; U.S. Department of Commerce, 1968] that formed a patchy winds- treak extending northeast of the crater (Figure 1). This prevailing wind may have persisted since crater formation [Breed et al., 1984]. Biostratigraphic correlations between crater-floor lake sediments and other southwestern lakes [Forester, 1987], however, suggest that other aspects of climate varied considerably over the crater's 50,000-year history. Wetter-than-present conditions existed during the late Wisconsin pluvial from 25-20 to 15-12 ka (thousands of years before present [e.g., Morrison and Frye, 1965; Smith and Street-Perrott, 1983; Street-Perrott and Harrison, 1985; Van De-

vender et al., 1987; Benson and Thompson, 1987; COHMAP, 1988]). An apparent absence of crater floor lacustrine sediments >25-30 ka (R. Forester, personal communication, 1983) implies precipitation at 50-30 ka averaged less than during the pluvial, but possibly exceed-

ed current values [e.g.,Morrison and Frye, 1965; Mifflin and Wheat, 1979; Smith, 1979]. Although there is a dearth of postpluvial crater lakebeds (R. Forester, personal communication, 1983) increased monsoonal precipitation may have occurred during the late Pleisto- cene/early Holocene [e.g., Van Devender, 1990; SpauMing and Graumlich, 1986; Betancourt, 1987]. Conditions since the early Ho- locene, however, averaged drier than present [e.g., Van Devender et al., 1987; Wells et al., 1985; Forester, 1987].

PREVIOUS ESTIMATES OF EROSION

Previous conclusions regarding the preservation state of ejecta vary (Table 2). For example, Nishiizumi et al. [ 1991 ] estimated con- siderable erosion of primarily Coconino and fallback ejecta with missing thicknesses decreasing from 12-20 m near the rim to 3-5 m at 1.0R [Shoemaker and Kieffer, 1974]. Their estimates are based largely on observations of the steep near-rim and the assumption that uniform exposure ages of large, more-distal ejecta blocks reflected rapid exhumation during early crater history. Similarly, Roddy et al. [ 1975] and Roddy [ 1978] estimated that 20-25% of the original ejec- ta volume was eroded based on the difference between predicted "preerosion" crater profiles and present thicknesses established by extensive drilling around the crater. They estimated that erosion had widened the rim by 30 m while lowering it by 20 m; farther out, ero- sion decreased to -2-3 m.

In contrast, Phillips et al. [ 1991 ] found that ejecta blocks between the rim and -1.0R possessed exposure ages matching the crater age and concluded that these blocks had never been buried. Relief of

GRANT AND SCHULTZ: EROSION OF EJECTA AT METEOR CRATER, ARIZONA 15,035

Meteor Crater Drainage Density

• • / ..... ,::.'...'•.:.':::.'.::..,::,' .... ;.. ,

'"'•"';':'/'""'?""'"'"'••'"'"";'"';"::;;'• Mete or C r a ,er R d '1 " I ß • ..... ..::.:..?' ','•:'.::,. :'•:. .'.:; /_.:' ,•-_ .>:.7. ....... :",<:.-1•:• .."::, .,,•.:::/-'-•.::' ,.: • .....'. =.:.v..'.-..'.'•:•o,.. ..... ß . :.::.".::::.:' /..-'i:'::-½" .' .... ,:, , 1,5, .:.....?..:, ,

•,e •" , .:Z.'.':./ :: ' ' ..,.:.';'"' ß ,._'.:.. ..... 2-'-/ ,

,,-'t:,....'..:.,?;;.•'•" •?.'•' • '::: .... ..'..:::.-.,:.:'•'. ...-, ß ........ . • ' • '":":"•" .... '•_i ' • ,,/ ,:-.".'?:"ii!:-: •-- .:i.

• '-8: i:::.::!:i:: !

j- ß ,•:,.:... /\ .:.:. ß %: ...... ,._,,..;•:.'•:.:.'-•: -::.'. •... ] <';':' ,':".:: • .:,0 • <':":':'• /::t:::

I '"' '::ii::.%.":::;.':.-..:l{..,,... • •iii: ...... 'g;iZ•?::: ...... _/' ..?.? I I ::::::::::::::::::::::::::::::::::::::::::: .•:• ":;i:::;;::::: .• ! / :::-:.:.•; ß :.-.v.v.v.--....v.v.v.,.--.v...: .; ,.: :..', / /" ..".:•.:/ I -.: .............. .",$•.'.:.: ....... :.:.:.:.:,...•:.: :. Chavez ::::::::::::::::::::::::::: .... ::;::::::'• i[ / ..:..i• ?

•/ Pass Rd .... •}!i::.;•: ....... :'¾.•Ji•O.... ?'• \ • ! :5!.?..? - •P -':':::::':!::::•:!:? • ::::i:!:i:i:i:i:•/ • :.;i:i:;.':' x, ....'.:\:..\:; • "-i:!:!:i:i:;.i:i:i:i!:

Density (0.6R beyond rim crest) 8.6 km/km 2

• Incised Gullies • Flow Direction

:.-'.• Diffuse Drainage/Fan [• Drainage Divide

(:-[.-) 35ø1'30",111ø1'15"

0 1 km

AMOUNTS OF EJECTA EROSION

Approach of Current Study

Quantitative estimates of the amount of ejecta eroded (beyond 0.25R-0.5R) in this study are based, first, on coarsening of surface lag deposits on the ejecta and, second, on the sediment budget in a drainage basin on the western crater flank. The first approach com- pares the coarse-grained ejecta lag deposits with those of underlying in situ ejecta. Ejecta lags are defined as surface accumulations of coarse, angular to slightly rounded fragments averaging -2-3 cm thick (excluding large blocks) that directly overlay their parent mate- rial. Such lags do not exhibit the underlying silty, vesicular, clast- poor horizon typical of most stone pavements in the arid southwest [McFadden et al., 1987]. Lags can be used to accurately estimate erosion if (1) unweathered ejecta grain size properties are fairly uni- form; (2) lag formation reflects predominantly vertical erosion; and (3) in situ weathering is accounted for. When these conditions apply, the eroded ejecta thickness (Do, in cm), is approximated by

Do = (CllCo)D1 (1)

where Cl is the maximum possible concentration of coarse frag- ments in lag deposits relative to unweathered ejecta, Co (based on a normalized sample percentage determined from the standard devi- ation of grain size distributions), and Dl is the average thickness (in cm) of the lag-layer being considered (Figure 4). The second ap- proach assesses the sediment budget of alluvial/colluvial deposits in a largely enclosed drainage basin west of the crater. This method yields values for average local erosion when the total volume of missing sediment is divided by the area of the drainage basin. Addi- tional corrections are required, however, in order to account for sedi- ment transport from the basin through minor divide breaches, eolian deflation, and vertical dissolution.

TABLE 1. Estimated Volumes of Rocks Excavated

During Crater Formation

Fig. 2. Drainage map of gullies and intermittent streams eroded into the ejec- ta and surrounding plains. Also mapped are areas of diffuse drainage, major drainage divides, and flow directions. Note the pronounced southwest to northeast orientation of the precrater systems. Gullies on the ejecta have a drainage density of 8.6 km/km 2 and are generally less than 1-2 m deep; how- ever, a few gullies incising Coconino ejecta south of the crater are 3-4 m deep. Divides between individual basins are generally defined by topograph- ic variations in the primary ejecta, and basins possess an average relief ratio of 0.07. To the south of the crater, one of the precrater drainages has been in- terrupted by the crater/ejecta and diverted into the adjacent precrater fluvial network to the southeast. The letter B marks the semienclosed drainage basin in Figure 9, whose northern and southern divides are mostly defined by low, subradial ejecta ridges except where ascending an alluvial fan apex (north- em) or crossing a diffuse drainage (southern). R marks the location of the ground penetrating radar transect in Figure 11. A museum and apartments are on the north crater rim.

, , . ,

Formation Minimum, m 3 Maximum, m 3

Moenkopi 0.5 X 106 * 6.4 X 106 + Kaibab 45.6 X 106+ 49.5 X 106 *

Coconino/Toroweap 11.0 X 106 + 26.0 X 106 *

*From Roddy et al. [1975].

+From Croft [ 1980].

TABLE 2. Previous Estimates of Erosion of Ejecta at Meteor Crater

NR, m +

12-20 m

Study

Nishiizumi et al. [ 1991 ] and

Shoemaker and Kieffer [1974]

OC, m++

-3-5 m*

these blocks limited erosion of surrounding ejecta to only 1-2 m (Table 2). Phillips et al. [ 1991 ] felt that it was unlikely that: (1) all blocks were uniformly buried during impact but then exhumed by identical erosion rates; or (2) the blocks were variably buried but then simultaneously exposed by different erosion rates. Their lower estimate of erosion is similar to that reported by Grant and Schultz [ 1989, 1990] and Pilon et al. [ 1991 ], where variations in outer-con- tinuous-ejecta thickness detected by ground penetrating radar were ascribed to superposition on preimpact surface topography, rather than erosion from an assumed preerosion profile (Table 2).

Roddy [ 1978] and

Roddy et al. [1975]

Grant and Schultz [ 1989, 1990] •

Phillips et al. [ 1991 ] •

Pilon et al. [ 1991 ] • , ,

+NR=Near Rim (<0.5R).

++OC=Outer-Continuous (>0.5R).

*Inferred from data presented in published results.

up to 20 m

-0.5-0.6

<1-2 *

<2-3 *

15,036 GRANT AND SCHULTZ: EROSION OF EJECTA AT METEOR CRATER, ARIZONA

SELECTED CROSS-SECTIONS FROM WEST OF CRATER

N4

(B)

5E [CD 845W N45W lB} S45E

,.•,,.•. .. .• ,.'.• .... .• ...•. '• ':. :H:..•, ....,•..'.•,:.....,•. •.. ß .... ..•/'., •, ....

i I i i 15m 0 5 m

Diffuse Drainage N20W S20E •o??•.•, .•..... :•:•- T ...... D i f f u s e D r a i n a g e

Dis t a I F a n ( D ) i•%'o•i•'-•"q• ß •..• '•••••.•.:•7.;• • ,,, ,, ...... ß •'.;'•.•'•a9• ...... •'.•'" a•:•.'.: ' '.•a..9• .............. •:• ß t•' -•'

Ejecta 10x Vertical Exaggeration

Fi•. 3. Selected cross sections through diffuse drainage deposits, distal alluvial fans, and thi, colluviu• i,side the semienclosed basin. The location of each cross section is indicated in Figure 9. F•s and diffuse drainages •e typicMly • 1.•1.5 m •d 0.3-0.? m thick, respectively (Figure 2). Locally thicker f•s (up to 3 m thick) and diffuse drainages (up to ] .5 m thick) occur, but the lesser accumula- tio,s •e much more common. In 8eneral, the thicknesses of fans •d diffuse drainaBes decrease distally. Unit thicknesses were ly constrained usin• pits excavated t•oush to the underlyin8 ejecta. Burled colluvium beneath the diffuse drainage (•enerally not shown) is less than -20 cm thick.

(c)

Equation Used For Erosion Estimates Based On Grain Size Comparisons

Vertical Erosion (Do)=(Ci/Co)(Di)

Eroded Ejecta Thickness (Do)

-• ..... ,• ..... ..... v...,,, ........ •.•o.,......,- Lag Deposit : ,.;o:M .... Q':..O'"' :....-o:.•,... ....... .... ;.:•.. .......•.:a-...'...,.• ...... .•-...'•:'.•'.o .... '•, -- ,, .,,Lag Thickness (Di) • Coarse Fragment •'"&'.':•' •'.R;:•:"' :o:.'.. •', .... : ..:.•: .. :....'.;•.'.0.': 0.•..'•'.• :.',:O"-•:'i•;i'•.', •',-•::• •' .. ..... :•.....•..•.,...• ...•',. •.. •.,;. ,. ..... .•..•.'r•O'.½:•' .-.'•:o"."_.•:.;•O.A•;• Concentration ß '.•.0 .... 0'.•.½-.... "'6' '"-' ........... '" :':.•..:• •'._'Z•z. :.9';".' ",, ",..".;:ø.•:•.:.'<.'b'.'.'•

:•xo.-n'-........•,..h..•;'_•.::•...:o..•': .':':...'•. r .... .o 0 ] 'o . '.•'.'E[-:' .•: .... "•5:..:i.•.. ß .'. .... :.0"'..':0..?'o', -,.'"!•:' .'.":c•' •....o.•'.o...•-:.-d...,•.:•.0:•:..½.j. :&_o.'.?&q ..%',-,-':•:.: ", ..... '.:•. :• .... ß .' '.,•'.' :•.'a'.' "•..•.:-:,•.:0':'0:,.•.'. '..'.ø:'.%,v •P•.•.•:•!.'.'&9.: Q..:•...:...•:...<):•:•%t

•..:o:: .%.v...•:;4 .'o•, :_..•:•....•u.;0.:•..:,•'.•iq [•'.'•:'..½•:•:.•'::•L,::•2;•:...."•:'/.<•':'.";'•:..•'•i'i! t v,.o...'..•:.q..'.'•.-•.,k.e?..-•:•:•.x.: ':4•'" :.o' ' ' .:..? .' .' ß -•:•.': -- ß ß ';.• ."•z ß ß .... ..o..,•.•:o:..•,'.•.iS:o:O,.: :•... ,..•,.•: r:.',".:• .'(•. •,'. •:. n .,• .uf. •::•::• .%.kg.h' ',_85 :, '..•.l

Unweathered Electa • Unweathered Electa Coarse Fragment Concentration (Co) Coarse Fragment Concentration (Co)

Initial Ejecta Thickness Present Ejecta Thickness

Time = 0 Time = 1

Fi E. 4. Idealized block diaEram identifying terms used in calculation of vertical erosion from surface ]aE deposits versus subsurface cjecta •rain size comparisons. The equation does not include corrections for losses owinE to lateral transport or other lag evolution processes that are discussed in the text. Block diaErams are not to scale, and la• deposit thickness is exa88erated.

The following discussion details these two methodsß Resultant erosion estimates are then compared with data from ground penetrat- ing radar transects through alluvium on the ejecta, the scale of pre- served primary-ejecta features, and erosion of other catastrophic, unconsolidated landforms.

Surface Versus Subsurface Grain Sizes

A total of 110 samples (Figure 5) of subsurface ejecta, ejecta-lag deposits, alluvium, colluvium, and windstreak sediments were sieved using standard methods [Folk, 1980] and screen-mesh sizes of 16 mm (-4.0 phi) to 0.03 mm (5.0 phi). Subsurface samples (28)

generally from below 30 cm depths were used to characterize the un- weathered ejecta grain size distribution and then compared with ejecta lags from the surface in the same locations (Figure 5). Ejecta fragments coarser than --.3 cm required in-field sieving of ejecta ex- cavated from pits (Figure 5). Still larger blocks (>20 cm) involved measuring the block number/area in the walls of large pits and the block number/area observed along a gully incising the crater flank (Figure 5). The measured number/area clearly reflects in situ accu- mulation following fluvial erosion of fines because fans downstream of this and other gullies are devoid of blocks > 15-20 cm in diameter (Figure 6).

GRANT AND SCHULTZ: EROSION OF EJECTA AT METEOR CRATER, ARIZONA 15,037

Fig. 5. Location map marking the position from which individual samples (open circles), sample pairs (solid circles), and multiple samples from vertical transects (stars) were obtained for sieve analysis. Also marked are the positions of the excavated clast pits (triangles) and the gullies and ejecta pits/mine shafts used for block density determinations/comparisons (squares). Locations from which samples used to characterize the unweathered ejecta were obtained are marked by double circles and circled stars. Circles inside squares mark alluvial fan and diffuse drainage samples washed in a 20% HC1 solution and examined for the presence or absence of Coconino grains.

Sieve analyses. Grain sizes of the unweathered, subsurface Kai- bab (Figures 7a and 7b) and Coconino (Figure 7c) ejecta show modes at 0.074 mm (3.75 phi) and 0.21 mm (2.25 phi), respectively. All ejecta samples also exhibit modes (Figure 7c) at the smallest 0.031 mm (<5.0 phi) and largest sizes >4 mm (>-2.0 phi) that are probably artifacts of cutoff sieve diameters. The total Kaibab ejecta mass in each grain size bin is fairly uniform and has a standard devi- ation of 30-55%. The resulting contrast between subsurface ejecta samples and ejecta lags (e.g., Figure 7d) indicates that surface grains larger than 4 mm (-2 phi) are concentrated by a factor up to 6X. In situ field sieving of larger clasts reveals, however, that ejecta lag sur- faces contain a 3X concentration of coarse fragments (Figure 5) with the largest blocks (>20 cm) concentrated by only -2X (Figure 6).

The coarsening of fragments in the ejecta lags indicates only mi- nor vertical erosion. Removal of fines from only -20 cm of ejecta could produce the observed 6X concentration of coarse fragments in the surface lag measured by sieve analyses (Table 3a). Slightly great- er erosion (30 cm) is necessary to account for the 3X concentration noted during field sieving. The 2X concentration of blocks >20 cm requires 40 cm erosion in addition to that required for incisement of the associated gully (Table 3a). Hence the three size ranges yield generally comparable values of erosion; however, they do not in- clude adjustments for processes of lag evolution that might influence clast concentrations, thereby masking more significant erosion.

Evolution of coarse-grained surface lags. The evolution of most lag surfaces leading to stone pavements reflect a variety of processes including (1) deflation, (2) surface wash, (3) colluviation, (4) physi- cal/chemical weathering, (5) eolian deposition, (6) soil creep, and (7) upward clast migration. At Meteor Crater, comparison between

surface lags and subsurface ejecta properties allows assessing the relative importance of each of these processes in ejecta-lag evolu- tion.

Strong southwesterly winds at Meteor Crater deflate fine- grained ejecta and form the windstreak, thereby contributing to lag evolution as in other areas of the desert southwest [Cooke and War- ren, 1973' Ritter, 1986; Dohrenwend, 1987]. Grain sizes within the- windstreak are smaller than 0.35 mm (< 1.5 phi), but are large enough to reflect saltation transport (Figure 8a). Because Canyon Diablo is upwind, it intercepts saltating sediments from distant sources; hence, the windstreak must be largely created by deflation of ejecta around Meteor Crater. '.

Surface wash on the ejecta (i.e., unconcentrated surface runoff) also selectively removes fine fragments and could contribute locally to lag evolution [Cooke and Warren, 1973; Ritter, 1986; Dohren- wend, 1987]. The presence of thin alluvial and colluvial deposits around the crater most likely reflects the consequences of surface wash. Equations given in Ernrnett [1970, 1978] yield surface wash velocities on the outer-continuous ejecta of-* 10 cm/s (based on flow depths of 1-2 cm and gradients of 0.1-0.01) and indicate that frag- ments larger than 1-2 mm accumulate in situ. Fluvial transport of larger blocks is confined only within gullies.

Colluviation (i.e., combined effects of surface wash and local

mass movement) also causes some lateral transport of surface frag- ments [McFadden et al., 1986, 1987]. At Meteor Crater this process creates colluvium deposits up to 40 cm thick that bury outer-contin- uous ejecta in certain locations. Although constituent grain size dis- tributions (Figure 8b) generally resemble unweathered Kaibab ejec- ta, larger and smaller sizes are notably depleted (Figures 7a and 7b). Based on the maximum 40-cm thickness of the colluvium, lateral

15,038 GRANT AND SCHULTZ: EROSION OF EJECTA AT METEOR CRATER, ARIZONA

POST i M E N . •?'' .. i

m (b)

6:4 Blocks/Meter 2

..

:'.;:• -•:,,....._...•.. • .... • ....--::•..•-..' ....... . .... .•t-.:,. '-"•,.'y'&'. :- .•::i•.:½;;i:.;:::•: b.. ...•:•5:..•. •i-•: •. :;4•'- ':' '.'•'..::..• ..... ' '-'•:t"•"• '•-•--••••':'•"• .... ..• %1: ------• -:::•t.;•-::• ::•? ½": •.:•::.::½•...,•. ...• •--- :•:• • ...... ?½•?:•?.::•,•...'...: .................. '":'"'"•'•'•'•'•:•:-;..:' .-:•,. "• ...... • i i :.--..•z•½:•:.,.:•.• ................... -'-:-•:...: :.;•: ".::::.' .......... • ,,::• •,;• ,:• •- :- ......... ............. -;•,'--•::•-•.:..:.?•:• ......... ....-.:.• ................ ...... ......... •::• ::..:. ................... . ............ •:•f--'•:. '::.:.:::•'.::•' ......... • .......... ' '•"•.<;•"•.•:•; ...... ;--:•: i'E• . . . '. ' ........ :';:"--'•½'::'i•'-::'*" ::;'.'•' "?": "" :•' ':•"½'• : i• "• ":: -"::<::;'"-'-::- ........ "' '" ' ...... ;- •½:•'--•:•:•;".*'a:.':':.• .......... :•"-.•'•'". "½•. ::•>... ::• -: I[ .:.:;:;, '.. '":.'"'"*'" ,.:..•.-'::-::g½::. "4½ : .,: .; ".-' .•.-•..::;-.•g.-..:•. ', "•,•:..•.. ':'.'.)•:..• ..... .½•'i .•½-i:::.•:.:i..--:• : .: '. ..-.-.'..:.'•. '-'• ....:'" '.t

.... ß -":•::'•.:•,-::" ?•.•i.;'-..-'" '-:' '" ': '!.,•5: .... '-•?"•' ,: *'::? .... ' ½:: ': •:;'• '" '" "' ...... ¾ ':?:;•2•..•;'::'*'• .. .. . . .... -• .-.,. •,+*2•:•..t. '-%:.:: •.:• •:- *-'• .... .--•:.½•:....:-: • ...... ß ............. .•::.:•, .•:: •.' .... :.. ..... ... - •: . . -: ..... :;g•:.....:,',-½• ..< '.:.; ....................... ..,.•.- ........ ..,,.,... •.. -:......?: :..•::::-* . --...,•::.•e- --•:: .. :•-•- •isg;•.. :...• .. • ... .:•:.::• . -•-•..-- .--.., :•-•..- . , .•.%..':0 .... : ..•-%::::•:; ,. % ..... % '. •-"•-,,-•½'Z•.,: ." •"--'-.:::.•%;-'½-"-:;:::7'"•';"•'"'.•5•:'-•-• '• ........ -. •.-" :-.'.'•q ' •..::'•"'•J• ..... .:?;•%::•:':• ........ 'L;:?.•:.5:':......."-..:.::.;•;" •....,'•:......::...'.::•..: '-:•:¾:.5'.,-.. '.;.%-•...4• ..... ...:::....•

.;::,',"..•½.-:Z•:.:--'::,• ....... • .:.:•.-.. ...... .'::'•"--".•5 .... •--:.::-:'""8't'::•'::-.:".;:•,• ..... •'?•. :½-:.i:;';;•[" .'-

. ........ •,•g' .:5 '... •-?'•. '?.- ';.:.•. :e?•;,<.'"•:".•'..::.; -.. ... ?.•: --':t" .•"-'•'•--•"-•••••'-•..::.' ............... . ......... •.•::..• ...•:....::.:':•-• ' ..•:.•::; :...'5 ..... '.•.. :-..:..':.•::"'..-•:2•-.• .... '"½•'•••••'•-t•:' .... .'•':'•:.•:' ........ ....... •. :•;•,"'-'- ----•-•:•'/•:-;t*...½ ..... ½ .:. •?:?...•:.:, ::•;..%•.F• ..... :•:½•,....•--:---•; •.

.Fig. 6. Comparison between (a) the block densities per m 2 exposed in the walls of pits and mines dug into the pristine ejecta and (b) the block density exposed on the floor of a gully on the northwest-crater flank. Block numbers derived from Figure 6b required measure of the fairly uniform density of blocks in 1 m wide cross sections (normalized to cross-sectional area) along the gully. The bottom portion . of Figure 6b depicts the in situ accumulation of blocks that are too large to be transported as surrounding finer-grained material is removed. The number of blocks exposed across segments of the gully floor is only slightly larger than expected from gully incisement; hence, little erosion has occurred in addition to that required for gully incisement.

transport of fragments by colluviation correspond to vertical lower- ing of up to that value (40 cm).

Weathering (e.g., vertical dissolution, frost wedging) should sys- tematically reduce large ejecta fragments to sizes transportable by surface wash and deflation [Cooke and Warren, 1973; Ritter, 1986; Dohrenwend, 1987]. Several observations, however, limit the effec- tiveness of this process at Meteor Crater. First, dissolution of large Kaibab blocks can be constrained to less than 5 cm since their sur-

faces preserve exposure ages matching that of the crater [Phillips et al., 1991]. Second, small (-15-25 cm) unfragmented, distal ejecta blocks are preserved beyond the outer-continuous ejecta. Impact and survival at this range (up to 5.3R) require an original block size within a factor of~2X that observed [see Schultz and Gault, 1979] and limits combined physical and dissolution weathering to ~20 cm. Third, observations of weathered fragments (~0.1-0.5 m) around

very large ejecta blocks (~ 1-3 m) reveal that they can still be easily traced to their pre-weathered position. The large size of the weath- ered fragments precludes significant lateral transport by surface wash and is a more qualitative indicator of only minor weathering. Based on these observations, contributions to lag evolution by weathering processes corresponds to ~20 cm.

The remaining processes of eolian deposition, soil creep, and up- ward clast migration make negligible contributions to lag formation at Meteor Crater in the areas sampled. First, the underlying silty zone signifying eolian deposition in other settings [Wells et al., 1985; McFadden et al., 1986, 1987] is not present under the ejecta lag out- side the windstreak. Second, not only are contributions to stone pavement formation by soil creep questionable [Selby, 1982; Wells et al., 1990], but features (e.g., surface terracettes) thought to be diag- nostic of the process [Denny, 1965; Hooke, 1972, 1990; Cooke and

GRANT AND SCHULTZ: EROSION OF EJEL-'TA AT METEOR CRATER, ARIZONA 15,039

Average Near Rim Koibob Ejecto All Sample= From 0.SR or Less

(a) 40.

•}0. •0.

O.

Grain-Size (phi unit=)

50.

40.

20.

10.

Average Distal Koibob Ejecto All Samples From 0.SR or Greater

(b)

Grain-Size (phi unit=)

E

50.

Average Coconino Ejecto All Samples From 0.SR or Less

40.

20.

(c)

Grain-Size (phi units)

Fig. 7. Histogram plots of the average percent ejecta mass versus grain size for unweathered Kaibab ejecta samples from (a) less than 0.5R and (b) great- er than 0.5R. All plots present grain size in phi units (grain diameter in mm to the-log2). Modes occur at <5.0, 3.75, and >-2.0 phi in both plots. The stan- dard deviation of the mass within any grain size interval is generally 30-50% of the average interval mass. (c) Histogram plot of unweathered Coconino ejecta with modes at 5.0, 2.25, and -2.0 phi. Closer examination reveals the 2.25 phi mode reflects the breakdown of the Coconino ejecta into constituent grains. The standard deviations of mass/grain size interval are similar to those in the Kaibab ejecta. (d) A typical comparison plot of the surface-lag versus unweathered subsurface grain size distributions from the Kaibab ejecta. Fragments in the surface lag material that are coarser than -2.0 phi are concentrated 2X relative to the unweathered subsurface ejecta. (e) Grain size distribution of in situ-weathered Moenkopi Formation from beyond the limit of the continuous ejecta. Note the distinct differences between the ejec- ta grain size (Figures 7a, 7b and 7c) and the Moenkopi Formation grain size distributions.

Typical Surface vs. Subsurface Comparison Plot From Distal Kaibab œjecta 1.2 km North of Crater

40.

20. 10.

Grc:in.,-Size (phi units)

In Situ Weathered Moenkopi Formation Subsurface Sample 1.3 km North of Crater

20.

10.

(e)

ß % %

Grain-Size (phi units)

Warren, 1973] are also not found on the ejecta. And third, the fairly uniform grain size with depth indicates minimal upward clast migra- tion [Cooke and Warren, 1973; Ritter, 1986; Darrenwend, 1987] through the ejecta.

Modified erosion estimates. Ejecta lags at Meteor Crater princi- pally reflect 20-40 cm vertical lowering and in situ accumulation of coarse fragments following selective removal of fines by eolian and

fluvial (surface wash) processes. Colluvial and weatlaering pro- cesses further contribute to ejecta removal and account for an equiv- alent lowering of at most 40 and 20 cm, respectively. When erosion estimates from sieve analyses (20 cm), field sieving (30 cm), and block comparisons (40 cm) also include the effects of colluvial and weathering processes, the total erosion of the outer-continuous ejec- ta is now modified to at most ~ 1 m (Table 3a).

15,040 GRANT AND SCHULTZ: EROSION OF EJECTA AT METEOR CRATER, ARIZONA

Sieve analyses

Clast pits

Block density ,

Uncorrected

Estimate, cm

2O

3O

40 ,

TABLE 3a. Values Used in Estimating Erosion'

Surface Versus Subsurface Grain-Size Comparisons , ,

Corrections, cm

C•)11uviation Weathering , ....

40 20

40 20

40 20 ,

Corrected

Estimates, cm

80

90

100 ,

i

Ejecta(deflation)

Ej ecta(Colluviurn)

Buried ejecta(dcflation)

Colluvium

Colluvium(Buried)

Alluvial fan

Diffuse drainage

Uncorrected

Thickness, cm

20

30

15

variable

variable ,

TABLE 3b. Values Used in Estimating Erosion:

Sediment Budget for Semienclosed Drainage Basin

Deflation/Weathering Corrections, cm ., . -

Used Maximum

45 70

25 55

45 70

25 55

3O 70

30 70

, ,

"Best" Estimate of

Corrected Thickness, cm _

45

20

25

75

40

variable

variable

Sediment Budget Within Semienclosed Basin

Field mapping delineates an enclosed drainage basin on the west- ern crater flank (Figure 2). Surfaces within this basin can be divided into four surficial units on the basis of both field studies of the depo- sitional environments and analyses of grain size and shape (Table 3b; Figure 9). First, an ejecta unit is distinguished in the map where Kai- bab ejecta deposits are exposed or buried by less than 20-cm collu- vium; this unit comprises 38% of the basin area. Second, a colluvium unit is recognized by thicker accumulations of deposits (20 cm to 40 cm deep) containing slightly rounded clasts (Figure 8b); colluvium covers 26% of the intrabasin surfaces. Third, distinctive, fine-

grained alluvial fan deposits with coarsest fragments rarely >5 cm (Figure 8c) cover 19% of the basin, principally occurring between low, subradial, ejecta ridges. Fourth, fine-grained deposits extend- ing (Figure 8d) from the alluvial fans are labeled diffuse drainage de- posits and represent 17% of intrabasin surfaces. Most colluvial and alluvial fan surfaces in the basin are capped by coarse lag deposits.

The sediment volume represented by these four depositional units provides a minimum estimate for the volume of ejecta eroded and transported locally within the basin. Any sediment transported outside of the basin via fluviation, deflation, and dissolution must be

constrained as well. In spite of the number of necessary correction factors, the amounts and processes of degradation can be constrained using field observations and can provide an independent estimate of overall erosion of the outer-continuous ejecta. As a sensitivity test for our derived erosion estimates, the maximum possible volume of sediment transported outside the basin is also used to calculate ero- sion.

Initial calculation of the sediment budget within the basin. Most ejecta in the basin (72%) occur on the steep upper flank (Figure 9). Fans bury ejecta and thin distally to 0.6R from the present rim-crest, whereas diffuse drainages extend to about 1.25R (Figures 2, 3, 8c 8d, 9). Therefore, erosion dominates the upper flank and leads to deposi- tion over the outer-continuous ejecta. With this in mind, the total drainage basin area (AT) containing the volume of intrabasin depos- its (VT) was divided into two parts: an upper flank, near-rim area dominated by erosion (AR) that retains minimal alluvium and collu- vium (VR); and a more distal area of the continuous ejecta (AD) that

accumulates significant alluvium and colluvium (Vr)) derived local- ly and from the upper flank of the raised rim (Figure 10). Hence,

Ar)=Ae + Ac + Af + As (2a)

l/D= V•+ V• + Vf+ Vd (2b)

where the subscripts e, c, f, and d correspond to the area/volume of ejecta, colluvium, alluvial fan, and diffuse drainage units, respec- tively. Unweathered ejecta originally characterized both near-rim and distal areas (VRo and Vr)0). Therefore, average erosion in cm (E) beyond the near rim is

E = VD•/AD. (3)

At present, however, minimal alluvium and colluvium near the rim (VR) indicate that the total deposit accumulated in the basin (VT) should be equivalent to the more distal accumulation of alluvial/co!- luvial deposits (VD). These mass balance requirements yield the fol- lowing two equations (including the density differences between ejecta and alluvium). First, the volume of distal continuous ejecta (VD0) must equal the total volume of ejecta eroded from (AD) since crater formation

VD0 = VT- VR0 (4a)

AD = AT-- AR (4b)

This relationship is illustrated in Figure 10. The quantities VT, AR, AD can be measured using the distribution and thickness of transported ejecta accumulated in the drainage basin. Subtraction of VR0 from VT gives the volume of sediment eroded (VD0) in equation (4a) from the distal-continuous ejecta area, AD, in terms of values constrained by field observations. Therefore,

v,= wo + (AV•) + (AW) + (AW) + (AW) (5)

where A refers to the current inventory of the each depositional envi- ronment minus the contribution derived from near-rim sources. Pre-

vious studies provide working values for the initial, preerosion pro- file and the original volume of near-rim ejecta, Vh• [Roddy et al., 1975; Roddy, 1978; Shoemaker and Kieffer, 1974].

Field mapping provides the inputs for each term in equation (5) leading to the average erosion in equation (3). Excavation and sam- pling of some portions of the ejecta in the basin demonstrate that the

GRANT AND SCHULTZ: EROSION OF EJECTA AT METEOR CRATER, ARIZONA 15,041

Grain-Size of Depositional Units at Meteor Crater

Windstreak

Proximal (solid) and Distal (open and cross-hatched) Samples

50,

40. -

30. -

20. -

O. •[•n. ru r• ..... • • \•.00 \, X, X X• X X• 'øo qøø 'øo %0 % %0 'øo

Grain- Size (phi units)

Colluvium

Inside Semi-Enclosed Basin West of Crater

50.

40.

50.

20.

10.

x z x z xe xd. x; x•, "e.o o .% qøo .% .% .o o .% Grain-Size (phi units)

Alluvial Fan Surface vs. Subsurface Sample Plot Inside Semi-Lnclosed Basin West of Crater

50.

40. I •: 30ø

in 20. 10.

0.

\• X! X Xli , X•$,

Grain-Size (phi units)

Diffuse-Drainage

Inside Semi-Enclosed Basin West of Crater

50.

40,

30.

Grain-Size (phi units)

Fig. 8. (a) Grain size distributions for proximal and distal windstreak sediments. All windstreak deposits at the crater are well sorted, have a mode at 2.5 phi, and show little variability with location. Windstreak grain size distributions from different composition source regions (i.e., Kaibab versus Coconino) are also remarkably similar. (b) Colluvial deposit grain size distribution from inside the semien- closed basin west of the crater. This and other colluvium samples are depleted in coarse fragments (>20-30 cm) and lack the prominent modes of in situ ejecta samples. In addition, the colluvium samples are enriched by a factor of 1/3 in sediment susceptible to deflation relative to the in situ Kaibab ejecta. Colluvium fragments are typically more rounded than in situ ejecta fragments.(c) Grain size distri- butions of coarse-grained surface-lag deposits from an alluvial fan and subsurface alluvium from inside the semienclosed basin. Coar- se-grained fragments are concentrated on fan surfaces by 5X-18X relative to the subsurface material and suggest in situ accumulation from deflation of between - 15 cm and 30 cm alluvium. (d) Grain size distribution of sediment from diffuse drainage inside the semien- closed basin. Except for the largest sizes, the distribution is similar to that in the alluvial fans. Modes occur between 3.0 and 4.0 phi. Coarse-grained surface lags are less well developed or absent on most diffuse drainage surfaces.

average colluvial cover is close to the maximum 20 cm used in defin- ing the mapped unit; therefore, Ve is 10,400 m 3. The colluvium unit comprises 36,900 m 3 (Table 3b) based on an average 30-cm thick- ness; however, additional colluvium beneath both fans and diffuse

drainages is up to 15 cm thick and could contribute an additional 25,500 m 3 of sediment. Hence, Vc is the sum of these two volumes, 62,400 m 3 (Table 3b). The volume of sediment contained in the allu- vial fans (Vf) and diffuse drainages (V•) is 69,400 m 3 and 51,400 m 3, respectively (Table 4). These estimates include adjustments for pos- sible fluvial transport out of the basin and into adjacent alluvial fans covering 13,000 m 2 and diffuse drainages over an area of 55,000 m 2. An absence of deposits from well-defined integrated drainages at minor drainage breaches limits more distal transport and qualitative- ly supports the validity of this adjustment. The following discussion

focuses on additional corrections for deflation from basin surfaces.

Each correction further includes a 5--cm adjustment for dissolution. Corrections for deflation. Deflation from ejecta surfaces is esti-

mated by dividing the volume of the windstreak by its source area. Losses from alluvium are estimated using the concentration of coarse fragments in alluvial fan-lag deposits relative to subsurface alluvium. Similarly, losses by colluvium deflation are estimated from the observed change in grain size distribution from the un- weathered ejecta to the colluvium. Deflation values for each deposi- tional unit are then given as "best" and "maximum" with alternative values for deflation given by their sums.

The windstreak area/average thickness downwind of Kaibab and Coconino ejecta source regions is 320,000 mV5-10 cm and 410,000 m:/15 cm representing -24,000 m 3 and 61,500 m 3 eroded sediment,

15,042 GRANT AND SCHULTZ: EROSION OF E/F_,CTA AT METEOR CRATER, ARIZONA

Fig. 9. Depositional environment map of the semienclosed drainage basin on the west side of the crater. Drainage divides that define the basin (dashed lines) are breached on the north and south. To compensate for material that may have been lost from the basin through these breaches, large areas of allu- vial fan and diffuse drainage deposition adjacent to the basin were included in the calculations. Total basin area is 475,000 m 2. Mapped units include ejecta (182,000 m2), of which 72% lacks a thin colluvial cover and is located on the upper rim flank; thin colluvium (123,000 m2); alluvial fans (90,000 m2); and diffuse drainage (80,000 m2). Alluvial fans and diffuse drainage ad- jacent to the basin included in calculations cover an additional 13,000 m 2 and 55,000 m 2, respectively. Solid lines mark the location of selected cross sec- tions in Figure 3.

respectively. These volumes, however, only reflect the saltating component of deflation and require corrections to account for the un- der-represented fine fraction lost via suspension transport. Compar- ison between Kaibab and Coconino ejecta and windstreak grain sizes indicates a deficiency of windstreak sediment smaller than 0.009

no/,•j// ' •,1 AD I

• 'i•...'• ..... ......:7.'.'

VD =Ve + Vc +

i• Diffuse Drainage [V,•} _

!i• Alluvial Fan (Vf }

'"':;":• Colluvium (:V•)

E jecta [Ve

Bedrock

Fig. 10. Idealized block diagram of the crater flank (not to scale) in the vic.in- ity of the semienclosed basin west of the crater. View is to the south-sou- theast. The dashed line on the low ridge near the right side of the figure repre- sents a portion of the divide between the basin and drainage into Canyon Dia- blo. Near-rim areas of high erosion (A•) that originally contained a volume of sediment (V•0) as well as more outer-continuous portions of the ejecta (AD) containing sediments (VDD are delineated. Examples of the volume of sediments contained within the various more distal depositional environ- ments (Ve•, Vq, Vh, VdD are also shown. See text for discussion.

mm (3.5 phi) that amounts to 95% (22,800 m 3) and 30% ( 18,500 m3), respectively, of the sediment represented by the saltating fraction (Figures 7a 7c, 8a). Hence, the total volumes of sediment deflated from the Kaibab and Coconino ejecta source regions to form the present windstreak are 46,800 m 3 and 80,000 m 3, corresponding to a deflated sediment mass of 8.4 X 107 kg and 1.4 X 10 s kg, respective- ly, for an assumed windstreak bulk density of 1800 kg/m 3.

Dividing the derived mass of deflated sediment by a bulk ejecta density of 2150 kg/m 3 [Regan and Hinze, 1975; Roddy et al., 1975] reveals that 39,100 m 3 and 66,900 m 3 of the Kaibab and Coconino

ejecta has been deflated. Dividing these volumes of deflation by their source areas (720,000 m 2 for the Kaibab and 200,000 m 2 for the Co- conino) yields 13 cm and 34 cm eolian lowering of the Kaibab and Coconino ejecta, respectively. Because prevailing winds at the crater may be unchanged since impact [Breed et al., 1984], such estimates could reflect total deflation. The poor preservation potential of a windstreak suggests, however, that additional unrecorded deflation may have occurred as well.

TABLE 4. Semienclosed Drainage Basin Field Mapping

•Iapped ...... Area, m 2 % Total Uncorrected Best Estimate Unit Basin Area Volume, m 3 of Corrected Volume, m 3

i_ , , . i,, [ [. [ i [ , i [ , ,,

Ejecta 182,000 38 10,400 164,700++

Colluvium 123,000 26 36,900 158,700õ

Alluvial Fans 90,000 19 69,400' 99,900'

Diffuse Drainage 80,000 17 51,400+ 91,900+ I 11 I I , I IlL I ,l [llll[ . I.[lll I J I ] ! I I Ill, J , I1 [.

Total basin area is 475,000 m 2 (see Figure 9).

* Includes volume of additional 13,000 m 2 alluvial fan adjacent to basin. + Includes volume of additional 55,000 m 2 diffuse drainage adjacent to basin.

++ Includes volume of thin colluvium on ejecta, deflation from exposed and presently buried ejecta.

õ Includes volume estimate for buried colluvium, deflation from exposed and buried colluvium.

Alluvium, colluvium, and deflated ejecta densities are 1800 kg/m 3 (measured), 2000 kg/m 3 (estimated), and 2150 kg/m 3, respectively.

GRANT AND SCHULTZ: EROSION OF F. JECTA AT METEOR CRATER, ARIZONA 15,043

Several observations further constrain total deflation since crater

formation. The windstreak is presently inactive as confirmed by su- perposition of 900-year-old-volcanic ash from nearby Sunset Cra- ter (E.M. Shoemaker, personal communication, 1989), a paucity of bedforms, and the widespread occurrence of vegetation. Because the regional climate has been relatively uniform and caused little erosion since the early Holocene [Nishiizumi et al., 1991 ], the windstreak probably dates back to at least 8-10 ka or of postpluvial or monsoon- al age. A maximum of about ~70 cm deflation from Kaibab ejecta can be estimated from the current windstreak inventory, the mini- mum age, and the assumption of uniform (but unrecorded) deflation over crater history. A lesser value of 45 cm total deflation from the Kaibab ejecta, however, provides a more realistic value based on the likelihood that the intensity of eolian activity is variable and climate

sensitive (Table 3b). The calculation of AVe in equation (5) further requires the approximate deflation from Kaibab ejecta prior to burial beneath alluvium and colluvium. If a similar rate of eolian stripping and an average burial age of half the crater age are assumed, a "best" estimate of 25 cm is derived. Ejecta burial by alluvium/colluvium as recently as 10 ka, however, yields a "maximum" estimate of 55 cm (Table 3b).

Alluvial fans in the drainage basin are mantled by coarse-lag de- posits that can be used to estimate deflation more directly. Coarse- clast concentrations in fan-lags relative to underlying alluvium indi- • cate that 30cm deflation has occurred, a value adopted as a"best" for fans and diffuse drainages based on their similar fine-grained com- ponent. Because most alluvial surfaces in the basin are relatively young but relict, additional deflation from now buried surfaces is possible. A pluvial or monsoonal age for most alluvium is implied by (1) soil development [Shoemaker, 1960]; (2) exposure ages around the crater [Nishiizumi et al., 1991; Phillips et al., 1991 ]; (3) uniform, fine-grained, down-drainage grain size [Mayer et al., 1984]; (4) rare bar and swale topography [Denny, 1965; Wells et al., 1987]; (5) widespread occurrence of surface lags; and (6) occurrence of collu- vium beneath many deposits. If the alluvium predates the monsoon- al and/or pluvial period, however, total deflation is constrained by a paucity of exhumed surfaces and limits maximum deflation to 2-3 times the "best" value, or ~70 cm.

Deflation from colluvial surfaces can be constrained by contrast- ing constituent grain sizes and ages of the colluvium with the Kaibab ejecta. This approach reveals that the younger colluvium has a larger deflatable fraction. The younger age of the colluvium should ap- proximately offset its greater deflation potential. Hence, working values previously derived for ejecta deflation ("best"= 45 cm; "max- imum" = 70 cm) also should reasonably characterize the colluvium. Colluvium up to 15 cm in thickness is buried by alluvium that is ~ 1/2 the average thickness of the exposed colluvium; therefore, a "best" estimate of deflation prior to burial is 25 cm. Arguments similar to those used to estimate maximum deflation from buried ejecta limits deflation from buried colluvium to 55 cm at most.

Modified erosion estimates. Corrections for deflation in volume estimates for deposits in the basin allow calculation of "best" and "maximum" values for the various terms in equation (5), leading to an estimate for overall erosion. Deflation of 30 cm alluvium equates to 71,000 m 3 for a total alluvium volume of 191,800 m 3 (Table 4). "Best" estimated deflation from exposed and buried colluvium rep- resents 96,300 m 3 sediment for a total of 158,700 m 3 colluvium (Table 4). Deflation from exposed and buried ejecta removed 154,300 m 3 sediment (Table 4) for a total of 164,700 m 3. Summation of the corrected values and conversion to an equivalent unweathered ejecta mass/volume yields 1.0 X 109 kg/465,000 m 3 for the appropri- ate densities (Table 4). The total equivalent ejecta mass/volume of

eroded sediment calculated using all "maximum" estimates be- comes 1.6 X 109 kg/744,000 m 3.

Before the summed volume of eroded sediment can be used to es-

timate average outer-continuous ej ecta erosion in the basin, the ejec- ta (Vv, o) eroded from the near-rim (AR) and deposited in more distal reaches (Ar•, Figure 10) should be constrained further. First, a simpli- fied model of near rim erosion predicts that erosion over the first 100 m outside the present rim crest decreases to values matching the more outer-continuous ejecta (Figure 10). This model is based on the power-law decay of elevation and slope away from the rim due to underlying structural uplift and ejecta thinning [Roddy et al., 1975; Garvin et al., 1989]. Widening of this high-erosion zone near the rim results in increased estimates of the volume of ej ecta eroded from the near rim (Vv, o), which in turn cause estimates of the volume eroded more distally (Vr•0) to decrease. Therefore, minimizing the near-rim high-erosion zone causes estimated erosion on the more outer-con- tinuous ejecta to be maximized. Second, the 12 m to 15-20 m verti- cal erosion at the original rim crest [Shoemaker and Kieffer, 1974; Roddy et al., 1975; Roddy, 1978] is decreased to ~ 10 m and ~ 15 m lowering at the current rim position to account for 30 m erosional widening of the crater [Roddy et al., 1975; Roddy, 1978]. These val- ues yield a near-rim area of the basin (A•t) of 57,000 m 2, a distal area (At0 of 418,000 m 2, and near-rim volumes (V•t0) of 429,000 m 3 and 286,000 m 3 for 15 m and 10 m rim erosion, respectively. The remain- ing volume, Vr•0, is the total-deposit volume minus V•0, or 36,000 m 3 and 179,000 m 3 for 15 m and 10 m rim crest lowering, respectively. The resultant "best" estimate of average erosion ofouter-continuous ejecta (Vr•o/Ar•) is from ~ 10 cm to ~40 cm for 15 m and 10 m rim crest lowering, respectively. Similar calculations of all "maximum" vol- umes limit average erosion of the outer-continuous ejecta to at most 110 cm.

Although erosion estimates derived from the •ntrabasin sediment budget incorporate a number of losses, such values can be validated by field observations and all are similar to estimates from the first approach involving grain size comparisons. In addition, denudation of the outer-continuous ejecta is maximized by minimizing the width and volume of the near-rim high-erosion zone. This point is illustrated by the difference in estimated erosion obtained using a near-tim volume based on a rim lowering of 10 m versus 15 m. Be- cause the scale of depositional features in th,e basin and elsewhere at the crater is similar, the 0.4-1.1 m estimate of erosion on the outer-

continuous ejecta should be representative.

DISCUSSION

The minimal amount of erosion that we derive for Meteor Cra-

ter's outer-continuous ejecta leads to the possibility that much of the ejecta might retain a relatively pristine form partly masked by vege- tation, an alluvial and colluvial veneer, removal of finer scale tex-

tures, and loss of color contrasts on ejecta surfaces. As discussed next, this inference can be supported by the results of ground-pene- trating radar studies [Grant and Schultz, 1991 ] of ejecta surfaces be- neath the alluvium; by the scale of primary ejecta features still pre- served; and by comparisons with preservation states of other land- forms created by catastrophic events.

High-resolution ground-penetrating radar allows comparison of exposed and buried surfaces (Figure 11). Radar transects show that gradients on buried ejecta/alluvium contacts extrapolate continu- ously to exposed ejecta surfaces. Consequently, erosion has been limited since deposition of most of the overlying alluvium. In addi- tion, ~ 1 m local relief on both buried alluvium/ejecta contacts and exposed ejecta is comparable. This observation indicates that incise-

15,044 GRANT AND SCHULTZ: EROSION OF EJECTA AT METEOR CRATER, ARIZONA

0 -

(ns)

15

Blocked Drainage South of Crater GPR Data

0 (•cate approx ) 2 m

Interpretation

(ns)

15

Ejecta (cm)

150

Surface•

Alluvium

Kaibab Ejecta • •

ß .

Kaibab Ejecta

0

Alluvium

(cm)

150

South

Fig. 11. Ground penetrating radar transect across the alluvium that is filling a blocked drainage on the south side of the crater (see Figure 2 for location). Radar was used to confirm depths of alluvium around the crater and to provide additional information regarding the erostonal history. Depth scales differ between ejecta and alluvium due to varying dielectric properties. An analog SIR-3 GPR from Geophysical Survey Systems, Inc., was used with a 500-mHz antenna to obtain data. The gradient along the contacts between alluvium and underlying ejecta is close to that on adjacent, exposed surfaces, thereby suggesting minimal vertical erosion since emplacement. In addition, the contact shows little evidence for fluvial incisement and therefore erosion prior to burial.

ment prior to burial has been minimal (Figure 11 ). Hence, the maxi- mum 1-m estimate of erosion on the outer-continuous ejecta is con- sistent with evidence from in situ features revealed by radar (Table 5).

As part of a separate study focusing on the styles and mechanics of ejecta emplacement (Schultz and Grant, 1989; P.H. Schultz and J.A. Grant, manuscript in preparation), we also sought to identify key ejecta signatures associated with emplacement. Careful exami- nation of the crater exterior reveals the preservation of subtle, but un- mistakable primary ejecta features. For example, several •- 1 m thick distal ejecta lobes superpose low buttes --2.0R north of the crater [Tilghman, 1905; Barringer, 1910] (Figure 1). These lobes persist despite exposure to prevailing winds (Figure 12). Their preservation in such an erosionally sensitive setting is inconsistent with wholesale 2-3 m removal of the outer ejecta as implied by some previous esti- mates. Moreover, preservation of scattered Kaibab ejecta blocks (•- 15 to 25 cm in diameter) beyond the continuous ejecta at distances up to 5.3R also implies minimal erosion (Table 5).

Although this study emphasizes the amount and styles of degra- dation, the total average erosion since crater formation provides an

effective long-term erosion rate of 2 cm/ka that can be calibrated by derived rates for other diverse landforms. Derived long-term rates measured on other southwestern U.S. surfaces include basaltic cin-

der cones (1.1-2.8 cm/ka), dolomitic rocks (1.7-2.1 cm/ka), some plutonic rocks (0.2 cm/ka), and piedmont surfaces (0.5 cm/ka) as cited in other studies [Marchand, 1971; Turrin and Dohrenwend,

TABLE 5. Summary of Estimated Erosion of

Distal-Continuous Ejecta at Meteor Crater ,,

' ' 'NR', m +

Coarsening of lags

Ba,;in sediment budget 10-15*

Ground penetrating radar

Primary ejecta features

+NR=Near Rim (<0.SR).

++OC=Outer-Continuous (>0.5R).

*Modified estimates of near-rim erosion cited above; see

text for discussion.

Min Max

0.8 1.0

0.4 1.1

<1

<1 ,

1984; Dohrenwend et at., 1986: Dohrenwend, 1987]. Average ero- sion rates along the steep near rim of Meteor Crater, however, are both expectedly higher [Shoemaker and Kieffer, 1974; Nishiizumi et at., 1991 ] and consistent with the paradigm that erosion is influenced by local slope and relief. Such time-averaged estimates, however, may not accurately reflect actual erosion rates at any given time.

The pristine state of ejecta around Meteor Crater is also compara- ble with the preservation of four other catastrophic landforms of sim- ilar age that contain abundant coarse fragments and have experi- enced generally dry (time-averaged) climate histories. First, exten-

GRANT AND SCHULTZ: EROSION OF EJECTA AT METEOR CRATER, ARIZONA 15,045

Fig. 12. One of several preserved distal-ejecta lobes on top of preexisting, elongate Moenkopi buttes located approximately 2.0R north-northwest of the crater. The lobe persists despite an elevated position and direct exposure to prevailing winds. The crater is to the left. View is to the west-northwest.

sive ejecta deposits are preserved around the >7000 year old [O 'C- onnell, 1965; Milton, 1968; Baker, 1981 ] Henbury Craters in Austra- lia and include ray/lobe features and closed impact depressions as small as 6 m in diameter with relief less than 0.5 m [Hodge, 1965; Hodge and Wright, 1971; Alderman, 1932; Milton, 1968]. Second, low-gradient, blocky surfaces on the ~ 18 ka [Stout, 1975; Johnson, 1978] Blackhawk landslide in California are also well preserved [Shreve, 1968]. Third, the 13-18 ka gravel/cobble bedforms in the Channeled Scabland of Washington [Baker, 1978] remain clearly recognizable and are not deeply eroded. Finally, some alluvial fans in the southwest United States preserve small, late Pleistocene/middle Holocene sheetflood bedforms [Wells andDohrenwend, 1985; Wells et al., 1990]. Due to the catastrophic nature of events forming these deposits, the larger size fractions are less sensitive to further trans- port under normal conditions, a condition analogous to the observed preservation of outer--continuous ejecta at Meteor Crater.

The maximum 1-m erosion of the outer--continuous ejecta esti- mated here (Table 5) is consistent with amounts inferred by Phillips et al. [ 1991 ] and Pilon et al. [ 1991 ], but significantly less than the amount of erosion deduced by Shoemaker and Kieffer [ 1974], Roddy et al. [1975], Roddy [ 1978], and Nishiizttmi et al. [ 1991 ](Table 2). This discrepancy may arise because workers citing higher denuda- tion also concluded that considerable Coconino ejecta has been lost. Although transported sediments in the windstreak indicate up to 2-3 m of Coconino ejecta has been deflated from existing localized out- crops, more widespread erosion of Coconino ejecta seems unlikely for four reasons. First, the amount of Coconino that must be removed to account for this erosional state exceeds the maximum calculated

volume of ejected Coconino sandstone [Roddy et al., 1975] by at least 2X. Second, sediments containing Coconino sands within allu- vial and eolian depositional sinks in and around the crater are rare ex-

cept in deposits directly downslope and/or downwind of existing Co- conino ejecta exposures. The absence of an appreciable Coconino component in most alluvium and drainage systems/traps precludes its widespread occurrence initially. There is also a paucity of Coconi- no sediments in samples from a transect crossing stratigraphy on the crater floor. Any possible unsampled Coconino component at great- er depth in the crater center must be volumetrically small and indi- cates that widespread deflation of Coconino ejecta has not occurred. Third, some near-rim exposures of Coconino ejecta occur at eleva- tions lower than the adjacent alluvial terraces but are not mantled by washed down alluvium as would be expected for more extensive degradation. Finally, crater flanks remain poorly incised and inter- fluve areas remain unfounded, thereby indicating only incipient flu- vial erosion [e.g., Ritter, 1986].

One implication of our results is that the volume of Coconino sandstone ejected probably lies between that predicted by Roddy et al. [ 1975] and Cro• [ 1980](Table 1). Our results may be most con- sistent with those of Grieve and Garyin [ 1984], who model crater

formation and predict a depth of excavation for Meteor Crater that is slightly greater than that of Cro• [ 1980]. Such models likely require further refinement before they can be considered to reflect accurate- ly all details of the impact process and crater formation.

The ground truth provided by field studies provides the necessary quantitative reference for identifying the signatures of degradation processes in different geologic and planetary settings. Comparisons of degradational states on more densely cratered surfaces can be used to infer the intensity and processes of degradation with allow- ances for different substrates. On Mars, such knowledge could be achieved using high-resolution Mars Observer data and would pro- vide a general indicator of regional erosion. Conversely, the results might allow understanding past climates.

15,046 GRANT AND SCHULTZ: EROSION OF ]•JECTA AT METEOR CRATER, ARIZONA

CONCLUDING REMARKS

The amount and styles of erosion around rapidly created land- forms such as impact craters can be constrained quantitatively using two techniques based on field studies: comparisons between subsur- face ejecta and surface ejecta-lag grain size properties and evalua- tion of the sediment budget within local drainage basins. At Meteor Crater, these methods place an upper limit of 1 m average erosion on widespread outcrops of Kaibab outer-continuous ejecta located more than 0.25R-0.5R from the present rim. This new estimate is consistent with the results of ground penetrating radar studies of bu- ried ejecta structure and evaluation of the scale of preserved prima- ry-ej ecta features, including distal ejecta lobes and blocks. Although our estimate of erosion on the ejecta is consistent with the conclusion of some previous studies [Phillips et al., 1991; Pilon et al., 1991 ], it is less than that of Shoemaker and Kieffer [ 1974], Roddy et al. [ 1975], and Roddy [ 1978]. Studies calling for higher erosion must invoke a missing component of Coconino ejecta, but field evidence for this component in depositional sinks inside and around the crater is ab- sent. Instead, erosion of low-gradient, outer-continuous Kaibab ejecta surfaces ironically helps to preserve and limit further denuda- tion as fluvial and eolian processes remove finer fragments to pro- duce a protective lag. Because degradational morphology on ejecta surfaces reflects both the intensity and processes of erosion, the field methods described here could provide a new basis for remotely eva- luating degradation in response to contrasting climates around other craters on the Earth, Mars, and Venus.

Acknowledgements. Thanks are expressed to C. Halfen, J. Mustard, P. NieverL P. Fisher, O. Barnouin, S. Gjos, Nancy Grant, and Mary Grant. We thank Richard Grieve and an anonymous individual for constructive reviews of the manuscripl. Gene Shoemaker provided stimulating discussions and in- sight during time in the field. Geophysical Survey Systems, Inc. graciously provided training and access to a GPR. Finally, this work would not have been possible without the support of George Shoemaker, Jon Gordon, A1 and Beth Masse, and the rest of the Meteor Crater staff. We also thank the Bar- ringer family for continuing to make the crater accessible to study. This re- search was supported by NASA grant NAGW-705 and a travel grant from the Barringer Crater Company.

REFERENCES

Alderman, A.R., The meteorite craters at Henbury, Central Australia, Miner- al. Mag., 23, 19-32, 1932.

Baker, V.R., Paleohydraulics and hydrodynamics of Scabland floods, in The ChanneledScabland, edited by V.R. Baker and D. Nummedal, pp. 58-79, NASA, Washington, D.C., 1978.

Baker, V.R., Australian analogs to geomorphic features on Mars, Reports of the Planetary Geology Program, NASA Tech. Memo., 84211, 58-79, 1981.

Barringer, D.M., Coon Mountain and its crater, Proc. Acad. Natl. Sci. Phil., 66. 861-886, 1905.

Barringer, D.M., Meteor Crater (Formerly called Coon Mountain Butte) in Northern Central Arizona, 45 pp., Published by D.M. Barringer, 1910.

Barringer, D.M., Further notes on Meteor Crater, Arizona, Proc. Acad. Natl. Sci. Phil., 57, 556-565, 1914.

Benson, L.V., and R.S. Thompson, The physical record of lakes in the Great Basin, in The Geology of North America, vol. K-3, North America and Adjacent Oceans During the Last Glaciation, edited by W.F. Ruddiman and H.E. Wright, Jr., pp. 241-260, Geological Society of America, Boul- der, Colo., 1987.

Betancourt, J.L., Late Quaternary vegetation history of the southwestern U.S.: The packrat midden record, paper presented at 4th Workshop on Climate Variability of the Eastern North Pacific and Western North America, PACLIM, Pacific Grove, Calif., March 22-26, 1987.

Breed, C.S., J.F. McCauley, W.J. Breed, C.K. McCauley, and A.S. Cotera, Jr., Eolian (wind formed) landscapes, in Landscapes of Arizona: The Geo- logical Story, 505 pp., edited by T.L. Smiley, J.D. Nations, T.L. Pewe, and J.P. Schafer, University Press of America, New York, 1984.

Cart, M.H., The Surface of Mars, 232 pp., Yale University Press, New Ha- ven, Conn., 1981.

COHMAP Members, Climatic changes of the last 18,000 years: Observa- tions and model simulations, Science, 241, 1043-1052, 1988.

Cooke, R.U., and A. Warren, Geomorphology in Desserts, 394 pp., Universi- ty of California Press, Los Angeles, 1973.

Croft, S.K., Cratering flow fields: Implications for the excavation and tran- sient expansion stages of crater formation, Proc. Lunar Planet. Sci. Conf, llth, 2347-2378, 1980.

Denny, C.S., Alluvial fans in the Death Valley region, California and Ne- vada, U.S. Geol. Surv. Prof Pap., 466, 62 pp., 1965.

Dohrenwend, J.C., Basin and Range, in Geomorphic Systems of North Amer- ica, Centennial Special Volume 2, edited by W.L. Graf, pp. 303-342, Geological Society of America, Boulder, Colo., 1987.

Dohrenwend, J.C., S.G. Wells, and B.D. Turrin, Degradation of Quaternary cinder cones in the Cima volcanic field, Mojave Desert, California, Geol. Soc. Am. Bull., 97, 421-427, 1986.

Emmett, W.W., The hydraulics of overland flow on Hillslopes, U.S. Geol. Surv. Prof Pap., 662-A, 68pp., 1970.

Emmett, W.W., Overland flow, in Hillslope Hydrology, pp. 145-176, John Wiley, New York, 1978.

Folk, R.L., Petrology of Sedimentary Rocks, 185 pp., Hemphill, Austin, Tex., 1980.

Forester, R.M., Late Quaternary paleoclimate records from lacustrine ostra- codes, in The Geology of North America, vol. K-3, North America and Adjacent Oceans During the Last Deglaciation, edited by W.F. Ruddi- man and H.E. Wright, Jr., pp. 261-276, Geological Society of America, Boulder, Colo., 1987.

Gatwin, J.B., J.L. Buffon, B.A. Campbell, and S.H. Zisk, Terrain analysis of Meteor Crater ejecta blanket (abstract), Lunar Planet. Sci., XX, 333-334, 1989.

Gile, L.H., F.F. Peterson, and R.B. Grossman, Morphological and genetic se- quences of carbonate accumulation in desert soils, Soil Sci., 101, 347-360, 1966.

Grant, J. A., and P. H. Schultz, The erosional state and style of Meteor Crater, Arizona (abstract), Lunar Planet. Sci., XX, 355-356, 1989.

Grant, J. A., and P. H., Schultz, Amounts and styles of ejecta erosion at Me- teor Crater, Arizona (abstract), Lunar Planet. Sci., XXI, 433-434, 1990.

Grant, J.A., and P.H. Schultz, Characteristics of ejecta and alluvial deposits at Meteor Crater, Arizona and Odessa Craters, Texas: Results from

ground penetrating radar (abstract), Lunar Planet. Sci., XXII, 481-482, 1991.

Green, C.R., and W.D. Sellers, Arizona Climate, 503 pp., University of Ari- zona Press, Tucson, 1964.

Grieve, R. A. F., and J.B. Garyin, A geometric model for excavation and modification at terrestrial simple impact craters, J. Geophys. Res., 89, 11,561-11,572, 1984.

Hodge, P.W., The Henbury Meteorite Craters, Smithsonian Contrib. Astro- phys., 8, 199-213, 1965.

}lodge, P.W., and F.W. Wright, Meteoritic particles in the soil surrounding lhe Henbury Meteorite Craters, J. Geophys. Res., 76, 3880-3895, 1971.

Hooke, R.LeB., Geomorphic evidence for late-Wisconsin and Holocene tectonic deformation, Death Valley, California, Geol. Soc. Am. Bull., 83, 2073-2098, 1972.

Hooke, R.LeB., Discussion of"Use of multiparameter relative-age methods for age estimation and correlation of alluvial fan surfaces on a desert piedmont, Eastern Mojave Desert, California", Quat. Res., 34, 262, 1990.

Johnson, B., Blackhawk landslide, California, U.S.A., in Rockslides and

Avalanches, vol.l, Natural Phenomena, edited by B. Voight, Elsevier, New York, pp. 481-504, 1978.

Machette, M.N., Calcic soils of the southwestern United States, Geol. Soc.

Am. Spec. Pap., 203, 1-21, 1985. Marchand, D.E., Rates and modes of denudation, White Mountains, eastern

California, Am. J. Sci., 270, 109-135, 1971.

Mayer, L., R. Gerson, and W.B. Bull, Alluvial gravel production and deposi- tion: A useful indicator of Quaternary climate changes in deserts, Catena, 5, suppl., 137-151, 1984.

McFadden, L.D., S.G. Wells, and J.C. Dohrenwend, Influences of Quaterna- ry climate changes on processes of soil development an desert loess de- posits of the Cima volcanic field, California, Catena, 13, 361-389,1986.

McFadden, L.D., S.G. Wells, and M.J. Jercinovich, Influences of eolian and

GRANT AND SCHULTZ: EROSION OF EJF_•A AT METEOR CRATER, ARIZONA 15,047

pedogenic processes on the origin and evolution of desert pavements, Geology, 15, 504-508, 1987.

McFadden, L.D., J.B. Ritter, and S.G. Wells, Use of multiparameter relative- age methods for age estimation and correlation of alluvial fan surfaces on a desert piedmont, eastern Mojave Desert, California, Quat. Res., 32, 276-290, 1989.

McKee, E.D., The Coconino sandstone-its history and origin, Carnegie Inst. Washington, Publ. 440, 77-115, 1934.

McKee, E.D., The environment and history ofthe Toroweap and Kaibab For- mations of northern Arizona and southern Utah, Carnegie Inst. Washing- ton, Publ. 492, 268 pp., 1938.

McKee, E.D., Small-scale structures in the Coconino sandstone of northern

Arizona, J. Geology, 53, 313-325, 1945. McKee, E.D., Stratigraphy and history of the Moenkopi Formation of Triass-

ic age, Geol. Soc. Am. Mem., 61,133 pp., 1954. Mifflin, M.D., and M.M. Wheat, Pluvial lakes and estimated pluvial climates

of Nevada, Bull. Nev. Bur. Mines Geol., 94, 1979.

Milton, D.J., Structural geology of the Henbury Meteorite Craters Northern Territory, Australia, U.S. Geol. Surv. Prof. Pap. 599C, 17 pp., 1968.

Moore, R.T., E.D. Wilson, and R.T. O'Haire, Geologic map of Coconino County, Arizona, Arizona Bureau of Mines, Tucson, 1960.

Morrison, R.B., and J.C. Frye, Correlation of the middle and late Quaternary successions of the Lake Lahontan, Lake Bonneville, Rocky Mountain (Wasatch Range), southern Great Plains, and eastern Midwest areas, Nev. Bur. Mines Rep., 9, 45 pp., 1965.

Mutch, T.A., R.E. Arvidson, J.W. Head, III, K.L. Jones, and R.S. Saunders, The Geology of Mars, 400 pp., Princeton University Press, Princeton, N.J.,1976.

Nininger, H.H., Arizona • Meteorite Crater- Past, Present, Future, 232 pp., American Meteorite Laboratory, Denver, Colo., 1956.

Nishiizumi, K., C.P. Kohl, E.M. Shoemaker, J.R. Arnold, J. Klein, D. Fink,

and R. Middleton, In situ 1øBe-26A1 exposure ages at Meteor Crater, Ari- zona, Geochim Cosmochim. Acta, 55, 2699-2703, 1991.

O'Connell, E., A catalog of meteorite craters and related features with a guide to the literature, 218 pp., RAND Corporation, Santa Monica, Calif., 1965.

Phillips, F.M., M.G. Zreda, S.S. Smith, D. Elmore, P.W. Kubik, R.I. Dom, and D.J. Roddy, Age and geomorphic history of Meteor Crater, Arizona, from cosmogenic 36C1 and 14C in rock varnish, Geochim. Cosmochim. Acta, 55, 2695-2698, 1991.

Pilon, J.A., R.A.F. Grieve, and V.L. Sharpton, The subsurface character of Meteor Crater, Arizona, as determined by ground-probing radar, J. Geo- phys. Res., 96, 15,563-15,576, 1991.

Regan, R.D. and W.J. Hinze, Gravity and magnetic investigations of Meteor Crater, Arizona, J. Geophys. Res., 80, 776-788, 1975.

Ritter, D.F., Process Geomorphology, 603 pp., Wm. C. Brown, Dubuque, Iowa, 1986.

Roddy, D.J., Pre-impact geologic conditions, physical properties, energy calculations, meteorite and initial crater dimensions and orientations of

joints, faults, and walls at Meteor Crater, Arizona, Proc. Lunar Planet. Sci. Conf., 9th, 3891-3930, 1978.

Roddy, D.J., J.M. Boyce, G.W. Colton, and A.L. Dial, Jr., Meteor Crater, Ari- zona, rim drilling with thickness, structural uplift, diameter, depth vol- ume, and mass-balance calculations, Proc. Lunar Planet. Sci. Conf., 6th, 2621-2644, 1975.

Schultz, P.H., Moon Morphology: Interpretations Based on Lunar Orbiter Photography, 626 pp., University of Texas Press, Austin, 1976.

Schultz, P.H., and D.E. Gault, Atmospheric effects on martian ejecta em- placement, J. Geophys. Res., 84, 7669-7687, 1979.

Schultz, P.H., and J.A. Grant, Styles of ejecta emplacement, Meteor Crater (abstract), Lunar Planet. Sci., XXII, 972-973, 1989.

Selby, M.J., Hillslope Materials and Processes, 264 pp., Oxford University Press, New York, 1982.

Shoemaker, E.M., Impact mechanics at Meteor Crater, Arizona, Ph.D. dis- sertation, Dep. of Geol., Princeton Univ., Princeton, N.J., 55 pp., 1960.

Shoemaker, E.M., Meteor Crater, Arizona, in Centennial Field Guide, vol. 2,

Rocky Mountain Section of the Geological Society of America, edited by S.S. Beus, pp. 399-404, Geological Society of America, Boulder, Colo., 1987.

Shoemaker, E.M., and S.E. Kieffer, Guidebook to the geology of Meteor Crater, Arizona, Pub. 17, 66 pp., Ariz. State Univ. Cent. for Meteorite Stud., Tempe, 1974.

Shreve, R.L., The Blackhawk landslide, Geol. Soc. Am. Spec. Pap., 108, 47 pp., 1968.

Smith, G.I., Subsurface stratigraphy and geochemistry of late Quaternary evaporites, Searles Lake, California, U.S. Geol. Surv. Prof Pap., 1043, 130 pp., 1979.

Smith, G.I., and F.A. Street-Perrott, Pluvial lakes of the western United

States, in Late-Quaternary Environments of the United States, vol. 1, The Late Pleistocene, edited by H.E. Wright, Jr., pp. 190-212, University of Minnesota Press, Minneapolis, 1983.

Spaulding, W.G., and L.J. Graumlich, The last pluvial climatic episodes in the deserts of southwestern North America, Nature, 320, 441-444,1986.

Stout, M.L., Age of the Blackhawk landslide, southern California, Geol. Soc. Am. Abstr. Programs, 7, 378, 1975.

Street-Perrott, F.A., and S.P. Harrison, Lake levels and climate reconstruc-

tion, in Paleoclimate Analysis and Modeling, edited by A.D. Hecht, pp. 291-340, John Wiley, New York, 1985.

Sutton, S.R., Thermoluminescence measurements on shock-metamor-

phosed sandstone and dolomite from Meteor Crater, Arizona, 2, Thermo- luminescence age of Meteor Crater, J. Geophys. Res., 90, 3690-3700, 1985.

Tilghman, B.C., Coon butte, Arizona, Proc. Acad. Natl. Sci. Phil., 57, 887-914, 1905.

Turrin, B.D., and J.C., Dohrenwend, K-Ar ages of basaltic volcanism in the Lunar Crater volcanic field, northern Nye County, Nevada: Implications for Quaternary tectonism in the central Great Basin, Geol. Soc. Am. Abstr. Programs, 16, 679, 1984.

U.S. Department of Commerce, Climatic Atlas of the United States, 80 pp., U.S. Govt. Printing Office, Washington, D.C., 1968.

Van Devender, T.R., Late Quaternary vegetation and climate of the Sonoran Desert, United States and Mexico, in Packrat Middens: The Last 40,000

Years of Biotic Change, edited by J.L. Betancourt, T.R. Van Devender, and P.S. Martin, pp. 134-165, University of Arizona Press, Tucson, 1990.

Van Devender, T.R., R.S. Thompson, and J.L. Betancourt, Vegetation history of the deserts of southwestern North America: The nature and timing of the late Wisconsin-Holocene transition, in The Geology of North Ameri- ca, vol. K-3, North America and Adjacent Oceans During the Last De- glaciation, edited by W.F. Ruddiman and H.E. Wright, Jr., pp. 323-352, Geol. Soc. Am., Boulder, Colo., 1987.

Wells, S.G., and J.C. Dohrenwend, Relict sheetflood bed forms on late Quat- ernary alluvial-fan surfaces in the southwestern United States, Geology, 13, 512-516, 1985.

Wells, S.G., J.C. Dohrenwend, L.D. McFadden, B.D. Turrin, and K.D. Mah- rer, Late Cenozoic landscape evolution on lava flow surfaces of the Cima volcanic field, Mojave Desert, California, Geol. Soc. Am. Bull., 96, 1518-1529, 1985.

Wells, S.G., L.D. McFadden, and J.C. Dohrenwend, Influence of late Quater- nary climate changes on geomorphic and pedogenic processes on a desert piedmont, Eastern Mojave Desert, California, Quat. Res., 27, 130-146, 1987.

Wells, S.G., L.D. McFadden, and J.B. Ritter, Reply to R.LeB. Hooke's dis- cussion of "Use of multiparameter relative-age methods for age estima- tion and correlation of alluvial fan surfaces on a desert piedmont, Eastern Mojave Desert, California", Quat. Res., 34, 263-265, 1990.

Zreda, M.G., F.M. Phillips, D. Elmore, P.W. Kubik, P. Sharma, and R.I. Dom, Cosmogenic chlorine-36 production rates in terrestrial rocks, Earth Planet. Sci. Lett., 105, 94-109, 1991.

J.A. Gram and P.H. Schultz, Brown University, Box 1846, Geological Sciences, Providence, R102912.

(Received May 12, 1992; revised June 9, 1993;

accepted June 10, 1993.)