The geophysical signature of terrestrial impact craters

The Geophysical Signature of Terrestrial ImpactCraters M. Pilkington R. A. F. Grieve Geophysics Division GeologicalSurveyof Canada,Ottawa,Ontario Abstract.A majortoolin theinitialrecognition andstudy gravityhighrestricted to the cratercenterandextending of terrestrialimpact craters,-20% of which are buried out to <0.5D. The magneticsignatureof cratersis more beneathpostimpact sediments, is geophysics. The general varied.The dominanteffect is a magneticlow due to a in susceptibility. Largestructures (D > 40 km) geophysicalcharacter of terrestrial impact craters is reduction compiledand outlinedwith emphasison its relationto the tend to exhibit central high-amplitudeanomalies,with of <0.5D, due to remanentlymagnetized impactprocessand as an aid to the recognitionof addi- dimensions tionalimpactcraters.The mostcommonandconspicuous bodiesin the targetrocks.The sourcesof thesebodiesare geophysical signatureis a circulargravitylow. For simple wide rangingand includethe effectsof shock,heat, and bowl-shapedcraters, gravity models indicate that the chemical alteration. Thefew studies overcratersinvolving methods indicateresistivity lowscoinciding with anomaly is largely due to the presenceof an interior electrical allochthonous breccialens. In complexcraters,modeling the extent of the potentialfield anomaliesand related to indicates thatthemaincontribution to thegravityanomaly fracturing.Seismic techniques,particularlyreflection is fromfracturedparautochthonous targetrocksin thefloor surveys,have provideddetailsof the subsurfacestructure of the crater.The gravity signatureof both simpleand of craters. Incoherent reflections and reduced seismic due to brecciation and fracturingare expected, complexcraterformscan be modeledwell, usingknown velocities morphometricparametersof impactstructures. The size of the degreeof coherency of reflections increasing away the gravity anomalygenerallyincreaseswith increasing from and below the center of the structure. From the techniques a setof generalcriteriacan craterdiameterreachinga maximumof--20-30 mGal at variousgeophysical diameters D of -20-30 km. Further increases in D have a negligibleeffecton the magnitude of thegravityanomaly due to lithostaticeffectson deep fractures.The general gravitysignature of a simplelow canbe modifiedby target rock and erosionaleffects,and there is a tendencyfor beestablished thatcorrespond to thegeophysical signature of impactcraters.Thesecriteriacanbe usedto evaluatethe hypothesis thatanyparticularsetof geophysical anomalies is due to impact.Confirmationof an impactorigin, however,is basedon geologicevidence. largercomplexstructures(D > 30 km) to exhibita relative INTRODUCTION structuralsignatureof terrestrialimpact craters.Conversely, sedimentationserves to protect craters but removesthem from direct observation. Approximately Impact of solid bodiesis the most fundamentalof all processes that have takenplaceon the terrestrialplanets 20% of the known terrestrial craters are buffed beneath [Shoemaker, 1977]. An integralcomponent in understand- postimpactsediments.Thus the recognition,which is ing the geological and geophysicalconsequences of based on surface morphologicaland morphometric large-scale impactis thedatasuppliedby terrestrialimpact observationsof terrestrialimpact craters,differs from craters.Basiccharacterization studiesof terrestrialimpact planetaryimpact craters.Terrestrialimpact cratersare craters provide much needed constraintson models of recognized largelyby lithologicalandstructuralevidence, crateringmechanics,serveas analogstudiesfor various with a major tool in the recognitionof a lithostmctural types of lunar and meteoriticsamples,and aid in the anomaly being geophysics.Many terrestrialimpact interpretationof the nature and significanceof remote structureswere identified initially as geophysical observationsof planetary surfaces.The very active anomalies andtheirimpactoriginestablished laterthrough terrestrialgeologicenvironment, however,resultsin the geologicstudies, for example,EagleButte,Canada. modification andpartialdestruction of muchof therecord Terrestrialdata are currently the main source of of impact.Erosion quickly removesthe original mor- information onthenatureof thethirddimension of impact phologicalelementsand, with time, the geologicaland cratersranging up to over 100 km in diameter. This Copyright1992by theAmericanGeophysical Union. 8755-1209/92/92RG-00192 Reviews of Geophysics, 30, 2 / May 1992 pages161-181 Papernumber92RG00192 $15.00 '161' 162 ß Pilkington andGrieve:SIGNATURE OF TERRESTRIAL IMPACTCRATERS 29, 1 /REVIEWS OF GEOPHYSICS informationcan be obtaineddirectlyby observations at a their geophysicalsignatures,but where no confirmatory numberof cratersof approximately the samedimensions geologicevidenceof impactis available.All thestructures characteristics discussed here have buterodedto differentlevels,by drilling,andby geophysi- and their geophysical cal methods.Even with direct observations, geophysical diagnosticevidence of impact in the form of shock effects. dataprovidean additionalperspective of the thirddimen- metamorphic sion.Geophysics anddrillingdataarea powerfulcombination for derivingstructuralinformationat depthand for OF IMPACT identifyingtargetsat impact structureswith resource GENERAL CHARACTERISTICS CRATERS potential,suchas thosewith hydrocarbons, for example, Red Wing Creek,NorthDakota,or significantminemlizaBeforediscussing thegeophysical signatureof terrestrial tion, for example,Sudbury,Canada. In thiscontribution we haveattempteda compilationof impactcraters,it is informativeto providea brief context thegeneralgeophysical characteristics of terrestrial impact as to their morphologyand structure.Impactcratershave craters. The geophysicalmethods discussedinclude two basic morphologicalforms: simple and complex. gravity,magnetic,seismic,andelectrical.Most emphasis Thesetwo forms occuron virtually all planetarybodies is placedon gravitybecauseof its utility,beingdirectly witha solidsurfaceanddifferonlyin thediameterrangeat related to impact-induceddensity changes, and the whichthe transitionfromoneform to anothertakesplace. considerablegravity data base that exists for impact To first order, this transition diameter is a function of craters.The compilationof the geophysicalsignatureof surfacegravityand targetrock type [Pike, 1980]. On the terrestrialimpactcratersis designedto serveas an aid in Earth, simple cratersoccur up to a diameterof 4 km in the recognitionof possibleimpactcratersin existingand crystallineand 2 km in sedimentarytargetrocks [Dence, futuregeophysical surveydata.The discovery of additional 1972]. Above these diameters, terrestrial craters have a impactcratersandtheirsubsequent studyaddsnotonly to complexform. Simple craters are characterizedby a bowl-shaped theknowledgebaseof theterrestrialimpactrecordbut also to the overallunderstanding of the fundamental processof depressionwith a structurallyupraisedand fracturedrim impactand its consequences for planetaryevolution.The area,which,when fresh,is overlainby an exteriorejecta presentrateof discoveryis threeto five new structures per blanketof breccia.The depthfrom the rim to the floor of year [Grieve, 1991]. As the moreobvioussurfacestruc- thecraterisreferred to astheapparent depth(d,).Drilling dataindicatethatthe apparentfloor of the tures have been recognized,geophysicshas played an andgeophysical increasinglyimportantrole in the initial recognitionof simple craters is underlainby a breccia lens, which is approximately parabolicin crosssection(Figure 1). This structures of possibleimpactorigin. or displacedand transported Whereappropriate, we haverelatedgeophysical models breccialensis allochthonous to the physicalnature of impact cratersand the impact material,which exhibitsa range of shockmetamorphic process. We havenotconsidered datafromthosestructures effectsbut, in general,consistsof relatively unshocked thathavebeensuggested to be of impactorigin,basedon material.The breccialens is flooredby parautochthonous Fallout ejecta dG dt Fractured & Brecciated Target• A • A /x /x Breccia A Lens Rocks Shocked Target Rocks Figure1. Schematic cross section of a simple crater. Thevalued•istheapparent crater depth, anddt isthe true craterdepth. 30, 2/REVIEWSOF GEOPHYSICS Pilkington andGrieve:SIGNATURE OF TERRESTRIAL IMPACTCRATERS ß 163 or essentiallyin situ targetrock material,whichformsthe walls and floor of the so-calledtrue crater (Figure 1). The of stratigraphicuplift (SU• undergoneby the beds now exposedin centralstructures[Grieveet al., 1981]. depthto the baseof the breccialensis referredto as the truedepth(dr).Theparautochthonous target rocksbeneath the breccia lens are fractured and in the center show GRAVITY SIGNATURE petrographic evidenceof shockpressures >15 GPa. The formation of simple craters is relatively well General Character The most conspicuousgeophysicalsignature over understood[Grieveet al., 1989;Melosh, 1989]. Briefly, a crateringflow field is setup in thetargetrocksastheresult impactcratersis a residualnegativegravityanomaly,that of the shock wave and trailing rarefactionwave fronts is, the anomalyleft after the regional field has been propagating from the point of impact.This resultsin the removed.In plan view these anomaliesare generally by thepresence of intrinsic excavationand displacementof material to form the circularbutcanbe complicated so-calledtransientcavity. The transientcavity is believed lateraldensitycontrastswithin the targetrocks.The cause to have an original depth-diameter ratio of-0.3 and is of thenegative anomalies is low-density materialresulting unstable.Its walls collapseinward to producethe final from both lithologicaland physicalchangesassociated crater,which is partiallyfilled with allochthonous breccia with the crateringprocess.For example,relativelylower derivedfrom the slumpingtransientcraterwalls. densitypostimpact sediments may occupythe topographic Larger complex craters have shallowerfinal depth- depressionof the crater, and in complex craters the diameterratios than simplecratersand representa much presenceof slightlylower densitycoherentimpactmelt moremodifiedform of the transientcavity. Detailsof the sheets enhance thenegativegravityeffect.Theselithologimechanics of complexcratersare lesswell understood, but cal changesare, however,minor comparedto density the principleshavebeenestablished [Schultzand Merrill, contrastsinduced by fracturing and brecciationof the 1981; Melosh, 1989]. The downwarddisplacementof the targetrocks.Fragmentation and redistributionof target targetrocksduringtransientcavityformationis not locked lithologies during crater formation leads to increased in, as in simple craters,and the cavity floor rebounds porosity levels in the allochthonousbreccia deposits. upward.This resultsin the formationof upliftedcentral Similarly, beneath the allochthonousbreccia deposits, structures, which consistof targetrocksshockedto levels shock-induced fracturingof autochthonous or in situtarget similar to those in the floor of simple craters.Uplift is rock also leads to increasedporosityand hencedensities accompanied by collapseof the transientcavity rim area, lower than those of the undisturbed formations. Nearresultingin a relatively shallowstructure,with a faulted surfacefracturingalso occursand may extendout to a rim zone and a flat floor (Figure 2). The floor of the distanceof one crater diameter beyond the crater rim structureis coveredby allochthonous materialsthatdid not [Gurov and Gurova, 1982]. This extended zone is, escapethe transientcavity by ejection.They are generally however,shallow with negligibleeffect on the observed Generally,thegravitylow extendsto or relativelyhighly shockedmaterialsand take the form of gravityanomalies. brecciasand impactmelt rocks.As the transientcavityis only slightlybeyondthecraterrim. Althoughnumerousdensitymeasurements are available essentiallydestroyedin the modificationprocess,thereis no equivalentto the relatively deep, with respectto for lithologiesfound within impact craters, there are diameter,interiorbreccialens observedat simplecraters relatively few determinationsof the density contrasts (cf. Figures 1 and 2). Although intensedisturbanceof betweenfracturedand unfracturedtargetrocks(Table 1). preimpactlithologiesis confinedto the areaof the original As well as a variationin densitycontrastsbetweencraters, transientcavity,complexcraterformationis characterized samplesfrom a given craterexhibit a rangeof density by considerableoverall movementof target materials values,usuallyshowingan increasewith depth(Figure3). to the decreasein the level of comparedto simple craters. In structureswith good This increasecorresponds stratigraphic controlit is possibleto directly measurethe shock-induced stressand thusfracturingwith depth.We relativemovementof lithologies,in particular,the amount know of only onecaseof deepdensitymeasurements from Ejecta Impact melt rocks AIIochthonous "'---' breccia Limit major disruption t• shock effects in autochthonous rocks Figure 2. Schematiccrosssectionof a complexcrater.SU denotesthe amountof structural uplift. 164 ß Pilkington andGrieve:SIGNATURE OF TERRESTRIAL IMPACTCRATERS BULK DENSITY (Kg/m 3) 2000 300 2200 I 2400 • 2600 • 2800 ' 3000 400 500 600 29, 1 /REVIEWS OF GEOPHYSICS characterof the gravity anomaly associatedwith a particularstructure.The initial size and densitycontrastof thebrecciatedandfracturedzoneis dependent on thecrater diameterand preimpactdensitydistribution of the target rocks.Subsequent modificationof thezonemaythenresult frompostimpact processes, suchaserosion. To evaluatetherelativeimportanceof thesefactors,we usethe valueof the maximumnegativegravityanomaly Ag over the crater.The useof only a singlevalue,Ag, to characterizethe complete crater anomaly is justified because ofthegenerally similar shape andsymmetry ofthe - crater gravity anomalies. Also, craters are surficial features.They have a small depthextentcomparedwith their areal size. Hence to first order, their gravitational 700 effect can be modeled as that due to a thin disk. Conse- quently, the maximum negative gravity anomaly over cratersis primarilydetermined by the densitycontrastand depthof thebrecciatedandfracturedzones. The distinctionin morphologybetween simple and 900complexcratersis notdirectlyreflectedin theform of their associated gravityanomaly.Over simplecraters,a circular bowl-shaped negativeanomalyis observed.This is alsothe t000casefor somesmallcomplexcraters,for example,Deep Bay, Canada.For larger complexcraters,however,the 1t00 gravitylow may be modifiedby the presenceof a central gravityhighwhich,in a few casesof erodedcraters,canbe field, for example,Vredefort, Figure 3. Loggedbulk formationdensity(averagedover 5-m greaterthanthe background intervals) for the Ntrdlingen 1973 boreholeat Ries, Germany SouthAfrica.In somecraterserodedto belowtheoriginal [after Ernstsonand Pohl, 1974]. Solid circles, suevite;open crater floor, only this central high may remain, and the circles,brecciatedcrystallinerocks with >10% suevite;pluses, crateris characterized by only a positiveanomaly,for 800- brecciatedcrystallinerockswith <10% suevite. example,UpheavalDome, Utah, andKentland,Indiana. Althoughthe existenceof a centraluplift in complex drilling. This is at Siljan, Sweden, where lower than cratersis nota sufficientconditionfor producinga relative normal densitiespersist to depthsof-5 km [Dyrelius, gravityhigh,thereis a tendencyfor mostlargercomplex 1988]. structures withD > 30 km to exhibitsucha centralgravity high (Figure5). For D = 30-100 km the corresponding structuraluplift (SU) rangesfrom 3.5 to 9.5 km (SU = Variationin GravityAnomalies Figure 4 showsthe relationbetweengravity anomaly and crater diameterD for 53 of the terrestrialimpact craterslisted in Table 2. The amplitudeof the negative gravity anomalyassociatedwith impactcratersincreases with the craterdiameter[cf. Dabizha and Fedynsky,1975, 1977]. Severalfactorsplay a role in determiningthe final 0.06D TM) [Grieve etal.,1981]. Therefore it ismore likely thatdensercrustalmaterialwill be broughtto the surface in large-impactevents.In thesecases,preimpactnearsurfacelithologicalvariationswill have a comparably minoreffecton the resultantcentralgravityanomaly.In addition,variationsin the degree of fracturingmay TABLE 1. DensityContrastsBetweenFracturedand UndisturbedTarget Rocksat Several Terrestrial Impact Structures DensityContrast, Target Rocks Gosses Bluff, Australia Brent, Canada Nicholson Lake, Canada Clearwater West, Canada Holleford, Canada Manicouagan,Canada Stderfj'firden, Finland sedimentary crystalline crystalline crystalline crystalline crystalline crystalline Average(crystalline) *Values include effects of sediments and breccias. kg/m $ 150 170-340' 70-140 170 240* 130 160 177.5 Reference Barlow [ 1979] Millman et al. [1960] Denceet al. [1968] Plante et al. [1990] Beals[1960] Sweeney[1978] Laurenet al. [1978] Pilkington andGrieve:SIGNATURE OFTERRESTRIAL IMPACTCRATERS ß 165 30, 2 / REVIEWSOF GEOPHYSICS TABLE2. MaximumNegative Residual GravityAnomalies for58Terrestrial ImpactStructures Diameter, Crater Actaman, Australia Aouelloul, Mauritania mGal C.P. ion E Age,m.y. -14 -1.25 ... ... 160 0.39 7 3 >570 3.1 2 1 3 <130 Ernstson etal.[1987] 0.049 Regan andHinze [1975] 88 Dabizha andFedynsky [1977] Azuara, Spain Barringer, Arizona Boltysh, Ukraine -10 --0.6 -30 Carswell, Canada -11 Brent, Canada Clearwater East,Canada Clearwater West,Canada -5 ... ... * . '1• 30 1.18 25 3 4 22 32 4 5 7 6 39 -13 -16 ... Crooked Creek,Missouri -2.5 .. Deep Bay,Canada -15 ... 13 El'gygytgyn, Russia -10 "• 18 Connolly Basin,* Australia --0.3 • Decaturville,* Missouri Dellen,Sweden DesPlaines, Illinois Flynn Creek, Tennessee Glasford,* Illinois 0 -7 -10 -1 0 9 i ... •'.') 7 Millman etal. [1960] 290 290 Planteetal. [1990] Planteetal. [1990] 115 <60 6 4 <300 15 8 6 7 110 <280 4 360 3.55 4 6 100 <430 -3.5 ... Haughton, Canada -9.5 ... 20.5 2 5 550 Janisj'arvi, Russia Kaluga, Russia -13 -12 ... 14 15 6 3 698 380 13 8 8 7 6 7 <300 430 400 Kara, Russia Kentland,* Indiana LacCouture, Canada LacLaMoinerie, Canada -20 -1 -5 --4 ... 2.35 -i• 4 ... 65 Lappaj'arvi, Finland -10 Lonar, India -3.6 Manicouagan, Canada -10 Middlesboro, Kentucky Mien,Sweden --3.5 -5 '") 17 . 1.83 '1• 100 Mistastin, Canada -15 ... -6 ... NewQuebec, Canada ... ... 6 9 28 3.44 5 6 5 2 7 6 6 2 142.5 <250 21.5 73 <300 121 38 1.4 ... ... 12.5 1.13 Ries,Germany -18 ... 24 2 6 186 Ragozinka, Russia -13 ... 9 4 55 5 7 5 220 3.1 13 2 7 12 <100 55 7 368 RoterKarnm, Namibia Sa'fiksj'firvi, Finland -9 -9.3 -6.5 ... ... ... SaintMartin,Canada -13.5 -11 Shunak, Kazakhstan Sierra Madera, Texas -2.5 -1.5 ... ... Siljan, Sweden -15 S6derfj'•den, Finland Steinheim, Germany -6 -2 ---•i• ... 23 2.5 40 6 3.8 SteenRiver,Canada -11 ... 25 Sudbury, Canada -30 ... 200 Tenoumer, Mauritania -10 Upheaval Dome,* Utah 0 . '• Vredefort, SouthAfrica -25 Wanapitei Lake,Canada -15 . WestHawkLake,Canada WolfeCreek,Australia -6 -2 ... Wells Creek, Tennessee -3 Zhamanshin, Kazakhstan-6 5 '6 --•i• 1.9 2 7 4 <400 0.2 14.8 5 514 550 14.8 4 95 6 1850 3 Shoemaker etal.[1989] Fox[1970] Offield and Pohn [1979] Dent[1973] Wickman [1988] Langan[1974] Roddy [1977] Buschback and Ryan [1963] Barlow[1979] Thomas andInnes[1977] Pohletal.[1988] Beals[1960] Dabizha andFeldman [1982] Dabizha andFedynsky [1977] Maslov [1977] Tudor[1971] thisstudy Thomas etal. [1978] 77.3 Elo [1976] 0.052 Fudalietal. [1980] 212 Sweeney [1978] Nicholson Lake,Canada -7.5 Pretoria SaltPan,SouthAfrica -5 Rouchouart, France 6 2 Innes [1964] 3.5 Dabizha and Feldman [1982] -5.5 -3 -1.8 7 6 320 Gosses Bluff,Australia GowLake,Canada Holieroral, Canada 22 4 3 Williams [1990] FudaliandCassidy [1972] 450 7 5 Reference 2.5 thisstudy Innes[1964] Denceetal. [1968] Fudalietal. [1973] Pohletal. [1977] Pohletal. [1978] Vishnevsky andLagutenko [1986] Fudali[1973] Eloetal.[1990b] thisstudy Feldman etal. [1979] VanLopikandGeyer[1963] Dyrelius [1988] Lauren etal.[1978] Ernstson [1984] thisstudy Popelar [1972] FudaliandCassidy [1972] 5 7 7 1970 7.5 5 37 Dence andPopelar [1972] 3.15 .875 4 2 100 <0.3 HallidayandGriffin[1963] Fudali[1979] 140 14 13.5 6 3 <65 Steinemann [1980] tlenkel[1982] 200 0.9 Joesting and Plouff[1958] Slawson [1976] Stearns etal.[1968] Florenskii etal.[1979] Residual gravity anomalies arein milligals. C. P.denotes themagnitude of thecentral gravity highif present. E denotes erosional level[Grieve andRobertson, 1979]:1,ejecta largely preserved; 2, ejecta partlypreserved; 3, ejecta removed, rimpartly preserved; 4, rimlargely eroded, crater-fill products preserved; 5,crater-fill products partly preserved; 6,remnants ofcrater-fill preserved, crater floorexposed; 7,crater floorremoved, substructure exposed. *The exact value is unknown. *Heavily eroded structures overwhich onlythecentral highremains. These values arenotplotted inFigure 4. 29, 1 / REVIEWSOF GEOPHYSICS 166ß Pilkington andGrieve:SIGNATURE OFTERRESTRIAL IMPACT CRATERS [] • [][] 01_+ x TARGET ROCKS CRATER CENTRAL HIGH PRESENT NO CENTRAL HIGH CRYSTALLINE + m SEDIMENTARY X ß DTAMETER (km} Figure 4. Variationin themaximumnegativegravityanomalywith craterdiameter.Line marked BAZ showsthe variationpredictedby the simplehemispherical cratermodelof Basilevsky et al. []983]. contribute to thecentralgravityanomalies.As thetransient cavityfloorrebounds in complexcraters, particlevelocities in the central portion are directed upward and inward, WOLFE leadingto a stateof compression in the resultingcentral CREEK D:O.875kmuplift. Hence initial impact-induced porosityis reduced, and densityincreasedwith respectto the surrounding mGals craterfloor [Grieve, 1988]. For smaller structures(D < 30 km) the existenceof a -'15 • WANAP ITEl LAKE D=7.5km mGoIs _J-•• -10 -.__.i_ • centralrelativegravityhigh is primarilydetermined by the preimpacttargetlithologies.For example,at Lake Lappaj'•vi, Finland,the 3-mGal centralhigh is causedby upliftedshockedPrecambrian granodiofites and gneisses havinghigherdensitiesthanthe surrounding brecciasand sediments[Elo, 1976]. At Carswell,Canada,the gravity highof 11 mGalis theresultof a centralcoreof fractured LAPPAJARVI basement gneiss (p=2610 kg/m 3)rin4ged byupto1500m of D=17km Athabasca sandstone (p = 2490 kg/m•) andsmalleramounts ofdolomite (p= 2850kg/m 3)[Innes, 1964].Theopposite mayalsooccur.Forexample,thecentralupliftat Steinheim, Germany,has broughtlower-densityDogger and Lias WEST claystones and sandstones to the surface,resultingin a CLEARWATER LAKE greatercontributionto the negativeanomalythan that D=32km caused by fracturing andbrecciation alone[Ernstson, 1984]. As crater diameters increase past 20-30 krn, the maximum negativegravity anomaly appearsto reach a limitingvalueof-30 mGal (Figure4). Increasingcrater MAN!COUAGAN diameterthen has a negligibleeffect on the anomaly D='lOOkm amplitude,and we can replaceour simpledisk with an infinite slab model (Figure 6). The limiting gravity Figure 5. Residualgravityanomalyprofilesover impactcraters anomalyvaluecan be interpretedin termsof a maximum scaledto crater diameterand maximum gravity anomalyvalue. depthto thebaseof the fracturedzoneassociated with the Arrows indicate crater rim. All but Wolfe Creek, Westem crater [Basilevskyet al., 1983]. Using the infinite slab Australia,are complexcraters.Profile for Lappaj'firvi,Finland, formula fig = 2tcGzAp,where G is the gravitational does not crossthe 3-mGal central high noted in the text and constant, Ap is the densitycontrastbetweenfracturedand Table 2. undisturbed targetrocks,and z is the depthextentof the mGols • mo,s ._/ Pilkington andGrieve:SIGNATURE OF TERRESTRIAL IMPACTCRATERS ß 167 30, 2 / REVIEWSOF GEOPHYSICS havinglargeranomalies for a givencraterdiameter.Thiscan be explainedpartiallyin termsof the initial densitiesof the targetrocks.If crateringcauses10%voidsin a sandstone (p = 2400kg/m 3)ora gneiss (p= 2700kg/m3), theresulting densitycontrasts betweennormalandfracturedmaterialare INFINITE 220 and250 kg/m 3, respectively. This interpretation, SLAB 100 ;• • • •_•110' • however,cannotbe supportedby actual sample density measurements becauseof the paucityof availableinformation.Nevertheless, the singlesedimentaryexamplein Table 1 doeslie at thelowerendof thesampledensityrange. The erosionallevel of cratersappearsto have only a minoreffecton thecorrelation betweengravityanomalyand craterdiameter.Figure7 showsthatin the caseof complex craterstheheavilyerodedstructures, with an erosionallevel E = 6-7 (seeTable 2 captionfor an explanationof E), tend to have smaller gravity anomalies.For a given diameter range,however,anomalyamplitudeshowssomevariation :• • •_•110• CRATER DIAMETER within a particular erosional level(Figure 7, inset). In the (km) definition, however, of the observed erosional level Figure 6. Comparison betweenmaximumgravityanomalies due [Grieve and Robertson, 1979] the amount of crater fill to an infinite slab and single-diskmodels.The disk modelshave productsremoved and/or the degree of removal of the constantdepths(z) in kilometersand densitycontrasts(Ap) in disturbedzone beneaththe crateris poorly constrainedfor kg/m 3. highly erodedstructures(E = 6-7) and allows for significantvariationin thecorresponding gravityeffect. fractured zone, fora slabwithAp= 100kg/m 3anda depth extentof 8 km, the maximumgravityanomalyproducedis Modeling From the precedingdiscussionwe have seen that the -33 mGal, similar to the maximum crater anomaly. Differentdensitycontrasts wouldresultin differentdepth primarycontrolson anomalyamplitudeare the depthof extents.Nevertheless, 8 km is the depthbeyondwhichit is the brecciatedand fracturedzones(which scalewith crater expectedthatfracturesare essentially closedby lithostatic diameter)and their densitycontrastswith the surrounding targetlithologies.Effectsdue to the exactcratergeometry, pressure [PerrierandQuiblier,1974]. erosionallevel, and preimpact target lithologiescan be considered secondary. Consequently, simple source Effectof TargetRockandErosion canbe usedin modelingthe gravitationaleffect Figure 4 also shows a distinctionbetween craters geometries Basilevskyet al. [1983] showedthatif formedin sedimentary andcrystallinelithologies,thelatter of impactstructures. 3 Figure 7. Variationin the maximum negative gravity anomaly 6 7 with crater diameter. Each crater 5 , 7 3 is labeledby its erosionallevel. See Table 2 for an explanation 5 2 of erosional 2 2 level values. Inset shows the maximum negative gravity anomalyversuserosional level for complexcraterswith 6 4 3 diameters between 6 and 16 km. Symbolsasin Figure4. 4001 i 2 :• • .•3•I,10 ' • CRATER DTAMETER 3 i 4 EROSIoNAL I :• • • •'•3•1,10"• (km} i 5 6 LEVEL 168 ß PilkingtonandGrieve: SIGNATUREOF TERRESTRIAL IMPACTCRATERS thefracturedzoneis modeledas a hemisphere with a radius equal to that of the crater and a densitycontrastof 100 kg/m 3,thentherelation between AgandD issimply-Ag (mGal) = D (km). This approximation providesa reasonable fit for craterdiameterslessthan--5 km, althougha significant amountof scatteraroundthe line occurs(Figure4). Factorscontributingto the misfit includeambiguitiesin regional/residual separationof the gravity anomaly,the preimpactdensity distribution,and post-impacttectonic history.The hemispherical fracturevolumemodel,however, has no directphysicalbasisand clearlyoverestimates the sizeof thegravityanomalyat D > 20-30 km (Figure4). A morerealisticapproachto predictingthegravitational effects of impact craters should incorporate known observational dataconcerningcraterstructureand make,if possible,the distinctionbetween simple and complex forms.In this studywe useknown morphometric scaling relationshipsto develop simple models relating crater diameterD and gravity effect. Simplecratersare characterizedby a bowl-shapedform, where the apparentcrater Crystalline 29, 1 /REVIEWSOF GEOPHYSICS dt=0.200 ø'3 da=0.150 0'4 dt=0.520 0'2 (4) (5) (6) wheretheda valuesarefromGrieveandPesonen [1992] andthedt valuesarefromGrieveandRobertson [1979]. Figure9 alsoshowsthe variationin Ag versuscratersize for our simpletwo-diskmodelusingthe complexcrater scalingrelations(3)-(6). The modelclearlyunderestimates thegravity anomaly, even forthelarger-density 3contrasts considered (APt = 250kg/m 3,Apo= 150kg/m ), which represent an upperboundto theexpectedrange(Table 1). Moreover,the existenceof significantgravityanomalies overheavilyerodedcomplexcraters(i.e., craterswherethe allochthonous depositshave been removedand the crater floor exposed)indicates that the volume of fractured/ disturbed target rocks involved in the formation of complexcratersis muchgreaterthanthatindicated by the depthtothecrater floor(dt). A measure of thedepthof thedisturbed zoneatcomplex (depth da)maybefilledwithpostimpact sediments andthe truecraterfloor(depthdt) is markedby thebaseof the cratersis givenby the structuraluplift (SU) definedas the allochthonous breccialens. From morphometricdata over amount of uplift undergoneby the deepesthorizon a number of simple craters the following empirical presentlyexposedin thecentraluplift [Grieveet al., 1981]. relationships,which are independentof target lithology, Figure 10 showsthe amountof structuraluplift versus havebeen determined[Grieve et al., 1989]. da=0.13D 1'06 (1) dt= 0.2801'02 (2) craterdiametercomparedwith the maximumdepthsof reduceddensityfrom previousgravity modelsof several complexcraters.The good agreementbetweenthe two suggests thatSU providesa usefulestimate of thedepthof the fractured zone. For diameters above-30 km the lack of dependenceof Ag on D suggeststhat the infinite slab model is appropriate.As D increasesabove~30 km, the where dimensions are in kilometers. We modela simplecraterusingtwo disksof diameterD depthof the fracturedzoneappearsto reacha limiting andD - 2da, thickness da and(dt -da) , anddensity value,andthegravityeffectis essentiallyconstant. contrasts APt,Apb(Figure8). Our modelassumes postimpactsedimentary infill whichis foundevenin relatively youngcraterssuchas the49,000-year-oldBarringercrater. MAGNETIC SIGNATURE Figure 9 shows the maximum gravity anomaly Ag calculatedover sucha modelusing(1) and(2) anddensity General Character The generalcharacterof magneticanomaliesassociated contrasts asindicated. ForAPtandApbbothequalto 100 kg/m 3 the variation in Ag is similarto thatof the with impactcratersis morecomplexthan the gravity Basilevskyet al. [1983] model.Figure 9 showsthat disk models with density contrasts between 100 and 200 kg/m 3 lie within the scatter ofthe observations for simple craters. GRAVITY Previous modeling of gravity data at simple craters [Fudali ANOMALY and Cassidy, 1972; Grieve et al., 1989] confirms that apparentand true craterdepthsfrom (1) and (2) provide goodestimates for modelbodydimensions, if theappropriate densitycontrastsare used.Thereforereduceddensities due to fracturingof autochthonous rocksbeneaththe crater floor do not appear to contributesignificantlyto the gravityanomalyat simplecraters. For complexcratersthe scalingrelationships between apparentand truecraterdepthanddiameterare Sedimentary da= 0.120 ø'3 (3) Figure 8. Geometry of the two-disk crater model and its associated gravityanomaly. 30, 2 / REVIEWS OF GEOPHYSICS Pilkington andGrieve:SIGNATURE OF TERRESTRIAL IMPACTCRATERS ß 169 C9 L• Z - _ <I: r'rC9 - • ••.o o • • ••_4•.o, CRATER • • ••.4•.½ DIANETER (km) relationships (equations (3-6)). The densitycontrasts of Figure9. Variationin themaximum negative gravityanomaly scaling with diameterof the two-diskmodel for simple and complex theupperandlowerdisksin complexcratersare 250 and 150 3,respectively. Verticaldashed linesdenote thediameters craters. AptandApt , arethedensity contrasts of theupper kg/m associated with the transition from simple to complex formsfor (sedimentary infill) andlower(breccia lens)disksfor thesimple (2 km) andcrystalline(4 km) targetrocks.Symbols cratermodelin kg/m 3. Simplecraterdiskthicknesses are sedimentary calculatedfrom scalingrelationships (equations(1) and (2)). Disk thicknesses for the complexcratersare calculatedfrom uj O - T - I-- _ . . MIEe • oCLW MIDo .•'LAP o eHAU 0_.. asin Figure4. signaturebecauseof the muchgreatervariation(ordersof magnitude)in the magnetic propertiesof rocks. The dominanteffectovercratersis a magneticlow [cf. Dabizha and Fedynsky,1975; Clark, 1983] ranging in amplitude from tensto a few hundrednanoteslas (nT). This type of signature is most easily recognized, particularly in crystallineenvironments, by the truncationand disruption of regional magnetictrends(Figure 11). The magnetic lows are best defined over simple craters, where the anomalyis smoothand simple, for example,Barfinger, Arizona,andWest Hawk Lake, Canada(Figure 12). Lower field intensifies arealsofoundoverlargerstructures, where the reducedfield can be modified by the presenceof shorter-wavelength anomalies.Theselocalizedanomalies are usuallyof large amplitude,for example,+ 1000 nT, and occur atornear thecrater center (Figure 13).Table 3 summarizesthe magnetic anomaly character over 37 impact structures.As in the gravity case, there is not a Figure 10. Resultsof modelingof gravity anomaliesover one-to-onecorrelationof the characterof the magnetic complexcratersby previousworkers(seeTable2 for references anomalyandcratermorphology.In addition,the presence and detailsof the modelsused).Depthto modelbaseis shown of a centralgravityhigh doesnot imply the existenceof an versus crater diameter. Also shown is the amount of structural centralmagneticanomaly.A broadcorrelation uplift(SU) calculated fromGrieveet al. [1981]versusdiameter. associated does, however, exist between anomaly form and crater SOD, S0derfj'•rden;MID, Middlesboro;MIE, Mien; LAP, with D < -10 km all havemagneticlows, Lappaj'firvi;DEE, Deep Bay; ELG, El'gygytgyn; HAU, size.Structures Haughton; RIE, Ries;SIL, Siljan;CLW, WestClearwater Lake; while all structures with D > 40 km exhibit central MAN, Manicouagan. high-amplitude anomalies. CRATER 0IAHETER (km) 29, 1 /REVIEWS OF GEOPHYSICS 170 ß Pilkington andGrieve:SIGNATURE OF TERRESTRIAL IMPACTCRATERS / / / /I ! \ / / / 100 nT 20nT • ... • / / / / .•-• / / / 1 } / 0 I / - • km I I)/ 2 I // Figure 11. Residualmagneticfield intensityover DeepBay, Saskatchewan, Canada.Contourintervalis 20 nT, flight height300 m. Contrary to the gravity case,severalimpact sites are marked by the absenceof significantchangesin the observedmagneticfield (Table3). For someof thesmaller structuresthis may be, in part, due to attenuationand undersampling of any anomalouseffectsrelatedto flight heightand line spacingin aeromagnetic surveys.Unfortunately,groundsurveydata are only availableat a small numberof craters.In the caseswheremagneticanomalies relatedto impactare present,severalmechanisms related to the impactprocessmay be operative,all of whichmay be sufficientto radicallychangethe magneticproperties of the targetrocks. Shock Effects Peak pressuresrecordedin the autochthonous rocksof the centraluplift at complexcratersmay reach-30 GPa, whichis sufficientto produceshockdemagnetization and remagnetization effects.Experimentalstudieshaveshown that shockpressuresof the order of 1 GPa can remove existingremanentmagnetizations [HargravesandPerkins, 1969; Pohl et al., 1975; Cisowskiand Fuller, 1978]. At pressuresof >10 GPa, Kumar and Ward [1963] have also detecteda reductionin magneticsusceptibilitydue to shock.Similarly, decreasesin susceptibilitylevels have beendetectedat nuclearexplosionsites[Short,1965].The effects of shock metamorphismcan also aid in the productionand modificationof magneticcarriers;for amphiboleand biotitedecompose to producemagnetite [Chao, 1968]. At lower pressures, titanomagnetite can resultfrom thebreakdownof ilmenite[Chao, 1968]. In additionto demagnetization, targetrockscan also acquirea shockremanentmagnetization (SRM) in the directionof the Earth'sfield at the time of impact.The intensityof SRM is proportionalto the ambientfield strength[Pohl et al., 1975] and decreaseswith distance fromthe pointof impact[Cisowski and Fuller, 1978]. Although thepresence of SRM hasbeendemonstrated for nuclearexplosionsites[Hargravesand Perkins,1969], its recognitionis rare at impactcraters.Cisowskiand Fuller [1978], however,detecteda secondarycomponentof magnetization probablyacquiredat the time of impactat Barringer,Arizona,and Lonar, India, while Halls [1979] hasdocumented the existence of a secondary magnetization at Slate Islands, Canada, which exhibits several propertiesin keepingwith thoseof SRM as determined experimentally[Cisowskiand Fuller, 1978]. Halls [1979] foundthatthe intensityof the SRM decreases awayfrom thepointof impactandis restricted to thelow coercivity fraction. The remanencewas also acquiredrapidly (between impactandtheformation of thecentraluplift),as evidenced by a small directional scatter. Studies at Charlevoix,Canada,suggestthat the large reductionin remanence intensityin samples fromthecentralupliftis shockrelated[Robertson and Roy, 1979]. However,the example,at pressures of >40 GPa and T > 1000øC, high coercivityof the carrierphase(titanohematite) at 30, 2/REVIEWSOF GEOPHYSICS Pilkington andGrieve:SIGNATURE OF TERRESTRIAL IMPACTCRATERS ß 171 95ø09'W Charlcvoixmay explainwhy an acquiredSRM is absent. SRM is mostlikely to occurin autochthonous targetrocks experiencing pressures greaterthan1 GPabuttemperatures lessthantheCuriepointsof themagneticphases present. 95o{5'W Thermal Effects For material occurringclose to the point of impact, postshock temperatures are greatenough(>1000øC)to causewholerock melting,whichcanresultin the production of nonmagneticimpact glasses[Pohl, 1971]. Unlike impactglasses,slowercooledcrystallineimpactmelt rocks canacquirea thermoremanent magnetizationCrRM) in the direction of the Earth's magnetic field at the time of impact.MagneticresettingthroughTRM in the ambient field directionhas led to a number of palaeomagnetic dating studiesat impactcraters[Angenheisterand Pohl, 49 ø 1964; Larochelie and Currie, 1967; Robertson, 1967; Currie and Larochelie, 1969; Pohl and Soffel, 1971; Bylund, 1974; Cisowski, 1988; Pesonen and Marcos, 1990]basedon samplesof impactmelt rocksandbreccia. lkm Figure 12. (Opposite) Total magneticfield intensityoverWest Hawk Lake, Manitoba, Canada [after Clark, 1980]. Contour intervalis 100 nT, flightheight300 m. Figure 13. Residualmagneticfield intensityover Ries, Germany[after Pohl et al., 1977]. Contourinterval is 5 nT, flight height 1000 m. Dashed line shows crater rim. The large positiveanomalyin the southwestern part of the crateris causedby basic and 43o80 ' 45ø90' 0 I krn 4 I 44ø00' 44øt0' ultrabasic basement rocks unrelatedto the impactevent[Pohl et al., 1977]. 172 ß PilkingtonandGrieve: SIGNATUREOF TERRESTRIAL IMPACTCRATERS A stableremanenceand low directionalscatterappearto be characteristics of impact melt rocks and most likely reflecttherapidacquisitionof themagnetization [Pohland Softel, 1971]. At lower temperatures, remagnetization of magneticmineralsthroughthe acquisitionof a TRM also occurs.Even below the Curie temperature, a partialTRM may still be acquired. Chemical Effects 29, 1 /REVIEWSOF GEOPHYSICS 1985]. Oxidationof magnetiteto hematitehas led to the formation of a highcoercivityCRM at Siljan,Sweden,asa resultof circulationof meteoricwater throughimpactinducedcracksand fissures[Elmingand Bylund,1991]. Althoughpostimpact processes, suchas chemicalweathering, leaching,and metamorphism, will furthermodify magneticpropertiesover large time intervals,the CRMs underdiscussion can be contemporary with the impact event. Low-temperature (<130øC) oxidation of titanomag- Following impact, elevatedresidualtemperatures and hydrothermal alterationcanproducenew magneticphases leadingto the acquisitionof a chemicalremanentmagnetization(CRM) in the directionof the ambientfield. The central magnetic anomaly at Saint Martin, Canada, is netireto maghemite at Ries,Germany,occurredat thetime of impact,resulting in boththeoriginalandalteredphases showingthe same magnetizationdirections[Iseri et al., 1989]. attributed to the formation of hematite from the breakdown Modeling of mafic silicatesin the floor of the centraluplift [Coles Clearly, severalmechanisms can resultin the formation and Clark, 1982].Fracturingof the targetrocksaidsin the of remanentmagnetizations both during and after the circulationof hydrothermalfluids, and the presenceof crateringevent. No one magneticphaseseemsto be oxygen favors higher magnetizationintensifies[Grant, characteristic of impactsites.Magnetizations havebeen TABLE 3. MagneticAnomalyCharacterfor 37 TerrestrialImpactStructures Crater Barringer, Arizona Bosumtwi, Ghana Brent,Canada Carswell, Canada Clearwater East,Canada Clearwater West,Canada Crooked Creek,Missouri Decaturville, Missouri DeepBay,Canada El'gygytgyn, Russia Gosses Bluff,Australia Haughton, Canada Holleford, Canada Janisj'grvi, Russia Kara,Russia LacCouture, Canada LacLaMoinerie, Canada Lappaj'firvi, Finland Logancha, Russia Manicouagan, Canada Manson, Iowa Mien,Sweden Mistastin, Canada NewQuebec, Canada Nicholson Lake,Canada PilotLake,Canada Ries,Germany Ragozinka, Russia SaintMartin,Canada Serpent Mound,Ohio Shunak, Kazakhstan Siljan,Sweden SUdbury, Canada Vredefort, South Africa Wanapitei Lake,Canada WestHawkLake,Canada Anomaly Diameter, Ion L/S C L/S C L/S L/S N L/S L/S L/S C C L/S L/S C L/S N L/S L/S C C C N N L/S N C L/S C N L/S C C C L/S L/S Zhamanshin, Kazakhstan L/S E 1.18 10.5 3 39 22 32 7 6 13 18 22 20.5 2.35 1 3 4 7 4 5 6 4 5 3 7 2 5 14 65 8 8 17 1.83 100 35 9 28 3.44 12.5 5.8 24 9 40 6.4 6 5 6 7 6 2 5 6 6 6 2 6 6 2 4 5 2 3.1 55 200 140 7.5 3.15 13.5 2 7 6 7 5 4 3 Age,rn.y. 0.049 1.3 450 115 290 290 320 <300 100 3.5 142.5 21.5 550 698 73 430 400 77.3 50 212 65.7 121 38 1.4 <400 445 14.8 55 220 <320 12 368 1850 1970 37 100 0.9 Reference Regan andHinze[1975] Jones etal. [1981] Millmanetal. [1960] Currie[1969] thisstudy thisstudy Fox[1970] OffieldandPohn[1979] Innesetal. [1964] Dabizha andFeldman [1982] Young [1972] thisstudy Beals[1960] Dabizha andFeldman [1982] Maslov[1977] thisstudy thisstudy Eloetal. [1990a] Feldman etal. [1983] Coles andClark[1978] Hartung etal. [1990] Henkel[1982] thisstudy thisstudy thisstudy thisstudy Pohletal. [ 1977] Vishevsky andLagutenko [1986] ColesandClark[1982] Sappenfield [1950] Dabizha andFeldman [1982] Pedersen etal. [1990] Gupta etal. [1984] Corner etal. [1990] Clark[1981] Clark[1980] Florenskii etal.[1979] E denotes erosional level(seeTable2 footnotes fordetails). L/S:magnetic low,subdued anomalies; C:central magnetic anomalypresent;N: no apparentanomalous effects. 30, 2 / REVIEWSOF GEOPHYSICS Pilkington andGrieve:SIGNATURE OFTERRESTRIAL IMPACTCRATERS ß 173 observedin carrierssuchas magnetite[Larochelieand magneticfieldsobservedover cratersin suchterranesare in susceptibility levels.As in Curtie, 1967; Pohl, 1971], hematite-ilmenite[Robertson primarilydue to a decrease the case of density data, there are relatively few studies and Roy, 1979;Colesand Clark, 1982],and pyrrhotite [Pesonenand Marcos, 1990]. As well as the effectsof thatcomparethe magneticpropertiesof fracturedautochoutside remanence,the intensity of induced magnetization thonousrocksand their undisturbedcounterparts (essentially dueto thepresence of magnetite) mustalsobe the crater. Masaitis et al. [1975], however, have found taken into account, when consideringthe magnetic nearly a tenfold decreasein susceptibilitybetween anomalies over craters. undisturbedand fracturedbasementgneissesat Popigai, The anomalous field is determined by thevectorsumof Russia.The dominantmechanismproducingthisdecrease thefieldsdueto boththeremanentandinducedmagnetiza- is notclear,althougha reductionfrom shockandenhanced alterationeffectsfrom fluids throughout tion, with the relativemagnitudeof the two givenby the low-temperature the fractured zone [Henkeland Guzman,1977] are likely Kanigsberger ratio Q. Brecciasand impactmelt rocks to be significant. generally showQ valuesmuchlargerthanunity:Q < 50 at Haughton,Canada[Pohl et al., 1988], Q > 10 at Rochechouart, France[Pohl and Soffel,1971], Q ~ 15 at Dellen,Sweden[Bylund,1974], Q - 10 at Mien, Sweden [Stanfors, 1973],andQ ~ 10at Ries,Germany[Pohlet al., 1977]. Therefore,at least in thesecases,inducedmagnetizationcanbe considered negligiblein thesecases,and the observedmagnetic anomaliesare mainly due to remanentmagnetization. This is obviouslythe casefor thoseanomalieswhoseshapeclearlyindicatesa causative ELECTRICAL SIGNATURE General Character Brecciationand fracturing of target rocks indirectly causeslarge changesin their electricalproperties.The conductivityof rocksis heavilydependenton their water content.A less than 1% change in water content can producemore than an order of magnitudechangein magnetization in a direction otherthantheambient field. The degreeof fragmentation determines the Quantitative analyses of therelativelyshortwavelength, conductivity. amount and distribution of fluids within the rock and hence centralmagneticanomalies at a limitednumberof craters The largeincreasein conductivity haveall successfully modeledtheseintenseanomalies as its electricalproperties. of fluid-filled fractured material results in electrical the productof remanentlymagnetizedbodies[Young, 1972; Coles and Clark, 1978, 1982; Henkel, 1982]. We methodsbeingpotentiallyusefulin mappingthe structure have found no documentedcasesof inductively mag- of impactcraters. netized bodies. The source of these central anomalies is wideranging.Magneticbasement maybe exposedin the central uplift at highly eroded structures(Carswell, Vredefort),wherethe anomalous field coincideswith the basementoutcrop.Alterationzonesof small areal extent may occurwithinthe centraluplift (SaintMartin)or the centralanomalymaybe causedby thepresence of highly magnetic impactmeltrocks/suevite/breccias (Mien,Ries). The sourceof the magneticlowsover cratersis more problematical. Undoubtedly, theimpactprocess results in a reductionin magnetization (remanent,induced,or both) intensityof thetargetmaterial.In thecaseof freshcraters (E < 3, Table3), postimpact sedimentary infill will tendto be nonmagnetic, if CP,M mechanisms are not operative [Wasilewski,1973], and so contributeto the reducedfield intensity.Similarly,Bealset al. [1963] suggestthat for relativelyunerodedcratersin nonsedimentary targets,the magneticlow may be causedin part by the random orientation (dueto disruption duringthecrateringevent)of magnetization vectorsin crystallinerocksin the breccia lens.Thesecausescannot,however,explainthe low fields overthe moreerodedstructures. By analogyto thegravity casean importantcontribution to themagneticfield must come from the autochthonous basement beneath the crater floor. In Precambriancrystalline terranes,most magnetic anomaliesare due to inductive magnetization[Grant, 1985], which leads us to suggestthat the diminished Resistivity Few examplesof electricalinvestigations of impact structuresexist, however, in the literature and most of these involve DC resistivitymethods.Where a distinct contrastexistsbetweenthe allochthonous brecciadeposits and the underlyingautochthonous targetrocks,electrical profilingusingresistivitysoundingcan map the structure of the truecraterfloor, for example,at Ragozinka,Russia [Vishevsky andLagutenko,1986]. At theRiesstructurethe resistivity contrast between lake sedimentsand the underlyingsuevitehasallowedthe boundaryof the inner craterto be determined[Ernstson,1974]. Drilling at Ries showsthe increasein resistivitywith depth, from lake sediments(<10 •m), to suevite (10-90 •m) and brecciatedbasement(100-300 •m). Resistivity soundings at Kaalijarvi, Estonia, also exemplifythe changein electricalpropertiesas a function of the degreeof fracturing [Aaloe et al., 1976]. Nearsurface,stronglyfractureddolomitesshowresistivitiesof 130-150 •m and are underlainby lessfracturedmaterial (.-250 •m) and relativelyundisturbed targetrocks(>300 •m). Florenskiiet al. [1979] were able to detectthe base of allochthonous depositsat the edgesof the Zhamanshin structure,Kazakhstan,basedon a resistivitycontrastof 20-100 •m at depth.As in the caseof refractionseismic surveysoutlininglower velocitiesoutsidethe craterrim, electrical profiling has detected lower than normal 29, 1 /REVIEWS OF GEOPHYSICS 174 ß Pilkington andGrieve:SIGNATURE OF TERRESTRIAL IMPACTCRATERS surveyat Barringercraterhasshownthat resistivities up to onecraterdiameterbeyondthecraterrim A reconnaissance [Aaloe et al., 1976; Clark, 1980]. For example,at West featuressuchas the craterwall and the baseof the ejecta Hawk Lake, Clark [1980] found resistivitiesincreasing blanketcan be mappedwith this approach[Pilon et al., from 50 to 300 f•m near the crater rim to ~8000 f•m at ~3 1991]. Figure 14 showsan exampleof a raw and interkm from the center of the structure. pretedsectionacrossthe exterior ejecta blanket at the Barringer structure.Indicationsof the ejecta-bedrock contactcan be seen, as well as subsurfacelithological Magnetotellurics In order to determine the deeper electrical structure contacts and faults. associated with impact,magnetotelluric (MT) surveyshave beencarriedout at Siljan [Zhanget al., 1988] and Charlevoix [Mareschaland Chouteau,1990].At Siljan theMT SEISMIC SIGNATURE resultsindicatea conductiveregion,1000f•m comparedto 10,000f•m for normalgraniticcrust,thatapproximates the General Character horizontalextentof thecraterbutoccursat a depthof 5-20 Seismicrefractionand reflectionsurveysprovide a km. No surfaceconductivezone was found.Zhang et al. detailedimage of the subsurfacestructureof impact [1988] invoked fluid migration throughimpact-induced craters.Both techniquesare capableof determiningthe fractures at these depths to explain their results. At velocitydistribution,which reflectsthe degreeof fracturCharlevoix, Mareschal and Chouteau [1990] found that ing as a resultof shockstress.The relativechangein culturalnoiseprecludeda definitiveinterpretation of the seismicvelocitiesdue to impact-induced fracturingand near-surface structure. They found, however, that a brecciation aremuchlargerthanthecorresponding change subhorizontal conductorwith resistivity100 f•m compared in density,for example,up to 50% for velocities compared to an uppercrustalaverageof 10,000f•m at 1.5 km depth to a 5% differencein densities at Barringer.For complex was suggestedby high-frequencydata. This featurewas craters,in particular,reflectionseismicsurveyscanlocate interpretedasa subhorizontal faultrelatedto impact. andmapmajorstructural featuressuchasthecentraluplift, peripheraldepression, distributionof craterfill, faultedrim and the possibletraceof the transientcavity. Due to the Ground-Penetrating Radar For detailed subsurfacemapping, ground-penetrating high resolving power of the reflection method, the radar has been shown to have utility at relatively small estimatedmorphological parametersin the seismicrecord impactcraters[Pilon et al., 1991]. As with conventional can be used in testing and refining models of crater electrical methods,the presenceof water within pore structureandcrateringmechanics. spacesstronglyinfluencesthe responseto the radarpulse. Fracturinghas a major effect on the propagationof In particular,the depthof penetrationis dependenton the seismicenergy,with the presenceof discontinuities and electromagnetic absorptionpropertiesof the subsurface. voidscausinglower thannormalvelocities.The effectsof Reflecting horizons are caused by dielectric contrasts, fracturingshouldbe distinguished from straightforward whichmay or may not correspond to lithologicalchanges. changesin porosity,which also lead to density and Distance (m) ..... .... ,. 7. I,t .... •- '•...::•...... •., .•....... ,• .•. .•,•.... ,.,•,•,., ..... 0.......,•,.,,•,••,,,,,,.•' •, ..•., Distance • 5o ¾ ,•o ,•o too I t•o I too ! •o •o I •o qo •o __ 109- -• I -3.8 - _ • 173E .-- I Moenkopi Formalion -7.0 Kaibab Formation _ _ 237- • •o1• / •'• "' / / /•/ • Geological contact / / Fault --16.6 ............. Parabolic echo frommain powerline Figure 14. Ground-probing radar data acrossexteriorejecta area seriesof subhorizontal subsurface lithologicalcontactsand blanketat Barringercrater,Arizona.The craterrim corresponds high-anglefaults. The large parabolic reflectorsat stations to station 708. (Top) Processedradar data. (Bottom) Line 580-660 are due to an overheadpowerline. The datahavenot interpretationof data. The ejecta-bedrock contactis visible, as beentopographically corrected. 30, 2/REVIEWSOF GEOPHYSICS PilkingtonandGrieve:SIGNATUREOF TERRESTRIAL IMPACTCRATERS ß 175 velocitychanges. The natureof the fracturingdetermines foundat the Kaalij'•vi structure[Aaloeet al., 1976]. At complex craters, such as Charlevoix, refraction the sizeof velocitychange.If two materialshavethesame porosityandbulk density,butonecontains a largenumber seismicshave revealedlower than averageupper crustal of smallmicrocracks as opposedto a few largefractures, velocities (6.2 versus 6.4 km/s) coincident with the disturbance[Lyonset al., 1980]. At Ries, seismicparticledisplacements are small enoughthat the impact-induced smallerfractureswill leadto a greaterdecrease in observed modelingof refractiondata has,basedon initial velocity velocities [Kuster and Toksoz, 1974]. If the crater is estimates fromreflectionvelocity-depth spreads[Pohland buried, then increasedlithostaticpressurewill reduce Will, 1974] that indicatedlower than averagevelocities fractureporosity,but velocitieswill still remainbelow down to at least 3 km, suggestedthe presenceof a their preimpactvalues[Gardneret al., 1974]. The final low-velocityzoneextendingdownto 5-6 km [Pohlet al., velocitydistribution of thefractured zoneis alsocontrolled 1977]. Velocity distributions,however, within impact are not alwayslower thannormal.At the highly by the type of fractureinfilling, with highervelocities structures eroded Vredefort structure, refraction seismics have beingproduced by thepresence of fluids. revealedhigherthanaveragevelocitiesat the centerof the structuredue to the presenceof an upliftedlower crustal Refractionsurveysover simpleimpactstructureshave basementcore [Green and Chetty, 1990]. Structural successfully delineatedthe depthand horizontalextentof informationfrom refraction surveys may also provide the low-velocityzone.Seismicstudiesof simplecratersin constraintson crater size and morphology.At Manson, crystallinetargets,suchas Brent,Canada,showa decrease Iowa, Smithand Sendlein[1971] mappedhighly variable (up to 50-m relief) withinthecraterin of up to 25% in compressional wavevelocitieswithinthe bedrocktopography craterand a depth extentapproximatingthe baseof the contrast to smooth variations outside the crater rim. allochthonousbreccialens [Millman et al., 1960; Sander et al., 1963]. Shock-inducedfracturingwithin sedimentary Reflection The most detailed geophysical information on the target rocks at Barringer results in a 50% velocity decrease,with the reducedvelocityzone corresponding to subsurfacestructureof impact cratershas been gathered sediment infill and the allochthonousbreccia lens [Acker- from reflection seismicstudies.These have, in most cases, man et al., 1975]. The fracturedzone at Barringeralso beencarriedout over complexcratersin bothsedimentary extendsup to onecraterdiameterbeyondthecraterrim as and crystallineenvironments.Figure 15 showsa repreevidencedby lower velocities,which increaseaway from sentativeseismicsectionand its interpretationover the the structure.Similar velocity distributionshave been Montagnaisstructure,occurring off the coast of Nova Refraction IMPACT CRATER 45Km MONTAGNAIS SW 1-94 ?J•.. o Km5 ""• SUEVITE •..,•METAMORPHIC BASEMENT HORIZONTAL SCALE'•l• MELT ROCK HORIZON • • (MEGUMA) Figure 15. Sixtyfold reflectionseismicsectionand corresponding interpretationacrossthe Montagnaisstructure, Nova Scotia,Canada[afterJansaet al., 1989]. 176 ß Pilkington andGrieve:SIGNATURE OF TERRESTRIAL IMPACTCRATERS Scotia, Canada. The section delineates all the major morphologicalfeaturesexpectedover sucha structure.It also exemplifies the dominant seismic signature of impact-inducedfracturingand brecciation,that is, the reducedcoherencyof seismicenergydue to disruptionof the reflecting horizons.The central uplift in a crater is usuallydistinguished by thegreatestamountof disruption, alongwith thrustfaultingwithin the uplift anddownfaulting at its edges[Brenanet al., 1975;Jansaet al., 1989]. Within the uplift the coherencyof the seismicresponse usuallyimprovesaway from and below the cratercenter [Ezeji-Okoye, 1985]. Reflectors within the uplift often show decreasingamountsof uplift with increasingdepth [Juhlin and Pedersen, 1987]. In sectionswhere good markerhorizonsare presentthe baseof the disruptedzone is marked by the reappearance of undisturbedreflectors (Figure 16), and the amount of structuraluplift can be determined[Brown, 1973], althoughintensebrecciationat depthcanprecludeany estimationof how muchuplift has takenplace [Scottand Hajnal, 1988]. In structures wherea clear termination 29, 1! REVIEWS OF GEOPHYSICS --49o45'N STRUCTURE CO ONFISHSCALES 150•-•_. CONTOUR INTERVAL 30m 0 5kin I of marker horizons exists toward the cratercenterthe point of transitionbetweencoherentand incoherentresponses providesan approximatemeasureof the extent of the transientcavity [Juhlin and Pedersen, 1987; Scott and Hajnal, 1988; Andersonand Ilartung, 1991]. The peripheral zone surroundingthe central uplift is againmarkedby somedisruptionof reflectors.Downward displacement of formationsis apparent(Figure 17), with the amountof displacement decreasingwith depth[Brenan et al., 1975]. Knowing the amountof uplift in the crater center and the downthrow of faults in the peripheral depression allowsthecalculationof theamountof material removed from the annular ring and that added to the central uplift [Wilson and Stearns, 1968; Juhlin and Pedersen,1987]. In relatively unerodedstructures, breccia Figure 17. Structurecontourson the baseof the Fish Scalesformation at the Eagle Butte structure,Alberta, Canada [after Sawatzky,1976]. The patternof structuralhigh (centraluplift) and surroundinglow (ring depression)is typical of structure mapsovercomplexcraters. occurringin the annular depressionmay appear as a mappableseismicallytransparent zone[Jansaet al., 1989]. Beyondthe annulardepressionis the craterrim, which is marked by inwardly directed downfaultingof blocks towardthe cratercenter.Target materialin this regionis only mildly deformed so that marker horizonscan be traced accurately [Brenan et al., 1975; Anderson and Hartung, 1991]. Interpretationof seismic sections at Haughton alsoshowan increase in boththespatialdensity and penetrationof faults toward the crater interior [Scott Figure 16. Reflectionseismic sectionthroughthe Red Wing Creek structure,North Dakota. Rim to rim distance is 9 krn. The Cretaceous Greenhorn (Kgh) and OrdovicianWinnipeg (Ow) reflectors are locatedat depthsof-715 and -3625 m, respectively.Be-4 Kgh neath the disrupted zone related to impact is the reap- pearanceof coherentreflectors [afterBrenanet at., 1975]. •Ow 30, 2 / REVIEWSOF GEOPHYSICS Pilkington andGrieve:SIGNATURE OF TERRESTRIAL IMPACTCRATERS ß 177 andHajnal, 1988],reflectingtheradialdistribution of the targets,which include near-surfacenonmagneticsedidisruptiveeffectsof impact. mentary environments, butwherethecentral uplifthasthe potential to bring up magnetic crystallinebasement material. Furthermore,as crater diameter increasesin both SUMMARY AND DISCUSSION Geophysicalinvestigationsat impact structuresreveal that a variety of signaturescan result from the impactinducedphysical changesin target rocks. As a result, attemptsat identifyinga structuresolelyon the basisof geophysicalevidencecan be ambiguous.Nevertheless, from the increasing amount of data that exist over terrestrialimpactcraters,somebroadconclusions can be drawnbasedon the geophysicalresponse(s) of a suspected impactstructure.The reliabilityof any suchconclusionis clearly dependenton the size of the data baseavailable overprobableimpactsfor a particulargeophysical method. In this regard,gravity and magneticobservations are the primary indicators, followed by seismic and finally, electrical methods. The characteristic geophysical signaturesused in determiningthe likelihood of an anomalousstructurebeing due to impact are mainly qualitative,althoughfor the gravitycase,whereanomaly magnitudecan be relatedto craterdiameter,somesemiquantitativecriteriacan be applied.Also, for the purposes of identifyingandcharacterizing possibleimpactstructures solelyfrom geophysical data,we will considerherein the discussiononly buried structuresfor which no surfacebased information exists. In most cases, burial will have sedimentary and crystalline targets, the maximum temperatures and pressuresreachedin the centraluplift increase [Grieve and Cintala, 1991] such that thermal, chemical,andshockmagnetization effectsare morelikely to occur. The electricalresponseof impactcratersappearsto be lesswell definedthan for potentialfields, althoughthe numberof structuresstudiedis, at present,quite small (<10). Resistivitysoundings and profilesshowa gradual increasein resistivityaway from thecratercenterand with depth.Magnetotelluricresultshave proved less informative. Fracturingandbrecciation duringthe impactprocess lead to decreasedseismicvelocitieswithin the crater confines,andthe form of the disruptionof targetmaterial canbe determined fromseismicreflectionsurveys. On the basisof the principalgeophysical anomaly characteristics it is possibleto outlinesomegeneralcriteria whichare consistent with thepresenceof a buriedimpact structure. Potential field anomalies will be near-circular or circularin shape.The gravity field will, in general,show an anomalylow of less than 30-35 mGal. An estimateof the crater diameter D can be made from the extent of the low and the observedAg comparedwith Figure4. The scatterof pointsand the logarithmicscaledoesallow for anappreciable variation in Agfor a givenD andviceversa. been rapid enoughto preservethe major morphological However, we can still reject certain combinations,for featuresof the crater,so consideration of the geophysical example,a 20-km crateris unlikely to have a •gravity signaturesof heavily erodedstructures(E > 6) will be anomalyof <5 mGal. At largediameters,D > 30 km, and avoided. particularlyin geologicsituationswheredensityincreases Considering,first, gravityand magneticmethods,from with depth, such as sedimentsoverlying crystalline our previousdiscussionit is clear that no singularrule basement,there may be a central gravity high. The existsfor relatingcratersize (throughthe diameterD) to diameterof this centralhigh is generally<0.5D and its anomalycharacteror magnitude.Nonetheless,the depth amplitudelessthan the overallgravitylow. A magnetic extentof the low-densityallochthonous brecciazone can low or subduedzone will likely accompanyand coincide be estimated,in the case of simple craters, from the with the overallgravity low. Intensemagneticanomalies morphometricscalingrelationships(1 and 2) and of the of eithersign,however,mayoccurat thecenterof thelow, fractured,structurallydisturbedzone for complexforms, butthediameterof thesewill be generally<0.5D andmost by theamountof structuraluplift.Thusthereis a meansof likelymorerestricted thananycentralgravityhigh. predictinga gravityresponse (Ag) basedon estimatesof D. Accompanyingthe potential field lows will be a If the targetrocksare sedimentary, then it is more likely low-velocityzone that will approximatethe craterarea,as the gravity anomalymay fall below that predictedby the definedby the gravity and magneticanomalies.This zone morphometric scaling relations. Considering crater will be characterizedby incoherentseismicreflections, morphology,the currentdata basesuggeststhat for both with thedegreeof coherency of thereflections increasing gravity and magnetic fields, the dominant effect of a away from and below the anomalousarea. If there arle simple crater is a simple concentricanomaly low. For sufficient high-quality reflectors to resolve subsurface complex craters the transitionfrom only a low to the structure, suchasa centralareaof upliftedlithologies, then possiblepresenceof centralanomaliestakesplace in the the structuralelementsshouldconformto generalmor10-30 km diameterrangefor both typesof potentialfield phometricelementsof terrestrialcraters.For example, data, with cratersof diameters> 30 km tendingto have anomalies with D > 4 km should show evidence of an area central anomalies.Gravity central highs tend to rel'lect of uplift in the center,with the amountof uplift a known denser rocks brought nearer the surface during the functionof D. Finally, a low-resistivityzone coincident crateringprocess.This is alsoalmostcertainlythecasefor with the aboveanomaliesis expected.If theserequiremagneticdata observedover complexcratersin mixed mentsare satisfiedby the geophysicalobservations, then 178 ß Pilkington andGrieve:SIGNATURE OF TERRESTRIAL IMPACTCRATERS 29, 1 /REVIEWS OF GEOPHYSICS the presenceof an impact structuremay be indicated. cumulatively,they will provide importantdata for the Confirmationof an impactorigin for the structureproduc- recordof cratering. ing the geophysicalanomalies,however,can only be ACKNOWLEDGMENTS. The authors thank Alan Goodacre achieved through additional observationswhich are and Pierre Keating for helpful commentson an earlier versionof geologicalin nature. the manuscript. L. Jansa kindly providedthe originalfigureof the Fifty percentof theknowncratersin Commonwealth of Montagnaisstructure.Formalreviewsby S. Honjo, W. McKinIndependentStates are buried and were presumably non, M. Neugebauer,and R. Phillipsare appreciated. This work detected by geophysical methods andsubsequently drilled. is contribution of theGeologicalSurveyof Canada28891. Unfortunately,details of the geophysicaldata are not H. J. Melosh was editor for this paper. He thankstechnical available and could not be used here. Geophysical reviewersW. B. McKinnonandR. J. Phillips,andan anonymous exploration anddrillingfor economic reasons alsoleadsto crossdisciplinereviewerfor their assistance. restricteddata.However,with a greaterappreciation of the REFERENCES relevanceof impactin creatingstructures with economic potential,moreinformation andinterpretation is forthcomAaloe, A., A. Dabizha, B. Kamaukh, and V. Starobdubtsev, ing. For example,Sawatzky[1977]listedsix structures in Geophysicalinvestigationson the main Kaali crater (in the Williston basin,North America (Dumas,Eagle Butte, Russian),Eesti NSV Tea& Akad. Tolim. Keem. Geol., 25, Elbow Lake, Hartney,Red Wing, and Viewfield), which 58-65, 1976. he considersto have geophysical(seismic) signatures Ackermann, H. D., R. H. Godson, and J. S. Watkins, A seismic refraction techniqueused for subsurfaceinvestigationsat consistent with an impactorigin.Red Wing andViewfield MeteorCrater,Arizona,J. Geophys.Res.,80, 765-775, 1975. are activeoil producers, while Red Wing andEagleButte Alvarez, L. W., W. Alvarez, F. Asaro, and H. V. Michel, havebeenconfirmedas impactstructures on the basisof Extraterrestrial causefor the Cretaceous/tertiary extinction, geologicaldata [Brenanet al., 1975;Ezeji-Okoye,1985]. Science, 208, 1095-1108, 1980. Recently,threegeophysical anomalies potentiallyrelated Anderson,R. R., andJ. B. Hartung,The structuralconfiguration of the Manson impact structure,Iowa, as interpretedfrom to impactstructures havebeenidentifiedthroughhydrocarseismicdata and confirmedby drill samples (abstract),in bon explorationin the Cooper and EromangaBasins, Australia [O'Neil, 1989]. The Pretoria Salt Pan, South Africa, was a well-known Lunar PlanetScienceXXII, pp. 25-26, Lunarand Planetary Institute, Houston, Tex., 1991. Angenheister,G., and J. Pohl, Remanent magnetizationof sueritefrom the Ries area(southernGermany),Z. Geophys., geophysical anomaly,with a-5-mGal gravity anomaly 30, 258-259, 1964. over a -1-km-diameter area, of suggestedimpact origin Barlow, B.C., Gravityinvestigations of the GossesBluff impact [Fudali et al., 1973], that hasrecentlybeenconfirmedby structure, central Australia, Bur. Min. Res. J. Aust. Geol. drilling[Reimoldet al., 1991]. Mostrecently,a ~ 180-kmGeophys.,4,323-339, 1979. diametergeophysicalanomalyon the Yucatanpeninsula Basilevsky,A. T., B. A. Iranov, K. P. Florensky,O. I. Yakovlev, V. I. Feldman,L. V. Granovsky,and M. A. Sandovsky, [Penfieldand Camargo, 1981] has been identifiedas a Impact Craters on the Moon and Planets (in Russian),200 buried impact crater [tlildebrand et al., 1991]. This pp., Nauk, Moscow,1983. structure,known as Chicxulub, is currently the most Beals,C. S., A probable meteoritecraterof Precambrian ageat promisingcandidatefor the site of the putativemajor Holleford,Ontario,Publ. Dom. Obs.Ottawa,24, 117-142, 1960. impacteventat the Cretaceous-Tertiary boundarythat is Beals, C. S., M. J. S. Innes, and J. A. Rottenberg,Fossil believedto haveresultedin a globalmassextinctionevent meteoritecraters,in The Moon, Meteorites and Cornets,edited by B. M. Middlehurst and G. P. Kuiper, pp. 235-284, Universityof ChicagoPress,Chicago,II1., 1963. Brenan,R. L., B. L. Peterson,andH. J. Smith,The originof Red States,havegeophysical (potentialfield) signatures similar Wing Creek structure:McKenzie County, North Dakota, [Alvarez et al., 1980]. Other structures, such as the ~100-km-diameter Can-Am structure, Canada-United to knowncraters[Forsythet al., 1990] but lack seismicor drilling information. Some magnetic anomalies, as detectedby Magsat, have been suggestedto be the remnantsof Precambrianimpactbasinsin the 1000-kmsize range, for example, the Bangui anomaly, Africa [Girdler et al., 1991], the SimpsonDesert, Australia [ReimoldandDuane, 1991]. Althoughthesegeophysically anomalousareas of the Earth's upper crust may have associatedgeological anomalies, there is no direct evidencethat theselargegeophysicalanomaliesare due to impact. Nevertheless,many impact craters remain undiscovered. For example,it is estimatedthat only 20% of the Phanerozoic-aged impactcratersin Australiahave been discovered[Shoemakerand Shoemaker,1988]. While most individual discoveries of a new crater will not have thepotentialsignificance of a structuresuchas Chicxulub, Earth Sci. Bull., 8, 1-41, 1975. Brown, A. R., A detailedseismicstudyof GossesBluff, Northern Territory, Rep. Bur. Miner. Res. Geol. Geophys.Aust., 163, 1-42, 1973. Buschbach, T. C., andR. Ryan,Ordovicianexplosionstructureat Glasford,Illinois, Bull. Am. Assoc.Pet. Geol., 47, 2015-2022, 1963. Bylund,G., Palaeomagnetism of a probablemeteoriteimpact,the Dellen structure, Geol. Fo. Stockholm Foh, 96, 275-278, 1974. Chao, E. C. T., Pressureand temperaturehistoriesof impact metamorphosed rocks--based on petrographicobservations, in ShockMetamorphism of NaturalMaterials,editedby B. M. French and N.M. Short, pp. 135-158, Mono Book, Baltimore, Md., 1968. Cisowski, S. M., Paleomagnetismof Manson structurecores inconsistentwith a K/T link (abstract), in Lunar Planet ScienceXIX, pp. 188-189, Lunar and PlanetaryInstitute, Houston,Tex., 1988. 30, 2 / REVIEWSOF GEOPHYSICS Pilkington andGrieve:SIGNATURE OF TERRESTRIAL IMPACTCRATERS ß 179 Cisowski, S. M., and M. Fuller, The effect of shock on the magnetism of terrestrial rocks, J. Geophys. Res., 83, 3441-3458, 1978. Ernstson, K., The structure of the Ries craterfrom geoelectric depthsoundings, Z. Geophys., 40, 639-659, 1974. Ernstson, K., A gravity-derived modelfor the Steinheimimpact crater, Geol Rundsch.,73, 483-498, 1984. Clark, J. F., Geomagnetic surveysat WestHawk Lake,Manitoba, Canada, Can. Dep. Energy, Min. Resourc.Set., 20, 1-11, Ernstson,K., and J. Pohl, Somecommentson the geophysical 1980. logging measurementsin the Nordlingen 1973 research borehole(in German),Geol. Bavarica, 72, 81-90, 1974. Clark, J. F., Geomagneticsurveysat four Canadiancraters,Can. Dep. Energy,Min. Resour.Geomagn.Rep.,81-1, 1-41, 1981. Ernstson,K., H. Feld, and J. Fiebag,Impacthypothesis for the Clark, J. F., Magneticsurveydata at meteoriticimpactsitesin Azuara structure(Spain) strengthened (abstract),Meteoritics, 22, 373, 1987. North America, Geomagn.Serv. Can. Open File Rep., 83-5, 1-30, 1983. Coles, R. L., and J. F. Clark, The centralmagneticanomaly, Manicouagan structure, Quebec, J. Geophys. Res., 83, 2805-2808, 1978. Ezeji-Okoye,S., The originof the EagleButtestructure, Eagle Butte, Alberta Canada, report, 75 pp., Pan Can. Petrol., Calgary,1985. Feldman,V. I., A. I. Dabizha,andL. B. Granovsky,The Shunak meteoritecrater(in Russian),Meteoritika, 38, 99-103, 1979. Coles,R. L., and J. F. Clark, Lake St. Martin impactstructure, Manitoba,Canada:Magneticanomaliesandmagnetizations, J. Feldman, V. I., L. V. Sazonova, Yu. V. Mironov, I. 13. Geophys. Res.,87, 7087-7095, 1982. Kapustkina,B. A. Ivanov, and O. Yu. Schmidt, Circular Comer,B., R. J. Durrheim,and L. O. Nicolaysen,Relationships structureLoganchaas possiblemeteoritecraterin basaltsof between the Vredefort structure and the Witwatersrand basin the Tunguskasyneclise(abstract),in Lunar Planet Science within the tectonic frameworkof the Kaapvaal craton as XIV, pp. 191-192, Lunar and PlanetaryInstitute,Houston, Tex., 1983. interpretedfrom regional gravity and aeromagneticdata, Florenskii,P. V., A. I. Dabizha, A. O. Aaloe, E. S. 13orshkov, Tectonophysics, 171, 49-61, 1990. Currie,K. L., Geologicalnoteson the Carswellcircularstructure, and V. I. Miklyayev,Geologic-geophysical characteristics of Saskatchewan (74K), Pap. Geol. Surv. Can., 67-32, 60 pp., the Zhamanshin meteoritecrater (in Russian),Meteoritika, 1969. 38, 86-97, 1979. Currie, K. L., and A. Larochelie,A palcomagneticstudy of Forsyth,D. A., M. Pilkington,R. A. F. Grieve,andD. Abinett,A volcanic rocks from Mistastin Lake, Labrador, Canada, Earth major circular structurebeneathsouthernLake Huron from Planet. Sci. Lett., 6, 309-315, 1969. potentialfield data,Geology,18, 773-777, 1990. Dabizha,A. I., and V. V. Fedynsky,The Earth's"star wounds" Fox, J. H., Geophysicalinvestigationof the Crooked Creek, and their diagnosisby geophysicalmethods(in Russian), Missouri, cryptoexplosionstructure (abstract), Eos Trans. AGU, 51,770, 1970. ZamlyaVselennaya, 3, 56-64, 1975. Dabizha,A. I., andV. V. Fedynsky,Featuresof thegravitational Fudali, R. F., Roter Kamm: Evidencefor an impactorigin, field of astroblemes(in Russian), Meteoritika, 36, 113-120, 1977. Dabizha,A. I., and V. I. Feldman,The geophysical propertiesof some astroblemesin the USSR (in Russian),Meteoritika, 40, 91-101, 1982. Meteoritics, 8, 245-257, 1973. Fudali, R. F., Gravity investigationof Wolf Creek crater, Western Australia,J. Geol. 87, 55-67, 1979. Fudali,R. F., andW. A. Cassidy,Gravityreconnaissance at three Mauritaniancratersof explosiveorigin,Meteoritics,7, 51-70, 1972. Dence,M. R., The natureand significanceof terrestrialimpact structures, 24thInt. Geol.Congr.,77-89, 1972. Fudali,R. F., D. P. Gold, andJ. J. Gumey,The PretoriaSaltPan: Astroblemeor cryptovolcano?, J. Geol., 81,495-507, 1973. Dence,M. R., and J. Popelar,Evidencefor an impactoriginfor Lake Wanapitei, Ontario, New Developmentsin Sudbury Fudali, R. F., D. J. Milton, K. Fredericksson,and A. Dube, Geology,editedby J. V. Guy-Bray,Geol.Assoc.,Spec.Pap., Morphology of Lonar crater, India: Comparisonsand I0, 117-124, 1972. implications,Moon Planets,23,493-515, 1980. Dence,M. R., M. Innes,and P. B. Robertson,Recentgeological Gardner,G. H. F., L. W. Gardner,andA. R. Gregory,Formation and geophysical studies of Canadian craters, in Shock velocityanddensityraThediagnostic basesfor stratigraphic Metamorphismof Natural Materials, editedby B. M. French traps,Geophysics, 39, 770-780, 1974. and N.M. Short,pp. 339-362, Mono Book, Baltimore,Md., Girdler,R. W., P.T. Taylor,andJ. J. Frawley,Continental satellite 1968. magnetic anomaliesand large impacts (abstract),paper Dent,B. E., Studiesof largeimpactcraters,Ph.D.thesis,109 pp., presented atIUGG 20thGeneralAssembly, Vienna,1991. Stanford Univ., Stanford,Calif., 1973. Grant, F. S., Aeromagnetics, geologyand ore environments, 1, Dyrelius, D., The gravity field of the Siljan ring structure,in Magnetitein igneous, sedimentary andmetamorphic rocks:An Deep Drilling in CrystallineBedrock,vol. 1, The Deep Gas overview,Geoexploration, 23, 303-333, 1985. Drilling in the Siljan Impact Structure, Sweden and Green,R. W., and P. Chetty,Seismicrefractionstudiesin the Astroblemes,edited by A. Boden and K. G. Eriksson,pp. basementof the Vredefortstructure,Tectonophysics, 171, 105-113, 1990. 85-94, Springer-Verlag, New York, 1988. Elming, S.-A., and G. Bylund, Palaeomagnetism and the Siljan Grieve, R. A. F., The formationof large impactstructures and impact structure,central Sweden, Geophys.J. Int., 105, constraints on the nature of Siljan, in Deep Drilling in 757-770, 1991. CrystallineBedrock,vol. 1, The Deep Gas Drilling in the Elo, S., A study of the gravity anomalyassociatedwith Lake SiljanImpactStructure,Swedenand Astroblemes, editedby Lappaj'firvi,Finland,Rep. Geol. Surv. Finland, Q20/21/1976/ A. BodenandK. G. Eriksson, pp. 328-348, Springer-Verlag, New York, 1988. 2, 20 pp., 1976. Elo, S., T. Jokinen,and H. Soininen,Geophysicalinvestigations Grieve, R. A. F., Terrestrialimpact:The record in the rocks, Meteoritics, 26, 175-194, 1991. of Lake Lappaj'firvi impact structure, western Finland (abstract), Prog. Abstracts Geol. Surv. Finland, Fennos- Grieve,R. A. F., andM. J. Cintala,Differentialscalingof crater candianImpactStructures, 21, 1990a. parameters: Implicationsfor the observedterrestrialrecord,in Elo, S., L. Kivekas,H. Kujala,S. I. Lahti, andP. Pihlaja,Recent LunarPlanetScience XXll, pp.493-494, LunarandPlanetary Institute, Houston, Tex., 1991. studiesof Lake S•i'fi_ksj:,irvi meteoriteimpact crater, southwestern Finland (abstract), Prog. Abstracts Geol. Surv. Grieve,R. A. F., and L. J. Pesonen, The terrestrialimpact Finland,Fennoscandian ImpactStructure,44, 1990b. crateringrecord,Tectonophysics, in press,1992. 180 ß Pilkington andGrieve:SIGNATURE OF TERRESTRIAL IMPACTCRATERS 29, 1! REVIEWS OF GEOPHYSICS Grieve, R. A. F., and P. B. Robertson,The terrestrialcratering Langan,R. T., Structuralanalysisof theregionnearDes Plaines, record, 1, Current statusof observations,Icarus, 38, 212-229, Illinois using geophysicalmethods,M.S. thesis,54 pp., NorthwesternUniv., Evanston,I11.,1974. 1979. Grieve, R. A. F., P. B. Robertson,and M. R. Dence, Constraints Larochelie,A., andK. L. Curtie,Paleomagnetic studyof igneous on the formationof ring impactstructures, basedon terrestrial rocksfrom the Manicouagan structure,Quebec,J. Geophys. Res., 72, 4163-4169, 1967. data, in Multi-Ring Basins,Proceedingsof the Lunar Planet Science,vol. 12A, editedby P. H. Schultzand R. B. Merrill, Lauren,L., J. Lehtovaara,R. Bostrom,and R. Tynni, On the geologyof the circulardepression at S6derfj'•xden, western pp. 37-57, Pergamon,New York, 1981. Finland, Bull. Geol. Surv. Fin., 297, 1-38, 1978. Grieve,R. A. F., J. B. Garyin,J. M. Coderre,andJ. Rupert,Test of a geometricmodel for the modificationstageof simple Lyons,J. A., D. A. Forsyth,andJ. A. Mair, Crustalstudiesin the La Malbaie region,Quebec,Can. J. Earth Sci., 17, 478-490, impactcraterdevelopment, Meteoritics,24, 83-88, 1989. 1980. Gupta,V. K., F. S. Grant,andK. D. Card,Gravityandmagnetic characteristics of theSudburystructure, The GeologyandOre Mareschal,M., andM. Chouteau,A magnetotelluric investiga- Depositsof the SudburyStructure,editedby E.G. Pye et al., Geol. Surv. Ont., 1,381-410, 1984. Gurov, E. P., and E. P. Gutova, Someregularitiesof the area spreading of fracturesaroundEl'gygytgyncrater(abstract),in Lunar and Planetary ScienceXIII, pp. 291-292, Lunar and PlanetaryInstitute,Houston,Tex., 1982. Halliday, I., and A. A. Griffin, Evidence in supportof a meteoriticorigin for West Hawk Lake, Manitoba,Canada,J. Geophys.Res.,68, 5297-5305, 1963. Halls, H. C., The SlateIslandsmeteoriteimpactsite:A studyof shockremanentmagnetization,Geophys.J. R. Astron.Soc., 59, 553-591, 1979. Hargraves,R. B., andW. E. Perkins,Investigations of the effect of shockon naturalreinanentmagnetism,J. Geophys.Res., 74, 2576-2589, 1969. tion of the structuralgeology beneath Charlevoix Crater, Quebec,Phys.Earth.Planet.lnt., 60, 120-131, 1990. Masaitis,V. L., M. V. Mikhailov,andT. V. Selivanovskaya, The PopigayMeteorite Crater (in Russian),124 pp., Nauka, Moscow, 1975. Maslov, M. A., On the origin of the Kara depression(in Russian),Meteoritika, 36, 123-130, 1977. Melosh,H. J., ImpactCratering:A GeologicProcess,245 pp., OxfordUniversityPress,New York, 1989. Millman,P.M., B. A. Liberty,J. F. Clark, P. Willmore,andM. J. S. Innes, The Brent crater,Publ. Dom. Obs. Ottawa, 24, 1-43, 1960. Offield, T. W., and H. A. Pohn, Geologyof the Decaturville impactstructure, Missouri,U.S. Geol.Surv.Prof. Pap., 1042, 1-48, 1979. Hartung, J. B., M. J. Kunk, and R. R. Anderson,Geology, O'Neil, B. J., (Ed.), The Cooper and Eromanga Basins, geophysics,and geochronologyof the Manson impact Australia,Proceedings of PetrologyExplorationSocietyof structure,Global Catastrophes in Earth History, editedby V. Australia,Soc.of PetrologyEngineers,AustralianSocietyof ExplorationGeophysics, Adelaide,1989. L. Sharptonand P. D. Ward, Geol. Soc.Am. Spec.Pap., 247, 207-221, 1990. Pedersen, L. B., T. M. Rasmussen, andD. Dyrelius,Construction Henkel, H., The Lake Mien structure,Sver. Geol. Unders., Ser. of component mapsfrom aeromagnetic total field anomaly maps,Geophys.Prospect.,38, 795-804, 1990. C, NFR 2638-3, 7 pp., 1982. Henkel, H., andM. Guzman,Magneticfeaturesof fracturezones, Penfield,G. T., and A. CamargoZ., Definitionof a major Geoexploration,15, 173-181, 1977. igneouszonein the centralYucUtanplatformwith aeromagHildebrand,A. R., G. T. Penfield,D. A. Kring,M. Pilkington,A. netics and gravity (abstrac0,paper presentedat Technical CamargoZ., S. B. Jacobsen,and W. V. Boynton,Chicxulub Program,AbstractsandBibliographies 51st AnnualMeeting Soc.Explor.Geophys.1981. crater:A possibleCretaceous-Tertiary boundaryimpactcrater on the Yucutan peninsula,Mexico, Geology, 19, 867-871, Perrier,R. and J. Quiblier,Thicknesschangesin sedimentary 1991. layersduringcompactionhistory:Methodsfor quantitative Innes, M. J. S., Recent advances in meteorite crater research at the Dominion Observatory,Ottawa, Canada,Meteoritics,2, 219-241, 1964. Innes,M. J. S., W. J. Pearson,and J. W. Geuer,The Deep Bay crater, Publ. Dom. Obs. Ottawa, 31, 19-52, 1964. evaluation,Am. Assoc.Pet. Geol. Bull., 58,507-528, 1974. Pesonen,L. J., and N. Marcos,Palaeomagnetism of the Lake Lappaj},irvi impactstructure,central-westFinland (abstract), Prog.AbstractsGeol.Surv.Finland, Fennoscandian Impact Structures,26, 1990. Iseri, D. A., J. W. Geissman,H. E. Newsom, and G. Graup, Pike, R. J., Controlof cratermorphology by gravityandtarget Paleomagneticand rock magneticexaminationof the natural type:Mars,Earth,Moon,Proc.Lunar Planet.Sci. Conf.11th, 2159-2189, 1980. remanentmagnetizationof suerite depositsat Ries crater, West Germany(abstract),Meteoritics,24, 280, 1989. Pilon,J. A., R. A. F. Grieve,andV. L. Sharpton,The subsurface Jansa,L. F., G. Pe-Piper,P. B. Robertson,and O. Freidenreich, character of MeteorCrater,Arizona,asdetermined by ground probingradar,J. Geophys.Res.,96, 15,563-15,576, 1991. Montagnais:A submarineimpact structureon the Scotian Shell eastern Canada, Geol. Soc. Am. Bull., 101, 450-463, 1989. Plante,L., M.-K. Seguin,andJ. Rondot,Etudegravimetrique des astroblemesdu Lac a l'Eau Claire, Nouveau-Quebec(in French),Geoexploration, 26, 303-323, 1990. Joesting,H. R., and D. Plouff, Geophysicalstudiesof the Upheaval Dome area, San Juan County, Utah, lnt. Assoc. Pohl,J., On theoriginof themagnetization of impactbrecciason Petrol. Geol.Annu.Field Conf.Guideb.,9, 86-92, 1958. Earth,Z. Geophys., 37, 549-555, 1971. Jones, W. B., M. Bacon, and D. A. Hastings,The Lake Pohl, J., and H. Soffel,Paleomagnetic age determination of the Bosumtwi impact crater, Ghana, Geol. Soc. Am. Bull., 92, Rochechouart impact structure(France), Z. Geophys.,37, 342-349, 1981. 857-866, 1971. Juhlin,C., and L. B. Pedersen,Reflectionseismicinvestigations Pohl, J., and M. Will, Comparisonbetweenvelocitymeasureof the Siljan impactstructure,Sweden,J. Geophys.Res.,92, mentsfrom researchboreholdN6rdlingen1973 with seismic 14,113-14,122, 1987. depths inside and outside of Ries (in German), Geol. Bavarica, 72, 75-80, 1974. Kumar, A., andJ. H. Ward, The effectof explosiveshockon the susceptibilityof magnetitegrout mixture, Rep. Mr-63-10, 8 Pohl,J., U. Bleil, andU. Homemann,Shockmagnetization and pp., Univ. of Calif., 1963. demagnetization of basaltby transientstressup to 10 kbar,J. Kuster, G. T., and M. N. Toksoz, Velocity and attenuationof Geophys., 41, 23-41, 1975. seismic waves in two-phasemedia, Part II, Experimental Pohl,J., D. Stoffier,H. Gall, andK. Emston,The Riesimpact results,Geophysics, 39, 607-618, 1974. crater,in Impactand ExplosionCratering,editedby D. J. 30, 2 / REVIEWS OF GEOPHYSICS Pilkington andGrieve:SIGNATUREOF TERRESTRIAL IMPACTCRATERS ß 181 Short,N.M., A comparison of featurescharacteristic of nuclear Roddyet at.,pp. 343-404, Pergamon, New York, 1977. explosioncratersand astroblemes, Ann. N.Y. Acad. Sci., 123, Pohl, J., K. Ernstson,and P. Lambert,Gravity measurements in 573-616, 1965. the Rochechouart impact structure (France) (abstract), Meteoritics, 13, 601-604, 1978. Slawson,W. F., Vredefortcore: A cross-section of the upper crust?,Geochim. Cosmochim.Acta, 40, 117-121, 1976. Pohl, J., A. Eckstaller, and P. B. Robertson, Gravity and magneticinvestigations in the Haughtonimpact structure, Smith,T. A., and L. V. A. Sendlein,Geophysicalstudyof the Devon Island, Canada,Meteoritics, 23, 235-238, 1988. Manson impact crater (abstract), Eos Trans. AGU, 52, 264-265, 1971. Popelar,J., Gravity interpretationof the Sudburyarea, New formation developments in Sudburygeology,editedby J. V. Guy-Bray, Stanfors,R., The Mien stmcture--A cryptoexplosion Geol.Assoc.Can., Spec.Pap., 10, 103-111, 1972. in the Fennoscandian basement,Ph.D. thesis,144 pp., Lunds Univ. Geol. Inst. Lund, Sweden, 1973. Regan,R. D. and W. J. Hinze, Gravity andmagneticinvestigations of Meteor crater, Arizona, J. Geophys.Res., 80, Stearns,R. G., C. W. Wilson, Jr., H. A. Tiedemann, J. T. Wilcox, 776-788, 1975. Reimold, W. U., and M. J. Duane, Discussionfor the criteria for and P.S. Marsh, The Wells Creek structure,Tennessee,in ShockMetamorphismof Natural Materials, editedby B. M. French and N.M. Short, pp. 323-337, Mono Book, Bal- recognition of multiringimpactbasins--withreferenceto the timore, Md., 1968. Simpsondesertdepression andthe Vredefortdome(abstract), in Lunar and PlanetaryScienceXXII, pp. 1115-1116, Lunar Steinemann, C. F., A gravitystudyof the MiddlesborocrypandPlanetaryInstitute,Houston,Tex., 1991. toexplosivestructure,M. S. thesis,41 pp., Univ. of Ky, Lexington,1980. Reimold,W. U., C. Koeberl,S. J. Kerr, andT. C. Partridge,The PretoriaSalt Pan--The first firm evidencefor an origin by Sweeney,J. F., Gravitystudyof greatimpact,J. Geophys. Res., 83, 2809-2815, 1978. impact(abstract),in Lunar and PlanetaryScienceXXII, pp. 1117-1118, Lunar and Planetary Institute, Houston, Tex., Thomas,M.D., and M. J. S. Innes,The Gow Lake impact structure, northern Saskatchewan, Can. J. Earth Sci., 14, 1991. 1788-1795, 1977. Robertson,P. B., and J. L. Roy, Shock-diminishing paleomagnetic remanenceat the Charlevoiximpactstructure,Quebec, Thomas,M.D., D. H. Halliday, and R. Stephenson, Gravity Can. J. Earth Sci., 16, 1842-1856, 1979. anomaliesand geologicalstructurein northernLabradorand Robertson,W. A., Manicouagan,Quebec,paleomagnetic results, northeastern Quebec,EarthPhys.Branch,Can.,GravityMap Can. J. Earth Sci., 4, 641-649, 1967. Ser., 157-161, 1-38, 1978. Roddy, D. J., Pre-impactconditionsand crateringprocesses at Tudor, D. S., A geophysicalstudyof the Kentlanddisturbedarea, the Flynn Creek crater,Tennessee,in Impact and Explosion Ph.D.thesis,111pp.,IndianaUniv., Bloomington,1971. Cratering, edited by D. J. Roddy et at., pp. 277-308, Van Lopik, J. R., and R. A. Geyer, Gravity and magnetic anomaliesof the Sierra Madera, Texas, "Dome," Science, 142, Pergamon,New York, 1977. Sander, G. W., A. Overton, and R. Bataille, Seismic and magneticinvestigationof the Deep Bay crater,J. R. Astron. Soc. Can., 58, 16-30, 1963. 45-47, 1963. Vishevsky, S. A., and V. N. Lagutenko,The Ragozinka astrobleme: An Eocenecraterin the centralUrals (in Russian), Akad. Nauk, SSœ,14, 1-42, 1986. Sappenfield,L. W., A magneticsurveyof the AdamsCounty cryptovolcanicstructure,M.S. thesis, 27 pp., Univ. of Wasilewski, P. J., Shock remagnetizationassociatedwith Cincinnati, Cincinnati, Ohio, 1950. meteoriteimpact at planetarysurfaces,Moon, 6, 264-291, 1973. Sawatzky,H. B., Two probablelate Cretaceousastroblemesin western Canada--Eagle Butte, Alberta and Dumas, Sas- Wickman,F. E., Possibleimpactstructures in Sweden,in Deep katchewan,Geophysics, 41, 1261-1271, 1976. Drilling in CrystallineBedrock,vol. 1, TheDeepGasDrilling Sawatzky,H. B., Buried impactcratersin the Williston Basin in the Siljan Impact Structure, Sweden and Astroblemes, and adjacentareas,in Impactand ExplosionCratering,edited edited by A. Boden and K. G. Eriksson,pp. 298-327, Springer-Verlag,New York, 1988. by D. J. Roddy et al., pp. 461-480, Pergamon,New York, 1977. Williams,G. E., The Acramanimpactstructure,SouthAustralia Schultz,P. H., and R. B. Merrill, (Eds.), Multi-Ring Basins,295 (abstract),InternationalWorkshopMeteoriteImpacton Early Earth, Perth, Australia, Lunar Planet Inst. Contrib., 746, pp., Pergamon,New York, 1981. Scott, D., and Z. Hajnal, Seismic signatureof the Haughton structure, Meteoritics, 23,239-247, 1988. Shoemaker,E. M., Why study impact craters?,in Impact and ExplosionCratering,editedby D. J. Roddyet al., pp. 1-10, Pergamon,New York, 1977. Shoemaker,E. M., and C. S. Shoemaker,Impact structuresof Australia, (abstract), in Lunar Planet Science XIX, pp. 1079-1080, Lunar and PlanetaryInstitute, Houston,Tex., 1988. Shoemaker,E. M., C. S. Shoemaker,and J. B. Plescia,Gravity investigationof the Connollybasinimpactstructure,Western Australia (abstract), in Lunar Planet Science XX, pp. 1010-1011, Lunar and PlanetaryInstitute, Houston,Tex., 1989. 60-61, 1990. Wilson, C. W., Jr., and R. G. Stearns,Geologyof the Wells Creek structure,Tennessee,Bull. Tenn. Div. Geol. 68, 1-236, 1968. Young,G. A., Gosses Bluff airbornemagneticsurvey,Geophys. Prospect.,20, 83-91, 1972. Zhang, P., T. M. Rasmussen,and L. B. Pedersen,Electric resistivitystructure of the Siljan impactregion,J. Geophys. Res., 93, 6485-6501, 1988. R. A. F. Grieveand M. Pilkington, Geophysics Division, GeologicalSurveyof Canada,Ottawa,Canada,K1A0Y3.