Venus: Analysis of the degree of impact crater deposit ... · DP = 0.15T, T CH = 0.3T, and T FH =...

Venus: Analysis of the degree of impact crater deposit degradation and

assessment of its use for dating geological units and features

Alexander T. BasilevskyVernadsky Institute of Geochemistry and Analytical Chemistry, Russia Academy of Sciences, Moscow, Russia

Department of Geological Sciences, Brown University, Providence, Rhode Island, USA

James W. HeadDepartment of Geological Sciences, Brown University, Providence, Rhode Island, USA

Received 4 September 2001; revised 21 February 2002; accepted 25 April 2002; published 30 August 2002.

[1] We use the degree of degradation of crater-associated radar-dark deposits on Venusto estimate the age of the crater and neighboring units. We analyzed craters �30 km indiameter superposed on regional plains (subpopulation 1; 138 craters) and on later units(subpopulation 2; 30 craters) and estimated percentages of craters with dark parabolas(DP), clear dark halo (CH), faint dark halo (FH), and no dark halo (NH). We constructedtheoretical models of the evolution of these percentages with time for two possibleinterpretations of the upper boundary of the regional plains: (1) globally synchronousand (2) nonsynchronous (diachronous). They show that in the synchronous case the darkdeposit lifetimes are proportional to the corresponding percentages observed forsubpopulation 1: TDP = 0.15T, TCH = 0.3T, and TFH = 0.3T, where T is the mean globalsurface age of Venus. In diachronous cases the percentages may or may not beproportional to the appropriate lifetimes, depending on the range of the regional plains. Ifthe time range of plains emplacement is not larger than ±0.5T, the use of TDP = 0.15T andTCH = 0.3T is appropriate. We consider the time �0.5T ago as the lower time boundaryfor the age of the now-observed CH craters and the time 0.1–0.15T ago as the lowerboundary for the age of DP craters. We propose that the Atlian Period of the geologichistory of Venus be subdivided into Upper and Lower Epochs, with the boundarybetween them at �0.5T ago, and the lower boundary of the Aurelian Period be placed at0.1–0.15T ago. We apply this approach to assess ages of activity of three volcanic-tectonic structures on Venus: Beta Regio-Devana Chasma (shows evidence of activityyounger than 0.5T ago), Mylitta Fluctus (also younger than 0.5T), and Atla Regio (showsevidence of activity younger than 0.1–0.15T ago). INDEX TERMS: 6295 Planetology: Solar

System Objects: Venus; 5475 Planetology: Solid Surface Planets: Tectonics (8149); 5420 Planetology: Solid

Surface Planets: Impact phenomena (includes cratering); 5470 Planetology: Solid Surface Planets: Surface

materials and properties; 5480 Planetology: Solid Surface Planets: Volcanism (8450); KEYWORDS: Venus,

impact craters, stratigraphy, dark parabola, dark halo

1. Introduction

[2] Determination of the absolute age of surface geologicunits of the planets and satellites is a powerful tool forunderstanding their geologic histories. This is usuallyaccomplished through determination of the areal densityof impact craters on different geologic units. Althoughsuccessfully applied to the studies of Mars, the Moon andother satellites, this approach meets difficulties in the caseof Venus. With about 1000 impact craters on the wholesurface of Venus [Schaber et al., 1992; Phillips et al., 1992],their areal density can be used only for determination of themean surface age of this planet and the mean surface ages ofglobal geologic units or terrain types occupying very large

areas (>15–20 � 106 km2), for example, tessera terrain,regional plains, and large volcanic constructs [Ivanov andBasilevsky, 1993; Basilevsky et al., 1999; Namiki andSolomon, 1994; Price and Suppe, 1994, 1995]. Estimationof the mean surface age of the geologic units on Venusimplies lumping together spatially separated areas of theunits and involves additional assumptions which arecriticized by some workers [Guest and Stofan, 1999; Camp-bell, 1999]. Individual structures, even as large as MaatMons or Ozza Mons volcanoes, Mylitta Fluctus lava fieldand Beta/Devana region or Ganis Chasma rifts, each containat best not more than several craters, and this precludes theirreliable dating based on determination of crater densities.[3] In this situation it was suggested by several workers

that the degree of degradation of individual craters could beused for estimation of surface age. Based on the observationthat impact craters with extensive dark parabolic deposits

JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 107, NO. E8, 10.1029/2001JE001584, 2002

Copyright 2002 by the American Geophysical Union.0148-0227/02/2001JE001584$09.00

5 - 1

(hereafter ‘‘dark-parabola craters’’) are among the youngestfeatures of local stratigraphic columns and compose about10% of the Venus crater population [Campbell et al., 1992], itwas suggested that they are younger than 0.1 T, where T is themean surface age of Venus [Arvidson et al., 1992; Basilevsky,1993; Strom, 1993; Herrick and Phillips, 1994]. In thisconcept, nonparabola craters are considered as old enough(>0.1 T) to lose parabolas, which they possessed initially, dueto surface degradation (mostly eolian) processes. Schultz[1992] pointed out the possible effect of the impact directionon the formation of the dark parabolas: impacts from the westare more favorable for parabola formation than impacts fromthe east. Izenberg et al. [1994] suggested an age sequence ofcrater degradation from youngest to oldest: 1) craters withparabola and dark halo, 2) craters with dark halo only, and 3)craters with partial or no dark deposits.[4] The age sense of this morphological sequence was

shown by Izenberg et al. [1994] through demonstration thatalong this sequence the percentage of volcanically embayed

and tectonically deformed craters progressively increased.Formation of crater-related radar-dark material of the parab-olas is believed to be a result of the deposition of the finefraction of crater ejecta sorted out during the settling ofejecta through the zonal (east-west) high-altitude winds ofthe atmosphere of Venus [Campbell et al., 1992; Vervackand Melosh, 1992]. The appearance of a nonparabolic radar-dark halo as a remnant from the degraded dark parabola is alogical conclusion supported by the observations of Izen-berg et al. [1994]. Eolian resurfacing of the dark material isusually suggested as the mechanism of dark deposit degra-dation [Arvidson et al., 1992; Izenberg et al., 1994; Herricket al., 1997]. A catalog sorting specific craters into the dark-parabola, dark halo and no halo classes corresponding todifferent degradation degrees can be found in the craterdatabase of R. Herrick (http://www.lpi.usra.edu/research/vc/vchome.html).[5] In this work we explore the same approach of using

the degree of degradation of individual craters (namely the

Figure 1. Craters Stuart (30.79�S, 20.22�E, D = 68.6 km) and Zenobia (29.35�S, 28.55�E, D =39.1 km). Crater Stuart is classified as a dark-parabola crater by Campbell et al. [1992], the R. Herrickdatabase and by us. Crater Zenobia is considered as a dark parabola crater by R. Herrick and in this paper,but is not in the list of dark-parabola craters of Campbell et al. [1992]. Portion of the global Magellanmosaic.

5 - 2 BASILEVSKY AND HEAD: IMPACT CRATER DEPOSIT DEGRADATION ON VENUS

degree of degradation of the crater-related radar-dark depos-its) for crater age estimates and apply it to the dating ofseveral prominent geologic structures on Venus. Becauseassigning discrete craters to different stages of degradationis based on the qualitative consideration of the crater imagesand resulting opinions, we decided not to rely solely on thecrater classification given in R. Herrick’s crater database butto do our own observations and the analysis, to classify thecraters, and then to compare our classification and sorting ofthe craters with that of Herrick.[6] We therefore undertook a photogeological analysis of

the Magellan images of all craters �30 km in diameter. Thissize limit is a compromise between morphologic visibility,which is better for craters of larger size, and statisticalreliability, which is favored by more numerous craters atsmaller size. As a guide for this study and as a source of thecrater names and sizes we used the updated version of theSchaber et al. [1998] crater database available on http://wwwflag.wr.usgs.gov/USGSFlag/Space/venus. Names ofother surface features were taken from the USGS planetarynomenclature site http://wwwflag.wr.usgs.gov/USGSFlag/

Space/nomen/vgrid.html. In our analysis we used all avail-able Magellan C1- and C2MIDRP’s as well as the globalSAR image of Venus made at Brown University by EmilyStewart, 2000. The latter is a mosaic of C1-MIDR browseimages at 1.6 km per pixel. Images from Cycles 2 and 3were used to fill gaps in Cycle 1 data from individual C1MIDRP’s. The advantage of this mosaic is that the level ofimage stretching is consistent across the globe. In the caseswhere high-resolution images of the craters studied wereespecially important for the analysis, we also used theFMAPs.

2. Crater Deposit Degradation Analysis

[7] In this work we studied images of 186 of 188 craters�30 km in diameter listed in the Schaber et al. [1998] craterdatabase. For two craters from the list there are no Magellanimages. Most of the studied craters (except 6) are also listedin Herrick’s database. We accomplished the study throughthe analysis of the degree of degradation of the crater,determining simultaneously which geologic units the

Figure 2. Craters Greenaway (22.9�N, 145.06�E, D = 93 km), Ban Zhao (17.14�N, 147.0�E, D = 39 km)and Maria Celeste (23.39�N, 140.38�E, D = 97.5 km). Crater Greenaway is classified as a dark-parabolacrater by Campbell et al. [1992] and by us and not classified as a dark-parabola crater by R. Herrick. Thereason is probably that the extended dark feature associated with crater Greenaway is not clearly parabolic(due to interference with dark deposits of other craters in the area). Crater Ban Zhao is classified as a dark-parabola crater by all (Campbell et al. [1992], R. Herrick, and us). Crater Maria Celeste is heavily obscuredby the dark parabola of the crater Greenaway. Portion of the global Magellan mosaic.

BASILEVSKY AND HEAD: IMPACT CRATER DEPOSIT DEGRADATION ON VENUS 5 - 3

studied craters are superposed on and which units aresuperposed on these craters. In most cases the relations ofcraters and the crater-related deposits with neighboringgeologic units were rather obvious and easily described.In typical cases, as well as in specific and complicatedcases, thorough photogeologic analysis and mapping wascompleted (for 37 of 186 craters). In mapping and descrip-tions we used geologic units suggested by Basilevsky et al.[1997] and Basilevsky and Head [1995, 1998, 2000]. Thismodel of the global stratigraphy of Venus is more updatedcompared to two other existing models [Tanaka et al., 1997](revised by Price [1995]). All three models follow thegeneral rules of identification of geologic (stratigraphic)units in planetary mapping summarized byWilhelms [1990].Recently, Hansen [2000] suggested modifications in plan-etary geologic mapping emphasizing a necessity to clearlydelineate tectonic structures from material units. This mightbe relevant for the analysis of the models of the Venusianstratigraphy mentioned above, but is not relevant for theapproach used in this work, that is, the usage of thedegradation degree of crater-related radar-dark deposits asan estimation of age.

[8] The geologic units used in this work are (from olderto younger): Tessera-forming material (Tt), material ofDensely fractured plains (Pdf ), material of Plains withfractures and (broad) ridges (Pfr), material of Fracture Belts(FB), material of Shield plains (Psh), material of Plains withwrinkle ridges (Pwr), material of Lobate and Smooth plains(Pl/Ps) and material of Rifted terrain (RT). Definitions anddescriptions of the units can be found elsewhere: Basilevskyet al. [1997] and Basilevsky and Head [1995, 1998, 2000].They are also clearly delineated from the following figuresof this paper and in the captions.[9] The following are a few notes clarifying some issues

of the stratigraphic model used. The Pwr plains are verywidespread (50–60% of the surface of Venus) so they areoften called ‘‘regional plains’’ by us and others. Thegeologic episode separating material of Pwr plains fromthe overlying materials of Lobate and Smooth plains wascompressional deformation resulting in a network of wrin-kle ridges. A significant (or even dominant) part of theShield plains material (Psh1) was also affected by wrinkleridging and is often considered as part of the regionalplains. Some small occurrences of Shield plains material

Figure 3. Crater Caccini (17.42�N, 170.43�E, D = 38.1 km) with a clear radar-dark halo superposed onregional Pwr plains. Outer boundary of the halo is diffuse. Listed in Herrick database as having dark halo.Portion of C1 MIDRP 15N163;1.


(Psh2) are superposed on the material of Pwr plains and onthe wrinkle ridge network and thus are generally contem-porary to the Lobate and Smooth (Pl/Ps) plains. Riftedterrain (RT) postdated the emplacement of Pwr plains andthe wrinkle ridging so it is broadly contemporaneous withPl/Ps and Psh2. The descriptive part of this work and mostof the conclusions do not depend on if the same units ofthe sequence mentioned are generally synchronous aroundthe planet, as is suggested by our stratigraphy model, or thesequences of units observed in different parts of the planetare diachronous, as suggested by Guest and Stofan [1999].Moreover, the results of this work bring some progress tothis controversy.

[10] It is well known that dark deposits associated withmany craters typically are poorly seen on highly faultedmaterials such as Tessera terrain, Densely fractured plains,Fracture Belts and Rifted terrain [Phillips et al., 1991;Arvidson et al., 1992; Campbell et al., 1992]. So weconcentrate in this study on craters superposed on regionalplains (Pwr + Psh1) and on Pl, Ps and Psh2 materialspostdating them. For the purpose of this work it is importantalso to determine the presence and the type of crater-relateddark deposits for craters sitting in the rift zones. This waspossible through observations of remnants of slightly frac-tured and unfractured blocks which are almost alwayspresent within the rift zones. In this work we determine

Figure 4. Crater La Fayette (70.19�N, 107.61�E, D = 39.6 km) with a clear radar-dark halo superposedon Pwr regional plains adjacent to Tusholi Corona. Outer boundary of the halo is sharp. Listed in Herrickdatabase as having dark halo. Portion of C1 MIDRP 75N119;1.


the number of craters and the type of the associated radar-dark deposits for two subpopulations: 1) those superposedon Pwr/Psh1 regional plains and 2) those superposed onpost-Pwr materials (Pl, Ps, Psh2 and RT). Craters super-posed on the regional plains and simultaneously on theolder units were considered as belonging to subpopulation 1because they postdated the regional plains and the wrinkleridge network. Craters superposed on the regional plainsand simultaneously on the younger units were considered asbelonging to the subpopulation 2 because they postdated thepost-Pwr materials. By studying craters superposed on Pwr/Psh1 regional plains and on post-Pwr materials we cover thetwo younger periods of the geologic history of Venus:Atlian and Aurelian [Basilevsky and Head, 1997; Basilevskyand Head, 1995, 1998, 2000]. Pre-Atlian periods of thegeologic history of this planet are not covered by this study.

[11] Among the 188 craters �30 km in diameter studied,138 craters were found to belong to subpopulation 1 (post-dating Pwr plains and, except for 2 cases, postdating theirwrinkle ridge network) while 30 craters were found to belongto subpopulation 2 (postdating post-Pwr units). Ten craterswere found to be superposed on the older units (mostly Tt)showing definite relations neither with regional plains norwith post-Pwr units, 7 craters were found to be heavilyflooded by regional plains or post-regional-plains materials,for 2 craters of the list there are no Magellan images, and forone crater Magellan imaged only a small part of it.[12] We classified craters sitting on regional plains and on

post-regional-plains units into five classes: 1) dark-parabola(DP) craters, 2) craters with clear dark halo (CH), 3) craterswith faint dark halo (FH), 4) craters with no dark halo (NH),5) craters with potential associated dark deposits obscured

Figure 5. Crater Isabella (29.81�S, 204.19�E, D = 175 km) with a clear radar-dark halo (see north,southeast and southwest of the crater) superposed on regional Pwr plains. This crater is listed in theHerrick database as having no halo probably because of the incompleteness of the circular halo. Portionof C2 MIDRP 30S181;1.


by other dark deposits. In its general idea this classification(classes 1–4) is practically the same as the classificationsuggested by Izenberg et al. [1994] and Herrick andPhillips [1994], and used in Herrick’s crater databasementioned above. In practice, our approach of imageanalysis and classification differed in some respects fromthat used in Herrick’s database and this difference will bedescribed in the following sections.

2.1. Dark-Parabola (DP) Craters

[13] In their definition and identification we follow thework of Campbell et al. [1992]. As a result of our photo-geological analysis we classified as dark-parabola craters(Figure 1) all 24 craters �30 km in diameter from their list.In addition we classified as DP two more craters of this sizecategory not listed by these authors. One of this craters isZenobia (29.35�S, 28.55�E, D = 39.1 km; Figure 1) andanother one is Bender (12.96�S, 327.35�E, 39.8 km).Zenobia has a degraded parabola that probably was thereason for not including it in the list by Campbell et al.[1992]. However, its parabolic feature is still very visible so

we decided to classify this crater as a dark-parabola one.Bender’s parabola is even more visible and its absence inthe Campbell et al. [1992] list is probably related to its late(cycle 3) imaging in the mission. So we have classified 26craters as DP, of which 21 belong to subpopulation 1(postdate regional plains) and 5 belong to subpopulation 2(postdate post-regional-plains units). Among our 26 DPcraters, 15 are classified by the Herrick database as certaindark-parabola ones, 3 as probably having dark parabolas, 3(Montessori, Bolein and Austen) as having dark nonpar-abolic haloes, and 4 (Dashkova, Cotton, Greenaway, Yonge)as having neither parabola nor dark halo (Figure 2).

2.2. Craters With Clear Dark Halo (CH)

[14] These differ from the dark-parabola craters in havinga typically smaller and nonparabolic associated dark feature.The halo was considered as prominent if it differed signifi-cantly in its darkness from the surrounding plains (Figure 3).In our approach, the size of the dark halo (large or small) wasnot a parameter of classification as well as the halo com-pleteness. The presence of a prominent dark feature even at

Figure 6. Crater Ermolova (60.30�N, 154.42�E, D = 60.9 km) with a clear radar-dark halo. Cratersuperposed on the field of Psh1 plains which are noticeably darkened by the crater-associated deposit.Even more clear are distant parts of the halo superposed on regional Pwr plains. Outer boundary of thehalo is diffuse. Listed in Herrick database as having no dark halo probably because surface neighboringthe crater looks brighter than at some distance. Portion of C1 MIDRP 60N153;1.


one side of the crater was sufficient for us to classify thecrater as CH (Figure 4). There are cases when a clear darkhalo surrounding the crater is seen within the crater-associ-ated dark parabola while other craters with a dark parabolashow no dark halo within. For us, the presence of a darkparabola was the dominant classificational parameter and weclassified these craters in the dark-parabola category. Weclassified as dark halo craters (CH and FH ones) only thosewhich have a dark halo and do not have a dark parabola.Herrick in his database describes whether or not a dark halois present within a dark parabola. So his category of dark-parabola craters includes part of his category of dark-halocraters.[15] As a result of our photogeological analysis we

classified 56 craters as having a clear dark halo, of which39 belong to subpopulation 1, and 17 belong to subpopu-

lation 2. Among our 56 CH craters, 36 are classified by theHerrick database as having dark haloes and 20 as having nodark haloes. The reason why those 20 craters were notclassified by R. Herrick as dark-halo was probably thecircular incompleteness of the halo (Figure 5) and theprominence of halo not at the boundary with crater ejectabut at some distance (Figure 6). One more reason for thedisagreement is that in some cases the crater under studywas significantly flooded by younger Pl/Ps plains, so thelink of the preserved dark halo remnants to the given craterwas not obvious and demanded special mapping andanalysis (crater Alcott, see below).

2.3. Craters With Faint Dark Halo (FH)

[16] Their halo is certainly visible but does not differsignificantly in darkness from the neighboring plains

Figure 7. Crater Deken (47.13�N, 288.48�E, D = 48 km) with a faint radar-dark halo superposed onregional Pwr plains and on the adjacent older units (mostly Pdf ). The halo, except the dark spot at SWtermination of the ejecta, is faint. Listed in Herrick database as having dark halo. Deken is one of a fewcraters on Venus superposed on Pwr plains and deformed by the wrinkle ridges (see EW-trending radar-bright feature in the northern part of the crater floor). Portion of C1 MIDRP 45N286;1.


(Figure 7). We classified 40 craters as FH, of which 34belong to subpopulation 1 and 6 belong to subpopulation2. Among 40 of our CH craters, 16 are classified in theHerrick database as having dark haloes and 24 as havingno dark haloes. The reasons why those 24 craters were notclassified by R. Herrick as dark-halo were probably lack oftheir prominence and the circular incompleteness of thehalo (Figure 8), as well as embayment by younger lavas(Figure 9).

2.4. Craters With No Dark Halo (NH)

[17] We classified 27 craters as NH (Figures 10 and 11),of which 26 belong to subpopulation 1 and one belongs tosubpopulation 2. Among our 27 NH craters all 27 areclassified by the Herrick database as having no dark haloes.It is necessary to say that some very faint remnants of darkdeposits can be found on stretched images on many cratersclassified as having no dark halo, so the boundary betweenthe craters with no halo and with faint halo sometimes maybe disputable.

2.5. Obscured Craters

[18] In this analysis we have found that in some casescraters under study and their close vicinities are mantled bydark deposits of other craters and by regional dark mantlesof unknown source (see Figure 2). This mantling was not anobstacle to determining on which geologic unit this givencrater was superposed, but it made it difficult or sometimes

impossible to determine if this given crater has the asso-ciated dark deposit and its type. We have found thatsubpopulation 1 (postdate regional plains) contains 18 suchcraters and subpopulation 2 (postdate post-regional-plainsunits) contains only one such crater. Of these 19 craters, 16are listed in Herrick’s database as having no dark haloes and3 craters are absent from the list.[19] The results of photogeologic analysis of the craters

which belong to subpopulations 1 and 2 are summarized inTable 1.[20] For the subsequent analysis we may ignore the

obscured craters and work with the remaining ones (option1). Or we may try to assess the possibility that they belongto classes DP to NH (option 2). With high probability, wemay suggest that among the obscured craters there are noDP craters because their areally extended parabolic featureswould be seen and found outside the obscured areas. Thus,potentially among the obscured craters may be those withclear halo, with faint halo and with no halo. Among the 18obscured craters of subpopulation 1, 7 are obscured bydense dark mantle, which may hide even a clear halo, and11 are obscured by a rather faint mantle, which can obscurefaint or no halo but a clear halo would be visible through it.So these observations suggest that we can subdivide the 7crater into classes CH, FH and NH based on the percentagesof these classes among the unobscured part of the subpo-pulation. Rounding the numbers to integers we come to theconclusion that among these 7 craters, 3 probably belong to

Figure 8. Crater Barrymore (52.34�S, 195.68�E, D = 56.6 km) with a faint radar-dark halo superposedon regional Pwr plains. Left is portion of C2 MIDRP 60S213;202. Right is portion of FMAP 54S189.The halo looks very faint but still visible on FMAP image and more prominent on C2 MIDRP image.Dark features seen south of the crater Barrymore on C2 MIDRP image are part of the dark parabolaassociated with crater Eudocia (59.07�S, 201.96�E, D = 27.5 km). Barrymore is listed in the Herrickdatabase as having no dark halo probably because of its circular incompleteness. This crater is anotherexample of very rare craters superposed on Pwr plains and deformed by the wrinkle ridging phase.


Figure 9. Crater Rhys (8.57�N, 298�E, D = 44 km) with a faint radar-dark halo. It was superposed onPwr regional plains and then embayed over 3/4 of its ejecta perimeter by the younger Pl lavas. Mottledfaint darkening is seen on the Pwr units and is absent on the crater-embaying Pl lavas. The latter wasprobably the reason why this crater is listed in Herrick’s database as having no dark halo. Portion of C1MIDRP 15N300;1 (top) and photogeologic map of the area (bottom).


class CH, 2 to class FH, and 2 to class NH. A similarprocedure with the 11 craters obscured by rather faint darkmantles leads us to the conclusion that 6 of them probablybelong to class FH and 5 to class NH. The only crater ofclass 5 in subpopulation 2 is obscured with a rather faintdark mantle so it may potentially belong to classes FH orNH. The procedure described above leads us to the con-clusion to consider it as belonging to class FH. In Table 2and Figure 12 we give the corrected distribution of cratersamong the classes in subpopulations 1 and 2:[21] As Table 2 and Figure 12 show, the distributions of

craters among classes DP through NH with class 5(obscured) ignored (option 1) and with class 5 divided andredistributed among classes CH, FH and NH (option 2), arerather similar. We believe, however, that it is more correct touse option 2 data and will use them in the subsequentanalysis and discussion. It is well seen from Table 2 and

Figure 12 that the percentages of craters with differentdegrees of degradation of crater-associated dark deposits insubpopulations 1 (postdate regional plains) and 2 (postdatepost-regional-plains units) are different. This difference maybe interpreted in terms of crater ages. But before we do thiswe need to discuss the possible natural and observationalbiases, such as the possible effects of crater geographiclatitude and crater size.

3. Analysis of the Distribution of Crater Classesas a Function of Latitude

[22] The presence and type of the observed crater-asso-ciated dark deposits may be a function of latitude for tworeasons. The first and most obvious reason is due to thespecifics of SAR imaging. With incidence angle <20� smallchanges in the surface slope produce large changes in radar

Figure 10. Crater Tsvetaeva (64.60�N, 147.40�E, D = 42.9 km) with no surrounding radar-dark halo.Crater is superposed on regional Pwr plains. Radar-dark deposit in lower part of the image is the distantpart of the extended halo of crater Ermolova (60.30�N, 154.42�E, D = 60.9 km). Portion of C1 MIDRP60N153;1.


backscatter, and the visibility of topographic relief on theimages is emphasized, while with the incidence angle >20�and <60�, the backscatter is dominated by surface roughness[Ford and Plaut, 1993]. Crater-associated dark deposits aredark mostly because their surface is smooth [Farr, 1993] sotheir visibility to some extent should be a function of theincidence angle, which in the case of the Magellan missionis a function of latitude. Most images analyzed in this workwere cycle 1 images and some from cycle 2. Magellanimaging was taken with incidence angles from 16.3� to45.7� in cycle 1 and with angles from 12.7� to 25.3� in cycle2. During cycle 1, imaging with incidence angle <20� wasmade poleward of 80�N and 60�S, and in cycle 2, polewardof 70�S [Plaut, 1993].[23] A second reason for a possible latitude dependence

may be the character of the circulation of the Venusianatmosphere, which in polar areas differs from that in themiddle and equatorial latitudes [Gierasch et al., 1997]. Wemay expect that the near-surface winds controlling eolianprocesses, which contribute to the degradation of the crater-associated radar-dark deposits, are latitude dependent. Thisis confirmed by the analysis of radar-dark wind streakspresented by Greeley et al. [1992]. We see the effect of theE-W superrotation of the Venusian atmosphere in the east-apexed crater-associated dark parabolas [Campbell et al.,

1992]. One may expect that close to the poles the atmos-pheric circulation is less favorable for the formation of darkparabolas.[24] In order to investigate the possible latitude depend-

ence of the crater-associated dark deposits, we subdividedthe surface of Venus into five latitudinal equal-area zones.Their boundaries are 36.9�N, 11.5�N, 11.5�S, and 36.9�S. InTable 3 are shown the numbers and percentages of thecraters with dark parabola (DP), clear dark halo (DH), faintdark halo (FH) and no halo (NH) belonging to the studiedsubpopulations 1 and 2 (combined) in these five zones (seealso Figure 13).

Figure 11. Crater Balch (29.90�N, 282.91�E, D = 40 km) with no surrounding radar-dark halo. Crater issuperposed on the regional Pwr plains and cut by the Devana Chasma rift (RT). Among the plains smallislands of tessera terrain (Tt) are present. Nonfaulted parts of Pwr plains look relatively dark with nodarkening increase around crater Balch. Tt and finely faulted RT units look relatively bright. Left isportion of C2 MIDRP 30N284;1. Right is portion of C1 MIDRP 30N279;1 (top) and photogeologic mapof this area (bottom).

Table 1. Superposition Relationships of Crater Classes in

Subpopulations 1 and 2

Crater Class

Superposed on Pwr Superposed onPost-Pwr Units

Amount Percent Amount Percent

1 Dark parabola 21 15.2 5 16.72 Clear dark halo 39 28.3 17 56.73 Faint dark halo 34 24.6 6 20.04 No halo 26 18.8 1 3.35 Obscured 18 13.1 1 3.3

Total 138 100.0 30 100.0


[25] As it is seen from Table 3 and Figure 13, a latitudedependence in areal distribution of craters with differentassociated dark deposits does exist. Higher latitude zones(poleward of 36.9�) compared to lower latitudes (11.5�–36.9�N and S) have fewer DP and CH craters and more(with one exception) FH and NH craters. This is the effectexpected from the poleward decrease in the Magellan radarincidence angle [Ford and Plaut, 1993; Plaut, 1993, Farr,1993]. The equatorial zone (11.5�N–11.5�S) shows mixedcharacteristics. It is closer to the higher-latitude zones inpercentages of DP and FH craters and closer to the lower-latitude zones in percentages of CH and NH craters. Thisunusual distribution is probably due to a stochastic effect.This zone includes the very large area of Aphrodite Terraand thus, compared to other zones, has fewer plains. This isan obvious reason for the noticeably smaller number of thecraters on plains in the equatorial zone that increases thestochastic noise. In studies of impact crater populations, ifthe number of craters of some sort is N, the value of

ffiffiffiffi

Np

isusually considered as its confidence level, which gives asense of possible stochastic error.[26] It is necessary to say, however, that the existence of

the latitude dependence in areal distribution of crater-associated dark deposits discussed does not mean that closeto poles (smaller incidence angles) there are no craters withdark deposits. Among 168 craters of subpopulations 1 and2, images of seven craters had been taken at incidence anglesmaller than the critical value (20�) cited above. Amongthese seven, there are two craters with no halo (Duse andHurston), two craters with a faint halo (Bickerdyke andLandowska), two craters with a clear halo (Ruslanovaand Marsch), and one crater with a readily visible darkparabola (Edinger). So we conclude that although thelatitudinal zonality seems to exist and we should keep itin mind, its existence does not negate the general approach

of using the degradation degree of crater-associated darkdeposits as an estimation of crater age.

4. Analysis of the Distribution of Crater Classesas a Function of Size

[27] The presence and type of crater-associated darkdeposits is expected to be a function of crater size becausethe larger the crater, the greater is the energy of its formationand the larger is the volume of the crater ejecta, which is thesource for the dark deposits. The effect of crater sizes can beseen if one compares size-distributions for all craters onVenus [Schaber et al., 1992] and the dark-parabola craters[Campbell et al., 1992]. The total crater population has amode in the 16–22.6 km interval while the subpopulation ofdark-parabola craters has a mode in the 32–45.2 kminterval.[28] To assess if the expected dependence does exist

within the size interval studied in this work (�30 km indiameter) we subdivided the 149 craters of the combinedsubpopulations 1 and 2, which belong to classes DP, CH,FH and NH, into four size categories each containingapproximately the same number of craters: the 37 largestcraters (57.6 to 270 km in diameter), the next 37 in sizecraters (42.9–57.0 km), 37 smaller craters (34.8–42.9 km)and the 38 smallest craters (30.2–34.7 km), see Table 4 andFigure 14. In addition, we subdivided these 149 craters intosix size categories (larger than 30, 40, 50, 60, 70 and 80 km)with different numbers of craters in them, see Table 5 andFigure 15.[29] The results presented in Table 4 and Figure 14 show

no clear trend, but instead display apparent stochasticvariations. The results presented in Table 5 and Figure 15show a slight trend of decrease of class CH percentage andincrease of class FH percentage along with decrease in

Figure 12. The corrected distribution of craters of classes DP (black), CH (dark gray), FH (light gray)and NH (white) in the subpopulations of craters superposed on regional plains and superposed on post-Pwr units. In each subpopulation the left histogram shows the distribution with the obscured cratersignored, while the right histogram shows distribution with the obscured craters redistributed amongclasses CH, FH, and NH.

Table 2. Corrected Superposition Relationships of Crater Classes in Subpopulations 1 and 2

Crater Class

Superposed on Pwr Superposed on Post-Pwr Units

Class 5 Ignored Class 5 Divided Class 5 Ignored Class 5 Divided

Amt Prcnt Amt Prcnt Amt Prcnt Amt Prcnt

1 Dark parabola 21 17.5 21 15.2 5 17.2 5 16.72 Clear dark halo 39 32.5 42 30.4 17 58.7 17 56.73 Faint dark halo 34 28.3 42 30.4 6 20.7 7 23.34 No halo 26 21.7 33 24.0 1 3.4 1 3.3

Total 120 100 138 100 29 100 30 100


crater size, but it is obviously within the possible stochasticvariation. So we may conclude that within the 149 craters ofsubpopulations 1 and 2 (combined) studied we do not see aclear dependence of the percentages of classes DP throughNH on crater size.

5. Degree of Degradation of Crater-AssociatedDark Deposits as a Measure of Crater Age

[30] Table 2 and Figure 12 give the corrected distribu-tions of craters among classes 1 to 4 in subpopulation 1(superposed on Pwr plains) and subpopulation 2 (super-posed on post-Pwr units). As previously mentioned, weprefer option 2 of the correction (obscured craters are notignored but divided between classes CH, FH and NH). Thepercentages of classes which represent different degradationdegrees of the crater-associated radar-dark deposits forsubpopulation 1, are as follows: dark-parabola craters,15%; craters with clear halo, 30%; craters with faint halo,30%; craters with no halo, 25%. This subpopulation of 138craters has been undergoing accumulation since the time ofemplacement of Pwr plains, which we designate as TPwr.Below we consider the meaning of the different percentagesof classes for two interpretations of the geological evolutionof Venus: 1) the so-called synchronous interpretation, inwhich evidence has been cited supporting the idea thatspecific tectonic events marking boundaries between thestratigraphic units (e.g., wrinkle-ridging) occur in differentareas of the planet at approximately the same time [Basi-levsky et al., 1997; Basilevsky and Head, 1995, 1998,2000]; and 2) the diachronous interpretation, in which suchevents are considered to happen in different areas of theplanet at different times [Guest and Stofan, 1999].[31] In the synchronous interpretation, the emplacement

of Pwr plains (and their wrinkle ridging) is considered ashappening around the planet approximately simultaneously.The age of these plains (TPwr), is approximately the same asthe mean age of the Venus surface, often designated as T, soTPwr � T (see arguments for that by Basilevsky et al. [1997,1999] and Basilevsky and Head [1998, 2000]). Let us

assume that during time period T, the rate of degradationof the crater-associated dark deposits was approximately thesame all around the planet and did not depend on the cratersize. In this case, the different crater-associated dark depos-its should have characteristic lifetimes: TDP for dark parab-olas, TCH for clear halo, and TFH for faint halo. The totallifetime for all dark deposits is to be TDD = TDP + TCH +TFH. These lifetimes should be proportional (see below) tothe observed percentages of craters having a differentdegree of degradation of the associated dark deposits(PDP, PCH and PFH). For the first 0.15 T since the momentof crater formation it has a dark parabola, which is beinggradually degraded during this time (TDP = 0.15T). Then forthe time period 0.3T, from 0.15T until 0.45T from craterformation, the crater has a clear halo (TCH = 0.3T), whichafter that becomes a faint halo and exists in this form foranother 0.3T, from 0.45T until 0.75T from the craterformation, (TFH = 0.3T). Then the crater loses its faint haloand becomes a nonhalo crater. Such is the process ofdegradation of the crater-associated dark deposits.[32] In the diachronous interpretation, the emplacement

of Pwr plains occurred in different places on Venus atdifferent times [Guest and Stofan, 1999]. In this case, timeTPwr has the meaning of the mean age. Comparable to thesituation in the synchronous option, it is close to the meanage of the surface (TPwr � T). Under the same assumptionsas above, the total lifetime for all dark deposits is also TDD =TDP + TCH + TFH. In the diachronous option, the earlier-formed areas of Pwr are older than T, while the later-formedones are younger than T. If the age of some area of Pwrplains is greater than the total lifetime of crater-associateddark deposits (twr > TDD), this area should have somenumber of nonhalo craters. Their number should be propor-tional to the time interval (twr – TDD), while percentages ofthe craters having a different degree of degradation ofassociated dark deposits (PDP, PCH and PFH) should becorrespondingly proportional to TDP, TCH, and TFH. If theage of some area of Pwr plains is equal to TDD, this area, inthe ideal case, should have no nonhalo craters and PDP, PCHand PFH should be correspondingly proportional to TDP,

Table 3. Distribution of Crater Classes in Subpopulations 1 and 2 as a Function of Latitude Zones

Crater Class

36.9–90N 11.5–36.9N 11.5N–11.5S 11.5–36.9S 36.9–90S

Amt % Amt % Amt % Amt % Amt %

1 DP 4 13.3 7 22.6 2 8.3 9 30.0 4 12.52 CH 9 30.0 12 38.7 11 45.9 16 53.4 8 25.03 FH 8 26.7 3 9.7 8 33.3 4 13.3 11 34.44 NH 9 30.0 9 29.0 3 12.5 1 3.3 9 28.1

Total 30 100 21 100 24 100 30 100 32 100

Figure 13. Distribution of craters of classes DP (black), CH (dark gray), FH (light gray) and NH (white)in the combined subpopulations 1 and 2 by five equal-area latitude zones.


TCH, and TFH. If the age of some area of Pwr plains issmaller than the dark-deposit total lifetime (twr < TDD), thisarea should have no nonhalo craters and should showviolations of proportionality of PDP, PCH and PFH to TDP,TCH, and TFH correspondingly. Depending on how youngthis area is, it may have lowered PFH, then no craters withfaint halo, then lowered PCH, then only craters with a darkparabola.[33] These considerations are illustrated by Figure 16,

which shows on the left the example of the model of thesynchronous interpretation, and on the right the example ofthe model of the diachronous interpretation. The verticalupward-pointing arrow shows the time with marked age T,and the present time at the end of the arrow. For simplicity,lifetimes of dark parabolas (TDP), clear halo (TCH) and fainthalo (TFH) are 0.25 T each (TDD = 0.75 T), and the timeinterval when craters now observed as nonhalo ones (TNH)is also 0.25 T. ‘‘Stars’’ show impact craters formed along thetime (one crater per 0.125 T). At the bottom of Figure 16there are histograms showing percentages of DP, CH, FHand NH craters resulting from the model cases considered.[34] The synchronous model is exemplified by seven

geologic provinces (1 through 7). In each of them, Pwrplains have been emplaced at time T, and in each of them 8craters formed for the time from T until present. Amongthem, the youngest two are observed at the present time asdark-parabola craters, the two preceding them have clearhalos, the two formed even earlier have faint halos, and thetwo oldest craters are nonparabola craters. It is well seenthat the percentages of craters of classes DP, CH, FH andNH are 25% of the total population for each class, thusbeing proportional to the appropriate lifetimes.[35] The diachronous model is exemplified by 11 prov-

inces (1 through 11). The lifetimes TDP, TCH, and TFH andthe rate of crater formation are the same as in the synchro-nous option. In province 6, Pwr plains have been emplacedat time T, they contain 8 craters, two of each class. Inprovince 1, Pwr plains have been emplaced at time 1.625T,they contain 13 craters, of which two have dark parabola;

two have clear halo; two have faint halo and seven have nohalo. In province 11, Pwr plains have been emplaced at time0.375T, they contain only three craters, of which two aredark-parabola, and one has a clear halo. Other provincesrepresent the intermediate cases.[36] We consider as examples of the diachronous model:

1) set of all provinces from 1 through 11; 2) set of provinces2 through 10, 3) set of provinces 3 through 6, 4) set ofprovinces 4 through 8, and, finally, 5) set of provinces 5through 7. All these five model cases are similar in that themean time of emplacement of Pwr plains is T. But they aredifferent in the range of the earliest-to-latest times ofemplacement of Pwr plains. The largest range is 1.250 T(for set 1), the smallest is 0.25 T (for set 5). The followingTable 6 and the bottom of Figure 16 summarize the effect ofthis range on the observed percentages of craters belongingto different classes.[37] The results presented in Table 6 and Figure 16

demonstrate visually what was discussed above. In the caseof the synchronous model, percentages of craters having adifferent degree of degradation of associated dark depositsare proportional to the lifetimes of the appropriate deposittypes. So if the synchronous model represents the reality ofVenus geology, then the direct application of numbers givenin the beginning of this section leads to the conclusion thatcraters with a dark parabola are younger than about 0.15T,craters with a clear halo have an age range from 0.15 to0.45T, craters with a faint halo have an age from 0.45 to0.75T, and nonhalo craters are older than 0.75T.[38] In the case of the diachronous model, depending on

how large the difference is in the times of emplacement ofPwr plains in different geological provinces, the percentagesdiscussed may be either proportional to the appropriatelifetimes or not. Analysis of the results given in Table 6and Figure 16 shows that the loss of proportionality for thisgiven class of degradation degree occurs when the youngestPwr emplacement happens within the time of formation ofcraters which are now observed in the form of this class. Inthe case of the extreme difference in times of Pwr emplace-

Table 4. Distribution of Crater Classes in Subpopulations 1 and 2 as a Function

of Crater Size

Crater Class

57.5–270 42.9–57 34.8–42.9 30.2–34.7 �


1 DP 6 16.2 5 13.5 8 21.6 7 18.4 26 17.42 CH 14 37.9 19 51.4 9 24.3 14 36.9 56 37.73 FH 8 21.6 8 21.6 11 29.8 13 34.2 40 26.84 NH 9 24.3 5 13.5 9 24.3 4 10.5 27 18.1

Total 37 100 37 100 37 100 38 100 149 100

Figure 14. Distribution of craters of classes DP (black), CH (dark gray), FH (light gray) and NH (white)in the combined subpopulations 1 and 2 by four size classes with total distribution for reference (thehistogram on the far right).


ments (set of provinces 1 through 11 in which the youngestPwr plains is emplaced during the time when craters nowpossessing a clear halo formed), the observed percentage ofDP craters is proportional to the model dark-parabola life-time (25%) while the observed percentages for CH (23.9%)and FH (19.3%) craters are not proportional to the appro-priate model lifetimes. In the case when the youngest Pwremplacement was during the time when craters now pos-sessing a faint halo formed (provinces 3 through 9), thepercentages of DP and CH craters are proportional to theappropriate lifetimes (25 and 25%). The proportionality ofpercentages of DP craters and CH craters to the appropriatelifetimes occurs even in the case of the set of provinces 2through 10, when the youngest Pwr emplacement happensat the lower boundary of CH time. If so, then referring to thepercentages of the DP, CH. FH and NH craters in post-Pwrsubpopulation (15, 30, 30, 25%) we may conclude that ifdifferences in the times of emplacement of Pwr plains indifferent provinces of Venus were not larger than 0.5T,where T is the mean of the Pwr emplacement times, then atleast for the DP and CH craters the observed percentages areproportional to their lifetimes.[39] The above analysis (see section 4 of this paper)

showed no clear dependence of dark-deposit degradationdegree on crater size. This analysis was done for craters�30 km in diameter. Although smaller craters have not beendirectly studied in this work we believe that it is alsopossible to use the degree of their dark-deposit degradationas a measure of their age too. Formation of smaller cratersproduces a smaller amount of ejecta, so their dark depositsare expected to have smaller volumes and areas, and thusshould be degraded faster than in the case of larger craters.So it is logical to expect that the above age estimates for thedark-parabola and clear halo craters �30 km in diameter arethe maximum margins for the smaller craters: smaller (howmuch smaller is unknown) than 0.5 to 0.1–0.15T for theclear-halo craters and less than 0.1–0.15T for the dark-parabola craters.

6. Dark-Deposit Classes in the Subpopulation ofCraters Superposed on Post-Pwr Units

[40] As was shown earlier (see Table 2 and Figure 12),the distribution of classes DP, CH, FH and NH for thecraters of subpopulation 2 (superposed on post-Pwr units)differs from that of population 1 (superposed on regionalplains). To better understand the meaning of this, weconsider the model cases as was done in the previoussection of the paper (see Figure 16 and Table 7). First wewill do this for the synchronous model. Figure 17 showsthree synchronous model cases. They are similar in thesame time of emplacement of the Pwr plains (TPwr = T), inthe rate of crater formation (one crater per 0.1T), and in thelifetimes of different types of the dark deposits. The lattercorrespond to the corrected PDP, PCH, PFH and PNH of thesubpopulation 1 (15, 30, 30 and 25%). On the left is showncase 1, in which emplacement of the post-Pwr units indifferent geologic provinces is evenly distributed with time.The mean age of the post-Pwr units in this case is 0.5T. Inthe center is shown case 2, in which emplacement of thepost-Pwr units is concentrated in the beginning of timeperiod T. The mean age of the post-Pwr units in this case is0.82T. On the right is shown case 3, in which emplacementof the post-Pwr units is concentrated in the upper part oftime period T. The mean age of the post-Pwr units in thiscase is 0.22T.[41] At the bottom of Figure 17 is shown the model-result

percentages of the DP, CH, FH and NH classes for sub-population 2 (upper row of histograms) and the correspond-ing percentages for subpopulation 1 (lower row). Dashedboxes on the upper row histograms show percentages ofVenus craters of subpopulation 2 studied above (Table 2 andFigure 12). Figure 17 shows that percentages of the DP, CH,FH and NH classes of subpopulation 1 differ from those ofsubpopulation 2. In model case 1, when emplacements ofthe post-Pwr units in different geologic provinces are evenlydistributed with time, the population 2 percentages PDP andPCH are significantly higher, while PFH, and especially PNH,are lower than those of subpopulation 1. In model case 2,when emplacement of the post-Pwr units is concentrated inthe beginning of time period T, the population 2 percentagesPDP, PCH and PFH are somewhat higher, while PNH issignificantly lower, than those of subpopulation 1. In modelcase 3, when emplacement of the post-Pwr units is con-centrated in the upper part of time period T, population 2percentages PDP and PCH are much higher (especially PDP)than those of subpopulation 1, while FH and NH craters areabsent. Comparisons with the percentages for real subpo-pulation 2 show that the latter most resembles model case 1,

Table 5. Distribution of Crater Classes in Subpopulations 1 and 2

as a Function of Crater Size

CraterClass

>80 >70 >60 >50 >40 >30

Amt % Amt % Amt % Amt % Amt % Amt %

1 DP 2 13.3 4 16.7 6 17.7 7 14.6 13 15.5 26 17.42 CH 9 60.1 11 45.8 14 41.2 20 41.7 34 40.5 56 37.73 FH 2 13.3 5 20.8 6 17.6 11 22.9 18 21.4 40 26.84 NH 2 13.3 4 16.7 8 23.6 10 20.8 19 22.6 27 18.1Total 15 100 24 100 34 100 48 100 84 100 149 100

Figure 15. Distribution of craters of classes DP (black), CH (dark gray), FH (light gray) and NH (white)in the combined subpopulations 1 and 2 by six size classes, the sixth class (the histogram on the far right)is the total distribution.


with some similarity (DP) to case 2, and a prominentdissimilarity to case 3.[42] Three typical cases for the diachronous model are

shown on Figure 18 and in Table 8. They are similar in thetimes of emplacement of the post-Pwr units: mean time ofemplacement is T, extreme variations in the emplacementtimes tPwr are ±0.5T, variations of tPwr in different geologicprovinces are symmetrical in relation to T. As in the case ofthe synchronous model, these three diachronous cases arealso similar to each other in the rate of crater formation (onecrater per 0.1T), and in the lifetimes of different types ofdark deposits, which are proportional to the corrected PDP,PCH, PFH and PNH of the subpopulation 1 (15, 30, 30 and25%). On the left is shown the case when emplacement ofthe post-Pwr units in different geologic provinces are evenlydistributed with time. The mean age of the post-Pwr units inthis case is 0.48T. In the center is shown the case whenemplacement of the post-Pwr units is concentrated in thevery beginning of time period T. The mean age of the post-Pwr units in this case is 0.9T. In the geologic sense this isthe case when post-Pwr units represent the terminal stagesof those magmatic-tectonic cycles which formed theregional plains of the appropriate provinces. On the rightis shown the case when emplacement of the post-Pwr units

is concentrated in the upper part of time period T. The meanage of the post-Pwr units in this case is about 0.22T.[43] At the bottom of Figure 18 is shown the results of

model percentages of the DP, CH, FH and NH classes forsubpopulation 2 (upper row) and the corresponding percen-tages for subpopulation 1 (lower row). Dashed boxes on theupper row histograms show percentages of Venus craters ofsubpopulation 2 studied above. For subpopulation 1 theresulting percentages of the DP and CH classes are the sameas those in the synchronous model, while the percentages ofFH and NH classes are different. Compared to the synchro-nous model, the percentage of FH decreased while that of

Figure 16. Diagrams illustrating model evolution of the crater populations superposed on regionalplains in different geologic provinces (top) and the resulting percentages of craters of classes DP, CH, FH,and NH (bottom). See text for more details.

Table 6. Distribution of Crater Classes as a Function of Provinces

CraterClass

Prov.1–11

Prov.2–10

Prov.3–9

Prov.4–8

Prov.5–7


1 DP 22 25.0 18 25 14 25 10 25.0 6 25.02 CH 21 23.9 18 25 14 25 10 25.0 6 25.03 FH 17 19.3 15 20.8 13 23.2 10 25.0 6 25.04 NH 28 31.8 21 29.2 15 26.8 10 25.0 6 25.0

Total 88 100 72 100 56 100 40 100 24 100


NH proportionally increased. In the diachronous model case1, when emplacement of the post-Pwr units in differentgeologic provinces is evenly distributed with time, thepopulation 2 percentages PDP and PCH are significantlyhigher while PFH and especially PNH are lower than thoseof subpopulation 1. In model case 2, when emplacement ofthe post-Pwr units is concentrated in the beginning of timeperiod T, the population 2 percentages PDP, PCH and PFH areslightly higher, while PNH is slightly lower, than those ofsubpopulation 1. In model case 3, when emplacement of thepost-Pwr units is concentrated in the upper part of time

period T, the population 2 percentages PDP and PCH aremuch higher (especially PDP) than those of subpopulation 1,while FH and NH craters are absent. Comparisons with thepercentages for the real subpopulation 2 show that the lattermost resembles model case 1, with some similarity (DP andFH) to case 2, and a prominent dissimilarity to case 3.[44] These model considerations show that the percen-

tages of dark-parabola, clear-halo, faint-halo and nonhalocraters determined in subpopulation 1 (superposed onregional plains) and subpopulation 2 (superposed on post-Pwr units) are in agreement with the generally (on average)

Table 7. Typical Cases for the Synchronous Model of the Regional Plains Emplacement

CraterClass

Case 1,tP-Pwr Distributed Evenly

Case 2,tP-Pwr in the Beginning of T

Case 3,tP-Pwr in the End of T

Subpop.1 Subpop.2 Subpop.1 Subpop.2 Subpop.1 Subpop.2

N % N % N % N % N % N %

DP 7.5 15.0 7 28.0 7.5 15.0 7.5 18.3 7.5 15.0 7 63.6CH 15 30.0 10.5 42.0 15 30.0 15 36.5 15 30.0 4 36.4FH 15 30.0 6 24.0 15 30.0 14.5 35.4 15 30.0 – –NH 12.5 25.0 1.5 6.0 12.5 25.0 4 9.8 12.5 25.0 – –Total 50 100 25 100 50 100 41 100 50 100 11 100

Figure 17. Diagrams illustrating model evolution (synchronous cases) of the crater populationssuperposed on regional plains (wavy black bars at the bottom of each column) and on the post-Pwr units(black bars above) in different geologic provinces (top) and the resulting percentages of craters of classesDP, CH, FH, and NH (bottom). The bottoms of all columns are at the same time-level T, showing thesynchronous emplacement of the regional plains. See text for more details.


higher stratigraphic position of subpopulation 2 comparedto subpopulation 1. Although at the moment we do notknow whether the synchronous or diachronous modelcorresponds better to the reality of Venus geology, the

observed percentages imply rather constant rates of volcanicand tectonic activity during post-Pwr time with the possi-bility of somewhat higher rates in the lower part of thisperiod. If the variations of times of emplacements of

Figure 18. Diagrams illustrating model evolution (diachronous cases) of the crater populationssuperposed on regional plains (wavy black bars at the bottom of each column) and on the post-Pwr units(black bars above) in different geologic provinces (top) and the resulting percentages of craters of classesDP, CH, FH, and NH (bottom). The bottoms of the columns are at different time levels, showingdiachronous emplacement of the regional plains. See text for more details.

Table 8. Typical Cases for the Diachronous Model of the Regional Plains Emplacement

CraterClass

Case 1,tP-Pwr Distributed Evenly

Case 2,tP-Pwr in the Beginning of T

Case 3,tP-Pwr at the End of T

Subpop.1 Subpop.2 Subpop.1 Subpop.2 Subpop.1 Subpop.2

N % N % N % N % N % N %

DP 7.5 15.0 7 29.2 7.5 15.0 7.5 16.7 7.5 15.0 7 63.6CH 15 30.0 11 45.9 15 30.0 15 33.3 15 30.0 4 36.4FH 12 24.0 4.5 18.7 12 24.0 11 24.4 12 24.0 – –NH 15.5 31.0 1.5 6.2 15.5 31.0 11.5 25.6 15.5 31.0 – –Total 50 100 24 100 50 100 45 100 50 100 11 100


regional plains are less than 0.5T (it includes synchronousoption as the extreme case), the percentages of dark-parab-ola and clear halo craters should be proportional to thelifetimes of these types of crater-associated dark deposits.[45] Keeping in mind the assumptions mentioned above,

possible uncertainties in classification of specific craters,and the observed latitudinal zonality, it is probably appro-priate to consider CH craters as having formed during thetime period between about 0.5T and 0.1–0.15T, and DPcraters as having formed between 0.1–0.15T and thepresent. For FH and NH craters we may conclude that theyare older than about 0.5T. We believe that these ageestimates can be equally applied to craters of subpopula-tions 1 and 2.[46] Earlier we proposed that the period during which

craters now having dark parabolas formed be designated asthe Aurelian Period [Basilevsky and Head, 1995, 1998,2000; Basilevsky et al., 1997]. The period between emplace-ment of the regional plains (and their wrinkle-ridging) andthe beginning of Aurelian time was defined as the AtlianPeriod. The Atlian/Aurelian boundary (estimated at thattime as �0.1T) is controlled by surficial processes and toa first approximation does not depend on the synchronousvs. diachronous controversy, while the meaning of the lowerAtlian boundary (emplacement of the wrinkle ridge net-work) does depend on it. The results of the present workprovide the basis to subdivide the Atlian Period into twoparts (Figure 19). Their boundary is the beginning of thetime during which craters now observed as having a clearhalo formed. This boundary is also controlled by thesurficial processes and evidently does not depend on thesynchronous vs. diachronous controversy. Thus, the timeboundaries for the Upper Atlian time are interpreted to be�0.5T and 0.1–0.15T and for the Lower Atlian time, �0.5Tand T. In the case of diachronous option the lower boundaryhas a meaning of the value being averaged from local timeboundaries.

7. Dating Geologic Structures

[47] In this section we apply the estimates of the lifetimesfor the dark-parabola and clear-halo craters described aboveto assessing the ages of several concrete geologic regionsand structures.

7.1. Beta-Devana Structure

[48] Beta Regio is a 2000 � 3000 km dome-like topo-graphic highstanding 4–5 km above the surrounding terrainand centered at 30�N, 282�E [Solomon et al., 1992; Senskeet al., 1992]. It is cut by the several kilometer deep, N-Strending trough of Devana Chasma consisting of a set ofnormal faults. A shield-like volcanic edifice, Theia Mons,350 km across and 5 km high, is located in the southern partof Beta Regio. It is superposed on the Devana Chasma rift,partly filling it with lava flows. In the northern part of BetaRegio there is another shield-shaped topographic high ofapproximately the same size and height, Rhea Mons. Basedon Pioneer Venus and Arecibo data Rhea Mons wasconsidered as volcano too [Masursky et al., 1980; Campbellet al., 1984; Stofan et al., 1989] but the Magellan datashowed that area within and around Rhea Mons is com-posed of rift-faulted tessera terrain with no distinguishablelava flows, thus disproving its volcanic origin [Solomon etal., 1992; Senske et al., 1992]. Beta Regio has a substantialgravity anomaly with a large apparent depth of compensa-tion [Phillips et al., 1997; Sjogren et al., 1997; Smrekar etal., 1997]. On the basis of topographic, morphologic andgravity characteristics Beta Regio is interpreted as a regionof uplift, rifting and volcanism overlying a site of upwellingof hot mantle material [Solomon et al., 1992; Senske et al.,1992; Smrekar et al., 1997].[49] We have found that among 10 impact craters in this

region [Schaber et al., 1998], five craters are directlyaffected by Beta-Devana volcanic-tectonic activity and thusthey can potentially be used for constraining the age of thisstructure. Figure 20 shows the general morphology of theBeta-Devana structure and positions of the mapped areas.Figure 21 shows images of all the craters [portions of thesame Magellan mosaic (C1MIDRP 30N279;1)] that permitsone to compare the darkness for all five of them. The first isthe 40 km crater Balch in the central part of the structureshown above (see Figure 11) as an example of a nonhalocrater. It was superposed on Pwr regional plains andseverely cut by the Devana rift faults. The absence of ahalo places formation of this crater in the first half of Atliantime (that is, older than 0.5T) so one can conclude that therifting occurred after emplacement of the regional plainsand within or later than the first half of the Atlian period.[50] Second is the 83.6 km crater Sanger superposed on

Psh1 regional plains and tessera (Tt) on the NE flank of theBeta uplift (Figures 20, 21, and 22). This crater has aprominent hummocky ejecta deposit and extended ejectaoutflows, and beyond them, a clear dark halo. The presenceof the clear halo places the formation of this crater in thesecond half of the Atlian Period (the time from �0.5T to0.1–0.15T). To the NW of the crater, its ejecta outflows areobviously cut by long and narrow fractures branching fromthe main set of faults of Devana Chasma. This providesevidence that within the time from �0.5T to 0.1–0.15T, oreven later, Devana rifting was active.

Figure 19. Diagram showing types of crater-associateddark deposits and related time-stratigraphic units.


[51] Third is the 13.5 km crater Raisa. It lies 350 kmnorth of the summit of Theia Mons volcano among Pl lavaflows and is heavily flooded by them (Figure 23). Geologicmapping shows that close to the crater, among the Pl flows,are ‘‘windows’’ of Pwr plains with recognizable wrinkleridges on them. These areas are dark and close enough tothe crater to be interpreted as remnants of the faint dark halo

associated with the crater Raisa. We distinguish here twounits of Pl flows. The older ones (Pl1) are observed to thewest and south of the crater not being in direct contact withit. The younger ones are seen almost everywhere. Theyflood the crater leaving unflooded only part of its rim and asmall remnant of hummocky ejecta on the north-east part ofthe crater. The mapped Pl1unit appears noticeably darker

Figure 20. Image showing the general morphology of the Beta-Devana structure and areas whereimpact craters show age relations with post-Pwr volcanism and rifting (white boxes). Portion of C1MDRP 30N279;1.


than Pl2 although not so dark as the neighboring window ofPwr. We interpret its darkness to be part of the crater darkhalo. If this is correct, then one can conclude that the craterRaisa was superposed on the Pwr plains, darkened them andprobably the early unit of the Pl flows while the later flows(Pl2), which are most abundant here, postdated the crater.[52] Crater Raisa is smaller that the craters of the sub-

populations directly studied by us (>30 km in diameter). Asit was suggested above we may apply the age estimates tothe smaller crater as the maximum possible values. In thesynchronous interpretation the maximum possible age esti-mate for faint-halo craters is about 0.75T to 0.5T. In thediachronous interpretation it depends on how large is therange of the times of emplacement of the Pwr plains. Sokeeping in mind that crater Raisa is more than two timessmaller than the lower limit of the crater subpopulationstudied by us, we may suggest that the crater Raisa formedin more recent time than �0.75T and maybe even morerecently than 0.5T. This is the estimate of when the TheiaMons volcano was active.[53] Fourth is the 15.5 km crater Olga (Figure 24). It lies

in southern Beta at the eastern flank of the Devana rift. Thecrater is superposed on heavily faulted terrain (RT) withinwhich are moderately faulted remnants of Pwr plains andponds of less faulted to unfaulted Pl flows. To the north andeast of the crater there are two areas of moderately faultedplains which bear a few features appearing similar towrinkle ridges. Thus, these are probably Pwr plainsalthough we can not exclude that they are early subunitsof Pl flows. These two areas are dark enough to beconsidered as remnants of a clear halo. The crater is locallycut by faults of a younger phase (RT2). If the observed darkareas are indeed remnants of the crater-associated halo, thecrater maximum age should be not greater than �0.5T. Thiscrater was superposed on major episodes of the Devanarifting and cut by the later one.[54] Fifth is the 10.5 km crater Tako (Figure 25). It is

superposed on Pwr plains and on faults branching fromDevana Chasma. The crater has a faint dark halo. If weconsider the synchronous interpretation, its maximum age isless than 0.75T and, keeping in mind its small size, it maybe significantly less than this estimate.

[55] In summary, in reference to the analysis of ageestimates for the Beta-Devana structure, we may concludethat the observations put major constraints in the vicinity ofthe craters Sanger (rift-associated faulting occurred after0.5T), Raisa (Theia lava flows are younger than 0.75T), andOlga (some rift-associated faulting occurred after 0.5T),Balch (faulting after emplacement of Pwr plains, that inthe synchronous interpretation means after T), and Tako(faulting after 0.75 T, synchronous interpretation). Thegeneral conclusion is that at least some part of the volcanicand tectonic activity of the Beta-Devana structure occurredwithin the Upper Atlian time (<0.5T) or even later, withinAurelian time.

7.2. Kalaipahoa-Tarbell-Mylitta Structure

[56] Mylitta Fluctus is the 400 � 1200 km volcanic flowfield at the SE edge of Lavinia Planitia (Figure 26). Itssource is within Tarbell Patera which sits on the KalaipahoaLinea rift zone. Roberts et al. [1992] mapped the timesequence of the Mylitta volcanic flows showing that the firstphase of volcanism was characterized by flows spreadingradially from the source centered at 58�S, 351.5�E. Most ofthe Mylitta lava field was produced during later phases,when lavas flowed northward flooding regional Pwr plainsof Lavinia Planitia. All observed lavas of Mylitta Fluctusbelong to the Atlian-aged unit of Lobate Plains (Pl).[57] In the southern part of the volcanic field, within the

early phase lavas mentioned previously, occurs the 66 kmcrater Alcott, whose relations with the surrounding geologicunits help to date the Mylitta volcanism (Figure 27). Exceptfor its eastern and southeastern rim, the crater is heavilyflooded by Pl flows emanating from Tarbell Patera. Theeastern part of the crater is embayed from inside and outsideby Smooth plains (Ps), another member of the AtlianSystem. The relationship between the Ps unit and TarbellPatera is unclear. The Ps unit is moderately deformed byfaults of the southern flank of the Kalaipahoa Linea rift (ourunit RT). Within the rift, where deformation is generallyheavy, locally there are blocks of moderately deformed Pwrplains. The latter are more abundant north of the rift and tothe northwest of Tarbell Patera. The Mylitta Pl flows floodthe Pwr plains, the rifted terrain (RT), and the Ps unit.

Figure 21. Portions of C1 MDRP 30N279;1 images of five craters showing age relations with post-Pwrvolcanism and rifting of the Beta-Devana structure. Each fragment covers an area 70 � 115 km.


[58] The areas of Pwr plains close to Alcott crater arevery dark. At larger distances, 400–500 km NE and NWfrom the crater, there is seen a prominent darkening of Pwrplains which progressively decreases with increasing dis-tance from the crater. We interpret these darkened areas of

Pwr plains surrounding the crater Alcott to be unfloodedremnants of a clear halo associated with the crater Alcott.For the area northeast of the crater, where the darkening isspatially continuous from the closest to farthest vicinities ofthe crater, this interpretation seems quite evident. For the

Figure 22. Crater Sanger (33.77�N, 288.56�E, D = 83.6 km) having a clear dark halo. Ejecta outflowsof this crater are cut by the long fault (white arrows) branching from the Devana Chasma rift. Portion ofC1 MIDRP 30N279;1 (top) and photogeologic map of the area (bottom).


darkened area to the northwest of the crater, this interpre-tation is less certain. The darkening in this specific placemay also be a remnant of a halo of some crater completelyflooded by Mylitta Fluctus, or even a pyroclastic deposit.[59] But even in the case that this interpretation is correct

only for the area northeast of the crater, the Alcott-associ-ated halo affects Pwr plains, probably RT of Kalaipahoa

Linea, but does not affect units Ps and Pl. Thus one canconclude that both the Mylitta lavas and the Ps unit in thisvicinity are younger than �0.5T and this puts their emplace-ment into the Upper Atlian period, or even younger, withinAurelian time. A late phase of Kalaipahoa rifting, whichaffected the Ps unit, would also be Upper Atlian. The majorphase of deformation, which produced the RT unit of

Figure 23. Crater Raisa (27.95�N, 280.30�E, D = 13.5 km) heavily flooded by lavas of Theia Monsvolcano. Darker areas are remnants of Pwr plains affected by the faint halo associated with this crater.Portion of FMAP 30N282 (left) and photogeologic map of the area (right).

Figure 24. Crater Olga (26.1�N, 283.8�E, D = 15.5 km) heavily fractured by the Devana Chasma rift.Darker area of moderately deformed segment of Pwr plains north and northeast of the crater is interpretedas affected by the clear halo associated with this crater. Notice that unfractured blocks of the Pwr unit aredarker closer to the crater, and less dark at distance. Portion of FMAP 30N282 (left) and photogeologicmap of the area (right).


Kalaipahoa Linea, although it seems to predate the forma-tion of Alcott, could be both younger and older than 0.5T.

7.3. Atla Regio Structure

[60] Atla Regio is a volcanic-tectonic structure centeredon the 1000 km � 1000 km upland where five rift zones areseen to converge (Figure 28). This structure includes severallarge volcanic mountains. Among them, the largest are OzzaMons, a 7.5-km-high volcano, which dominates the centralpart of Atla, and Maat Mons, a 9-km-high volcano, which isoffset from the Atla structure center to the west [Senske etal., 1992]. Smaller volcanoes and lava fields associated withrift zones are frequent here. Atla Regio has a substantialgravity anomaly with a large apparent depth of compensa-tion [Phillips et al., 1997; Sjogren et al., 1997; Smrekaret al., 1997]. This large region has 18 impact craters[Schaber et al., 1998]. Among them 9 craters show relationswith post-Pwr units which may help to date the relativelyyoung volcanic and tectonic activity of this area (Figure 29).[61] One of these craters is the 38.9 km crater Uvaisi

(Figure 30), which has already been used for estimating theage of volcanism in this area [Basilevsky, 1993; Basilevskyand Head, 2002]. It has a dark parabola, so it was formedwithin the Aurelian Period (<0.1–0.1.5T). The crater issuperposed on the lavas emanating from Ozza Mons vol-cano which are significantly deformed by the faults of theNE part of Dali Chasma rift. The western part of Uvaisi’sejecta and part of its floor are embayed by distal flows ofMaat Mons volcano. These relations show that the MaatMons volcano was active as recently as 0.1–0.15T, or evenless. Emplacement of lavas whose source was Ozza Monsvolcano, and their deformation by the Dali Chasma struc-

tures, predated formation of the crater Uvaisi, so they maybe either also Aurelian or Atlian.[62] There are two craters, Fossey (30.4 km in diameter)

and Piscopia (26.2 km) which sit on Pwr regional plains andshow age relations with a lava flow unit emerging from thewestern flank of Maat Mons (Figure 31). Fossey has a well-developed clear dark halo, which affects Pwr plains within60–80 km outwards of the crater ejecta blanket and doesnot affect the Maat lava flow, which enters inside the halozone. We interpret this as evidence that formation of craterFossey predated this lava flow. The presence of a clear darkhalo associated with this crater dates it as being youngerthan 0.5T, which is the maximum age limit for the Maat lavaflow.[63] The crater Piscopia is 150 km east of the crater

Fossey (Figures 31 and 32). It has an associated faint darkhalo. If we apply to it the age estimates described above(section 6) for craters �30 km in diameter, Piscopia is olderthan 0.5T and, in the case of the synchronous interpretation,is younger than 0.75T. Piscopia is 26.2 km in diameter, thatis close to 30 km, so these estimates seem to be valid for ittoo. A protuberance of that lava flow which approachedcrater Fossey, enters into the southern part of Piscopia’sejecta blanket. This is in agreement with the age relationsconsidered above for this lava flow with crater Fossey,although it does not place any additional age constraints.[64] The 21.8 km crater Melba is superposed on the Pwr

regional plains north of Maat Mons (Figure 33). It has anassociated clear dark halo. The distal flow of Maat volcanoenters into the eastern and northeastern parts of the craterhalo and embays the crater hummocky ejecta blanket. Thediameter of crater Melba is smaller than 30 km so the

Figure 25. Crater Tako (25.11�N, 285.27�E, D = 10.7 km) superposed on Pwr plains and on faultsbranching from Devana Chasma. Portion of FMAP 30N282 (left) and photogeologic map of the area(right).


presence of a clear dark halo dates this crater as being UpperAtlian (<0.5T ago) and maybe even Aurelian (<0.1–0.15Tago). This implies that the flow from Maat volcano whichpostdates crater Melba is either Upper Atlian or Aurelian.

[65] The 29.1 km crater Von Schuurman is superposed onPl lavas associated with the northeastern segment of theDali Chasma rift zone (Figure 34). The crater has aprominent dark parabola which affects the Pl lavas, the

Figure 26. Image showing general morphology of Kalapahoi-Tarbell-Mylitta structure and the areawhere impact crater Alcott shows age relations with post-Pwr volcanism and rifting (white box). Portionof C1 MDRP 60S347;1.


rifted terrain (RT) and the older units Pwr and Pfr. Becauseof the presence of the associated dark parabola, the craterVon Schuurman is assigned an Aurelian age (<0.1–0.15T).So the Dali Chasma rifted terrain at this locality and theassociated Pl volcanism may be either Aurelian or Atlian.[66] The 32.8 km crater Sitwell is superposed on the

rifted terrain (RT) of Ganis Chasma which cuts here the Pwrregional plains (Figure 35). This crater was already used fordating the geologic units of this area [Basilevsky, 1993]. Ithas a prominent dark parabola which affects the riftedterrain and the Pwr regional plains. Because of the presenceof the associated dark parabola, the crater Sitwell isassigned an Aurelian age (<0.1–0.15T). So the GanisChasma rifting at this locality may be either Aurelian orAtlian.[67] The 36.2 km crater Bashkirtseff is superposed on the

Pwr regional plains and flooded from the north by the Pllavas emanating from the Yolkai-Estsan Mons volcano,sitting on the Ganis Chasma rift (Figure 35). South of thecrater the surface of the Pwr plains is darkened and weinterpret this as the remnant of a faint halo associated withthe crater Bashkirtseff. If so, this crater should be assigned aLower Atlian age. If true, this implies that the Pl lavas and

the rifting in this vicinity may be either Lower Atlian, orUpper Atlian or even Aurelian.[68] The 11.5 km crater Udagan is superposed on Pwr

regional plains moderately deformed by faults branchingfrom Tkashi-mapa Chasma rift zone (Figure 36). The craterhas a faint dark halo. If this crater was 30 km in diameter orlarger, the presence of the faint halo would suggest a date ofLower Atlian (> 0.5T). But it is almost three times smallerthan this size limit so it may be significantly younger than0.5T. Thus the age estimates for this crater and the rift-associated faults, which predated its formation, are ratheruncertain. Future studies involving age dating of craterssmaller than 30 km in diameter may put more constraints onthese estimates.[69] The 7 km crater Urazbike is superposed on the area

of rifted terrain (RT) between Parga Chasma and DaliChasma (Figure 37). The dominant trend of the RT faultshere is in agreement with that of Parga Chasma and indisagreement with the dominant trend of faults of DaliChasma. The RT unit is embayed by the Pl lavas. Therelation of the crater Urazbike with the Pl unit is not veryobvious. Small flows to the west of the crater, which may beinterpreted as ejecta outflows of Urazbike, seem to be

Figure 27. Crater Alcott (59.52�S, 354.39�E, D = 66 km) heavily flooded by Pl lava flows andembayed from the east by the Ps unit. Darker areas to the northeast of the crater are interpreted asremnants of the clear dark halo associated with this crater. Portion of FMAP 54S351 (left) andphotogeologic map of the area (right).


superposed on the Pl unit. But this relation, and theinterpretation of the flows as crater outflows, are notobvious. Crater Urazbike has a faint radar halo which seemsto affect not only the RT unit but the Pl unit as well. We thusprefer the interpretation that the crater Urazbike postdatesboth units RT and Pl. If Urazbike crater was 30 km indiameter or larger, the presence of a faint halo would

suggest a date of Lower Atlian (> 0.5T). But it is almostfour times smaller than this size limit so it may besignificantly younger than 0.5T. Thus the age estimatesfor this crater, as well as for the RT and Pl units, whichpredate its formation, are rather uncertain. This again showsthe necessity of future studies involving age estimates ofcraters smaller than 30 km in diameter.

Figure 28. Image showing general morphology of Atla Regio structure and areas where impact cratersshow age relations with post-Pwr volcanism and rifting (white boxes). Portion of C2 MDRP 00N183;1.


[70] In summary, in reference to the analysis of ageestimates for the Atla Regio structure, we may concludethat the observations put major constraints in the vicinity ofthe craters Uvaisi (Maat lava flows are younger than 0.1–

0.15T, no strong constraints for Ozza Lavas and Dalirifting), Fossey (Maat lava flow is younger than 0.5T),Melba (Maat lava flow is younger than 0.5T), Sitwell(Ganis Chasma rifting predated this crater of Aurelian

Figure 29. Portions of C2 MDRP 00N183;1 images of nine craters showing age relations with post-Pwrvolcanism and rifting of Beta-Devana structure. Each fragment covers an area 70 � 115 km.


age), and Von Schuurman (Dali Chasma rifting and asso-ciated volcanism predated this crater of Aurelian age).Observations in the vicinity of craters Piscopia (Maat flowpostdated this crater of Lower Atlian age) and Bashkirtseff(Pl lavas associated with Ganis rift postdated this crater ofLower Atlian age) are less conclusive. Even less conclusiveare observations in the vicinities of craters Udagan andUrazbike which are much smaller than the craters for whichthe age estimates have been worked out.

8. General Discussion

[71] This study is a continuation of the approach sug-gested earlier by a number of researchers to use the degreeof degradation of crater-associated radar-dark deposits forestimation of the crater age and the age of the neighboringgeologic units [Arvidson et al., 1992, Basilevsky, 1993;Strom, 1993; Herrick and Phillips, 1994; Izenberg et al.,1994]. Through the analysis of percentages of volcanicallyembayed and tectonized craters it was shown that themorphologic sequence: craters with parabola and dark halo! craters with dark halo only ! craters with partial or nodark deposits, is equivalent to the age sequence [Izenberg etal., 1994]. For craters >22.6 km in diameter, Herrick andPhillips [1994] determined the percentages of craters withdark parabolas and dark haloes (8 and 35% correspond-ingly) and suggested that the former can be dated as beingnot older than 0.08T and the latter, not older than 0.35T.Earlier, using the same logic, Arvidson et al. [1992],Basilevsky [1993], and Strom [1993] estimated the age ofcraters with dark parabolas as not older than 0.1T.[72] In these estimates, T is the mean age of the surface of

Venus. Based on the determination of the global mean craterdensity, it was estimated by Schaber et al. [1992] as �500m.y., by Phillips et al. [1992] as 400–800 m.y., and byMcKinnon et al. [1997] as �750 m.y., with any age between

�300 m.y. and 1 b.y. being possible. These large error barsare mostly due to factors which are beyond any issues ofVenus geology (e.g., the flux of Venus-crossing asteroids,contribution of cometary impacts, and atmospheric screen-ing, especially important for comets). This is why it is morepractical to estimate the absolute ages of Venus units andlandforms in fractions of T, rather than in terms of millionsor billion of years.[73] Although the classification of Venus craters accord-

ing to the degree of degradation of the associated darkdeposits can be found in the crater database of R. Herrick(http://www.lpi.usra.edu/research/vc/vchome.html), wehave made our independent photogeologic analysis of theMagellan images of craters �30 km in diameter superposedon Pwr + Psh1 regional plains (subpopulation 1, 138craters) and on post-Pwr units (subpopulation 2, 30 craters).We classified these craters into four classes: 1) craters withdark parabolas (DP), 2) craters with clear dark halo (CH), 3)craters with faint dark halo (FH) and 4) craters with no darkhalo (NH). Comparison with the Herrick database showedgeneral agreement between the two classifications. We alsofound, however, a difference in some cases caused mostlyby minor differences in classificational criteria used in thetwo efforts: the degree of circular completeness of the halo(mostly), and the influence of later volcanism.[74] Our photogeologic analysis has resulted in the esti-

mation of the percentages of craters with different degreesof degradation of the associated dark deposits: PDP = 15,PCH = 30, PFH = 30, and PNH = 25% for subpopulation 1and PDP = 17, PCH = 57, PFH = 23, and PNH = 3% forsubpopulation 2. In addition to this we have analyzed ourresults in terms of the possible influence of dark depositdegradation degree on crater latitude and size. A latitudinaldependence has been found; there is some deficit of DP andDH craters at higher latitudes. Size dependence variations,although not reliably detected, can not be excluded even

Figure 30. Crater Uvaisi (2.34�N, 198.25�E, D = 38.9 km). This dark-parabola crater is superposed onfaulted lavas of Ozza Mons (Pl_1) and embayed by lavas of Maat Mons (Pl_2). Mosaic of parts of FMIDRP 00N194;12 and FMIDRP 05N194;1 (left) and photogeologic map of the area (right). This figureis Figure 9 of Basilevsky and Head [2002].


Figure 31. Craters Fossey (2.02�N, 188.72�E, D = 30.4 km) and Piscopia (1.50�N, 190.91�E, D =26.2 km), both superposed on Pwr plains. Radar-bright lava flow emanating from Maat Mons (Pl) enteredinto the area of the clear halo of Fossey and shows no darkening. A protuberance of this flow covers partof the ejecta blanket to the south of the faint-halo crater Piscopia. Portion of C1 MIDRP 00N197;1 (top)and photogeologic map of the area (bottom).


inside the size range studied (30–270 km), and such a factoris very possible for craters whose diameters are significantlysmaller than 30 km.[75] It is obvious that the percentages of craters having a

different degree of degradation of their associated darkdeposits are generally indicative of the lifetimes of thesedeposit types. The difference in the percentages for sub-population 1 and 2 clearly demonstrates the age sense of thepercentages. However, transformation of the percentagesdetermined into estimates of the appropriate lifetimesrequires understanding of the nature of the upper boundaryof the regional plains: Does this boundary (marked byemplacement of the wrinkle ridge network) have approxi-mately the same absolute age all around the planet (synchro-nous interpretation) or not (diachronous interpretation)? Inour previous publications [Basilevsky and Head, 1998, 2000;Basilevsky et al., 1997] we have provided evidence in favor ofthe synchronous interpretation, but we believe that it is notyet possible to rule out completely the diachronous interpre-tation suggested, for example, by Guest and Stofan [1999].[76] To further test these ideas, we have constructed a

theoretical model of the evolution of the percentages of DP,CH, FH and NH craters for the cases of synchronous anddiachronous interpretations. This modeling showed that inthe synchronous case, the lifetimes of different types ofcrater-associated dark deposits are proportional to the cor-responding percentages of subpopulation 1: TDP = 0.15T,TCH = 0.3T, TFH = 0.3T. In the diachronous cases, thepercentages of craters with different degradation degreemay or may not be proportional to the appropriate lifetimes,depending on the range of absolute ages of the emplacementof the regional plains in different geological provinces ofthe planet. It was found that if this range is not larger than±0.5T, the proportionality of the percentages and lifetimesfor DP and CP craters is valid even in diachronous cases.[77] Keeping in mind possible differences in the classi-

ficational interpretations and the influence of the latitudinaldependence (observed) and the size dependence (not

excluded), time � 0.5T ago was interpreted as the lowerboundary for the age of clear-halo craters, and 0.1–0.15Tago as the lower time boundary for the age of dark parabolacraters. In our earlier work [Basilevsky and Head, 1995,1998, 2000; Basilevsky et al., 1997] the lifetime of dark-parabola craters (at that time 0.1T) was considered as thelower time boundary for our uppermost stratigraphic unit,the Aurelian. Based on the results of this work we interpretthis boundary to be 0.1–0.15T ago. We also suggest that theAtlian Period be divided into two parts: the Upper Atlianand Lower Atlian Epochs with a boundary between them at�0.5T (combined lifetimes of the DP and CH craters largerthan 30 km in diameter).[78] Comparing the percentages PDP, PCH, PFH and PNH

for subpopulation 2 (craters superposed on post-Pwr units)found in photogeological analysis, and results from thetheoretical modeling, suggests that rates of volcanic andtectonic activity were either constant through the post-Pwrtime (that is Atlian + Aurelian Periods) or that the activity inthe beginning of this time was somewhat higher than in latertime. If we consider the combined percentages (DP + CH)and (FH + NH), thus decreasing the stochastic variation, thesimilarity in the observed and modeled percentages is onlyin the case of constant rates. These conclusions are validboth for synchronous and diachronous interpretations. Theyare also in agreement with our earlier analysis of percen-tages of post-Pwr craters predating and postdating theneighboring post-Pwr units [Basilevsky and Head, 2002].[79] As it follows from general consideration and as

supported by the theoretical models presented above (see,for example, Figure 16), units which are older than T shouldhave larger percentages of nonhalo craters. At first glance,this looks promising for testing the absolute ‘‘antiquity’’ ofrelatively old (pre-Atlian) units of the stratigraphic column[e.g., Basilevsky and Head, 1995, 1998, 2000]. But obser-vations of craters superposed on and embayed by the Pwr/Psh1 regional plains (Rusalkan Period) [Collins et al.,1999], as well as observations of craters superposed on

Figure 32. Crater Piscopia. Protuberance of Pl flow enters into the hummocky ejecta blanket of thecrater. Portion of FMAP 06N186 (left) and photogeologic map of the area (right).


pre-Rusalkian units and showing evidence of postdating theregional plains, demonstrated that the timescale of pre-Atlian (down to the Fortunian) units is significantly shorter(maybe by an order of magnitude) compared to that of theAtlian + Aurelian [Basilevsky et al., 1999]. This is why

usage of the degradation degree of crater-associated darkdeposits seems not to be applicable to testing the absolute‘‘antiquity’’ of pre-Atlian units.[80] The tool of analysis of degradation degree of crater-

associated dark deposits was applied by us to the dating of

Figure 33. Crater Melba (4.71�N, 193.46�E, D = 21.8 km). Flow emanating from Maat Mons embaysthe crater ejecta blanket and covers the eastern and northeastern parts of the clear halo of this crater.Portion of C1 MIDRP 00N197;1 (top) and photogeologic map of the area (bottom).


relatively young volcanic and tectonic activity in threeregions of Venus. This analysis for the Beta-Devana struc-ture showed that at least part of the tectonic and volcanicactivity within this region occurred in the Upper AtlianEpoch or even later (after �0.5T). This is in agreement withthe presence of a substantial gravity anomaly with a largeapparent depth of compensation found for this region[Phillips et al., 1997; Sjogren et al., 1997; Smrekar et al.,1997]. The relatively young age of the Beta-Devana activityfavors models which explain the observed anomaly bycurrent thermal and/or dynamical support [see, e.g., Smre-kar et al., 1997]. The same time limit (after �0.5T) has beendetermined for the volcanic activity in the Mylitta Fluctuslava field.[81] The analysis within the Atla Regio structure was

most conclusive for Maat Mons volcano. Its distal lavaflows are estimated to be Aurelian (after 0.1–0.15T) in theeast of the volcano and Upper Atlian or Aurelian in the westand north. Estimates of a very young age of Maat Monsactivity are in agreement with the results of Klose et al.[1992], who found that it is the only high mountaintop onVenus that has not weathered to a radar-reflective mineralassemblage. On the basis of this, these authors concludedthat there has been very recent volcanic activity on MaatMons. Keeping in mind that theoretical calculations andanalogies with terrestrial hot spots predict 100–200 m.y.lifetimes for hot spots of Venus [Smrekar and Parmentier,1996; Smrekar et al., 1997], the hot spot supplying MaatMons volcano may be currently active. For other parts ofthe Atla structure (Ganis Chasma rift, Ozza Mons volcano,and the northeastern segment of Dali Chasma rift), our

analysis did not show evidence of such young activity,although the abundance of Pl and RT units in these parts ofthe region indicates that they are not older than the LowerAtlian, and does not exclude them from being Upper Atlianor even Aurelian.

9. Conclusions

[82] The above considerations lead us to the followingconclusions:- The degree of degradation of the crater-associated

radar-dark deposits can be used as a practical tool forestimation of the absolute age of impact craters and theneighboring geologic units and landforms.- For craters �30 km in diameter, the lifetime of crater-

associated dark parabolas is about 0.1–0.15T, and thelifetime of a clear dark halo, which is the next stage ofdegradation, is about 0.3T, where T is the global mean ageof the surface of Venus.- These estimates are valid both for the case of a globally

synchronous upper boundary of the regional plains(emplacement of the wrinkle ridge network), and for thecase of nonsynchronous boundary, if times of emplacementof the regional plains in different geologic provinces ofVenus differ from each other by not more than ±0.5T.- This information provides us with the possibility of

subdividing the Atlian Period of the geologic history ofVenus into the Lower Atlian and Upper Atlian Epochs, withthe age of the boundary between them at �0.5T ago.- Comparing observed percentages of craters with

different types of associated dark deposits, and modeled

Figure 34. Crater Von Schuurman (5.02�S, 191.00�E, D = 29.1 km). Its dark parabola covers riftedterrain of the NE segment of Dali Chasma and the associated Pl flows. Portion of C1 MIDRP 00N197;1(left) and photogeologic map of the area (right).


Figure 35. Craters Sitwell (16.64�N, 190.40�E, D = 32.8 km) and Bashkirtseff (14.68N, 194.04E, 36.3km). Dark-parabola crater Sitwell covers the rifted terrain of Ganis Chasma while faint-halo (see lower-right) crater Bashkirtzeff is embayed by Pl lavas emanating from rift-associated volcano. Portion of C1MIDRP 15N197;1 (top) and photogeologic map of the area (bottom).


percentages, suggests rather constant rates of volcanism andrifting during post-Pwr time.- Application of this dating technique provides the

possibility of estimating the age of at least part of thevolcanic and tectonic activity of the Beta-Devana structureas less than �0.5T, which is Upper Atlian or even younger.- Using this same technique, the same age was estimated

for the volcanism forming the Mylitta Fluctus flow field.- In the Atla Regio structure, the most recent event is the

volcanic activity of Maat Mons (less than 0.1–0.15T ago,that is Aurelian, for at least part of it) while the activity ofother parts of this structure may be both Atlian andAurelian.- Using this technique to date other structures and units

on Venus appears promising.[83] Among the questions and problems that need to be

addressed in future studies in order to refine and test thisapproach are the following:- Are the styles and rates of degradation of dark deposits

associated with craters smaller than 30 km in diameter thesame as those for craters >30 km?- What specific processes are involved in the observed

degradation of crater-associated dark deposits?- Can differences in the crater environment, for example,

in roughness of the underlying surface, be a cause ofanomalous degradation styles and rates?- How can one estimate more quantitatively the influence

of signal/noise ratio on halo detection and visibility?- How can one better understand the effects of impact

velocity and stone/iron vs. cometary impacts on thephenomenon studied?- Can radar-dark splotches be used for dating purposes in

a similar manner?

[84] Acknowledgments. We thank Peter Neivert for graphic support,Anne Cote for help in manuscript preparation, Steve Pratt for computersupport, Emily Stewart for preparation of the global SAR image of Venus,and Mikhail Kreslavsky, Boris Ivanov and Mikhail Ivanov for helpful

discussions. Thanks are extended to two reviewers whose comments helpedimprove the manuscript. This research was supported by NASA GrantsNAG5-4723 and NAG5-4585 to JWH from the NASA Planetary Geologyand Geophysics Program and by grant ‘‘Artemida’’ to ATB from theRussian Ministry of Science, which are gratefully acknowledged.

ReferencesArvidson, R. E., R. Greeley, M. C. Malin, R. S. Saunders, N. Izenberg, J. F.Plaut, E. R. Stofan, and M. K. Shepard, Surface modification of Venus asinferred from Magellan observation of plains, J. Geophys. Res., 97,13,303–13,317, 1992.

Basilevsky, A. T., Age of rifting and associated volcanism in Atla Regio,Venus, Geophys. Res. Lett., 20, 883–886, 1993.

Basilevsky, A. T., and J. W. Head, Regional and global stratigraphy ofVenus: A preliminary assessment and implications for the geologic his-tory of Venus, Planet. Space Sci., 43, 1523–1553, 1995.

Basilevsky, A. T., and J. W. Head, The geologic history of Venus: Astratigraphic view, J. Geophys. Res., 103, 8531–8544, 1998.

Basilevsky, A. T., and J. W. Head, Geologic units on Venus: Evidence fortheir global correlation, Planet. Space Sci., 48, 75–111, 2000.

Basilevsky, A. T., and J. W. Head, On rates and styles of late volcanism andrifting on Venus, J. Geophys. Res., 107, 10.1029/2000JE001471, in press,2002.

Basilevsky, A. T., J. W. Head, G. G. Schaber, and R. G. Strom, Theresurfacing history of Venusin Venus II—Geology, Geophysics, Atmo-sphere, and Solar Wind Environment, edited by S. W. Bougher, D. M.Hunten, and R. J. Phillips, pp. 1047–1086, Univ. of Ariz. Press, Tucson,1997.

Basilevsky, A. T., J. W. Head, M. A. Ivanov, and V. P. Kryuchkov, Impactcraters on geologic units of northern Venus: Implications for duration ofthe transition from tessera to regional plains, Geophys. Res. Lett., 26,2593–2596, 1999.

Campbell, B. A., Surface formation rates and impact crater densities onVenus, J. Geophys. Res., 104, 21,951–21,956, 1999.

Campbell, D. B., J. W. Head, J. K. Harmon, and A. A. Hine, Venus volcan-ism and rift formation in Beta Regio, Science, 226, 167–170, 1984.

Campbell, D. B., N. J. S. Stacy, W. I. Newman, R. E. Arvidson, E. M.Jones, G. S. Musser, A. Y Roper, and C. Schaller, Magellan observationsof extended impact crater related features on the surface of Venus,J. Geophys. Res., 97, 16,249–16,277, 1992.

Collins, G., J. W. Head, A. T. Basilevsky, and M. Ivanov, Evidence forrapid regional plains emplacement on Venus from the population ofvolcanically embayed impact craters, J. Geophys. Res., 104, 24,121–24,140, 1999.

Farr, T. G., Radar interactions with geologic surfaces, in Guide to MagellanImage Interpretation, by edited by J. P. Ford et al., JPL Publ. 93-24, pp.45–73, Jet Propul. Lab., Pasadena, Calif., 1993.

Ford, J. P., and J. J. Plaut, Magellan image data, in Guide to Magellan

Figure 36. Crater Udagan (10.7�N, 206.9�E, D = 11.5 km) superposed on Pwr plains and on faultsdeforming them and branching from the Tkashi-mapa Chasma rift. Portion of FMAP 06N210 (left) andphotogeologic map of the area (right).


Image Interpretation, by edited by J. P. Ford et al., JPL Publ. 93-24, pp.7–18, Jet Propul. Lab., Pasadena, Calif., 1993.

Gierasch, P. J., et al., The general circulation of the Venus atmosphere: Anassessment, in Venus II—Geology, Geophysics, Atmosphere, and SolarWind Environment, edited by S. W. Bougher, D. M. Hunten, and R. J.Phillips, pp. 969–1014, Univ. of Ariz. Press, Tucson, 1997.

Greeley, R., et al., Aeolian features on Venus: Preliminary Magellan results,J. Geophys. Res., 97, 13,319–13,345, 1992.

Guest, J. E., and E. R. Stofan, A new view of the stratigraphic history ofVenus, Icarus, 139, 55–66, 1999.

Hansen, V. L., Geologic mapping of tectonic planets, Earth Planet. Sci.Lett., 176, 527–542, 2000.

Figure 37. Crater Urazbike (9.00�S, 202.48�E, D = 7 km) superposed on rifted terrain (RT) andprobably on Pl lavas embaying RT. Portion of FMAP 06S198 (top) and photogeologic map of the area(bottom).


Herrick, R. R., and R. J. Phillips, Implications of a global survey of Venu-sian impact craters, Icarus, 111, 387–416, 1994.

Herrick, R. R., V. L. Sharpton, M. C. Malin, S. C. Lyons, and K. Feely,Morphology and morphometry of impact craters, in Venus II—Geology,Geophysics, Atmosphere, and Solar Wind Environment, edited by S. W.Bougher, D. M. Hunten, and R. J. Phillips, pp. 1015–1046, Univ. ofAriz. Press, Tucson, 1997.

Ivanov, M. A., and A. T. Basilevsky, Density and morphology of impactcraters on tessera terrain, Venus, Geophys. Res. Lett., 20, 2579–2582,1993.

Izenberg, N. R., R. E. Arvidson, and R. J. Phillips, Impact crater degrada-tion on Venusian plains, Geophys. Res. Lett., 21, 289–292, 1994.

Klose, K. B., J. A. Wood, and A. Hashimoto, Mineral equilibria and thehigh radar reflectivity of Venus mountaintops, J. Geophys. Res., 97,16,353–16,369, 1992.

Masursky, H., E. Eliason, P. G. Ford, G. E. McGill, G. H. Pettengill, G. G.Schaber, and G. Schubert, Pioneer Venus radar results: Geology fromimages and altimetry, J. Geophys. Res., 85, 8232–8270, 1980.

McKinnon, W. B., K. J. Zahnle, B. A. Ivanov, and H. J. Melosh, Crateringon Venus: Models and observations, in Venus II—Geology, Geophysics,Atmosphere, and Solar Wind Environment, edited by S. W. Bougher,D. M. Hunten, and R. J. Phillips, pp. 969–1014, Univ. of Ariz. Press,Tucson, 1997.

Namiki, N., and S. C. Solomon, Impact crater densities on volcanoes andcoronae on Venus: Implications for volcanic resurfacing, Science, 265,929–933, 1994.

Phillips, R. J., R. E. Arvidson, J. M. Boyce, D. B. Campbell, J. E. Guest,G. G. Schaber, and L. A. Soderblom, Impact craters on Venus: Initialanalysis from Magellan, Science, 252, 288–297, 1991.

Phillips, R. J., R. F. Raubertas, R. E. Arvidson, I. C. Sarkar, R. R. Herrick,N. Izenberg, and R. E. Grimm, Impact craters and Venus resurfacinghistory, J. Geophys. Res., 97, 15,923–15,948, 1992.

Phillips, R. J., C. L. Johnson, S. J. Mackwell, P. Morgan, D. T. Sandwell,and M. T. Zuber, Lithospheric mechanics and dynamics of Venus, inVenus II—Geology, Geophysics, Atmosphere, and Solar Wind Environ-ment, edited by S. W. Bougher, D. M. Hunten, and R. J. Phillips, pp.1163–1204, Univ. of Ariz. Press, Tucson, 1997.

Plaut, J. J., Stereo imaging, in Guide to Magellan Image Interpretation,edited by J. P. Ford et al., JPL Publ. 93-24, pp. 33–43, Jet Propul. Lab.,Pasadena, Calif., 1993.

Price, M., Tectonic and Volcanic Map of Venus, Dept. of Geol. Sci., Prin-ceton Univ., Princeton, N. J., 1995.

Price, M., and J. Suppe, Young volcanism and rifting on Venus, Nature, 72,756–759, 1994.

Price, M., and J. Suppe, Constraints on the resurfacing history of Venusfrom the hypsometry and distribution of volcanism, tectonism, and im-pact craters, Earth Moon Planets, 71, 99–145, 1995.

Roberts, K. M., J. E. Guest, J. W. Head, and M. G. Lancaster, MylittaFluctus, Venus: Rift-related, centralized volcanism and emplacement oflarge-volume flow units, J. Geophys. Res., 97, 15,991–16,015, 1992.

Schaber, G. G., R. G. Strom, H. J. Moore, L. A. Soderblom, R. L. Kirk,D. J. Chadwick, D. D. Dawson, L. R. Gaddis, J. M. Boyce, and J. Russell,

Geology and of impact craters on Venus: What are they telling us?, J.Geophys. Res., 97, 13,257–13,301, 1992.

Schaber, G. G., R. L. Kirk, and R. G. Strom, Data base of impact craters onVenus based on analysis of Magellan radar images and altimetry data,U.S. Geol. Surv. Open File Rep., 98-104, 1998. (http://wwwflag.wr.usgs.gov)

Schultz, P. H., Atmospheric effects on ejecta emplacement and crater for-mation on Venus from Magellan, J. Geophys. Res., 97, 16,183–16,248,1992.

Senske, D. A., G. G. Schaber, and E. R. Stofan, Regional topographic riseson Venus: Geology of western Eistla Regio and comparison to BetaRegio and Atla Regio, J. Geophys. Res., 97, 13,395–13,420, 1992.

Sjogren, W. L., W. B. Banerdt, P. W. Chodas, A. S. Konopliv, G. Balmino,J. P. Barriot, J. Arkani-Hamed, T. R. Colvin, and M. E. Davies, TheVenus gravity field and other geodetic parameters, in Venus II—Geology,Geophysics, Atmosphere, and Solar Wind Environment, edited by S. W.Bougher, D. M. Hunten, and R. J. Phillips, pp. 1125–1161, Univ. of Ariz.Press, Tucson, 1997.

Smrekar, S. E., and E. M. Parmentier, The interaction of mantle plumeswith surface thermal and chemical boundary layers: Application to hot-spots on Venus, J. Geophys. Res., 101, 5397–5410, 1996.

Smrekar, S. E., W. S. Kiefer, and E. R. Stofan, Large volcanic rises onVenus, in Venus II—Geology, Geophysics, Atmosphere, and Solar WindEnvironment, edited by S. W. Bougher, D. M. Hunten, and R. J. Phillips,pp. 845–878, Univ. of Ariz. Press, Tucson, 1997.

Solomon, S. C., S. E. Smrekar, D. L. Bindschadler, R. E. Grimm, W. M.Kaula, G. E. McGill, R. E. Phillips, R. S. Saunders, G. Schubert, S. W.Squyres, and E. R. Stofan, Venus tectonics: An overview of Magellanobservations, J. Geophys. Res., 97, 13,199–13,255, 1992.

Stofan, E. R., J. W. Head, D. B. Campbell, S. H. Zisk, A. F. Bogomolov,O. N. Rzhiga, A. T. Basilevsky, and N. A. Armand, Venus rift zones:Geology of Beta Regio and Devana Chasma, Geol. Soc. Am. Bull., 101,143–156, 1989.

Strom, R. G., Parabolic features and the erosion rate on Venus, Proc. Lunar.Planet. Sci. Conf., 24th, 1371–1372, 1993.

Tanaka, K. L., D. A. Senske, M. Price, and R. L. Kirk, Physiography,geomorphic/geologic mapping and stratigraphy of Venus, in Venus II—Geology, Geophysics, Atmosphere, and Solar Wind Environment, editedby S. W. Bougher, D. M. Hunten, and R. J. Phillips, pp. 667–694, Univ.of Ariz. Press, Tucson, 1997.

Vervack, R. J., and H. J. Melosh, Wind interaction with falling ejecta:Origin of the parabolic features on Venus, Geophys. Res. Lett., 90,525–528, 1992.

Wilhelms, D., Geologic mapping, in Planetary Mapping, edited byR. Greeley and R. Batson, pp. 208–260, Cambridge Univ. Press, NewYork, 1990.

��A. T. Basilevsky and J. W. Head, Department of Geological Sciences,

Brown University, Box 1846, Providence, RI 02912, USA. ( [email protected])


Venus: Analysis of the degree of impact crater deposit ... · DP = 0.15T, T CH = 0.3T, and T FH =...

Documents

Transcript of Venus: Analysis of the degree of impact crater deposit ... · DP = 0.15T, T CH = 0.3T, and T FH =...