Geo log i cal Quar terly, 2020, 64 (3): 611–625 DOI: http://dx.doi.org/10.7306/gq.1548
Intrastratal flow in the Cre ta ceous Gyeokpori For ma tion (SW South Ko rea)
Uk Hwan BYUN 1, A.J. (Tom) VAN LOON2, Yi Kyun KWON3 and Kyoungtae KO1, *
1 Ko rea In sti tute of Geoscience and Min eral Re sources (KIGAM), Ge ol ogy Di vi sion, Daejeon 34132, South Ko rea
2 Shandong Uni ver sity of Sci ence and Tech nol ogy, Col lege of Earth Sci ence and En gi neer ing, Qingdao 266590, Shandong, China
3 Kongju Na tional Uni ver sity, De part ment of Geoenvironmental Sci ences, Kongju 314-701, South Ko rea
Byun, U.H., Van Loon, A.J. (Tom), Kwon, Y.K., Ko, K., 2020. Intrastratal flow in the Cre ta ceous Gyeokpori For ma tion (SW South Ko rea). Geo log i cal Quar terly, 64 (3): 611–625, doi: 10.7306/gq.1548.
Intrastratal flow is a pro cess that is still poorly un der stood, rarely de scribed and dif fi cult to in ter pret in an cient rocks. Sed i - ments in the Cre ta ceous lac us trine Gyeokpori For ma tion of south west ern South Ko rea con tain some cha ot i cally de formed sand stone lay ers with de formed mudstone clasts that are as cribed to this pro cess. The in ter pre ta tion is based on the fact that these lay ers can not be ex plained as a re sult of sub aque ous de bris flows or mass trans port, whereas the sed i men tary con - text, in clud ing the pres ence of other soft-sed i ment de for ma tion struc tures, in di cates that intrastratal flow must have been phys i cally pos si ble. The sed i men tary set ting was a lake in which mainly siliciclastic rocks were de pos ited, with some interbedded volcaniclastics. The nearby vol ca nic ac tiv ity caused seis mic shocks that af fected the un sta ble lake mar gins re - sult ing in the dom i nance of grav ity-flow de pos its, but also in a high sed i men ta tion rate that fa cil i tated soft-sed i ment de for ma - tion partly caused by intrastratal flow. This must have hap pened fairly fre quently dur ing a prob a bly lim ited time-span, as sev eral lay ers show ing traces of intrastratal flow are pres ent within a suc ces sion of only <1 m thick. The com bined data on the geo log i cal set ting and our find ings re gard ing the or i gin of the var i ous soft-sed i ment de for ma tion struc tures may help to rec og nize the traces left by intrastratal flow else where in the geo log i cal re cord.
Key words: intrastratal flow, Gyeokpori For ma tion, soft-sed i ment de for ma tion struc tures, lac us trine en vi ron ment, Cre ta - ceous.
INTRODUCTION
Intrastratal flow in an cient rocks can be in ter preted only on the ba sis of traces left by the pro cess. Such traces con sist al - most ex clu sively of soft-sed i ment de for ma tion struc tures (SSDS; this ac ro nym is used here for both sin gu lar and plu ral).
This poses a prob lem as SSDS of nu mer ous types are known to have sev eral pos si ble or i gins. More over, even for sed i ments ac cu mu lated in spe cific sed i men tary en vi ron ments such as the lake de pos its un der study here, SSDS with largely dif fer ent or i - gins have been re ported fre quently in the past few de cades (e.g., Sims, 1973, 1975; Hempton and Dewey, 1983; Alfaro et al., 1997; Rodríguez-Pascua et al., 2000; Moretti and Sabato, 2007; TaêgÍn and Türkmen, 2009; Gibert et al., 2011; TaêgÍn et al., 2011). All these stud ies em pha size that the deformational pro cesses can not al ways be re con structed with cer tainty, be - cause small dif fer ences in sed i ment char ac ter is tics or depo - sitional pro cesses can lead to en tirely dif fer ent SSDS. To make in ter pre ta tions of SSDS even more com plex, dif fer ent causes might lead to the de vel op ment of very sim i lar SSDS.
Among the nu mer ous anal y ses of SSDS, only a few as cribe them to intrastratal flow. This is un der stand able, be cause such a pro cess re quires spe cific con di tions; these have been de - tailed first by Wil liams (1960), who men tioned that rapid de po si - tion of silt and fine sand from sed i ment grav ity flows can re sult in loose pack ing, a rel a tively high pore-fluid vol ume, and rel a tively low shear strength. Rapid load ing of stratigraphically con fined lay ers of loosely packed silt can in duce mass flows and can re - duce the orig i nal vol ume of the layer, re sult ing in in creased neu - tral stress (pore-fluid) and de creased ef fec tive stress (grain- to-grain con tact). Liq ue fac tion can oc cur if these be come equal, at which point the bur ied intrastratal layer be haves as a con cen - trated sus pen sion and flows downslope. Upon ces sa tion of flow, the layer re turns to solid state. This ex pla na tion by Wil - liams (1960) is still ac cepted by re cent re search ers (e.g., Auchter et al., 2016), but they also make clear that only a com - bi na tion of fea tures can make it pos si ble to as cribe de for ma - tions in a de formed layer to intrastratal flow.
We do so in the pres ent con tri bu tion for some lay ers in the Cre ta ceous Gyeokpori For ma tion, which is ex posed along the west ern coast of South Ko rea. The for ma tion con sists of lac us - trine sed i ments that ac cu mu lated on the slope of a sub aque ous fan-delta sys tem, a po si tion that is com monly con sid ered most suit able to in duc ing and pre serv ing SSDS (Gibert et al., 2005;
Moretti and Sabato, 2007; Tan ner and Lucas, 2007; Ko et al., 2015, 2017; Gladkov et al., 2016; Jiang et al., 2016). In deed, the for ma tion con tains nu mer ous SSDS in ad di tion to those that
* Corresponding author, e-mail: kkt@kigam.re.kr
Received: October 3, 2019; accepted: November 17, 2019; first published online: August 17, 2020
we as cribe to intrastratal flow. The most fre quent de for ma tion struc tures are slump-re lated folds, some times form ing en ve - lopes of duc tile fine-grained ma te rial de pos ited around more rig idly be hav ing grav elly parts (Byun et al., 2019).
GEOLOGICAL SETTING
Dur ing the Cre ta ceous, the Izanagi Plate was subducted along the east ern mar gin of east ern Asia (Chough and Sohn, 2010). At the same time, nu mer ous non-ma rine sed i men tary bas ins of dif fer ent sizes were formed (Chough, 2013), in clud ing the lac us trine strike-slip Gyeokpori Ba sin on the pres ent-day south west ern shore of the Ko rean Pen in sula. Im me di ately west of the Gyeokpori Ba sin, a large (~20 x 10 km) vol ca nic body is pres ent. The ba sin is filled with the lac us trine Gyeokpori For ma - tion, the de vel op ment of which was af fected by the vol ca nic com plex, in for mally known as the Buan Volcanics (Koh et al., 2013), in the form of erup tion-in duced earth quakes, lahar -de - rived hyperpycnites, and volcaniclastic lay ers.
The struc tures un der study here were stud ied in a sec tion lo cated near the vil lage of Jukmak along the west ern coast of Byeonsanbando Na tional Park (Kim et al., 1995, 2003; Fig. 1).
The Gyeokpori For ma tion is ex posed here and in nearby sec - tions in coastal cliffs and on wave-cut ter races. Palaeocurrent di rec tions ob tained from the ge om e try and dip di rec tion of large sed i men tary lobes, bed ge om e try, and sed i men tary struc tures such as flute casts, cur rent rip ples, and gravel clus ters, in com - bi na tion with an anal y sis of the sed i men tary fa cies, in di cate that the sed i ments were de pos ited on a fan-delta on the south ern mar gin of the ba sin; the delta prograded to ward the north-north - west, while a fan-delta on the north ern ba sin mar gin prograded to ward the south-south west (Kim et al., 2003). The depocentre of the ba sin must thus have been lo cated in the west ern part of the ba sin, and the ba sin floor must have been gently in clined in the same di rec tion.
MATERIALS AND METHODOLOGY
LITHOLOGY, FACIES AND DEPOSITIONAL ENVIRONMENTS
The Gyeokpori For ma tion is com posed of con glom er ates with clasts of var i ous sizes, grav elly sand stones, mudstones, and abun dant volcaniclastic de pos its (Fig. 2).
Most of these are grav ity-flow de pos its formed by de bris flows and tur bid ity cur rents (Kim et al., 2003). The to tal thick - ness of the for ma tion is ~290 m. Ten sed i men tary fa cies are dis tin guished. They have been de tailed be fore (Byun et al., 2019) and the reader is re ferred to that study for de tails. The ten fa cies can be grouped into five as so ci a tions (Byun et al., 2019), four of which rep re sent sub aque ous fan or fan-delta sys tems, and one the ba sin plain (Kim et al., 1995, 2003). A mi nor part of the sed i ments was de pos ited on the steep ba sin slope, but most ac cu mu lated on the ba sin plain in front of the lac us trine fan-del tas (Kim et al., 2003). Part of the lay ers de pos ited on the ba sin slope show cha otic soft-sed i ment de for ma tions. These lay ers are the main sub ject of the pres ent study.
Fig. 1. Geo graph ical and geo log i cal set ting of the study area A – lo ca tion in south west ern South Ko rea; B – sche matic geo log i cal map of the west ern mar gin of the mostly vol ca nic area of Byeon - sanbando Na tional Park (mod i fied af ter Koh et al., 2013); C – ae rial maps (from http://map.google.com) show ing the lo ca tion of the sec - tions near Jukmak. Sec tion J2 con tains the lay ers af fected by intrastratal flow
Intrastratal flow in the Cre ta ceous Gyeokpori For ma tion (SW South Ko rea)613
e h t
f
o n o i s s e c c u s y r a t n e m i d e S
.
2
.
g i
F GyeokporiFormation
A–com pos itesuc ces sionnea rJukmak(seeFig .1fo rlo ca tiono fthesec tionsJ1 ,J2andJ3) ;B–sed i men tarylogo fthein ter va lwithintrastratalflows(seeA);
t
a n o i s s e c c u s e r i t n e e h t n
i Jukmak ,18cha ot i callyde formedlay ers(CDL ;thisab bre vi a tionisalsousedinthefo llow ingfig ures )havebeenfoundandin ves ti gated;
,
d u m –
M fS–finesand ,mS –me diumsand ,cS–coarsesand ,C–con glom er ate
DESCRIPTION OF THE CHAOTICALLY DEFORMED LAYERS
The cha ot i cally de formed lay ers (CDL) that were de pos ited on the ba sin slope are each <1 m thick. They oc cur in the up per part of the J2 ex po sure and in the lower part of the J3 ex po sure of the Jukmak sec tion (Fig. 3). These lay ers (Figs. 4 and 5) oc - cur in ter ca lated be tween intraclast-rich sand stones that form part of a sand stone-dom i nated suc ces sion with rel a tively thin mudstone in ter ca la tions.
The lay ers un der study are re mark able in them selves be - cause of the – of ten cha ot i cally – de for ma tion within these lay - ers, but also be cause of the un com mon char ac ter is tics of the clasts that are pres ent in these lay ers. Both as pects will be de - tailed in the fol low ing sec tions, fol lowed by a sec tion de voted to the re la tion ship be tween these two fea tures.
CHARACTERISTICS OF THE CDL
The CDL have lower and up per bound aries par al lel to the bed ding planes of the other lay ers in the suc ces sion (Fig. 6), most of which do not show soft-sed i ment de for ma tion struc - tures or which show struc tures that are char ac ter is tic of sed i - ment grav ity flows such as slumps (with folds that in di cate at least par tially ro ta tional de for ma tion) as de scribed and de picted by Byun et al. (2019).
The ma trix of the CDL, which ex tend as far as the out crop al - lows, con sists mainly of clean me dium to coarse sand with out any grad ing or lam i na tion (Fig. 7). Most lay ers are 0.3–1 m thick, and do not show sig nif i cant lat eral changes in thick ness (Fig. 2).
The up per and lower bound aries of the de formed lay ers are gen - er ally flat and sharp with out any ero sional fea tures (Figs. 4–6).
CHARACTERISTICS OF THE CLASTS
The stud ied lay ers con tain nu mer ous lam i nated mudstone clasts (in clud ing rip-down clasts), which are lithologically and tex tur ally iden ti cal to the lam i nated mudstones that are found in the sed i ments de pos ited on the mid dle part of the fan-delta (Fig.
7). The con tacts be tween the clasts and the ma trix are sharp (Fig. 8). These mudstone clasts are com monly de formed them - selves in the form of bend ing (Fig. 9) or, more rarely, strong fold ing. The clasts do not show a grad ual change in the in ten sity of de for ma tion in ei ther a hor i zon tal or a ver ti cal di rec tion within one layer, but rather seem to show more or less ran dom de - grees of de for ma tion. How ever, the CDL tend to have some spe cific hor i zon tal lev els in which clasts oc cur in larger con cen - tra tions than in other lev els (see Figs. 7 and 8).
The shapes of the mudstone clasts vary from platy to slightly bent and they can even be strongly folded. Most of the platy clasts are elon gated, with a pre ferred ori en ta tion more or less par al lel to the bed ding (Fig. 10). They have max i mum vis i - ble sizes rang ing from roughly 2 to 50 cm, and elon gated clasts com monly have an gu lar outer ends (Fig. 11). The more folded or strongly bent clasts do not show a clearly pre ferred ori en ta - tion, but a slight ten dency ex ists (Fig. 12) that the con vex parts of the bent clasts are ori en tated in the same di rec tion. This is the downslope di rec tion, roughly the same as the di rec tion shown by the con vex parts of the slump heads pres ent in the Gyeokpori For ma tion.
RELATIONSHIP BETWEEN THE DEFORMED LAYERS AND THE CLASTS
The li thol ogy of the mudstone clasts is very sim i lar to that of the lay ered mudstones un der- and over ly ing the layer; some of
Fig. 3. The up per part of J2 in the Jukmak sec tion with sev eral lay ers de formed by intrastratal flow
the clasts are still partly con nected with the over ly ing mudstone layer (Fig. 7), but never with the un der ly ing mudstone layer. The mudstone clasts in the de formed lay ers must con se quently be de rived from the partly bro ken-up over ly ing con sol i dated mud lay ers or com pa ra ble thin and bro ken-up mudstone lay ers that oc cur in most of the top parts of the de formed lay ers (Fig. 13), whereas iso lated frag ments oc cur near both the up per and the lower bound aries (Fig. 14).
INTERPRETATION OF THE CHAOTICALLY DEFORMED LAYERS AND THEIR
CHARACTERISTICS
The most re mark able fea tures in the CDL are the de formed mudstone clasts that float in the sand stone ma trix. They must have re sulted from soft-sed i ment de for ma tion that even tu ally Intrastratal flow in the Cre ta ceous Gyeokpori For ma tion (SW South Ko rea) 615
Fig. 4. De tails of some lay ers af fected by intrastratal flow
A – ex am ple of a layer with long and slightly de formed float ing clasts of lam i nated mudstone;
B – ex am ple of a layer with nu mer ous smaller clasts of float ing lam i nated mudstone
re sulted in frag men ta tion of one or more thin mud lay ers. This can be de duced from the fact that the clasts show the same lam i na tion as the bro ken-up mudstone lay ers (Fig. 15). Their break ing-up points to the pres ence of an im per me able layer that led to a high pore pres sure in the sand un der the in flu ence of a thick en ing over bur den.
Liq ue fac tion oc curred when the thresh old value of the shear stress was passed, re sult ing in intrastratal flow. Once the sand layer liq ue fied, it flowed along the slope and broke up the mud layer above and/or be low it (or both); this ex plains the rel a tively high con cen tra tion of clasts in the up per and lower parts of the de formed lay ers. The frag men ta tion pro cess caused by the liq - Fig. 5. De tail of cha otic soft-sed i ment de for ma tion struc tures in one
of the lay ers af fected by intrastratal flow
Fig. 6. De formed layer af fected by intrastratal flow in ter ca lated be tween two undeformed lay ers, show ing par al lel, rel a tively undeformed
lower and up per bound aries
ue fied intrastratal flow was pos si bly due to the co he sive char ac - ter of the al ready con sol i dated muddy ma te rial. The same ar gu - men ta tion was used by Chough and Chun (1988) and Kawa - kami and Kawamura (2002) for the ex pla na tion of the mud - stone clasts men tioned in the above de scrip tion sec tion; they also ar rived at the con clu sion of intrastratal flow.
Fa cies anal y ses by pre vi ous re search ers (Kim et al., 2003;
Koh et al., 2013), as well as our anal y sis of other SSDS in the Gyeokpori For ma tion (see Byun et al., 2019), con vinc ingly dem on strate that these suc ces sions were al most all de pos ited by sed i ment grav ity flows on the in clined slope of a lac us trine deltaic fan. This strongly sug gests that the orig i nally ac cu mu - lated sandy lay ers were de pos ited at a high sed i men ta tion rate, Intrastratal flow in the Cre ta ceous Gyeokpori For ma tion (SW South Ko rea) 617
Fig. 7. Large mudstone clasts in a de formed sand stone layer still partly ad hered to their bro ken-up par ent mudstone layer
Fig. 8. De tail of clasts of lam i nated mudstone within the sand of a layer af fected by intrastratal flow
Note the sharp con tacts be tween the clasts and the ma trix
Fig. 9. De tail of de formed (slightly bent to strongly folded) intraclasts of lam i nated mudstone in a sand layer af fected by intrastratal flow
Fig. 10. Platy clasts of lam i nated mudstone pref er en tially ori en tated (sub)par al lel to the bed ding
A – de tail of the (sub)par al lel ori en ta tion of the clasts within a layer; B – over view of (sub)par al lel clasts near the con tact with the un der- and over ly ing lay ers
un der which con di tions the sed i ments, af ter hav ing be come bur ied un der new sed i ment, were likely to con tain much pore wa ter. The silty/clayey lay ers ac cu mu lat ing above the sandy de pos its must have been rel a tively im per me able, thus form ing a seal ing layer, pre vent ing the pore wa ter in the sands to seep out grad u ally.
Con sid er ing the largely grav ity-driven depositional pro cess on the slope, the seal ing sed i ments be came cov ered by new grav ity-flow de pos its in ter mit tently but fre quently (Fig. 16A), so that the al ready high pore pres sure in the sandy lay ers be low the seal ing mud be came ever higher (Fig. 16B). These al ter nat - ing beds made the sed i ments eas ily sus cep ti ble to liq ue fac tion Intrastratal flow in the Cre ta ceous Gyeokpori For ma tion (SW South Ko rea) 619
Fig. 11. Clasts of dif fer ent sizes
Note the an gu lar outer ends in di cat ing that no trac tion trans port has taken place
Fig. 12. De formed mudstone clasts show ing a slight ten dency to have their con vex side ori en tated in the same di rec tion, re sem bling the bend ing of a slump head with the con vex side downslope, thus in di cat ing the di rec tion of the flow
The yel low dashed lines rep re sent the bound aries of the de formed layer, whereas the white dashed line in di cates a fault
and fluidization, which pro cesses may well have been trig gered by endogenic pro cesses (e.g., seis mic waves re sult ing from nearby vol ca nic erup tions; Fig. 16C, D). The liq ue fied wa - ter-sat u rated sed i ments flowed downslope within a con fined path way be tween the more co he sive lay ers that did not liq uefy (Fig. 16D). At the be gin ning of this flow, the sed i ments flowed down intrastratally, while im per me able, more con sol i dated mud lay ers be low and above acted as bound ing lay ers (Fig. 16D).
Dur ing a sub se quent stage, more lay ers of liq ue fied and fluidized sand started to move downslope intrastratally, even tu - ally break ing up an im per me able mud layer be tween the two flow ing masses of liq ue fied sand with mudstone clasts that con - tin ued to move downslope (Fig. 16E); when the slope be came too gen tle, the sed i ments stopped, thus form ing lay ers with de - formed mud clasts (Fig. 16F).
Fig. 13. Com par i son of the mudstone clasts and the mudstone de pos ited on the mid dle part of the fan-delta
A – autochthonous lam i nated mudstone;
B – char ac ter is tic mudstone clast in a sand stone layer af fected by intrastratal flow
DISCUSSION
As de scrip tions of intrastratal flow are rare in the lit er a ture, and be cause the pro cess of intrastratal flow is still poorly un der - stood, it is ap pro pri ate to in ves ti gate why other pro cesses could not have re sulted in the CDL de scribed above. Post- depositio - nally de formed lay ers in ter ca lated be tween undeformed lay ers formed subaqueously have most com monly been in ter preted to have re sulted from grav ity flows (par tic u larly slumps and tur bid - ity cur rents), from drag ging by (mostly sed i ment-laden) cur - rents, and from seis mic shock waves.
COMPARISON WITH GRAVITY-FLOW DEPOSITS
Grav ity-flow de pos its that show fre quent de for ma tions mainly con sist of slumps, co he sive mudflows, and turbidites (e.g., Haughton et al., 2003; Alsop and Marco, 2013). All of these can con sist of sandy ma te rial with mud clasts, but the char ac ter is tics of such de pos its dif fer fun da men tally from those un der study here.
Slumps oc cur abun dantly in the Gyeokpori For ma tion, com - monly as sets of al ter nat ing sandy and fine-grained turbidites.
How ever, these lay ers were de formed dif fer ently be cause of the dif fer ence in co he sion that re sulted in brit tle be hav ior of the con sol i dated sands and in plas tic be hav ior of the muds. The pro cesses in volved and the re sult ing de for ma tions have been de scribed in de tail by Byun et al. (2019). Con se quently, the muddy lay ers were bent (some times lo cally cha ot i cally be cause of the space prob lem cre ated dur ing the ro ta tional move ment of the slump ing mass), whereas the more sandy lay ers were com - monly bro ken up into larger or smaller frag ments, de pend ing on their pre cise place in the slump (see Byun et al., 2019). The pri - mary lam i na tion (and some times also cur rent rip ples) in the slumped sandy turbidites re mained well-pre served as a rule.
These char ac ter is tics are fun da men tally dif fer ent from the CDL un der study here, where no pri mary struc tures are pre served in the sand stones, where the sand stone is not bro ken up into small frag ments, and where the muddy lay ers have been bro - ken up into small frag ments. Con se quently, the dis tor tions in the CDL in ques tion can not be at trib uted to slump ing.
Mudflows can, if their downslope move ment takes place rel - a tively qui etly, pre serve frag ments with pri mary struc tures if the flow is suf fi ciently co he sive. How ever, this re quires a high amount of silt and/or clay (as al ready in di cated by the term
“mud flow”), but the lay ers un der study mainly con sist of fairly clean sand stone. More over, trans port within a mud flow that was able to break up con sol i dated mud lay ers should have re sulted in some round ing of the mud clasts (see, for in stance, Pisarska - -Jamroźy et al., 2019), but these are an gu lar. Fur ther, if both con sol i dated sand and mud lay ers had been in volved in the mud flow, not only mud clasts but also sandy clasts should be pres ent. For all these rea sons, an or i gin as mud flow de pos its can also be ex cluded.
Turbidites can con tain a level (the Tc di vi sion of Bouma) that is strongly de formed. The Gyeokpori For ma tion does con tain nu mer ous turbidites, in deed, but these show the “clas si cal”
char ac ter is tics (though mostly in com plete) of Bouma di vi sions, in clud ing con vo lu tions, sole marks, and grad ing. More over, these sandy turbidites show abun dant pri mary struc tures, while the CDL have none of these char ac ter is tics and there fore can - not be as cribed to de po si tion by tur bid ity cur rents.
Con sid er ing the above fun da men tal dif fer ences be tween the CDL and slumps, mud flow de pos its and turbidites, we must con clude that the sed i ments un der study can not be con sid ered as slumps, debrites, turbidites or any other form of mass-flow de posit.
An other prob lem to be solved is the ques tion whether the de formed lay ers can have been de pos ited be fore the over ly ing sed i ments were de pos ited. Such a sit u a tion would im ply liq ue - fac tion of the up per most sed i men tary suc ces sion of thin mud, Intrastratal flow in the Cre ta ceous Gyeokpori For ma tion (SW South Ko rea) 621
Fig. 14. Con cen tra tion of mudstone clasts in a sand stone layer near the mudstone lev els be low and above
sand and mud. If such a sed i men tary unit would start mov ing downslope (which must, con sid er ing the po si tion and ori en ta - tion of the mudstone clasts, have taken place), break ing up of the mud lay ers can have oc curred only if they had al ready been con sol i dated. Such a con sol i da tion is highly un likely, how ever, be cause of the high sed i men ta tion rate: con sol i da tion would have re quired too much time. An even more con vinc ing ar gu - ment against sub aque ous rather than intrastratal flow is that sub aque ous flow must have taken place in some form of grav ity flow (slump, de bris flow, turbidite). The CDL, how ever, does not show the char ac ter is tics that be long to one of these types of grav ity flow, which im plies that grav ity flowage with out a sed i - men tary cover must be ex cluded.
COMPARISON WITH DRAG STRUCTURES
Al though rel a tively rare, nu mer ous sand stone lay ers with fairly ex ten sive de for ma tion caused by drag ging have been de - scribed in the lit er a ture (e.g., Rana et al., 2016). Such drag ging tends to be caused by shear stress ex erted on the slightly con - sol i dated sandy sur face layer due to the move ment of some
mass. The mech a nism in volved is com monly a sed i ment-laden bot tom cur rent but can also be some thing dif fer ent, such as ice (e.g., Curran, 2007; Frey and Dashtgard, 2012; Salamon, 2015;
Mazumder et al., 2016; Rana et al., 2016). How ever, such drag - ging re sults in the bend ing of foresets (if pres ent) in a reg u lar (not cha otic) way, so the CDL un der study can not be at trib uted to this pro cess.
COMPARISON WITH SEISMITES
Earth quakes with a mag ni tude of over 4.5–5 pro duce shock waves with an in ten sity that can cause de for ma tion of spe cific near-sur face lay ers (oc ca sion ally a set of lay ers); such de - formed beds are called seismites. These lay ers can be sandy or finer-grained, or con sist of al ter na tions of finer and coarser ma - te rial. Their char ac ter is tics have been de scribed in nu mer ous stud ies, in clud ing ex am ples in which frag mented ma te rial of bro ken up lay ers is pres ent (e.g., Van Loon et al., 2016).
Al though no traces of tec tonic ac tiv ity dur ing the ac cu mu la - tion of the Gyeokpori For ma tion are known, ex ten sive vol ca nic ac tiv ity did take place (Kwon et al., 2017). This must have been Fig. 15. Sim i lar ity of the bro ken-up lam i nated mudstones and the mudstone clasts
A – de tail of bro ken-up autochthonous lam i nated mudstone in a de formed layer;
B – de tail of a clast of lam i nated mudstone in the same sand stone layer
ac com pa nied by earth quakes, some of which could cer tainly have had a suf fi ciently high mag ni tude to trig ger the for ma tion of seismites. In deed, some seismites are known in the for ma - tion, but their char ac ter is tics are (like most seismites de scribed in the lit er a ture), com pletely dif fer ent from those of the lay ers un der study here. If both sandy and silty ma te rial is in volved in seis mic de for ma tion, the first pro cess tak ing place is load ing, of ten ac com pa nied by small-scale fault ing. This de for ma tion pro cess is re stricted to the layer (or the set of lay ers) through which the seis mic wave (com monly a Ra leigh wave) trav els; the un der- and over ly ing lay ers re main undeformed. In the Gyeokpori For ma tion, how ever, there is no such well-de fined level of de for ma tion (Fig. 16F). More over, seismites tend to con tain load casts that were them selves de formed again by after shocks. None of these char ac ter is tics are pres ent in the lay ers un der study, so a seis mic or i gin for the de for ma tions must be ex cluded.
ARGUMENTS
IN FAVOUR OF INTRASTRATAL FLOW
Some struc tures that have been pre served in the CDL (rip-down clasts, some folds, and par tial fluidization) have, as will be de tailed be low, been de scribed ear lier from sed i ments that un der went intrastratal flow. One more such char ac ter is tic is that the intrastratally de formed lay ers have a par tially liq ue fied part near their lower bound ary, which has re sulted in semi-de - tached clasts near the top of the un der ly ing lam i nated
mudstone layer. It is note wor thy that the char ac ter is tics of the mudstone clasts show many sim i lar i ties with the clasts (called
“rip-down clasts”) de scribed by Chough and Chun (1988): “The clasts in clude small, fi brous shale chips and large disc forms (0.5–65 cm long and 0.1 mm 8 cm thick). The ex is tence of large, semi de tached clasts in the up per part sug gests that they were de rived from the over ly ing bed rather than the un der ly ing bed.” They also show sim i lar i ties with the clasts in de formed lay ers de scribed by Kawakami and Kawamura (2002): “Closely packed sheet-form or el lip soi dal mud clasts are em bed ded in a poorly sorted fine- to coarse-grained sandy ma trix. Some mud clasts are tightly folded, or show imbricated stack ing”. Sim i lar fea tures have also been as cribed to intrastratal flow by Whitmore and Strom (2010).
Rip-down clasts with com pa ra ble char ac ter is tics have also ear lier been re ported from lac us trine sed i ments, where they have been in ter preted as the re sult of intrastratal flow (Chough and Chun, 1988; Chun and Chough, 1992). These stud ies also con sid ered this as strong ev i dence for intrastratal flow be cause platy mud clasts with sharp edges can not re sult from wave-in - duced scour ing with out re sult ing in an ero sional sur face, nor mal grad ing by set tling of fines stirred up by waves, and hummocky cross-strat i fi ca tion. No such fea tures are pres ent in the lay ers un der study. More over, the mudstone clasts must have been de rived from the over ly ing lam i nated mudstone layer be cause some large clasts are still partly at tached to this over ly ing layer (Fig. 7).
The fold ing of the un der ly ing mudstone bed (Fig. 16F) is ad - di tional ev i dence for intrastratal flow. Such folds can only be as - Intrastratal flow in the Cre ta ceous Gyeokpori For ma tion (SW South Ko rea) 623
Fig. 16. Sche matic model of the de vel op ment of intrastratal flow
A – high sed i men ta tion rate re sults in a suc ces sion of al ter nat ing sandy and muddy wa ter-sat u rated sed i ments; B – com pac tion un der the in - flu ence of a thick en ing over bur den re sults in in creased pore-wa ter pres sure; C – some pro cess (here prob a bly a seis mic shock re lated to nearby vol ca nic ac tiv ity) re sults in liq ue fac tion of two sandy lay ers be tween still-co he sive mud lay ers; D – two sandy lay ers be come fluidized and the sand/wa ter mix tures start to flow downslope; E – two intrastratally flow ing lay ers break up the thin mud layer in be tween them and sand frag ments of the mud layer be come em bed ded in the flow; F – mud frag ments be come de formed dur ing flow, and the de formed mud clasts re main em bed ded in the ho mog e nized sand when the intrastratal flow co mes to rest
cribed to plas tic de for ma tion re sult ing from shear at the bot tom bound ary. The ex ten sive shear ing and frag men ta tion of the lam i nated mudstone sug gest that the mud was sub jected to penecontemporaneous de for ma tion with liq ue fac tion and sub - se quent intrastratal flow. If the fold ing of the un der ly ing layer would rep re sent slump ing, a more or less un du lat ing sur face would have been formed, and this would have been eas ily trun - cated by a fol low ing grav ity flow. Con sid er ing that such an ero - sional sur face is ab sent, the fold ing of the un der ly ing mudstone bed can only be due to intrastratal flow in a liq ue fied state. This pro cess thus dif fers from liq ue fac tion in wa ter-sat u rated, cohe - sionless silts and fine sands where the liq ue fac tion re sults in the loss of strength, but not in fold ing. The un der ly ing shale bed thus was par tially af fected by the flow, as proven by the fold struc tures and the par tially de tached mudstone lay ers.
FLOW DIRECTION
One of the im por tant ques tions con cern ing pre sumed intra - stratal flow is whether there was a slope, and – if so – whether it can be de ter mined that the flow took place in downslope di rec - tion. The first ques tion can be an swered eas ily: as all fa cies anal y ses (e.g., Kim et al., 1995, 2003) in di cate that the sed i - ments un der study were de pos ited on the mid dle part of a fan-delta, a slope must have been pres ent. Both the lower and up per bound aries of the beds un der study are par al lel to the bed ding of all other lay ers, and this must rep re sent the in clined sed i men tary sur face. A slope is also ev i denced by the pres ence of nu mer ous slumps.
An im por tant ques tion is, ob vi ously, in which di rec tion the intrastratal trans port took place. This could not be es tab lished di rectly from the CDL, be cause they do not show any fea tures that pro vide un am big u ous in for ma tion. How ever, the bend ing of the clasts shows, as in di cated above, a roughly iden ti cal ori - en ta tion as the bent heads of most of the slumps. For this rea - son we will pay here some at ten tion to the flow di rec tions found by mea sur ing the folds and fold axes of the slumps in the for ma tion.
The Jukmak slump data show a small est interlimb an gle in the 275° fold hinge az i muth, and fold tight en ing tends to in - crease along the 275° trend sup port ing that this is the slump di - rec tion. Con sid er ing the sim i lar ity in the ori en ta tion of the bent mudstone clasts in the CDL and the ori en ta tion of the slump heads, it must be con sid ered highly prob a ble that the intrastratal flow was di rected downslope, i.e. to ward 275°.
CONCLUSIONS
Cha ot i cally de formed sand stone lay ers are pres ent in the Cre ta ceous Gyeokpori For ma tion. They con tain float ing intraformational mudstone clasts that are also de formed. The char ac ter is tics of these sand stones can be ex plained sat is fac - to rily only by intrastratal flow. Two sandy lay ers, sep a rated by a mud layer, be came liq ue fied and started to flow downslope intrastratally. They broke up the mud layer in be tween, and the frag ments of this mud layer be came em bed ded in the flow. The mud frag ments be came de formed dur ing the flow, and the de - formed clasts re mained em bed ded in the ho mog e nized sand when the intrastratal flow came to rest. This re con struc tion shows that the na ture of the sed i ments (muddy or sandy) and par tic u larly their com bi na tion (thick or thin lay ers) plays an im - por tant role in intrastratal flow.
Ac knowl edge ments. This work was sup ported as a Ba sic Re search Pro ject (GP2020-003; Geo log i cal sur vey in the Ko - rean Pen in sula and pub li ca tion of the geo log i cal maps) of the Ko rea In sti tute of Geoscience and Min eral Re sources (KIGAM), funded by the Min is try of Sci ence, ICT (In for ma tion, Com mu ni ca tion and Tech nol ogy), and Fu ture Plan ning, Ko rea.
We thank the re view ers, G. Kawa kami and an anon y mous re - viewer, for nu mer ous sug ges tions that helped im prove both the ex pla na tion of the sci en tific pro cesses in volved and the in for - ma tion pre sented in the pho to graphs.
REFERENCES
Alfaro, P., Moretti, M., Soria, J.M., 1997. Soft-sed i ment de for ma - tion struc tures in duced by earth quakes (seismites) in Plio cene lac us trine de pos its (Guadix-Baza Ba sin, cen tral Betic Cor dil - lera). Eclogae Geologicae Helvetiae, 90: 531–540.
Alsop, G.I., Marco, S., 2013. Seismogenic slump folds formed by grav ity-driven tec ton ics down a neg li gi ble sub aque ous slope.
Tectonophysics, 605: 48–69.
Auchter, N.C., Romans, B.W., Hub bard, S.M., 2016. In flu ence of de posit ar chi tec ture on intrastratal de for ma tion, slope de pos its of the Tres Pasos For ma tion, Chile. Sed i men tary Ge ol ogy, 341:
13–26.
Byun, U.H., Van Loon, Kwon, Y.K., Ko, K., 2019. A new type of slump ing-in duced soft-sed i ment de for ma tion struc ture: the en - ve lope struc ture. Geologos, 25: 111–124.
Chough, S.K., 2013. Ge ol ogy and Sedimentology of the Ko rean Pen in sula. Elsevier, Am ster dam.
Chough, S.K., Chun, S.S., 1988. Intrastratal rip-down clasts, Late Cre ta ceous Uhangri For ma tion, south west Ko rea. Jour nal of Sed i men tary Pe trol ogy, 58: 530–533.
Chough, S.K., Sohn, Y.K., 2010. Tec tonic and sed i men tary evo lu - tion of a Cre ta ceous con ti nen tal arc–backarc sys tem in the Ko -
rean Pen in sula: new view. Earth-Sci ence Re views, 101:
225–249.
Chun, S.S., Chough, S.K., 1992. Depositional se quences from high-con cen tra tion tur bid ity cur rents, Cre ta ceous Uhangri For - ma tion (SW Ko rea). Sed i men tary Ge ol ogy, 77: 225–233.
Curran, J.C., 2007. The de crease in shear stress and in crease in trans port rates sub se quent to an in crease in sand sup ply to a gravel-bed chan nel. Sed i men tary Ge ol ogy, 202: 572–580.
Debacker, T.N., Dumon, M., Matthys, A., 2009. In ter pret ing fold and fault ge om e tries from within the lat eral to oblique parts of slumps: a case study from the An glo-Brabant De for ma tion Belt (Bel gium). Jour nal of Struc tural Ge ol ogy, 31: 1525–1539.
Frey, S.E., Dashtgard, S.E., 2012. Sea weed-as sisted, ben thic gravel trans port by tidal cur rents. Sed i men tary Ge ol ogy, 265-266: 121–125.
Gibert, L., Sanz de Galdeano, C., Alfaro, P., Scott, P.G., Lopez Garrido, A.C., 2005. Seis mic in duced slump in Early Pleis to - cene deltaic de pos its of the Baza Ba sin (SE Spain). Sed i men - tary Ge ol ogy, 179: 279–294.
Gibert, L., Alfaro, P., García-Tortosa, F.J., Scott, G., 2011.
Superposed de formed beds pro duced by sin gle earth quakes
(Tecopa Ba sin, Cal i for nia): In sights into paleoseismology. Sed i - men tary Ge ol ogy, 235: 148-159.
Gladkov, A.S., Lobova, E.U., Deev, E.V., Korzhenkov, A.M., Mazeika, J.V., Abdieva, S.V., Rogozhin, E.A., Rodkin, M.V., Fortuna, A.B., Charimov, T.A., Yudakhin, A.S., 2016. Earth - quake-in duced soft-sed i ment de for ma tion struc tures in Late Pleis to cene lac us trine de pos its of Issyk-Kul lake (Kyrgystan).
Sed i men tary Ge ol ogy, 344: 112-122.
Hansen, E., 1971. Strain Fa cies. Springer, Berlin.
Haughton, P.D.W., Barker, S.P., McCaffrey, W.D., 2003. “Linked”
debrites in sand-rich turbidite sys tems; or i gin and sig nif i cance.
Sedimentology, 50: 459–482.
Hempton, M.R., Dewey, J.F., 1983. Earth quake-in duced deforma - tional struc tures in young lac us trine sed i ments, East Ana to lian Fault, south east Tur key. Tectonophysics, 98: T7–T14.
Jiang, J., Zhong, N., Li, Y., Xu, H., Yang, H., Peng, X., 2016. Soft sed i ment de for ma tion struc tures in the Lixian lac us trine sed i - ments, east ern Ti betan Pla teau and im pli ca tions for postglacial seis mic ac tiv ity. Sed i men tary Ge ol ogy, 344: 123–134.
Kawakami, G., Kawamura, M., 2002. Sed i ment flow and de for ma - tion (SFD) lay ers: ev i dence for intrastratal flow in lam i nated muddy sed i ments of the Tri as sic Osawa For ma tion, North east Ja pan. Jour nal of Sed i men tary Pe trol ogy, 72: 171–181.
Kim, S.B., Chough, S.K., Chun, S.S., 1995. Bouldery de pos its in the low er most part of the Cre ta ceous Kyokpori For ma tion, SW Ko rea: cohesionless de bris flows and de bris falls on a steep-gra di ent delta slope. Sed i men tary Ge ol ogy, 98: 97–119.
Kim, S.B., Chough, S.K., Chun, S.S., 2003. Tec tonic con trols on spatio-tem po ral de vel op ment of depositional sys tems and gen - er a tion of fin ing-up ward ba sin fills in a strike-slip set ting:
Kyokpori For ma tion (Cre ta ceous), south-west Ko rea.
Sedimentology, 50: 639–665.
Ko, K., Park, S., Kwon, C.W., 2015. Soft-sed i ment de for ma tion struc tures in the Cre ta ceous Gyeokpori For ma tion of the Buan area, Ko rea: Struc tural char ac ter is tics, re con struc tion of paleoslope and trig ger ing mech a nism of slump (in Ko rean with Eng lish ab stract). Jour nal of Geo log i cal So ci ety of Ko rea, 51:
545–560.
Ko, K., Kim, S.W., Lee, H.-J., Hwang, I.G., Kim, B.C., Kee, W.-S., Kim, Y.-S., Gihm, Y.S., 2017. Soft sed i ment de for ma tion struc - tures in a lac us trine sed i men tary suc ces sion in duced by vol - cano-tec tonic ac tiv i ties: an ex am ple from the Cre ta ceous Beolgeumri For ma tion, Wido Volcanics, Ko rea. Sed i men tary Ge ol ogy, 358: 197–209.
Koh, H.J., Kwon, C.W., Park, S.I., Park, J., Kee, W.S., 2013. Geo - log i cal Re port of the Julpo and Wido-Hawangdeungdo sheets (1:50,000) (in Ko rean, with Eng lish ab stract). Ko rea In sti tute of Geoscience and Min eral Re sources.
Kwon, C.W., Ko, K., Gihm, Y.S., Koh, H.J., Kim, Y., 2017. Late Cre ta ceous vol ca nic arc sys tem in south west Ko rea: Dis tri bu - tion, li thol ogy, age, and tec tonic im pli ca tions. Cre ta ceous Re - search, 75: 125–140.
Mazumder, R., Van Loon, A.J., Malviya, P., Arima, M., Ogawa, Y., 2016. Soft-sed i ment de for ma tion struc tures in the Mio-Plio cene Misaki For ma tion within al ter nat ing deep-sea clays and vol ca nic
ashes (Miura Pen in sula, Ja pan). Sed i men tary Ge ol ogy, 344:
323–335.
Moretti, M., Sabato, L., 2007. Rec og ni tion of trig ger mech a nisms for soft-sed i ment de for ma tion in the Pleis to cene lac us trine de - pos its of the Sant’Arcangelo Ba sin (south ern It aly): seis mic shock vs. over load ing. Sed i men tary Ge ol ogy, 196: 31–45.
Pisarska-Jamroży, M., Van Loon, A.J., Mleczak, M., Ro man, M., 2019. Enig matic grav ity-flow de pos its at Ujście (west ern Po - land), trig gered by earth quakes (as ev i denced by seismites) caused by Saalian glacioisostatic crustal re bound. Geo mor phol - ogy, 326: 239–251.
Rana, R., Sati, S.P., Sundrival, Y., Juval, N., 2016. Gen e sis and im - pli ca tion of soft-sed i ment de for ma tion struc tures in high-en ergy flu vial de pos its of the Alaknanda Val ley, Garhwal Himalaya, In - dia. Sed i men tary Ge ol ogy, 344: 263–276.
Rodríguez-Pascua, M.A., Calvo, J.P., de Vicente, G., Gomez Gras, D., 2000. Seismites in lac us trine sed i ments of the Prebetic Zone, SE Spain, and their use as in di ca tors of earth - quake mag ni tudes dur ing the Late Mio cene. Sed i men tary Ge ol - ogy, 135: 117–135.
Salamon, T., 2015. Sed i men tary re cord of a Scan di na vian Ice Sheet drain age sys tem and till de po si tion over subglacial ob sta cles pro mot ing basal slid ing (an ex am ple from south ern Po land).
Sed i men tary Ge ol ogy, 330: 108–121.
Sims, J.D., 1973. Earth quake-in duced struc tures in sed i ments of Van Nor man Lake, San Fernando, Cal i for nia. Sci ence, 182:
161–163.
Sims, J.D., 1975. De ter min ing earth quake re cur rence in ter vals from deformational struc tures in young lac us trine sed i ments.
Tectonophysics, 29: 141–152.
Tan ner, L.H., Lucas, S.G., 2007. The Moenave For ma tion:
Sedimentologic and strati graphic con text of the Tri as sic–Ju ras - sic bound ary in the Four Cor ners area, south west ern USA.
Palaeo ge ogra phy, Palaeoclimatology, Palaeo ec ol ogy, 244:
111–125.
TaêgÍn, C.K., Türkmen, I., 2009. Anal y sis of soft-sed i ment de for - ma tion struc tures in Neo gene fluvio-lac us trine de pos its of ÇaybÈaÏ For ma tion. East ern Tur key. Sed i men tary Ge ol ogy, 218:
16–30.
TaêgÍn, C.K., Orhan, H., Türkmen, I., Aksoy, E., 2011. Soft-sed i - ment de for ma tion struc tures in the late Mio cene elmo For ma - tion around AdÏyaman area, South east ern Tur key. Sed i men tary Ge ol ogy, 235: 277–291.
Van Loon, A.J., Pisarska-Jamroży, M., Nartišs, M., Krievâns, M., Soms, J., 2016. Seismites re sult ing from high-fre quency, high-mag ni tude earth quakes in Lat via caused by Late Gla cial glacio-iso static up lift. Jour nal of Palaeo ge ogra phy, 5: 363–380.
Whitmore, J.H., Strom, R., 2010. Sand injectites at the base of the Coconino Sand stone, Grand Can yon, Ar i zona (USA). Sed i men - tary Ge ol ogy, 230: 46–59.
Wil liams, E., 1960. Intra-stra tal flow and con vo lute fold ing. Geo log i - cal Mag a zine, 97: 208–214.
Wood cock, N.H., 1979. The use of slump struc tures as palaeoslope ori en ta tion es ti ma tors. Sedimentology, 26: 83–99.
Intrastratal flow in the Cre ta ceous Gyeokpori For ma tion (SW South Ko rea) 625
