**2. Disconformities at seismic resolution**

**1. Introduction**

are discussed below.

**1.2. Growth of the concept**

are excluded from further discussion.

of this contribution.

**1.1. Definition of the subject**

It is generally accepted that only about 10% of the geologic time is recorded in the sedimentary rocks, whereas 90% is collapsed into non-deposition, alteration, and erosion surfaces collectively called unconformities [1]. Of these diverse surfaces with time value ranging from minutes to hundred millions of years, only those of practical use, that is, traceable on a scale exceeding one outcrop and marked with distinct diagnostic features

102 Seismic and Sequence Stratigraphy and Integrated Stratigraphy - New Insights and Contributions

This subheading, borrowed from Dunbar and Rodgers [2], brackets 230 years of unconformity research counting from recognition of an angular unconformity in late 1780s [3]. The word *unconformity* was adapted from German geology 3 decades later [4, p. 48] and until the midnineteenth century pertained to angular stratal discordances. Awareness of geologic time gaps between parallel bed sets, normally accompanied by signatures of erosion, emerged in late nineteenth century (e.g., [5]) under the influence of Charles Darwin's conclusion on the principal incompleteness of the stratigraphic record [6]. As most recently reviewed by Miall [1, 7], such stratigraphic breaks between parallel strata were classified into "unconformities Type a" by Blackwelder [8] and shortly after that named *disconformities* [9]. The other two types of unconformities of Blackwelder [8] were (b) contact between rocks of wholly unlike origin (for example, sandstone resting upon granite); and (c) angular discordance of beds with or without difference in lithologic character. Type (c) is the classical *angular unconformity* of James Hutton, and type (b) was named *nonconformity*. The latter term was coined by Pirsson and Schuchert [10] and refined into modern usage by Dunbar and Rodgers [2]. Surfaces between parallel bed sets recording time gaps but not bearing signs of erosion were named *paraconformities*, as opposed to erosion-marked *disconformities* [2]. However, the difference between the disconformity and paraconformity more often appears in the ability to recognize erosion and evolves with tools and methods. Here, the term stratigraphic unconformity is used as an equivalent of disconformity. Barrell [11] also coined a term *diastem* that became adapted for the time value of a sedimentation gap at an unconformity. Being most easily identified features, angular unconformities and nonconformities

Disconformity-bounded packages of sedimentary rocks, called cycles, cyclites, cyclothems, allostratigraphic units, and most commonly sequences, remained in focus for many decades, generating an impressive development of concepts, terminology, and discussion on local vs. global controls of base or sea level fluctuations [1, 2, 12–24]. It should be noted that sequence stratigraphy significantly expanded definition of sequences by including both disconformities and their correlative surfaces (conformities) in more complete basincentered sections [14]. Sequence stratigraphy is reviewed in this book but is not the focus Disconformities mostly show concordant stratal relationships below and above the surface. They are identifiable on seismic sections if the subaerial exposure allowed for development of significant relief and/or seismic-scale incised fluvial channels [20, 25]. Incised valleys form during base level fall and become filled when base level rises [26]. Fluvial incisions are deeper and favorable for seismic mapping where they cut into an uplifted plain (**Figure 1A**), plateau, or across a shelf break (**Figure 1B**), but may not be identifiable in paleo-hinterlands with a shallow base of erosion. Transgressive tide and wave abrasion of coasts, estuaries, and shoreface are able to modify the configuration of subaerial surfaces and any terrestrial sediment accumulated on it. The surfaces produced by such an abrasion are called ravinement surfaces [17, 22, 23]. The depth of transgressive erosion greatly varies depending on induration of the exposed sediment, on the wave and tide energy of a transgressing sea, and on the slope angle of the eroded sediments. While oceanic abrasion may cut down to tens of meters into seashore cliffs, plain lands characteristic of epicontinental sedimentary environments may show negligible transgressive stripping and delicate topsoil parts of weathering profiles largely preserved (e.g., [27]).

Seismic and hands-on-rock unconformities are not the same, and the proportion of false seismic unconformities is greater than was thought by Vail et al. [15]. Situations where time lines converge into a condensed section but portrayed as an onlap-offlap surface, or pseudo-unconformities envisioned from a surface of major lithological contrast, are very common misinterpretations [30, 31]. Difficulty in recognition of subaerial unconformities in the subsurface led to proposal of an alternative *genetic stratigraphic sequences* bounded by "maximum flooding surfaces" or condensed sections [16]—however, the concept of very limited use today.

**Figure 1.** Examples of mature unconformity surfaces with fluvial incisions visualized in 3D seismic models: (A) A 225 m subsurface slice (above sea level) along sub-Cretaceous unconformity visualizing high-sinuosity channels cut into weathered Devonian limestone, eastern Athabasca oil sands, Alberta (formally modified from [28]). (B) Hibernia Canyon cut through shelf edge during latest Campanian-earliest Danian (Cretaceous-Paleocene), Jeanne d'Arc Basin offshore Newfoundland, formally modified from Deptuck et al. [29].

Disconformities are also invisible with conventional seismic surveys in carbonate successions if (1) no reef crest or carbonate mound relief was produced in slowly subsiding platform setting; (2) transgressive carbonates deposited upon regressive carbonates with no impedance contrast produced; (3) sequences are thinner than reading resolution at a given pulse frequency; (4) fluvial channels or sufficient erosional relief did not develop and drainage of meteoric waters was entirely underground; and (5) stacked karst systems from successive paleo-aquifers overprint with no chance to trace particular karst horizons. The best presentday example of a carbonate plain where the day surface has a chance to be buried in such a hidden way is the Nullarbor Plain of Southwestern Australia [32]. This vast (~240,000 km2 ) plain was exposed for the last 14 My since mid-Miocene time, yet remains exceptionally flat and riverless with extensive underground cave systems produced during several Tertiary lowstands, including the ongoing uplift [33].

#### **3. Evaluation of hiatuses**

Most terrains show a relief or a slope gradient where prolonged flooding is recorded in onlap patterns, as opposed to geologically momentary (rapid) inundation of plain lowlands. Hiatuses therefore tend to wedge toward basin centers on chronostratigraphic charts [15], as demonstrated by case studies where unconformities receive cross-basinal biostratigraphic control [29]. Hiatuses reveal more complex histories in settings of differential subsidence in areas of large-scale salt diapirism or in tectonically active regions (e.g., foreland flexural bulging and tilting). The eustatic vs. tectonic control over transgressions and regressions is a subject of long-lasting debate [13, 20]. Tectonic control of an unconformity between two parallel bed sets can be interpreted where unconformities show poor or no correlation to major lowstands of "global sea level curves" or where the hiatus is diachronous with bedrock and caprock younging in the same direction. For example, a major intra-Cretaceous disconformity of central-southern Italy is generally younging eastward from Late Albian to Late Turonian—earliest Coniacian as revealed with refined biostratigraphic control [34]. This unconformity hosts karst-associated economic bauxites and is locally composite with two bauxitiferous paleokarsts divided by Cenomanian limestone of various thickness and time value. This diachroneity was interpreted as the translation of the lithospheric bulge in response to compression from the distal orogeny along the Adria Plate margin [34].

Biostratigraphy is the oldest yet still master method of recognizing hiatuses by missing zones, which can be processed with a graphic correlation technique [7]. Other absolute dating methods like U-Pb ID-TIMS and cyclostratigraphy are reviewed in [7, 35]. Resolution of biostratigraphy varies with the group employed, paleogeographic position, and the geologic age. The latter controls biostratigraphic resolution to a significant extent by cosmopolitism vs. provincialism of marine faunas. High cosmopolitism is characteristic of greenhouse periods with circum-tropical seaway connections, whereas provincialism is favored by forcing of the Earth into icehouse mode and shutdowns of low-latitude seaway connections, as likely happened during Pennsylvanian-Permian assembly of Pangaea [36, 37].

Noteworthy here are historically recognized but apparently non-existent disconformities. Usually, these "legacy hiatuses" heavily rely on biostratigraphy. An example is given by the "Late Middle Devonian unconformity" of the Mackenzie Corridor of Northwestern Canada. This unconformity was interpreted by Hume and Link [38] from the sharp thickness fluctuations and restricted spatial distribution of the Hare Indian and Ramparts formations, which was seen as a result of erosion prior to deposition of the black siliceous shale of the Canol Formation (**Figure 2**; [46]). A debate on the validity of this hiatus lasted ever since. The hiatus has been supported by the assignment of the upper Hare Indian Formation to the undifferentiated *varcus* conodont zone (=*rhenanus-ansatus* in **Figure 2**), whereas the lower part of Canol Formation was dated by conodonts as the Lower *asymmetrica* (≈*transitans-falsiovalis* on **Figure 2**) with speculative extension of the Canol base into the lowermost *asymmetrica* or present-day *norrisi* zone [39, 40]. *Hermanni-disparilis* interval was allegedly missing (**Figure 2**). However, scarce conodont data from the Ramparts limestone suggested its age range from the upper Hare Indian equivalent

Disconformities are also invisible with conventional seismic surveys in carbonate successions if (1) no reef crest or carbonate mound relief was produced in slowly subsiding platform setting; (2) transgressive carbonates deposited upon regressive carbonates with no impedance contrast produced; (3) sequences are thinner than reading resolution at a given pulse frequency; (4) fluvial channels or sufficient erosional relief did not develop and drainage of meteoric waters was entirely underground; and (5) stacked karst systems from successive paleo-aquifers overprint with no chance to trace particular karst horizons. The best presentday example of a carbonate plain where the day surface has a chance to be buried in such a hidden way is the Nullarbor Plain of Southwestern Australia [32]. This vast (~240,000 km2

104 Seismic and Sequence Stratigraphy and Integrated Stratigraphy - New Insights and Contributions

plain was exposed for the last 14 My since mid-Miocene time, yet remains exceptionally flat and riverless with extensive underground cave systems produced during several Tertiary

Most terrains show a relief or a slope gradient where prolonged flooding is recorded in onlap patterns, as opposed to geologically momentary (rapid) inundation of plain lowlands. Hiatuses therefore tend to wedge toward basin centers on chronostratigraphic charts [15], as demonstrated by case studies where unconformities receive cross-basinal biostratigraphic control [29]. Hiatuses reveal more complex histories in settings of differential subsidence in areas of large-scale salt diapirism or in tectonically active regions (e.g., foreland flexural bulging and tilting). The eustatic vs. tectonic control over transgressions and regressions is a subject of long-lasting debate [13, 20]. Tectonic control of an unconformity between two parallel bed sets can be interpreted where unconformities show poor or no correlation to major lowstands of "global sea level curves" or where the hiatus is diachronous with bedrock and caprock younging in the same direction. For example, a major intra-Cretaceous disconformity of central-southern Italy is generally younging eastward from Late Albian to Late Turonian—earliest Coniacian as revealed with refined biostratigraphic control [34]. This unconformity hosts karst-associated economic bauxites and is locally composite with two bauxitiferous paleokarsts divided by Cenomanian limestone of various thickness and time value. This diachroneity was interpreted as the translation of the lithospheric bulge in response to compression from the distal orogeny along the Adria

Biostratigraphy is the oldest yet still master method of recognizing hiatuses by missing zones, which can be processed with a graphic correlation technique [7]. Other absolute dating methods like U-Pb ID-TIMS and cyclostratigraphy are reviewed in [7, 35]. Resolution of biostratigraphy varies with the group employed, paleogeographic position, and the geologic age. The latter controls biostratigraphic resolution to a significant extent by cosmopolitism vs. provincialism of marine faunas. High cosmopolitism is characteristic of greenhouse periods with circum-tropical seaway connections, whereas provincialism is favored by forcing of the Earth into icehouse mode and shutdowns of low-latitude seaway connections, as likely happened

during Pennsylvanian-Permian assembly of Pangaea [36, 37].

lowstands, including the ongoing uplift [33].

**3. Evaluation of hiatuses**

Plate margin [34].

)

**Figure 2.** Legacy "Late Middle Devonian unconformity" on a simplified table of formations of Mackenzie Valley and Peel area; the unconformity advocated (right column) vs. discarded (left column).

to the *asymmetrica* zone [41]. Other workers argued that the Canol base is a conformity [42, 43] and indicated Ramparts-Canol interfingering in allochthonous debris units [41]. Nevertheless, this hiatus survived in the territorial table of formations until recently [44]. Decisive in retiring this hiatus are (1) updates in conodont data showing Canol base time gliding from the Frasnian *transitans-punctata* on top of Kee Scarp carbonate banks to the upper Givetian *norrisi* in offbank depressions [44]; (2) carbonate-bank slope depositional setting of allochthonous bioclastic debris interfingering with laminated black shales; and (3) absence of any evidence of subaerial exposure or vadose processes, like oxidation of pyrites and organic matter and characteristic redistribution of Fe and Mn, prior to the onset of Canol deposition [45].

An absolute majority of disconformities are subzonal or do not bear index fossils immediately below or above. Relative proxy for the duration of a hiatus is the maturity of a paleosol profile, e.g., progression from entisols to any of mature soil profiles defined by soil taxonomy [47], or stages of calcrete development [48], but the ability to deconvolute time is quite limited: paleosol appearance is a multivariate product of exposure duration, precipitation regime, temperature, relief, availability and type of vascular vegetation (and other soil biota), and transgressive truncation, with variable masking of paleonvironmental signals by burial diagenetic overprints. Most tools of radiogenic dating used to reveal soil age are not applicable to deep-time examples because of short isotope decay lifetime. U-Pb dating of soil carbonates, based on U adsorbed in calcite lattice, was demonstrated to provide quantitative estimate of pedogenic processes as old as Carboniferous [49, 50]. Also, the production of He isotopes by α-decay of U, Th, and their intermediate decay species was used to develop a (U-Th)/He geochronometer that is able to date materials in the range of a few thousands of years to 4.5 Ga (see review in [51]).
