Onshore–offshore correlations of AUTHORS fl William J. Schmelz ~ Department of Cretaceous uvial-deltaic Earth and Planetary Sciences, Rutgers University, Piscataway, New Jersey; and sequences, southern Baltimore Institute of Earth, Ocean, and Atmospheric Sciences, Rutgers University, Piscataway, Canyon trough New Jersey; [email protected] William J. Schmelz received a B.A. in William J. Schmelz, Kenneth G. Miller, environmental science from Connecticut Gregory S. Mountain, James V. Browning, and College in 2010 and is working toward his Kimberly E. Baldwin Ph.D. within the stratigraphy–earth history group in the Department of Earth and Planetary Sciences at Rutgers University, studying the tectonostratigraphic history of the ABSTRACT mid-Atlantic margin and the interconnected global mean sea-level record using We evaluate Cretaceous depositional sequences on approxi- integrated sequence stratigraphic, statistical, mately 4400 km (~2700 mi) of newly released multichannel and numerical modeling methods. seismic profiles and five wells on the continental shelf in the Kenneth G. Miller ~ southern Baltimore Canyon trough and tie the data to three wells Department of Earth and Planetary Sciences, Rutgers drilled onshore in the Maryland coastal plain. Seismic geometries University, Piscataway, New Jersey; and coupled with facies and biostratigraphy from the wells are used Institute of Earth, Ocean, and Atmospheric – to delineate mid-Cretaceous (Aptian Turonian) depositional Sciences, Rutgers University, Piscataway, sequences and paleogeography. Beneath these sequences, New Jersey; [email protected] – – 400 1000 m (1300 3300 ft) of Lower Cretaceous sedimentary Kenneth G. Miller received his Ph.D. from the rocks underlie the modern shelf. They thicken along strike to the Massachusetts Institute of Technology and the southwest, implying a southern sediment source. Aptian to Woods Hole Oceanographic Institution. He is a Cenomanian sediments were deposited in shelf to nearshore distinguished professor at Rutgers University settings. A landward movement of the depocenter and a shift working on stratigraphy, paleoceanography, toward facies indicative of deeper paleodepths marks a 107-yr and micropaleontology, focusing on Late – mid-Cretaceous transgression, within which we identify five se- Cretaceous Holocene sea-level and quences. A composite maximum flooding surface (MFS) within cryospheric evolution. the uppermost of these retrogradational units is associated Gregory S. Mountain ~ Department of with the Cenomanian–Turonian boundary and ocean anoxic Earth and Planetary Sciences, Rutgers event 2. Shingled, lower Turonian seismic reflections prograde University, Piscataway, New Jersey; and across the outer shelf, downlapping onto the composite MFS, Institute of Earth, Ocean, and Atmospheric and are truncated by a mid-Turonian sequence boundary. The Sciences, Rutgers University, Piscataway, New Jersey; [email protected] Upper Cretaceous section thickens seaward and along strike to the northeast, implying a northern source and little Late Cre- Gregory S. Mountain earned his Ph.D. at ’ taceous accommodation beneath the modern shelf. Mid-Cretaceous Columbia s Lamont-Doherty Earth Observatory and is the chair of the Department strata offshore Maryland are likely sand-prone, considering their fl of Earth and Planetary Sciences at Rutgers proximity to the correlative uvial facies of the onshore Potomac University. His research interests include Group. These potential reservoir sands are capped by re- application of seismic stratigraphy to 7 gional confining units generated by 10 -yr global mean sea-level understanding the impact of sea-level change and the history of deep-ocean circulation.

James V. Browning ~ Department of Copyright ©2020. The American Association of Petroleum Geologists. All rights reserved. Earth and Planetary Sciences, Rutgers Manuscript received September 6, 2018; provisional acceptance October 24, 2018; revised manuscript University, Piscataway, New Jersey; and received February 19, 2019; final acceptance May 6, 2019. DOI:10.1306/05061918197

AAPG Bulletin, v. 104, no. 2 (February 2020), pp. 411–448 411 Institute of Earth, Ocean, and Atmospheric flooding events and are excellent targets for supercritical carbon Sciences, Rutgers University, Piscataway, storage. New Jersey; [email protected] James V. Browning earned his Ph.D. at Rutgers University and is currently an assistant INTRODUCTION research professor and core curator in the Department of Earth and Planetary Sciences at Geometries observed on seismic reflection profiles allow objec- Rutgers University. His research focuses on tive recognition of genetically related strata bound by uncon- sea-level rise and fall, and the sedimentology of the Atlantic Coastal Plain. formities and their apparent correlative surfaces (Mitchum et al., 1977b). Integrating reflection seismic data with well logs, cut- Kimberly E. Baldwin ~ Department of tings, and core samples (including biostratigraphy and other Earth and Planetary Sciences, Rutgers chronological control) can significantly improve temporal reso- University, Piscataway, New Jersey; and lution and paleoenvironmental interpretations (Van Wagoner Institute of Earth, Ocean, and Atmospheric Sciences, Rutgers University, Piscataway, et al., 1990). Depositional sequences may be objectively recog- New Jersey; [email protected] nized in seismic, well-log, cutting, and core data and do not need to be uniquely tied to sea-level curves (Neal and Abreu, 2009; Kimberly E. Baldwin received her B.S. in earth and environmental sciences from Lehigh Miller et al., 2018). University and M.S. in geological sciences from Sequence stratigraphy provides a tool to place paleoenvir- Rutgers University, where she studied onmental reconstructions and basin evolution into a local and sediment wave formation as a result of long- regional context through mapping depositional environments term bottom-water circulation in the western and evaluating their trajectories through space and time as a Pacific.Sheiscurrentlyaresearchproject result of changes in subsidence, global mean sea level (GMSL), assistant at Rutgers University working and sediment supply. Further, sequence stratigraphy allows re- on the Cretaceous sequence stratigraphy construction of the stratigraphic record from its geomorphological offshore New Jersey. origins and provides the most reliable way of examining strata for the purpose of basin scale correlation and lithological prediction for the ACKNOWLEDGMENTS distribution of reservoirs and seals (Posamentier and James, 1993). This work was supported by National Science The geologic record on the mid-Atlantic margin off the coast Foundation grant OCE14-63759 and the US of New Jersey has been intensely studied, providing a “natural Department of Energy under award nos. laboratory” for understanding passive-margin evolution and sea- DE-FE0026087 (Mid-Atlantic US Offshore level change (e.g., Poag, 1985b; Sheridan and Grow, 1988; Poag Carbon Storage Resource Assessment and Sevon, 1989; Miller et al., 1998, 2005; Mountain et al., Project) and DE-FC26-0NT42589 (Midwest 2010). The 1970s–1980s saw exploration for hydrocarbon res- Regional Carbon Sequestration Partnership ervoirs in the thick (2–16 km [6600–52,500 ft]) postrift (upper Program) managed by Battelle. Guidance and support were provided by N. Gupta and Lower Jurassic and younger) strata of the Baltimore Canyon L. Cummings of Battelle. The organizational trough (BCT) (summary in Poag, 1985b; e.g., Grow and Sheridan, work of M. KunleDare of the Delaware 1988). Recently, efforts have focused on the Quaternary Geological Survey assessing legacy data that stratigraphy, as part of the Strata Formation on Margins includes the paper copies of the logs used (STRATAFORM) initiative (Nittrouer et al., 2007), and the herein is much appreciated. B. Dunst of the Miocene, with drilling by the Integrated Ocean Drilling Program Pennsylvania Department of Conservation Expedition (IODP) 313 on the New Jersey shallow shelf (e.g., and Natural Resources was immensely Mountain et al., 2010; Miller et al., 2013a, b). Early studies helpful in digitizing thousands of feet of well logs. We thank D. Andreasen and looked at the offshore Cretaceousaspartofstudiesrelated P. McLaughlin for their insights on the to hydrocarbon exploration that delineated Cretaceous res- Cretaceous sediments of the Maryland and ervoir sandstones (e.g., Libby-French, 1984; Poag, 1985b; Delaware coastal plain and D. Monteverde Prather, 1991). Later studies on the onshore Cretaceous coastal for his insight on seismic interpretations. plain used a sequence stratigraphic approach (Olsson et al., 1988; We also acknowledge the fundamental Kulpecz et al., 2008). With increased interest in supercritical carbon storage, recent studies have begun to evaluate the

412 Cretaceous Fluvial-Deltaic Sequences of the Southern Baltimore Canyon Trough offshore Cretaceous of the mid-Atlantic margin, focusing on the contributions from C. Lombardi (deceased), northern BCT (NBCT) and the adjacent coastal plain (Miller who identified the need for an updated et al., 2017, 2018) because it contains numerous industry wells framework for the Cretaceous stratigraphy of and thick reservoir sands. the Baltimore Canyon trough and developed a computer application used to display our The mid-Cretaceous (Aptian to Cenomanian) section within ~ – well-log data. We thank C. Fulthorpe and the offshore southern BCT (SBCT; 38° 39°N; offshore north- J. Neal for their thoughtful and constructive ern Virginia, Maryland, Delaware, and southern New Jersey; comments. Figure 1) has been sparingly studied (e.g., Prather, 1991; Klitgord et al., 1994) because of a paucity of wells and a previously sparse database of low-quality (by modern standards) multichannel seismic (MCS) profiles collected by the US Geological Survey (USGS) in the 1970s. As a result, there are a number of out- standing questions regarding the spatial variability and general continuity of Cretaceous sedimentation across the basin. In particular, the strata comprising the continental shelf, slope, and rise off the coast of Maryland have been given little attention, aside from the aforementioned regional correlation of Klitgord et al. (1994) that characterized the entire sedimentary section from basement to sea floor from Virginia to Canada using the USGS MCS profiles. Moreover, the knowledge of the stratigra- phy from onshore rotary wells (Doyle, 1982; Hansen, 1982; Andreasen et al., 2015; Miller et al., 2017) was not considered in the seismic interpretations, and, by extension, depositional se- quences have not been correlated from the coastal plain to the continental shelf. These data from the onshore coastal plain provide an important constraint on any seismic stratigraphic in- terpretation of the continental shelfoffthecoastsofMarylandto southern New Jersey. For this region, the nearest offshore wells are the outer continental shelf (OCS) wells off the coast of New Jersey over 100 km (>62 mi) to the north and Shell 93-1, which is located on the rise in over 2 km (>6600 ft) of water (Figure 1), on a margin where shelf-to-slope facies are difficult to correlate (Mountain and Tucholke, 1985; Poag, 1985a). With the recent release of hundreds of thousands of kilo- meters of industry seismic data covering the BCT (Triezenberg et al., 2016), it is now possible to examine the sedimentation history of this margin in greater detail. The seismic profiles not only provide data for interpretation of the Cretaceous sedi- mentary structure within the SBCT but also the opportunity to integrate information from the extensive literature and industry wells of the NBCT and the mid-Atlantic Coastal Plain. To address the opportunity for new insight on the larger-scale evolu- tion of sedimentation on this margin, this work examines fluvial- deltaic sedimentary sequences within Cretaceous strata of the SBCT. We rely primarily on a dense grid of marine MCS reflection profiles integrated with biostratigraphical and geophysical data from drill sites on the mid-Atlantic Coastal Plain, OCS, and con- tinental slope (Figure 1). Integrated sequence stratigraphic methods are used to map the depositional sequences of the continental shelf

SCHMELZ ET AL. 413 Figure 1. Map of the seismic and drill site data within the Baltimore Canyon trough; black rectangles outline the southern and northern subareas discussed in the text. Generalized outcrop areas of coastal plain strata are distinguished by color. Depth contours to prerift basement are shown in dashed lines. The yellow lines identify seismic profiles shown in figures herein. COST = Continental Offshore Stratigraphic Test; DEM = digital elevation model; fig. = figure; MD = Maryland; USGS = US Geological Survey. from southern New Jersey to Maryland. We identify depositional environments through space and time. depositional sequences off the coast of Maryland that Because the cyclicity of shallow marine siliciclastic en- correlate to depositional sequences present on the vironments at a passive margin is a function of GMSL Maryland coastal plain (Miller et al., 2017) and to the change, subsidence within the basin, and sediment NBCT (Miller et al., 2018), generating basin-scale supply to the margin (e.g., Reynolds et al., 1991), a sequence stratigraphic maps. regional sequence stratigraphic framework provides The basin-scale depositional sequences we interpret insight into how these factors have contributed to have implications for the identification of variations in the development and preservation of the sedimentary

414 Cretaceous Fluvial-Deltaic Sequences of the Southern Baltimore Canyon Trough architectures of the SBCT, both from sedimento- (cf., Klitgord et al., 1988; Manspeizer and Cousminer, logical and tectonostratigraphic perspectives. The 1988), and because of the location of deep-penetration improved constraints on sedimentary architecture wells, there are few samples of Middle to lower Upper provide predictability not only for hydrocarbon re- Jurassic strata. An initial postrift period of evaporite serves but also for supercritical storage of carbon precipitation in the embryo of the present-day At- dioxide. lantic Ocean was followed by a carbonate shelf re- gime with a fringing progradational reef that was subsequently buried in the mid-Cretaceous by sili- BACKGROUND ciclastic deposition (Jansa, 1981). For the majority of the BCT, the reef was buried in the Barremian The BCT, an offshore basin containing as much as when sediments overtopped the reef structure and 16 km (52,500 ft) of Middle Jurassic to Holocene were deposited on top of the Neocomian (Hau- strata (Grow and Sheridan, 1988), is bounded to the terivian, Valanginian, and Berriasian) deep-sea re- west by the basement hinge zone (approximated by flector b identified beneath the modern continental the 5 km [16,400 ft] basement contour; Figure 1). rise (Mountain and Tucholke, 1985). The carbonate Accretion of Paleozoic terranes and subsequent up- platform was generally progradational (Poag and lift was followed by Late Triassic to earliest Jurassic Valentine, 1988), with continentally sourced silici- rifting and extensional separation of the North clastics locally interfingering with the Upper Jurassic American and African plates (Klitgord et al., 1988; and Lower Cretaceous carbonates (Poag, 1985b). Withjack et al., 1998; Seton et al., 2012). The Proximally, the subsurface Neocomian sediments of remnant heat from the extension and rifting phase the New Jersey coastal plain record the fluvial de- resulted in a subsequent period of thermal subsidence position of the Waste Gate Formation in Maryland (i.e., McKenzie, 1978) and flexural sedimentary and southern New Jersey (Doyle, 1982; Hansen, loading on an elastic lithospheric plate (Steckler 1982; Miller et al., 2017). Under the modern shelf, and Watts, 1978; Watts et al., 1982; Klitgord et al., the heterolithic Valanginian to Barremian interval 1988). The rift phase (ca. 230–198Mainthisre- contains facies equivalent to the Missisauga Forma- gion; Withjack et al., 1998; Schlische et al., 2003) tion of the Scotian Shelf (Libby-French, 1984). The was characterized by faulting, minor folding, and siliciclastic sediments that eventually buried the Late Triassic–Early Jurassic synrift sedimentary carbonate platform are coarser than the finer mid- to deposition that occurred in two sets of elongate Upper Cretaceous sediments they were buried by northeast-trending half-graben basins, one land- (Poag and Valentine, 1988); the upward-fining suc- ward and one seaward of the basement “hinge cession reflects an increase in base level that began in zone” (Figure 1; Manspeizer and Cousminer, 1988). the mid-Cretaceous. These transitions, including the Sedimentation was punctuated by the circa 201.6– death of the great barrier reef, reflect the structural 200.5 Ma central Atlantic magmatic province evolution of the margin within a changing global (CAMP; Blackburn et al., 2013) prior to the forma- climate (Jansa, 1981; Poag, 1985b). tion of the postrift unconformity (PRU) horizon that The mid-Cretaceous sandstones of the BCT, separates syn- and postrift sedimentary rocks across described as the lithological equivalent to the Logan the entire margin (Withjack et al., 1998). The PRU Canyon (LC) Formation of the Scotian Shelf (Libby- has an estimated age of circa 175 Ma that is poorly French, 1984; Poag, 1985b), were deposited follow- constrained by assuming a constant sea-floor spread- ing the mixed siliciclastic carbonates of the Barremian ing rate and extrapolating a regression of ages from to Neocomian MissisaugaFormationequivalent magnetic lineations on oceanic crust landward (Libby-French, 1984; Poag, 1985b). The sandstones (Klitgord and Schouten, 1986), although the PRU is were originally partitioned into an upper and lower diachronous across its extent (Klitgord et al., 1988). LC sandstone by Libby-French (1984). These two The uppermost Jurassic and Cretaceous sedi- sands were thought to be separated by a fine-grained mentation within the BCT, the focus of this paper, unit called the Sable Shale. Closer examination of began approximately 50 m.y. after CAMP. The the lithological transitions and biostratigraphy in the presence of postrift Lower Jurassic strata is uncertain NBCT led Miller et al. (2018) to conclude that the

SCHMELZ ET AL. 415 LC was actually composed of three distinct deposi- drilled on the OCS (Seker, 2012). It may be coeval with tional sequences, the LC3 (oldest), LC2, and LC1 any of several onshore sand units, from the Coniacian (youngest), each possessing (1) lower regressive, Magothy to the Campanian Mount Laurel Formation lowstand systems tract (LST) interbedded silts and of the New Jersey coastal plain. The highest Late sands; (2) transgressive systems tract (TST) silts; and Cretaceous global sea levels occur close to the (3) upper regressive, highstand systems tract (HST) Cenomanian–Turonian boundary (Miller et al., sands (Table 1). The revised stratigraphic packaging 2005; Haq, 2014), producing the prodelta and presents a more-detailed portrayal of the sand-prone shelfal sediments of the onshore Raritan and Bass units in the regressive LST and HSTs and offers a River Formations (Browning et al., 2008) that cap predictive framework for reservoir sand and confining the underlying, sand-prone fluvial sediments of the shale-prone capstone units associated with the major Potomac Formation. This onshore interval of Raritan flooding surfaces. These three depositional sequences and Bass River Formation shales corresponds to are best observed in the NBCT on a well log transect the Dawson Canyon Formation offshore, similarly between the Great Stone dome (GSD) and the OCS capping the deltaic sands of the LC sequences fi (Miller et al., 2018). In this location, they were ap- identi ed by Miller et al. (2018). Additionally, parently deposited within estuarine, delta front, and Maastrichtian strata are deposited during a second long-term rise in sea level that ultimately culmi- prodelta paleoenvironments (Miller et al., 2018) that nates with a regional sea-level peak in the Eocene follow a first-order deepening upsection. These off- (Miller et al., 2005). The strata deposited during shore sands are coeval with the predominantly terres- this base level increase also comprise a major con- trial Potomac Formation sequences of the mid-Atlantic fining unit both onshore (the composite confining Coastal Plain (Table 1; Miller et al., 2004, 2017, 2018; unit of Zapecza, 1989) and offshore (Table 1). Browning et al., 2008). The Upper Cretaceous of the BCT is charac- terized by a number of transgressive-regressive cycles METHODS that have been subdivided into depositional se- quences on the mid-Atlantic Coastal Plain, where Overview sedimentation was predominantly deltaic, with some terrestrial and nondeltaic marine sedimentation during We evaluate the subsurface of the SBCT using seis- major sea-level lowstands and highstands, respectively mic profiles and well data (sequence stratigraphy, (Miller et al., 2004; Browning et al., 2008). Offshore, biostratigraphy, geophysical logs, and velocity sur- the section is largely comprised of the shaly Dawson veys) to document the spatiotemporal variations of Canyon equivalent (Table 1; Libby-French, 1984). A Cretaceous strata. Seismic stratigraphic principles single regionally persistent sandstone unit, the Middle (Mitchum et al., 1977b) were applied to subdivide Sandstone, was identified by Libby-French (1984) the strata into relatively conformable stratigraphic that ranges from Coniacian to Campanian in age, ac- packages. The well data provided chronological cording to biostratigraphic reports associated with wells constraints using biostratigraphy for these seismic

Table 1. Correlation of the Variety of Names Used to Identify the Cretaceous Strata in the Baltimore Canyon Trough

Abbreviations: Bass R. = Bass River; DCx = Dawson Canyon “x”;HST= highstand systems tract; LC = Logan Canyon; LK = Lower Cretaceous; Marshal. = Marshaltown; Miss. = Missisauga; MK = mid-Cretaceous; SBCT = southern Baltimore Canyon trough; UK = Upper Cretaceous. *Marshaltown, Englishtown, Merchantville, Cheesequake, Magothy, and Bass River onshore sequences.

416 Cretaceous Fluvial-Deltaic Sequences of the Southern Baltimore Canyon Trough surfaces. Integrating the two data sources allowed for consisting of 36- to 240-fold MCS reflection profiles the extrapolation of chronological information available (Table 2). These data included 19,000 km (7300 mi2) at the drill sites across the more-extensive seismic grid. of trackline that cover approximately 62,000 km2 Specifically, we integrated the biostratigraphy (~24,000 mi2) of the margin, from which we used and well-log correlations with a sequence stratigraphic approximately 4400 km (~2700 mi), spanning framework established for the NBCT (Miller et al., 12,000 km2 (4600 mi2). The data were provided in 2018) to build a chronological framework of the standard Society of Exploration Geophysicists “Y” SBCT. Gamma-ray logs were used for correlation to single-trace common midpoint format and imported the LC sequences identified by Miller et al. (2018) onto a workstation running IHS Kingdom® software. and provided lithological information to supplement Processing applied to most surveys (and some lines the seismic stratigraphy. Velocity surveys facilitated within individual surveys) included standard filtering, the integration of the wells with the seismic profiles. spiking deconvolution, velocity analysis, normal moveout, Moreover, information from onshore wells, includ- stacking, and migration. ing deep wells on the Maryland coastal plain, were TheNAMSSseismicdatahaveapeakfrequency juxtaposed with the seismic interpretation. Similar to of approximately 14 Hz at all depths, whereas the the offshore wells, the onshore wells provided chro- spectral density in the 20- to 40-Hz bandwidth de- nological constraints through biostratigraphy (Anderson, creases with depth. Within the interval of interest for 1948; Doyle, 1982) and previous sequence strati- this study, the calculated seismic resolution (defined by graphic analyses (Miller et al., 2017) that were a quarter of a wavelength and calculated from peak calibrated to the Geological Time Scale 2012 frequency and interval velocity from check shot (Gradstein et al., 2012). data) ranges from 30 m (100 ft) at the top to 70 m Once the sequence stratigraphy was defined, (230 ft) in the deeper part of the section (~3000 m structural contour maps, isopach maps, and Wheeler [~9800 ft]). (age–distance) diagrams were constructed from the seismic data to evaluate the movement of sedimentary Mapping Seismic Stratigraphic Surfaces depocenters through the Cretaceous. The stratigraphic architectures within the depositional sequences allowed The seismic sections were mapped to subdivide the for assessment of lithological variations through this record into depositional sequences according to interval. procedures described originally by Mitchum et al. (1977b) and Mitchum and Vail (1977). These pro- Seismic Stratigraphy cedures emphasize the primary importance of con- cordant and discordant relationships, expressed by Data reflector terminations: downlap, onlap, toplap, or erosional truncation (e.g., Mitchum et al., 1977b; Seismic Data—The seismic data used were collected Vail, 1987; Catuneanu, 2006; Neal and Abreu, 2009; for hydrocarbon exploration purposes within the Abreu et al., 2010; Miller et al., 2018). Reflector BCT by a number of exploration companies from terminations objectively delineate significant regional 1975 to 1982. They were recently released as a part unconformities that result from erosion or nonde- of the National Archive of Marine Seismic Surveys position and bind the relatively conformable, ge- (NAMSS) packaged by Triezenberg et al. (2016). netically related units that comprise depositional Two groups of surveys from this database were used, sequences (Mitchum et al., 1977b). A secondary

Table 2. Seismic Surveys Used to Map the Cretaceous Strata of the Southern Baltimore Canyon Trough

Survey Originator Year Lines (Used) Distance (Used)

B-01-75-AT 1975 Offshore Atlantic Group Seismic Survey 1975 149 (52) 11,734 km [7291 mi] (2800 km [1740 mi]) B-16-76-AT 1976 Offshore Atlantic Group Seismic Survey 1976 80 (18) 6728 km [4181 mi] (1600 km [994 mi])

Surveys were collected for exploration purposes and made available to the public within the National Archive of Marine Seismic Surveys (Triezenberg et al., 2016).

SCHMELZ ET AL. 417 interpretation characterizes packages of strata as pro- to 3 wells (Maryland Esso 1, Mobil Bethards, and Ohio grading or retrograding, typically using the clinoform Oil Hammond) drilled onshore on the Maryland coastal rollover (the change in gradient that partitions the plain (Table 3). We used three types of data from these topset and the foreset) as a geomorphological bench- wells: (1) biostratigraphy to provide chronological mark. The newly released data allow imaging of mid- constraints; (2) geophysical logs to interpret transitions Cretaceous clinoforms in the SBCT for the first time. in lithologies and depositional environments and to Prograding strata are assigned within the upper re- generate synthetic seismograms to make well–seismic gressive HST or lower regressive LST, whereas trans- ties; and (3) velocity surveys to export the biostrati- gressive surfaces are placed within the TST. Assigning graphic and lithological data to the seismic section and stratal packages within the depositional system and test accuracy of the regional time–depth conversion. systems tract framework and identifying their strati- Biostratigraphic data were obtained from technical graphically significant bounding surfaces (Brown and reports and paleontological summaries produced by the Fisher, 1977; Posamentier et al., 1988; Abreu et al., (since-dissolved) US Department of the Interior Minerals 2010) has implications for predicting depositional Management Service (MMS) or other published work environments and lithological characteristics within (e.g., Anderson, 1948; Doyle, 1982; Poag, 1985b). The a depositional sequence (Mitchum et al., 1977a; sources of biostratigraphic picks are provided in Table 3. Mitchum and Vail, 1977; Sangree and Widmier, The gamma-ray (GR) well log was used as the 1977; Posamentier and James, 1993). preferred lithological indicator, but the spontaneous Following the procedure of Vail (1987) and Abreu potential (SP) log was used if GR was not available. et al. (2010), we (1) identified reflection terminations The SP log was used to represent lithology in the by type: onlap, downlap, toplap, or truncation; (2) Maryland coastal plain wells. Sonic logs and bulk identified stratal configurations as progradational, ret- density were available for each of the OCS wells used rogradational, or aggradational; (3) mapped strati- herein. The velocity surveys were primarily used to graphically significant bounding surfaces; (4) assigned convert log measurements and biostratigraphic picks biostratigraphic ages; and (5) inferred depositional from depth to two-way traveltime (TWTT). environment and lithology based on all of the above.

Regional Time–Depth Conversion for Well Stratigraphy and Integration with Integration of Onshore Drill Site Data Seismic Data with Seismic Sections and Creation of Isopach Maps Well Data We used data from 5 of 32 shelf exploration wells drilled The conversion of TWTT to depth is required to off the mid-Atlantic Coast from 1978 to 1984, in addition generate structural contours and isopach maps in

Table 3. Locations of the Wells in the Southern Baltimore Canyon Trough Used for This Study, Geophysical Logs Available for Each Site, and Citations for the Biostratigraphy

Well Latitude Longitude Total Depth, m (ft) Biostratigraphy Velocity Survey Well Logs

Maryland Esso 1 38.408 -75.063 2346 (7697) Doyle (1982) — SP, SN Mobil Bethards 38.304 -75.487 2195 (7201) Doyle (1982) — SP, SN Ohio Oil Hammond 38.346 -75.275 1665 (5463) Doyle (1982) — SP, SN COST B-2 39.375 -72.734 4890 (16,043) Scholle (1977) Yes GR Mobil 17-2 38.968 -73.049 4260 (13,976) Steinkraus (1981) Yes GR, RHOB, DT Shell 272-1 38.702 -73.540 4114 (13,497) Steinkraus et al. (1985b) No GR, RHOB, DT Shell 273-1 38.716 -73.456 5334 (17,500) Steinkraus and Bebout (1985) Yes GR, RHOB, DT Tenneco 495-1 38.466 -73.378 5578 (18,301) Steinkraus et al. (1985a) Yes GR, RHOB, DT

The Maryland Esso 1, Ohio Oil Hammond, and Mobil Bethards wells were drilled in the 1940s on the eastern Maryland coastal plain. The Mobil 17-2, Shell 272-1, Shell 273-1, and Tenneco 495-1 wells were drilled in 1978 and 1979 on the outer continental shelf offshore southern New Jersey. Abbreviations: COST = Continental Offshore Stratigraphic Test; DT = sonic; GR = gamma ray; RHOB = bulk density; SN = short normal resistivity; SP = spontaneous potential.

418 Cretaceous Fluvial-Deltaic Sequences of the Southern Baltimore Canyon Trough meters. We initially used the velocity–depth grids generated by Klitgord et al. (1994), as described in Appendix 1. We supplemented their database to incorporate a more-detailed time–depth profile for the upper second (TWTT) of sediments generated by Mountain and Monteverde (2012) based on seismic data acquired from the research vessel (R/V) Oceanus expedition OC270 and well data from IODP expedition 313 (Figure 2). The accuracy of the conversion of the seismic data (TWTT) to depth was checked against values obtained from OCS well velocity surveys at the well locations (check shot surveys and velocity logs).

Study Area

The northern half of the study area is off the coast of southern New Jersey (Figure 1), containing the B-01-75-AT (survey nomenclature from NAMSS; Triezenberg et al., 2016) seismic survey and the offshore wells used for chronological control. The southern part of the study area contains the B-16-76- AT seismic survey, with control provided by deep stratigraphic test wells on the Maryland coastal plain. We projected the depth-registered chronological control from the wells in the northern part of the – – study area onto the B-01-75-AT seismic grid. Strati- Figure 2. (A) One of 4400 resampled time depth (T D) curves that comprise a three-dimensional (3-D) T–D model used to convert graphic horizons were then interpreted across this seismic profiles and interpretations on them from a reference in grid to establish a seismic stratigraphic framework. two-way traveltime (TWTT) to depth. The resampled curve is Lines from the B-16-76-AT data that intersected the an interpolation at 0.1-s intervals of data from Mountain and B-01-75-AT grid were used to correlate seismic data Monteverde (2012) in the uppermost second of sediments and the from the northern to the southern subareas. Seismic Geophysical Database of the United States North Atlantic Margin stratigraphy in the southern subarea was correlated (Klitgord et al., 1994) below 1 s. The accuracy of the depth model at to the Maryland wells onshore by extrapolating the a given location, represented by the thin red lines bounding the dips of onshore sequence boundaries (i.e., Miller interpolated curve, was determined through a statistical comparison of the T–D model to check shot data at well locations. (B) All of the 3-D et al., 2017) seaward. This required a projection of T–D model data points from locations within the Baltimore Canyon approximately 20 km (~12 mi) from the most sea- trough (BCT) with less than 200 m (<650 ft) of water depth. The ward well to the most landward traces of the seismic average T–D relationship for the continental shelf within the BCT is sections. plotted with a solid black line, and the dotted black lines bound 95% of In the area offshore southern New Jersey, well– the variation at each TWTT interval, approximated by two standard seismic ties to survey B-01-75-AT were made at the deviations (2s) from the mean. mbsl = meters below sea level. Mobil 17-2, Shell 272-1, Shell 273-1, and Tenneco 495-1 OCS exploration wells (Figure 3). The data interpretations and geophysical log data facilitated from these sites were projected onto the nearest correlations among Shell 272-1, Shell 273-1, Mobil seismic profiles from B-01-75-AT using the correla- 17-2, and Tenneco 495-1, guided by the seismic tions of acoustic travel time to depths established profiles that connect the wells (Figures 4–6). The with check shot surveys at each of these wells. Pro- seismic sections were interpreted using stratal ter- jected onto the seismic data, the MMS biostratigraphic minations to identify the significant seismic surfaces.

SCHMELZ ET AL. 419 Figure 3. Well-log correlation of Continental Offshore Stratigraphic Test (COST) B-2, Mobil 17-2, Shell 272-1, and Shell 273-1. Well-log correlations of Logan Canyon (LC) 1, 2, and 3 from the well-log transect of Miller et al. (2018) are plotted as red, dashed lines. The seismic surfaces mid-Cretaceous (MK) 1, 2, and 3 were converted to depth using the velocity surveys associated with the individual wells and plotted onto the well logs as dashed blue lines. The lithostratigraphic Sable Shale and Naskapi Shale units of Libby-French (1984) are identified by blue and purple shading, respectively. Boundaries of biofacies zones in the COST B-2 and Shell 272-1 wells (Sikora and

420 Cretaceous Fluvial-Deltaic Sequences of the Southern Baltimore Canyon Trough Biostratigraphic age assignments that bound or fall below 1.75 s of TWTT (~1750 m [~5750 ft] depth within these sedimentary packages provide prelimi- below mean sea level [MSL]). The underlying LK2– nary chronological control. In all, 7 distinct horizons LK1 interval is composed of shallow-water heterolithic and stratal packages were mapped, within 10 bio- siliciclastics of the Missisauga Formation equivalent, stratigraphic datum levels. overlying the similarly heterolithic Upper Jurassic To the south, two shelf-parallel seismic lines, Mic-Mac Formation (Poag, 1985b). As a result, the ma-005 (Figure 7) and ma-007 of survey B-16-76-AT Jurassic–Cretaceous boundary occurs within hetero- tie seven dip lines to the northern B-01-75-AT grid lithic strata within Shell 272-1 and Shell 273-1 (Figure1).Diplinema-004(Figure8)originatesap- (Figure 4) and Mobil 17-2 (Figure 5), exemplified by proximately 20 km (~12 mi) seaward of Ocean City, a serrated GR log expression. The upper bounding Maryland, and was projected onshore (Figure 9). Dip surface (LK1) is the basal surface to the LC3 se- lines ma-004 and ma-006 were used to generate quence of Miller et al. (2018) and is more prominent Wheeler diagrams (Figures 10, 11) to evaluate changes and traceable than the underlying LK2 reflection. in Cretaceous depocenters on the Maryland shelf. The LK2 reflection is associated with the top Ti- Finally, we combined the stratigraphic interpretations thonian biomarker in the OCS wells. Because of the on the northern B-01-75-AT and southern B-16-76-AT discontinuous nature of the LK2 reflection and re- seismic grids with sequence stratigraphic interpreta- flections contained within this sequence, the identifica- tions made on the coastal plain (Miller et al., 2017) to tion of this surface relied heavily on reregistering the generate basin-scale sequence stratigraphic maps. correlations at each well. Coherently stacked internal reflections representing systems tracts are difficult to identify. On the OCS, the LK2 package is also highly RESULTS faulted, which obscures the genetic basis for terminating reflections (i.e., terminations may be because of dis- Upper Jurassic and Lowermost Lower placed strata rather than by erosion or nondeposition). Cretaceous (Tithonian–Berriasian to A best effort was made to loop correlate LK2 around Barremian) the seismic grid; however, tracings of this deepest stratigraphic surface have the greatest uncertainty of Two distinct reflections, Lower Cretaceous (LK) 2 all those prepared in this study (see Appendix 2). and LK1, were traced throughout the northern The Missisauga Formation (Libby-French, 1984), subarea with varying confidence. These surfaces which comprises the 107-year scale LK2 sequence bound the LK2 sedimentary package (named for its between LK2 and LK1 (Table 1), is likely estuarine basal reflection; Table 1). The LK2 reflection corre- to fluvial coastal plain deposits. The offshore LK2 sponds with the top of the Jurassic strata, and the LK1 sequence correlates to the Waste Gate Formation reflection (Miller et al., 2018) falls beneath a Bar- onshore, based on biostratigraphic age estimates remian biostratigraphic marker in the offshore wells (Doyle, 1982) and our correlations (Figure 9). At Ten- and an unconformity in the seismic data in the NBCT neco 495-1 (Figure 6), the Lower Cretaceous interval and is a likely mid-Barremian hiatus (Miller et al., 2018). is no longer overlain by shale but rather by an interval This interval between the Barremian unconformity of low-GR values interpreted to be carbonate, po- (LK1) and the approximate Jurassic–Cretaceous tentially correlative to the Aptian “blanketlike facies” boundary (LK2) is generally 600–1000 m (~2000–3300 identified by Edson (1987a) in Shell 586-1 that ft) thick on the shelf (Figure 12). The LK1 reflection lies consists of limestone and sporadic interbedded shale

Figure 3. Continued. Olsson, 1991) are plotted and identified with blue lines and text on the y-axes. The parasequences of COST B-2 are represented by black arrows that point in the fining direction, whereas the large yellow triangles represent coarsening-upward para- sequence sets, and the blue triangles represent those that are fining upward. The biostratigraphy is compiled from technical reports associated with each well (Table 3). The biofacies indicate a deepening paleodepth within LC2 from COST B-2 to Shell 272-1. Paleodepth of biofacies are as follows: (1) biofacies 3: 60–130 m (200–430 ft); (2) biofacies 4ac: 130–600 m (430–2000 ft); and (3) biofacies 5 overlaps and replaces 4ac: 130–600 m (430–2000 ft) (Sikora and Olsson, 1991). The log depths are plotted in feet below kelly bushing (ftbKB), and a metric scale is provided for reference. SB = sequence boundary.

SCHMELZ ET AL. 421 Figure 4. Seismic correlation between the Shell 272-1 and Shell 273-1 wells, located on the outer continental shelf offshore southern New Jersey, using the a-207 and a-142 seismic profile lines of the B-01-75-AT seismic survey. The biostratigraphic picks within the wells are used to assign preliminary ages to the mapped seismic packages. A fault surface is mapped in light blue, near the intersection of lines a-207 and a-142. TWTT = two-way traveltime. deposited on a carbonate platform. Poag (1985b) Similar to the northern subarea, it is difficult correlated the Missisauga in Continental Offshore to trace the lowest stratigraphic horizons around Stratigraphic Test (COST) B-2 and COST B-3 to the B-16-76-AT seismic grid offshore Maryland the Potomac I sequence at Island Beach. However, in the southern subarea. The horizon attributed to the based on correlations to onshore Maryland (Figure 9; Jurassic–Cretaceous boundary (LK2) is particularly Miller et al., 2017), we favor correlation of the Mis- challenging to trace across the 100-km (62-mi) ma- sisauga to the onshore Waste Gate Formation. 005 seismic line (Figure 7). The likely heterolithic Offshore New Jersey, this package is influenced by the character of the strata and low vertical resolution “Gemini” fault system (Poag, 1987). (>50 m [>160 ft]) of the seismic data contributes to

422 Cretaceous Fluvial-Deltaic Sequences of the Southern Baltimore Canyon Trough Figure 5. Seismic correlation between the Shell 272-1 and Mobil 17-2 wells, located on the outer continental shelf offshore southern New Jersey, using the a-207, a-136, a-183, and a-150 seismic profile lines from the B-01-75-AT seismic survey. The biostratigraphic picks within the wells are used to assign preliminary ages to the mapped seismic packages. TWTT = two-way traveltime. the chaotic reflections in this interval. On the dip lines, converted ma-004 profile (Figure 9). The projection these strata produce similarly chaotic reflections, with of this basal Cretaceous surface seaward results in a few clearly terminating reflections (Figure 8). The LK1 correlation with LK2 that is within approximately reflection, at the top of the Missisauga Formation and 200 m (~650 ft) of the seismic reflection (at ~2400– the base of the LC3 sequence, is easier to trace because 2600 m [~7900–8500 ft] below MSL; Figure 9). it, and the reflections above it, are more continuous. Projecting the top of the Waste Gate Formation, Having traced these horizons around the B-16- which has a pre–Zone I Early Cretaceous palyno- 76-AT grid offshore Maryland, we correlated the logical age (Doyle, 1982) identified in the well cross Lower Cretaceous reflections to the wells located section offshore, correlates very well with the LK1 onshore Maryland. The Jurassic–Cretaceous bound- reflection. Therefore, the “blocky” log pattern typical ary identified in the cross section of onshore wells of fluvial sands and interbedded varicolored shales (Miller et al., 2017) was projected onto the depth- that characterize the onshore Waste Gate Formation

SCHMELZ ET AL. 423 Figure 6. Seismic correlation between the Shell 272-1 and Tenneco 495-1 wells, located on the outer continental shelf offshore southern New Jersey, using the a-217, a-132, and a-207 seismic profile lines from the B-01-75-AT seismic survey. The biostratigraphic picks within the wells are used to assign preliminary ages to the mapped seismic packages. TWTT = two-way traveltime.

(Hansen, 1982; Andreasen et al., 2015; Miller et al., 2017) (Miller et al., 2017). The LK2 strata that correlate with transition downdip to the heterolithic shallow-water the Waste Gate Formation similarly thickens (>1km siliciclastics of the Missisauga Formation in the [>3300 ft]) to the southeast (basinward) and to the offshore. southwest (along strike) on the continental shelf The Lower Cretaceous sequences are thicker in (Figure 12). the southern subarea than they are in the northern subarea, both onshore and on the continental shelf (Figure 12). The onshore Waste Gate sequence is Middle Cretaceous (Aptian to Lower approximately 500 m (~1600 ft) thick at the modern- Cenomanian) day shoreline on the Maryland coastal plain and thins along strike to the northeast, apparently pinching This interval is bound by the LK1 and mid-Cretaceous out south of the Island Beach well in New Jersey 1 (MK1) seismic sequence boundaries and is composed

424 Cretaceous Fluvial-Deltaic Sequences of the Southern Baltimore Canyon Trough Figure 7. Correlation of Cretaceous seismic surfaces along strike on the outer continental shelf using seismic line ma-005 from survey B-16-76-AT. This line connects the northern subarea (offshore from southern New Jersey) and seismic survey B-01-75-AT with the southern subarea (offshore Maryland) and the rest of seismic survey B-16-76-AT. The stratal character of the reflections transition from chaotic to parallel and subparallel upwards through the section. TWTT = two-way traveltime. of the LC3 (equivalent to the LK1 seismic se- via COST B-2 (Figure 3). Projecting the well-log quence), LC2 (equivalent to the mid-Cretaceous 3 correlations onto the seismic data, the three LC se- [MK3] seismic sequence), and LC1 (equivalent to quences (LC3, LC2, and LC1) on the well logs cor- the mid-Cretaceous 2 [MK2] seismic sequence) relate with identified seismic surfaces (MK3, MK2, sequences (Table 1) identified to the north of the and MK1, respectively). This correlation is supported study area by Miller et al. (2018). The GR logs and by seismic tracing between COST B-2 (where these biostratigraphy from the Shell 272-1, Shell 273-1, sequences were originally identified by Miller et al., and Mobil 17-2 wells allow for the direct correlation 2018) and Mobil 17-2 (Baldwin et al., 2017) and by to the stratigraphic sequences of Miller et al. (2018) biostratigraphic constraints. The entire interval between

SCHMELZ ET AL. 425 Figure 8. Seismic stratigraphic interpretation of Cretaceous strata using seismic profile ma-004, a dip line that originates approximately 20 km (~12 mi) downdip from Ocean City, Maryland, and extends southeast to the continental rise. The seismic surfaces were correlated to the northern subarea using line ma-005. TWTT = two-way traveltime. the LK1 and MK1 reflections in the northern subarea is (Miller et al., 2018), and the top of the Missisauga between 400 and 800 m (1300–2600 ft) thick on the Formation. The package of strata above this surface OCS, where it thins from north to south. The upper- is largely undifferentiated shale at Shell 272-1 and most MK1 surface is between 1000 and 2250 m Shell 273-1 (Figures 3, 4); the GR log within this (3300–7380 ft) below MSL. The mid-Cretaceous interval records only minor variations, and the ben- interval is thicker in the middle shelf, whereas the thic foraminiferal biofacies of Sikora and Olsson LowerCretaceousstratathickensignificantly off- (1991) indicates paleodepths of 130 to more than shore toward the shelf edge and slope. 200 m (430 to >650 ft). The base of the LC2 se- The base of the LC3 sequence is reflection LK1, quence within the Shell 272-1 and 273-1 wells oc- the middle Barremian seismic sequence boundary curs within this deep-water facies that comprises

426 Cretaceous Fluvial-Deltaic Sequences of the Southern Baltimore Canyon Trough Figure 9. Correlation of the Potomac and Waste Gate sequences (Miller et al., 2017) from a well-log cross section composed of the three deep stratigraphic test wells on the Maryland coastal plain, interpreted by Miller et al. (2017), to the ma-004 seismic section (Figure 8) converted from two-way traveltime to depth using a splice of the geophysical database of the United States North Atlantic margin (Klitgord et al., 1994) and the time–depth function of Mountain and Monteverde (2012) (in the upper 1 s of sediments). The onshore Potomac and Waste Gate sequences (Miller et al., 2017) are projected onto the seismic line through extrapolation of a spatial linear regression through the sequence picks in each of the wells across the transect. The resulting correlation matches the Potomac and Waste Gate sequence boundary surfaces with the Cretaceous seismic sequences interpreted herein. ftbKB = feet below kelly bushing; MSL = mean sea level; OCS = outer continental shelf; RES = resistivity. S HEZE AL. ET CHMELZ 427 Figure 10. Uninterpreted and interpreted seismic profiles and Wheeler diagram for mid-Cretaceous depositional sequences identified on line ma-004 (Figure 8). Units bound by the reflections marked with blue lines are assigned to the lowstand systems tract (LST), green lines to the transgressive systems tract (TST), and yellow lines to the highstand systems tract (HST). The red lines are sequence boundaries (SB). LK = Lower Cretaceous; MK = mid-Cretaceous; UK = Upper Cretaceous.

428 Cretaceous Fluvial-Deltaic Sequences of the Southern Baltimore Canyon Trough Figure 11. Uninterpreted and interpreted seismic profiles and Wheeler diagram for mid-Cretaceous depositional sequences identified on line ma-006, which is 10 km (6 mi) south of ma-004. Units bound by the reflections marked with blue lines are assigned to the lowstand systems tract (LST), green lines to the transgressive systems tract (TST), and yellow lines to the highstand systems tract (HST). The red lines are sequence boundaries (SB). LK = Lower Cretaceous; MK = mid-Cretaceous; UK = Upper Cretaceous.

SCHMELZ ET AL. 429 430 rtcosFuilDlacSqecso h otenBlioeCno Trough Canyon Baltimore Southern the of Sequences Fluvial-Deltaic Cretaceous

(A) (B) (C)

Figure 12. Isopach and contour maps of the uppermost Upper Cretaceous (UK) (A), mid-Cretaceous (MK) (B), and lowermost Lower Cretaceous (LK) (C) intervals in the southern Baltimore Canyon trough. These maps were created from the depth-converted seismic stratigraphic surfaces interpreted herein, and coeval depositional sequences correlated between 19 wells on the Mid-Atlantic Coastal Plain (Miller et al., 2017). The arrows indicate likely sediment source regions and pathways of sediment transport. nearly the entirety of the LC2 and LC3 sequences. A pattern that overlies the top of the LC1 sequence conservative approximation of the basal LC2 boundary and coincides with a Cenomanian biomarker (Figure 3). relies on the biostratigraphic pick for the top of (or This surface produces a high-amplitude reflection in within) the Aptian in the Shell 272-1 and Shell 273-1 the seismic data (MK1) that can be loop correlated wells (Figure 3). Seismic correlation of MK3, the through the study area. The LC1 is a thin interval seismic LC2 equivalent of Miller et al. (2018), be- within this subarea (~75–150 m [~250–490 ft] thick), tween COST B-2 and Mobil 17-2 (Figure 3; Baldwin and the bounding upper surface (MK1) is character- et al., 2017) supports this correlation within the ized by downlap. limitations of seismic resolution. The biofacies tran- Farther south, offshore Maryland, strata coeval sition between coeval LC2 strata in COST B-2 and with the LC sequences of the NBCT (Table 1; Miller Shell 272-1 (Figure 3; Sikora and Olsson, 1991) in- et al., 2018) have more continuous, curvilinear to parallel dicates a deepening to the south toward the SBCT. or subparallel internal reflections than in the NBCT. Libby-French (1981, 1984) identified a composite This is particularly evident on the dip line ma-004 shale that comprises the majority of the LC2 and LC3 (Figure 8), where clinoforms can be seen in the strata sequences as a single unit (i.e., the purple shading in coeval with the LC3 sequence (bound by the LK1 Figure 3), assigning it as the equivalent to the Scotian and MK3 reflections). We interpret this as reflecting Shelf Naskapi Formation and correlating it to a shale closer proximity to deltaic sources compared to off- of predominantly Barremian age (Scholle, 1977) shore New Jersey. The LC depositional sequences on (compared to the Aptian and Albian age of this in- this part of the shelf appear to contain onlapping terval in Shell 272-1). In contrast, we correlate this transgressive strata that thin landward below vertical interval of shale to the two individual depositional seismic resolution (Figures 10, 11). This creates sequences that become sandy to the north on the composite maximum flooding surface (MFS) and GSD (Miller et al., 2018), based on (1) biostrati- sequence boundary reflections. Upsection, thick re- graphic constraints provided by both well reports gressive HSTs prograde over the thin TSTs and MFSs. and re-evaluation by Miller et al. (2018) and (2) a Three prominent transgressive-regressive pack- depositional model of a deltaic unit transitioning from ages (i.e., above reflections MK3, MK2.1, and MK2) sand to shale, consistent with increasing paleodepths can be identified on the seismic data in the strata at COST B-2 and Shell 272-1 (Sikora and Olsson, 1991). bounded by reflections MK3 and MK1 (Figures The base of the LC1 sequence can be correlated 10, 11). These packages appear to be three distinct from COST B-2 to the southern wells (Figure 3). It seismic sequences on the Maryland continental is marked by the top of an approximately 50-m shelf. The interval between MK3 and MK1 thickens (~160-ft)-thick progradational unit composed of from north to south alongshelf (Figures 7, 12), and the three condensed, coarsening-upward parasequences. internal reflections transition upsection from curvi- Both this study and Libby-French (1984) correlate linear clinoforms to planar and subplanar reflection the boundary at the top of our LC2 to the same geometries. The three sequences are coeval with the approximate surface in the COST B-2 well (i.e., the LC1 and the LC2 sequences on the GSD, with the upper LC sandstone in Libby-French, 1984). Moreover, intervening MK2.1 pinching out to the north, limiting the “top Albian” marker lies just above the identified its identified spatial extent to the shelf offshore sandstones in multiple wells, providing additional Maryland. support for this boundary as the base of LC1. Seis- The Potomac I, Potomac II, and Potomac III se- mically, the base of the LC1 sequence is marked by quences identified in the onshore wells (Miller et al., the MK2 seismic sequence boundary (Figures 3, 4). 2017) correlate to the LK1, MK3-MK2.1, and MK2 The MK1 seismic sequence (equivalent to the sequences in the offshore, respectively (Figure 9; LC1 log sequence) in the southern wells can be Table 1). The top of the Potomac I sequence projects correlated to the MK1 seismic sequence boundary seaward to the top of the LK1 sequence. Moreover, identified by Miller et al. (2018) on the GSD and the geometric projection of the top of the Potomac III OCS (their supplemental online material figure 1). seaward correlates with the top of the MK1 sequence Gamma-ray logs at the Shell 272-1, Shell 273-1, and boundary, which is likely mid-Cenomanian. The Po- Mobil 17-2 wells show a distinctive bell-shaped tomac III base projects seaward to MK2.

SCHMELZ ET AL. 431 The mid-Cretaceous interval (LK1, MK3, and in this northern subarea, this interval is heavily faul- MK2 and MK2.1 in the south) also thickens along ted near the shelf edge (Figure 6). Within the interval strike from northeast to southwest, both in the above reflector MK1 (Figures 3, 4), a local peak GR coastal plain and on the continental shelf (Figure 12). value in the Shell 272-1 GR log (at 2063 meters Coeval with the Potomac Formation in New Jersey below kelly bushing [mbKB] [6770 feet below kelly and the Potomac Group (excluding the Waste Gate bushing {ftbKB}] or 2038 mbsl [6686 ftbsl]) lies Formation) in Maryland, this unit is approximately within a fining-upward trend (1) above the top of 1000 m (~3300 ft) thick under the Maryland coastal well-log sequence LC1, (2) above the converted plain at the shoreline, thinning to approximately depth of seismic surface MK1, and (3) above the 300 m (~980 ft) updip between Cambridge, Mary- highest occurrence (HO) of Rotalipora cushmani that land, and Fort Mott, New Jersey. At the northern- likely corresponds to the GMSL high coincident most limit of our study area, 20 km (12 mi) south of with ocean anoxic event (OAE) 2 at or near the Island Beach, New Jersey, the Potomac Formation is Cenomanian–Turonian boundary (Kennedy et al., approximately 400 m (~1300 ft) thick. This interval 2005). The GR peak also lies within an interval of is 800–1000 m (2600–3300 ft) thick on the middle maximum upper Albian to upper Turonian paleo- to outer shelf off the coast of Maryland, compared to depth in Shell 272-1 and in COST B-2 indicated by 500–800 m (1600–2600 ft) thick on the southern the biofacies of Sikora and Olsson (1991) (Figure 3). New Jersey Shelf. Compared to the underlying LK2 The LC1 (equivalent to the MK2 seismic) sequence is strata, the mid-Cretaceous depositional sequences older than the OAE 2 and supports the correlation of were deposited in a more-proximal location, with a MK2 to the early to mid-Cenomanian Potomac III se- depocenter near the modern shoreline (Figure 12). quence onshore. The overlying uppermost Cretaceous seismic package is bounded by the top Turonian reflection Upper Cretaceous (Upper Cenomanian to (UK2) at its base and the prominent top Campanian Maastrichtian–Campanian) reflection (UK1 of Miller et al., 2018 and “Red” of Greenlee and Moore, 1988 and Greenlee et al., Two Upper Cretaceous horizons are identified on 1992, their “top Cretaceous”) at its top. Toplapping the seismic data, UK2 and UK1. They correlate or truncating internal reflections are associated with roughly to the top Turonian and the top Campanian the top Campanian (UK1) surface that generally co- biomarkers in the OCS wells, respectively (Table 1). incides with a Campanian biomarker in the OCS The Upper Cretaceous interval lies above the LC wells (Figures 5–8). In general, the Maastrichtian sequences and is 200–600 m (650–2000 ft) thick on strata on the shelf are quite thin (<30 m [<100 ft]; the shelf (Figure 12). The upper UK1 surface lies Olsson and Wise, 1987), and, therefore, Greenlee between 1000 and 1750 m (3300–5700 ft) below et al. (1992) equated the top Campanian with the top MSL offshore New Jersey. The Upper Cretaceous Cretaceous. This Campanian interval can likely be thickens significantly toward the shelf edge and subdivided again, especially because it thickens in the coarsens upward from the basal Dawson Canyon vicinity of Mobil 17-2 and seaward. There are Con- Formation to the Campanian Middle Sandstone iacian and Santonian biomarkers in the wells that Formation of Libby-French (1984). might indicate additional unconformities noted on- In addition to the Middle Sandstone, a sand within shore by Olsson et al. (1988) and Miller et al. (2004) the MK1 sequence appears in the well-log data at that are not resolved in our seismic stratigraphic data. Shell 272-1 (at 2000 m [6500 ft] in Figure 4) that In the southern subarea offshore Maryland, shales out toward the more-distal Shell 273-1 and shingled, offlapping strata that are bound at the Mobil 17-2 wells (Figures 4, 5). The interval above base by MK1 and at the top by UK2 are probably reflector MK1 consists of high-relief clinoforms that Cenomanian to mid-Turonian and directly overly the prograde across the OCS (Figure 6). These internal LC equivalents (Figure 8). On strike line ma-005, this structures make the upper bounding surface (UK2) interval is a package of downlapping, curvilinear reflec- easy to identify via toplapping and downlapping in- tions that generally dips gently to the north and thins to ternal reflections. Like the other Cretaceous packages the south (Figure 7). This interval is approximately

432 Cretaceous Fluvial-Deltaic Sequences of the Southern Baltimore Canyon Trough 500 m (~1600 ft) thick on the outer half of the New Lower Cretaceous (e.g., Figures 6, 7); (2) thin, poorly Jersey Shelf but only 100–300 m (330–980 ft) thick defined clinoform geometries (e.g., Figure 6) in the off the coast of Maryland where most of the internal mid-Cretaceous; and (3) progradational packages reflections dip to the south. Moreover, there are two within a generalized regression in the Upper Creta- distinct seismic lobes (inferred to be deltaic lobes), ceous section that are capped by the high-amplitude separated by a potential onlap surface on line ma-005 UK1 surface (Figure 6). The offshore well data from (at ~1.75 s in the center of the image in Figure 7). Shell 272-1, Shell 273-1, and Mobil 17-2 (Figure 3) This interval correlates to a thin Upper Cretaceous reveal (1) heterolithic facies in the Lower Cretaceous, section on the Maryland coastal plain that thickens to (2) a transition from shallow marine sandstone and the north into New Jersey (Owens and Gohn, 1985; shale parasequences in the mid-Cretaceous, and (3) thin Olsson et al., 1988; Miller et al., 2017). On the Upper Cretaceous sandstones deposited between Wheeler diagrams (Figures 10, 11), the strata bound predominantly offshore (>30 m [>100 ft] paleo- by MK1 and UK2 show a distinct LST and TST and a depth) shales. We infer that these sedimentary prograding HST. The HSTs of this sequence mark a rocks were deposited within deltaic or deltaically return of regressive sedimentation to the outermost influenced environments at a variety of water depths. continental shelf. The top Campanian (UK1) can be traced con- From oldest to youngest, they are as follows. fidently from the northern subarea to the south as a high-amplitude reflection, with overlying reflections 1. The heterolithic strata of the Lower Cretaceous terminating onto it as downlap. Similar to the northern are terrestrial silty, lignitic, coal-bearing shales fl subarea, the interval below this horizon thickens sea- (Poag, 1985b) to terrestrially in uenced coarsening- ward toward the shelf edge (Figures 8, 12). Seismically, upward packages of delta-front sandstones (Libby- this Upper Cretaceous package bound by UK2 and French, 1984) deposited beneath the modern OCS; UK1 appears to thin and seismically pinch out in the these heterolithic strata lie landward of the thick, landward direction before reaching the Maryland deltaic Neocomian section of interbedded siltstone coastal plain. There is also significant toplap or trun- and fine-grained sandstone containing glauconite, cation associated with the bounding upper reflection oxidized pyrite, and red shales in the Shell 93-1 (UK1). However, considering that the vertical seis- well drilled on the continental rise offshore Maryland mic resolution is approximately 35 m (~115 ft), (Amato, 1987). there may be a veneer of undetected sediment. To- 2. The mid-Cretaceous strata are interpreted to be gether, the combined MK1 and UK2 strata thin from primarily deltaic, delta-front to prodelta deposits ~ approximately 300 m ( 980 ft) thick at the shoreline sections in the NBCT on the basis of well-log in the New Jersey coastal plain to approximately character and lithological characteristics (Miller ~ 100 m ( 330 ft) thick in Maryland. The thicknesses et al., 2018). Additionally, the shale-prone Aptian on the OCS thins similarly, along strike, from 700 m and Albian strata of the northern SBCT were in- (2300 ft) in the northern subarea to 400 m (1300 ft) terpreted to be part of a prograding deltaic suc- in the southern subarea. cession by Libby-French (1981). Although much of the onshore mid-Cretaceous section consists of nonmarine fluvial deposits, portions of the correl- DISCUSSION ative Albian and Cenomanian Potomac Formation sediments drilled in New Jersey are deltaic or influ- Spatiotemporal Variations in Depositional Environments enced by deltaic processes (Browning et al., 2008). 3. The Upper Cretaceous sediments drilled on- Wells provide a lithological ground truth for seismic shore in New Jersey are also deltaic or deltaically facies in the northern subarea that were seismically influenced (Browning et al., 2008), and it would traced to the southern subarea that lacks shelf wells. follow that the predominately shale-prone in- In the seismic data of the northern subarea of the tervals seaward of these onshore sediments were SBCT we observe (1) chaotic seismic facies in the deposited in a prodeltaic to deltaically influenced

SCHMELZ ET AL. 433 shelf environment that is likely to be exclusively evidence of postdepositional deformation from faults marine downdip. (cf. Figures 6, 7). From the base of the mid-Cretaceous (the LK1 surface), the clinoforms flatten gradually The seismic facies of the southern subarea are largely upsection (Figure 8). The shoaling of reflection surface similar to those of coeval facies in the northern slopes may indicate a deepening from the relatively subarea but differ slightly in important ways. In the shallow paleodepths in the Early Cretaceous (Figure 13) southern subarea, the lowermost Lower Cretaceous and a transition from a deltaically influenced shelf interval similarly has a predominantly chaotic seismic environment to middle to outer shelf and deeper character (Figures 7, 8). This would imply a similarly deposits (e.g., between sequence LK1 and sequence heterolithic lithology to the Missisauga Formation of MK2 in Figures 10, 11; Sangree and Widmier, 1977). the LK2 sequence of the northern subarea. A coastal Moreover, the HST of sequence MK2.1 (Figures 8, plain or paralic depositional environment for the 10, 11) is truncated erosionally at the most-landward LK2 interval also fits appropriately with the coeval, point of any mid-Cretaceous sequence, suggesting proximally positioned, fluvial sands of the Waste that a base level fall eroded preexisting topography Gate Formation in the subsurface of the Maryland at an updip location relative to other sequences in coastal plain. Moreover, the Shell 93-1 well captured this succession. This landward position of the most the record of anomalously thick deltaic sedimentation basinal clinoform in sequence MK2.1, as well as the fl in the Neocomian section within a “foundered” part relatively atter geometries, is evidence indicating that MK2.1 contained the most-proximal depocenter of the Upper Jurassic and Lower Cretaceous Abenaki through this section. However, it is more likely that Formation (Poag, 1985b) reefal structure (Amato, the overlying strata, coeval with the onshore Potomac 1987). The LK2 package is thicker in the southern III and Bass River sequences (Table 1), which are than the northern subarea and is thickest toward more deltaic and marine than the underlying Potomac the shelf edge (Figures 7, 8, 12). Formation on the New Jersey coastal plain (Browning The isopach maps (Figure 12) indicate a pre- et al., 2008), were deposited in deeper water with an dominately southern source of sediment during the even more proximal deltaic depocenter that is not pre-Aptian Early Cretaceous. It is possible that the entirely captured within our seismic sections (Figures influx of sediment from the south and high rates of 9, 12). Further, upsection from sequence MK2.1 siliciclastic sedimentation were responsible for the (Figures 8, 10, 11), the clinoforms regain relief and “foundered” carbonate reef described by Amato (1987) display shingled geometries in the internal reflections under the continental rise offshore Maryland. Re- of sequences MK2 and MK1, suggestive of deposition gardless, it is logical that the deltaic Neocomian in a deltaic depocenter. sedimentation observed in the lower slope (i.e., Amato, The LC2 and LC3 sequences in the northern fl 1987) is connected to the uvial Waste Gate Forma- subarea shale out in the Shell 272-1, Shell 273-1, and tion in the Maryland coastal plain by the intervening Mobil 17-2 wells, with a few coarsening-upward, paralic sediments of the Missisauga Formation. sandy HST parasequences present at the top of the Upsection, the mid-Cretaceous interval on the LC2 (in all three wells) and LC3 (at Mobil 17-2) shelf in the southern subarea contains packages of sequences (Figure 3). This contrasts with thick LST fl onlapping, downlapping, and truncated re ections and HST sandstones in the LC3 and the LC2 on the that can be identified as distinct depositional se- GSD (Miller et al., 2018). Moving farther south, it is quences with transgressive and regressive compo- likely that there are sand-prone intervals within the nents (Figures 8, 10, 11). The transition from the LC3 and LC2 sequences offshore Maryland, based chaotic seismic reflections of the Missisauga Forma- on proximity to the fluvial facies of the Potomac tion to mid-Cretaceous clinoform reflection geom- Group in Maryland and the clinoform features ob- etries likely reflects a rise in relative sea level (RSL). served on the seismic data in our southern subarea Compared to the northern subarea, the mid-Cretaceous (Figure 9). in the southern subarea contains a greater thickness As noted above, the Potomac I, Potomac II, and under the modern middle shelf region (Figures 7, 12) Potomac III sequences of the Maryland coastal plain and contains better-developed clinoforms with less correspond to the LK1, MK3-MK2.1, and MK2 seismic

434 Cretaceous Fluvial-Deltaic Sequences of the Southern Baltimore Canyon Trough Prodelta/ shelf to

Delta plain shelf to delta front Delta plain/Delta front

UK2

OAE 2 Prodelta Prodelta to to shelf slope

MK1

Deltaic

OAE 1d MK2

MK2.1

OAE 1c

LC2/MK3

OAE 1b Delta front to shelf MK3 Anastamosing river systems Anastamosing river

OAE 1a

LK1 Carbonate Nonmarine/ platform/ backbank facies

Figure 13. Comparison of the Geologic Time Scale 2012 (GTS2012) Cretaceous sequences (Hardenbol et al., 1998), coastal onlap (Haq, 2014), and paleodepths in the Baltimore Canyon trough (BCT) (Poag, 1985b; Olsson et al., 1988) to the spatial and temporal positions of seismic reflections internal to the sequences defined on line ma-004 through the creation of the Wheeler diagram (Figure 10). To assign the seismic reflections a temporal value, the sequence boundary reflections were given a likely age according to the biostratigraphic constraints, and the internal reflections were then plotted in time by interpolating their position between the bounding surfaces. The onshore sequences are plotted according to Miller et al. (2004). Prominent Cretaceous ocean anoxic events (OAE) that are observed in more than one ocean basin (Katz et al., 2005) were plotted to demonstrate potential temporal correlation to flooding events in the BCT. All records were adjusted to the GTS2012. Al = Albian; Ap = Aptian; Barr = Barremian; Ce = Cenomanian; COST = Continental Offshore Stratigraphic Test; DCx = Dawson Canyon “x”;LC= Logan Canyon; LK = Lower Cretaceous; MK = mid-Cretaceous; Sa = Santonian; Tu = Turonian; UK = Upper Cretaceous. sequences, respectively (Figure 9; Table 1). The strata of the New Jersey coastal plain, which is com- Potomac Group in Maryland has been described posed of facies associated with an anastomosing river as transitioning from fluvial to deltaic sedimentation system (Browning et al., 2008). Nonmarine fluvial to (Anderson, 1948; Owens and Gohn, 1985), although terrestrially influenced deltaic strata on the Maryland the preponderance of evidence indicates that these coastal plain suggest delta-front to prodelta mid- are fluvial units (Miller et al., 2017). The nonmarine Cretaceous facies beneath the modern continental interpretation fits more reasonably with the Potomac shelf. This is supported by isopach maps (Figure 12)

SCHMELZ ET AL. 435 that indicate a depocenter landward of the modern erosion by the lateral movement of an ancestral shore shelfedgeandthedistalLK2depocenter.Like parallel current system (landward and seaward move- the LK2 depocenter, the LK1, MK3, MK2.1, and ment of the current’saxisassealevelrisesandfalls, MK2 strata also appear to have a southern source respectively) analogous to the modern Gulf Stream. (Figure 12). Alternatively, the erosion could be attributed to slope In contrast to the Lower and mid-Cretaceous failure associated with the shelf edge in response to intervals, the Upper Cretaceous section is thicker in large earthquake associated with the Chicxulub the northern subarea than it is in the south (Figures impact at the end of the Cretaceous (Norris and 7, 12). The seismic sections in the northern subarea Firth, 2002). appear to show steeply dipping clinoforms above Age control on the mid-Cretaceous sequences is reflection MK1 that are truncated by the UK2 re- particularly coarse (e.g., stage or age level) because of flection (Figures 5, 6). These are some of the most shallow-water facies, lack of marine biostratigraphic distinct clinoforms of the Cretaceous section despite markers, and widely spaced wells with samples taken the heavy growth faulting that disrupts the strati- largely from cuttings. As a result, we are not able to graphic continuity of the northern subarea (Figure 6). detect or quantify hiatuses associated with the se- In the southern subarea, MK1 is thinner and is com- quence boundaries from the biostratigraphy. The age posed of shingled clinoforms that appear to offlap as of the hiatus between the Missisauga and the LC3 they prograde across the shelf (Figure 8). The strike (LK1 equivalent) sequence is inferred from various line is particularly revealing for the section above geological constraints on the GSD to be middle reflection MK1 because it shows downlapping re- Barremian (ca. 126–127 Ma); the LC3 sequence ap- flections that dip to the south (Figure 7). In plan parently captured the MFS associated with early Ap- view, the MK1 sequence forms an apparent paleo- tian OAE 1a (ca.125 Ma) at COST B-2 supporting this clinoform rollover under the modern shelf with de- interpretation. Age breaks between the Albian and creasing relief as the interval thins to the south (Figure 12). Aptian sequences are currently a matter of conjec- These observations are consistent with other isopach ture; we adopt a conservative minimal 2-m.y. break observations of a more prominent northern source for each: (1) the break between the LK1 and the area for the coeval Cenomanian to Turonian packages MK3 spans the Albian–Aptian boundary and is as- mapped by Poag and Sevon (1989). sumed to be 112–114 Ma; (2) the MK3 sequence Like the MK1 sequence, the UK2 sequence thins is within the Albian (assumed to be 107–110 Ma), from north to south (Figure 7). However, in contrast whereas the MK2.1 sequence is only represented in to the mid-Cretaceous interval and the MK1 se- the SBCT, implying a significant hiatus within the quence, UK2 thickens significantly toward the shelf Albian (ca. 107–105 Ma); and (3) the MK2.1 sequence edge in the SBCT like LK2 (Figure 8). The lower is Albian and presumed to be 105–102.5 Ma, with a part of the UK2 sequence onlaps the UK2 surface, hiatus from 102.5–100.5 Ma. Study of onshore se- but progradational clinoforms are absent within the quences with biostratigraphic constraints suggests that interval under the modern shelf (Figure 8). It is possible the breaks between the Magothy and the Bass River that there was progradation seaward of the modern were 92.5–91.5 Ma (Miller et al., 2004) and ap- shelf to slope break or erosion of the upper part of proximately 98–97 Ma between the Bass River and this cycle or both. Poag and Sevon (1989) also show the Potomac Formation (Miller et al., 2017). We Coniacian to Campanian sedimentation had a more adopt these durations for the hiatuses between UK2 distally located depocenter on this margin relative to and MK1, respectively. their underlying Cenomanian to Turonian package. Additionally, the 1970s vintage seismic profiles There is significant toplap or truncation associ- that we rely upon have a relatively coarse vertical ated with the UK1 reflection (Figures 5–8), and this resolution, ranging from 30 to 70 m (100–230 ft) would indicate that at least some erosion occurred or from the top to the bottom of our section, and a period of nondeposition elapsed. The cause of this the boundaries of the depositional sequences we erosion is uncertain, but the widespread middle to identify are based on geometric criteria we observe outer shelf distribution of a major hiatus prompted on the seismic profiles. The “back to basics” per- Olsson and Wise (1987) to attribute it to submarine spective on sequence stratigraphy stresses that the

436 Cretaceous Fluvial-Deltaic Sequences of the Southern Baltimore Canyon Trough temporal resolution of the sequences that we identi- provide a mechanism for onshore–offshore migra- fied are tied to the resolution of the data (Neal and tion of depocenters. However, the provenance, per- Abreu, 2009; Miller et al., 2018) relative to their haps coupled with the flexural compensation and sedimentary thicknesses. Our work integrating compaction that occurs when sediments are de- lithological, paleontological, and well-log criteria posited locally in a sedimentary basin, is the variable with the seismic data strengthens the case for de- that is most likely to control the along-strike varia- fining the sequences we have identified and corre- tions in depocenter position that we observe in our lating them with established basin-specificsequence data. stratigraphic frameworks (Miller et al., 2017, Our observations showing the evolution of sed- 2018). However, because the appearance of “rela- iment sources to the BCT generally agree with those tively conformable” and “genetically related” strata shown by Poag and Sevon (1989), who examined according to the geometric criteria picked on the our study area and beyond, albeit at a coarser reso- seismic data are so closely tied to the vertical lution. They hypothesize that the shifts in depo- resolution, higher-resolution seismic images might centers were related to the magnitude of erosion and allow for the identification of higher-order depo- transport of material from different “highland source sitional sequences within the sequences that we have terrains.” Poag and Sevon (1989) inferred that the identified using the seismic profiles presently available spatial gradient in sediment supply was an effect of to us. differential uplift between the central Appalachians, Further, a primary assumption we make is that Adirondacks, and the New England Appalachians the along strike (northern to southern) movements distributed through southern (ancient Potomac, in depocenter location through the Cretaceous are Susquehanna, Delaware, and James), central (the representative of changes in sediment supplied from ancient Hudson), and northern (the ancient Con- various source regions (Figure 12). The factors con- necticut and Eastern Massachusetts) river dispersal trolling Cretaceous accommodation and sedimenta- routes. However, the interpretations of Poag and tion in the BCT include the following: (1) the North Sevon (1989) were also based upon stratigraphic American plate, and the developing BCT, were observations of the BCT, and more independent moving, in part, northward during the Late Jurassic information is needed on provenance, paleorivers, and into the Cretaceous (Muller¨ et al., 2016); (2) and Appalachian uplift history to test our interpre- regional sea-level variations on the order of 10–35 m tations of sediment sources. (30–115 ft) (short-term) to more than 100 m (>330 ft) Spatial variations in sediment source and the (long-term) (Miller et al., 2005); (3) the lithosphere “seesaw” differential preservation of sequences (e.g., beneath the deposited Jurassic strata was cooling, and the Lower Cretaceous is better developed in the the thermal subsidence rates were decreasing expo- south and the Upper Cretaceous in the north; see nentially (Parsons and Sclater, 1977; McKenzie, also Miller et al., 2017) could be explained by tectonic 1978; Steckler and Watts, 1978; Kominz et al., mechanisms, including mantle dynamic topography 2008); (4) flexural compensation and compaction (MDT), flexural loading changes through time, and/or began to generate a larger proportion of the long- more active tectonics. In particular, MDT, or the dy- term accommodation in the Cretaceous as the namic subsidence or uplift of topography from aniso- elastic plate thickness increased and the sediment tropic convection of the underlying mantle (Mitrovica column thickened (Steckler and Watts, 1978; Poag, et al., 1989; Gurnis, 1992; Spasojevic and Gurnis, 1985b); and (5) sediments are being generated 2012; Flament et al., 2013), might have been a factor from the erosion of highland source terrains and are in the shifting depocenters. This process has long being routed via Cretaceous rivers and deposited been understood to shift elevations of passive conti- in the marine realm in the BCT (Poag and Sevon, nental margins over long timescales (2–10+ m.y.; 1989). The northward component of the move- Moucha et al., 2008; Petersen et al., 2010), but can ment of the North Atlantic plate combined with MDT explain the movement of depocenters on the the decaying rate of thermal subsidence likely ex- relatively shorter timescale we examine herein? The plains the basin-wide reduction in deposited sedi- scale of the deformation from mantle-influenced pro- mentary volumes, and the variations in sea level could cesses is relatively large, on the order of approximately

SCHMELZ ET AL. 437 500–2000 km (~310–1200 mi) horizontally and ap- boundaries) mapped herein. Focusing on the mid- proximately 500–1000 m (~1600–3300 ft) verti- Cretaceous, we delineate at least five stratigraphic cally (Mitrovica et al., 1989; Gurnis, 1992; Liu and sequences (Figure 13) that could be identified offshore Nummedal, 2004); the vectors of movement of zones Maryland, that is, the LK2, MK3, MK2, MK2.1, and of subsidence and uplift related to this process can MK1 sequences shown on the Wheeler diagrams be relatively slow, for example, the zone of topo- (Figures 10, 11). Using the biostratigraphic age graphical subsidence associated with the subducted constraints within our database and physical cor- Farallon plate traveled across the United States at a relations to the mid-Atlantic Coastal Plain, we have rate of approximately 25–40 km/m.y. (~15–25 mi/ integrated these sequences into a relatively coarse m.y.)inthetimesince90Ma(Spasojevicet al., 2009). stratigraphic framework for comparison with global So, presuming a topographical profile based on events, including cycle charts (Hardenbol et al., 1998; the aforementioned parameters and the potential Haq, 2014), Cretaceous OAEs (Figure 13), and re- rate of movement of that slope, this long-wavelength gional RSL records. Although the sea-level ampli- MDT subsidence could potentially span rates of tudes associated with the cycle charts have been approximately 6–80 m/m.y. (~20–260 ft/m.y.). For shown to be incorrect (i.e., too high by a factor of example, Browning et al. (2006) empirically deter- 2–3, with even the relative change uncertain), the mined Miocene differential subsidence in the mid- timing of the events has been shown to be generally Atlantic margin to be on the order of 30 m/m.y. correct for the last 100 m.y. (Miller et al., 2005). (100 ft/m.y.), and, similarly, Rowley et al. (2013) The 107-yr scale transgression we identify above calculated approximately 30 m/m.y. (~100 ft/m.y) matches both basin-specific RSL and GMSL varia- for the past 3 m.y. The temporal variations in isopach tions inferred from cycle charts (Figure 13). Within thicknesses we observe at certain locations within the BCT, both Poag (1985b) and Olsson et al. (1988) the BCT fall into that range, and the partial influ- identified shallower paleodepths in the Aptian and ence of this process cannot be discounted. However, Albian than the Cenomanian and Turonian, where depocenter movements appear to change direction, they both identified peak water depths. The global moving predominately landward from the Early Cre- cycle charts generally agree in placing a major global taceous into the mid-Cretaceous before moving seaward flooding event at the Cenomanian–Turonian bound- and north into the Late Cretaceous. These observa- ary (Haq et al., 1987; Miller et al., 2005; Haq, 2014) tions would require additional processes, like sea- that coincides roughly with OAE 2 (Friedrich et al., level change, sediment source variation, and local 2012). lithospheric flexure, considering the long wavelength Within the 107-yr interval of the aforementioned of MDT as well as the inertia associated with the plate transgression, we can track the cross-margin transi- tectonic processes that drive the most well-known tions in depocenter position and depositional facies mechanisms. as a record of the combined accommodation history that includes sea-level change. The isopach maps Sea-Level Variations and Integration with (Figure 12) show a transition from a distal Early Creta- Global Climate History ceous depocenter with maximum thicknesses beneath the modern slope to a more-proximal depocenter in The facies variations in this study indicate a major the mid-Cretaceous. Coupling this observation with RSL rise from the Early to the Late Cretaceous. In the landward mid-Cretaceous shift in facies (observed the offshore wells, this change is exemplified by the seismically and in the wells) indicates movement of the transition from the heterolithic Lower Cretaceous deltaic sediment source landward through this period facies to the more marine shale-prone Upper Cre- of time toward a composite MFS at the aforemen- taceous. The seismic facies appear to show that this tioned flooding event at the Cenomanian–Turonian generalized transition occurs within a wide swath of boundary. Within this retrogradational package, we the basin in the SBCT. Within this generalized RSL observe five transgressive-regressive cycles that com- rise, there are higher-order RSL rises and falls that, prise depositional sequences (LK1, MK3, MK2.1, in part, alter accommodation and contribute to the and MK2 and MK1). Moreover, we observe at least formation of the surfaces (MFSs and sequence two additional progradational Upper Cretaceous

438 Cretaceous Fluvial-Deltaic Sequences of the Southern Baltimore Canyon Trough sequences overlying it, with a combined depocenter In addition to the correlation of the OAE 2 to the located seaward of that of the mid-Cretaceous. We MK1 MFS (Figure 13), OAE 1a correlates with the infer that this seaward depocenter represents a period LK1 (LC3 equivalent) sequence. This early Aptian of regression succeeding the Early to mid-Cretaceous OAE 1a occurs at COST B-2 at approximately 2895 transgression. The higher-order variations in accommo- mbKB (~9500 ftbKB) and 2778 mbsf (~9114 ftbsf), dation in the mid-Cretaceous are relatively long period where total organic carbon spikes to 5.45% coupled events, with likely durations of several million years. with a d13C increase (Scholle, 1977)—within the The magnitude of the higher-order variations in TST of the LC3 sequence (Figure 3; Miller et al., sea level and the rate of sea-level change required 2018). One hypothesized mechanism for the pro- to form these sequences are unknown but is of sub- duction of OAEs is increased burial of organic carbon stantial interest considering the insight it would as a result of the increase in shallow shelf areas at provide regarding the mechanisms forcing the genesis continental margins related to marine transgression of these mid-Cretaceous sequences. (Schlanger and Jenkyns, 1976), which is associated In addition to the sedimentary and stratigraphic with and can be caused by increasing global tem- inferences, we show a probable tie of the MK1 MFS peratures. More-recent evidence suggests the addi- to OAE 2. We place the MK1 sequence boundary in tion of a magmatic trigger for the generation of OAEs the early Cenomanian within calcareous nannofossil (e.g., Turgeon and Creaser, 2008), which does not Zone CC9 (Figure 13; Table 1). We make this cor- exclude the original hypothesis because increased relation based on geometric projection of the top atmospheric CO2 concentrations would presumably of the Potomac III sequence boundary from the increase paleotemperatures and sea level. This pro- Maryland coastal plain to the MK1 boundary on the vides a causal basis for the coincidence of GMSL Maryland shelf (Figure 9). The Potomac Formation peaks, rising water levels in the BCT, and global extends into Zone CC9 and is separated from the OAEs, exemplified by their correlation to flooding overlying mid-Cenomanian to lower Turonian Bass events in our LK1 and MK1 sequences (Figure 13). River Formation by a regional unconformity (Miller The correlation of the other potential Creta- et al., 2004). On this basis, we correlate the MK1 ceous global OAEs 1b, 1c, and 1d (Arthur et al., surface to this regional unconformity with an 1990; Erbacher and Thurow, 1997; Lehmann, 2000) inferred mid-Cenomanian age. Benthic biofacies to our sequences is more speculative than the two studies of the COST B-2, Shell 272-1, and Shell major events. The OAE 1b appears to correlate to our 586-1 wells (Sikora and Olsson, 1991) show the sequence MK3, whereas OAE 1c and OAE 1d likely deepest water depths of the mid-Cretaceous and thus fall within hiatuses associated with seismic sequence indicate the general location of the composite MFS boundaries MK2.1 and MK2, respectively. These within this sequence, with a specific location identifi- other OAEs do not appear to be readily distinguish- able from the GR log. This composite MFS location able in the well data. However, the mechanisms for occurs just above the Cenomanian–Turonian bound- these less-pervasive OAEs may differ from the global ary in the wells, as indicated by the HO of R. cushmani; OAE 1a and OAE 2. Erbacher et al. (1996) identified this demonstrates that the regional RSL peak corre- OAE 1c as a “detrital” event caused by an increased lates with the GMSL peak near the Cenomanian– terrigenous supply of nutrients associated with a sea- Turonian boundary (Miller et al., 2005). A GR level fall supported, in part, by coincidence of this high above a bell-shaped GR expression in the upper event with type III kerogen. This characterization of Cenomanian section of many OCS wells (Miller the OAE 1c would fit its placement in a hiatus in a et al., 2018) is interpreted as the deepest paleo- shallow marine section. However, OAE 1b and depths within the MK1 (Dawson Canyon “x” equiv- OAE 1d may well correlate to flooding surfaces in alent) sequence and tie in roughly with OAE 2 sequences MK3 and MK2. The OAE 1b occurs (Figures 10, 11). The seismic facies above the during a period of rising water levels in the BCT and MK1 MFS appear to be shingled and progradational likely corresponds to the transgressive package we and may be related to the base level lowerings that observe in MK3 (Figure 13). The age of the base generate late Cenomanian to early Turonian se- of the MK2 (LC1 equivalent) sequence is loosely quence boundaries. constrained, and MK2 might actually include a

SCHMELZ ET AL. 439 GMSL flooding event coincident with OAE 1d. The correlation, the lower half of this unit, the Waste development of a higher-resolution chronostratig- Gate, would connect seaward with the Missisauga raphy would help to resolve the relationship be- equivalent of Libby-French (1984), who character- tween the “minor” OAE and sea-level fluctuations ized the unit as containing alternating thick-bedded on this margin. sandstone, siltstone, and shale. The heterolithic We did not examine the Lower Cretaceous, LK2, strata are terrestrial (Poag, 1985b) to terrestrially section in as much detail as the mid-Cretaceous, in influenced delta-front (Libby-French, 1984) deposits part because the apparent sedimentary characteristics beneath the modern OCS. Seaward, there are thick and depositional environments of this interval make Neocomian deltaic sandstones in the Shell 93-1 well the Wheeler diagram and trajectory analyses less in- drilled on the continental rise offshore Maryland formative. We suspect that this interval is composed (Amato, 1987). However, farther north, the seaward of shallow marine to estuarine siliciclastics, deposited margin of the paleoshelf is predominantly carbonate landward of a fringing carbonate platform. We expect into the Barremian, evidenced by the carbonate the transitions of sedimentation, and the preser- structure found in the Shell 586-1 (Edson, 1987a) and vation of those deposits, to be driven by a three- Shell 587-1 (Edson, 1987b) wells that were drilled at a dimensional (3-D) movement of distributary channels present-day water depth of nearly 2000 m (6500 ft). and might expect relatively chaotic and stochas- The isopach of the combined LK2 reservoir tic depositional patterns in the dip lines of two- suggests large volumes for storage. The LK2 measures dimensional seismic data. In contrast, the overlying at approximately 15,000 km3 (~529,000 BCF) vol- strata appear to be relatively neatly stacked lobate umetrically within the boundaries of the digital ele- clinoforms that produce observable transgressive and vation model (DEM) in Figure 12. This interval is regressive patterns on dip lines ma-004 and ma-006 sand prone throughout the region, although the in- (Figures 10, 11). dividual sandstone bodies are likely to be discontin- uous with uncertain hydrologic connectivity. These deeper, lower Lower Cretaceous rocks of our LK2 Implications for the Reservoir sequences are also more likely to contain thermally Characterization of the Waste Gate and mature deposits than those of the overlying mid- Potomac Formations and the Missisauga Cretaceous. Lower Cretaceous rocks in some wells and Logan Canyon Sandstones have been buried to reach borderline maturity (e.g., Shell 586-1 [Edson, 1987a] and Shell 93-1 The work herein supports the correlations of the [Amato, 1987]) but are largely immature in others onshore Waste Gate Formation and Potomac Forma- (e.g., Shell 587-1 [Edson, 1987b] and Murphy 106-1 tion sequences (Miller et al., 2017) with the offshore [Adinolfi, 1987]). These strata are also more likely to Missisauga and LC units, respectively (Libby-French, contain gas-prone kerogen than oil (Post et al., 2012), 1984; Miller et al., 2018). Each of the onshore and considering the terrestrial nature of the deposits. The offshore units has been evaluated in terms of potential underlying Jurassic strata, also likely to contain ter- to be either an aquifer (Hansen, 1982; Andreasen et al., restrially derived kerogen, are buried to sufficient depths. 2015), a hydrocarbon reservoir (Libby-French, 1984; As a hydrocarbon reservoir, the economic viability would Prather, 1991), or a carbon storage unit (Sugarman et al., depend, in part, on the primary concern identified for 2011; Miller et al., 2017, 2018). Our mapping and carbon storage potential—how continuous and hydro- facies interpretations have implications for the logically connected are the individual sandstone bodies? extent and continuity of these previously identified Considering the offshore alone, the best reservoir reservoirs. targets are the sandstones within the HSTs of the LC Miller et al. (2017) identified the fluvial Waste sequences in the vicinity of the GSD (Miller et al., Gate and Potomac I sequences as an onshore target 2018). In particular, the HSTs of the LC3 and the for carbon storage on the basis that it is thick and LC2 sequences, representing two of the five to seven confined in the New Jersey coastal plain while noting sand-prone zones identified by Miller et al. (2018) that the confinement of the unit is less certain updip within the original LC sandstones (Libby-French, in the Maryland coastal plain. On the basis of our 1984), are highly porous and permeable at the

440 Cretaceous Fluvial-Deltaic Sequences of the Southern Baltimore Canyon Trough COST B-2 well. These two sequences have sand- prone reservoir is likely primarily a CO2 storage prone zones within the HSTs and LSTs that extend reservoir. across the OCS in the NBCT. However, the LC1 sandstones shale out into the basin (Miller et al., 2018) and to the SBCT, as shown here. Similarly, the CONCLUSION LC3 and LC2 sequences appear to thin and shale out toward the wells south of the GSD (i.e., Mobil 17-2, We generated an integrated sequence stratigraphic Shell 273-1, and Shell 272-1 in Figure 3), repre- framework for the uppermost Jurassic and Cretaceous senting the lateral limitation of these porous lower strata of the SBCT using recently released seismic LC sequences. In these southern locations, the LC1 is data (Triezenberg et al., 2016), legacy well data, and also a thin shale deposited in greater than 80 m (>260 objective criteria to recognize the bounding surfaces ft) of water (Sikora and Olsson, 1991). of depositional sequences. We update and build upon Although they shale out in the northern part of the relatively limited evaluations within the litera- the study area, LC sequences may provide significant ture (Libby-French, 1981, 1984; Poag and Sevon, storage potential in the southern part of the study area. 1989; Prather, 1991; Klitgord et al., 1994) and rec- Although there are no well data south of Shell 272-1, oncile our observations with the regional sequence the section thickens to the south, offshore Maryland, stratigraphic framework established for the onshore where the isopach maps (Figure 12) indicate a nearby mid-Atlantic Coastal Plain (Miller et al., 2017) as mid-Cretaceous sediment source. These strata are well as the more-recent evaluation of mid-Cretaceous likely to be sand prone, particularly closer to the sedimentation on the OCS of the NBCT (Miller modern shoreline where the Lower Cretaceous sedi- et al., 2018). mentation was nonmarine. This is particularly true of We identify seven seismic stratigraphic surfaces LK1, which correlates to the Potomac I sequence that that can be traced from a northern subarea offshore Miller et al. (2017) documented as approximately 50% New Jersey to the south to the continental shelf off the sand in the downdip Ocean City, Maryland, well. All of coast of Maryland, constrainedinagebybiostratigraphic theLCsequencesonshoreandoffshoreareoverlainby data from five wells offshore southern New Jersey and a our MK1 that correspond to the shales of the Bass River cross section of three wells on the Maryland coastal and Dawson Canyon, onshore and offshore, respec- plain. A regional velocity–depth model is developed tively. Although this interval might not be entirely to convert the legacy seismic data to depth, and an homogenous (see sandstone bodies and clinoforms onshore–offshore correlation is made between the represented in the GR logs and seismic sections in depth-converted seismic surfaces of the Maryland Figures 4–6), the Upper Cretaceous RSL high likely Shelf and the onshore well-log stratigraphy (Miller forms a sufficient capstone. Offshore, this cap is of et al., 2017) within the three deep stratigraphic test sufficient depth for the purpose of storing supercritical wells in Maryland. The onshore–offshore correlation

CO2 (~800 m [~2600 ft]). However, the top of the provides additional, albeit coarse, constraints on Potomac Formation is more than 800 m (2600 ft) absolute age. below the ground surface onshore and may not be fully Weintegratetheonshorestratigraphyforthemid- confined, making the continuity of the seal overlying Atlantic Coastal Plain (Miller et al., 2004, 2017) with the Potomac I a consideration. In all, the LK1, MK3, the depth-converted seismic stratigraphy ultimately pro- and MK2 units comprise approximately 25,000 km3 ducing regional structural contour maps and isopachs for (~882,000 BCF) within the boundaries of the DEM the entire SBCT. These maps combine with Wheeler in Figure 12. In terms of potential as a commercial diagrams we generate for dip lines of the Maryland shelf reservoir, the LC sequences and its equivalents (Table to establish a better understanding of spatiotem- 1) are likely to contain the stratigraphic successions poral variations of Cretaceous sedimentation in the necessary to store hydrocarbon reserves. However, it is SBCT. We delineate the 3-D movement (planimetric unproven whether or not an underlying source rock space and time) of the sedimentary depocenter from a hasbeenburiedtoadepthsufficient to produce distal and southerly Early Cretaceous position to a thermallymatureoilornaturalgaswithamigration more-proximalbutstillsoutherlylocationinthe pathway to the deltaic sandstones, and thus this sand- mid-Cretaceous, followed by predominantly northerly

SCHMELZ ET AL. 441 and distal Late Cretaceous sedimentation. These spatial APPENDIX 1: CONVERSION OF SEISMIC variations in sedimentary thickness correspond to TWO-WAY TRAVELTIME TO DEPTH BELOW a deepening of depositional environment from the MEAN SEA LEVEL Early Cretaceous to a maximum paleodepth at the Cenomanian–Turonian boundary, followed by a Late A regional geoacoustic study that established a Geophysical Database of the United States Northern Atlantic Margin Cretaceous shoaling. (GDUSNAM; Klitgord et al., 1994) was conducted for the US Using the coarse chronostratigraphic framework de- Naval Oceanographic Office by the USGS. The velocity rived from the integrated onshore and offshore inter- structure was calculated using the stacking velocities pretations, we also correlate five of the mid-Cretaceous generated as a process step of the aforementioned USGS fi depositional sequences to global cycle charts and marine MCS pro les that canvas the margin and calibrating fi those velocities to sonic log and check shot surveys collected RSL records for the BCT. These ve sequences fall at the OCS industry wells along a few select seismic profiles 7 within a major 10 -yr transgression that spans the (Klitgord and Schneider, 1994). Geologically improbable mid-Cretaceous, reaching a GMSL high near the velocity variations (that were likely left in the original data fi Cenomanian–Turonian boundary (Haq et al., 1987; because they were not signi cant enough to affect the visual character of the finished seismic product) were smoothed by Miller et al., 2005; Haq, 2014), corresponding with Klitgord and Schneider (1994) and subsequently converted to the deepest paleodepths observed within the BCT interval velocities using the Dix formula (Dix, 1955; Taner (Poag, 1985b; Olsson et al., 1988; Sikora and Olsson, and Koehler, 1969). The velocity information was evaluated 1991) and the concomitant OAE 2. Seismic facies at approximately 3000-m (~9800-ft) spacing along the USGS corroborate this deepening trend and reveal higher- seismic lines and then subsequently gridded at 5-min spacing – in terms of both latitude and longitude, with a cubic spline order transgressive regressive packages that pro- interpolation routine. duce a retrogradational set of depositional sequences. The gridded database is organized around 19 strati- We propose that OAE 1a and OAE 2 are associated graphic horizons delineated across the margin. At each with MFSs of the lowermost and uppermost mid- spatial location within the database, each of the strati- graphic horizons representing a stratigraphic section present at Cretaceous sequences and suggest that these obser- that location contains the TWTT to that horizon and its depth vations support a causal relationship between rising in meters below sea level, in addition to information about GMSL and these OAE. density, seismic velocities, attenuation, and broad sedi- Based on the seismic facies we observe and our mentary classifications. This stratigraphically oriented da- correlation to coeval fluvial depositional facies in well tabase was reorganized into a database of monotonically increasing TWTTs and associated depths across the margin. data from the Maryland and New Jersey coastal plains We extracted the depth and TWTT information from (Browning et al., 2008; Miller et al., 2017), we suggest the original database structure for each location in the da- there are likely mid-Cretaceous, sand-prone deltaic tabase and replaced the values in the first second of TWTT facies offshore Maryland. This interval is largely shale below the sea floor with a time-depth relationship derived offshore southern New Jersey but thickens to the from modern data collected during the R/V Oceanus 270 cruise and tested against data from IODP Exp. 313 south, indicating a location more proximal to a deltaic (Mountain and Monteverde, 2012). Having made this re- sediment source. This package is thick and is likely placement, at every location, an interpolation was then per- to contain sedimentary facies that comprise a viable formed to calculate the time and depth information at consistent reservoir for storage of supercritical CO or oil and intervals of travel time, 0.1 s (Figure 2). 2 Therefore, the stratigraphically oriented GDUSNAM natural gas. It is also capped by muds and shales on- was transformed to a record of depths at monotonically in- shore and offshore. However, a viable source rock for creasing TWTT isochrons; that is, the depth data were no hydrocarbon accumulation in these potential reservoir longer tied to TWTT through the 19 horizons that are var- sandstones has yet to be identified, relegating this to a iably present throughout the margin. Instead, the data were CO storage reservoir. Beneath this mid-Cretaceous organized into 90 surfaces of depth at TWTT isochrons 2 spanning, at most, 0–8.9 s at 0.1-s intervals. interval, there is also a thick package of Lower Cre- For import to IHS Kingdom software and to facilitate taceous strata that likely contain sand-prone silici- tests of accuracy, the data were converted from the geo- clastics deposited behind a fringing carbonate graphic coordinate system that it was provided in (latitude platform. These sandstones might also present attrac- and longitude) to a Cartesian Mercator projection, using the Universal Transverse Mercator 18N coordinate system. This tive targets for CO2 storage or hydrocarbon explora- conversion was carried out using the open-source python- tion if they are locally amalgamated and hydrologically basedsoftwarepyproj.Theconversionresultedinirregu- connective. larly spaced data points that cannot be imported into an IHS

442 Cretaceous Fluvial-Deltaic Sequences of the Southern Baltimore Canyon Trough Kingdom project. Therefore, a linear interpolation was depositional sequences are geomorphological in their genesis. So applied to make a regularly spaced grid for each of the depth the identification of their characteristic forms, and the surfaces surfaces. The new depth grids with 2500-m (8200-ft) bounding them, mitigates the limitation of low-resolution data. spacing were imported into IHS Kingdom and used to Although, in attempting to trace single reflectors, there were convert the entire seismic database, including interpreted instances in which depositional sequences thinned to a thickness horizons from TWTT to depth. that is near or within the vertical resolution and preclude iden- tification of internal reflections. Loop correlation was applied to resolve uncertainties. Within the Cretaceous section of the BCT, APPENDIX 2: UNCERTAINTY OF DATA AND vertical seismic resolution typically ranged from 30 to 70 m – INTERPRETATIONS (100 230 ft). However, in most areas, and at the scale of analysis, internal reflections were observable, and many of the high- amplitude reflections were continuous, indicating a genetically The largest sources of uncertainty are the paucity of wells related origin for the consistent impedance contrasts (Figure 7). in the southern subarea, the interpretation of seismic data with Assuming a –10% error for the time–depth conversion a coarse vertical resolution, the time–depth conversions ob- obtained from the velocity surveys, errors associated with tained from velocity surveys for the projection of drill site data projecting the biostratigraphy and GR logs onto the seismic onto the seismic grid, and the accuracy of the regional data would be, very conservatively, within –75–175 m time–depth model. Compared to the northern subarea, there (–250–570 ft), with the better accuracies higher in the sec- are relatively few constraints on accuracy of the seismic in- tion. This conservative estimate of error is relatively large, and terpretations in the southern area because of the absence of well it is likely a maximum constraint because we did not account locations on the shelf. The coarse vertical resolution of the for the contribution of frequency content higher than the seismic data most strongly affects the seismic stratigraphic in- peak frequency at a given depth interval. Still, the coarse res- terpretations. Seismic artifacts, not limited to those generated by olution inherently associated with seismic stratigraphic inter- low vertical resolution, might alter perceptions of thickness and pretation of deeply buried (>1–1.5 km [>3300–4900 ft]) continuity for some stratal packages. The accuracy of the site strata can influence the assignment of age to a particular pack- velocity surveys affects the projection of data (e.g., biostratig- age. To mitigate adverse effects from potential inaccuracy, raphy) from the wells onto the seismic data. Inaccurate pro- sequences were correlated between multiple wells and ages jections of biostratigraphic levels within the wells because of assigned based on information from each. The correlations be- inaccuracies associated with the velocity surveys could result in tween wells were generally within the seismic resolution of the problematic absolute age constraints on the seismic surfaces, well-log correlation and biostratigraphic markers (Figure 3). along with complications because of cavings and other issues. Using the check shot data as a reference, the regional Finally, localized artifacts within the regional time–depth model velocity–depth model was calculated to have an accuracy of might create anticlines or synclines within structural contours better than –150 m (–490 ft) for sedimentary depths up to that do not exist in reality. Additionally, because this time– 3000 m (9800 ft). Like the velocity survey error, the accuracy depth model was used to correlate the onshore and offshore of the regional velocity–depth model could affect stratigraphic stratigraphy, artifacts could also affect the onshore–offshore correlations. The onshore–offshore correlation is dependent correlations. The magnitudes of each of the identified errors are upon converting the seismic data to depth, but a –150-m likely small enough to retain reasonable confidence in our re- (–490-ft) offset would not appear to significantly change the sults, especially considering the scales associated with this proposed correlations (Figure 9). analysis (i.e., the mapped seismic surfaces are typically sepa- rated by hundreds of meters of sediment and are mapped re- gionally). Moreover, the integrated nature of our analysis (that includes seismic stratigraphic interpretation, well-log correla- tion, and biostratigraphic constraints) produces checks and re- REFERENCES CITED dundancies that minimize the likelihood of miscorrelations. The vertical resolution of the seismic data is a product of the Abreu, V., J. E. Neal, K. M. Bohacs, and J. L. Kalbas, 2010, wavenumber of the seismic data through the section. As interval Sequence stratigraphy of siliciclastic systems - The velocities increase moving down through the section, the wave- ExxonMobil methodology: Atlas of Exercises: Tulsa, number decreases for a given source frequency, and the distance Oklahoma, SEPM Concepts in Sedimentology and between beds that can be resolved increases. The increasing dis- Paleontology 9, 226 p. tance between resolvable bedding variations naturally generates Adinolfi, F., 1987, Murphy Wilmington Canyon 106-1 well: challenges to the application of seismic stratigraphic methods. Geological and operational summary: Well report series: However, even if many bedding interfaces are contained within a Vienna, Virginia, US Department of the Interior, Min- canonical 1/4–1/8 l (Widess, 1973), the largest impedance con- erals Management Service, Atlantic OCS Region, 33 p. trasts through the section are likely to be preserved and recorded as Amato, R. V., 1987, Shell Baltimore Rise 93-1 well: Geological a response to the source signal, whereas minor variations sur- and operational summary: Well report series: Vienna, rounding the larger contrast will be overshadowed. This lends Virginia, US Department of the Interior, Minerals credibility toward tracing continuous, high-amplitude reflections Management Service, Atlantic OCS Region, 55 p. even if the seismic resolution is poor. Additionally, the seismic Anderson, J. L., 1948, Cretaceous and Tertiary subsurface geol- geometries associated with internal reflections of deltaic ogy - The stratigraphy, paleontology, and sedimentology of

SCHMELZ ET AL. 443 three deep test wells on the Eastern Shore of Maryland: Vienna, Virginia, US Department of the Interior, Min- Baltimore, Maryland, State of Maryland Board of Natural erals Management Service, Atlantic OCS Region, 46 p. Resources,MinesAndWaterResourcesBulletin,v.2,113p. Edson, G. M., 1987b, Shell Wilmington Canyon 587-1 well: Andreasen, D. C., A. W. Staley, and G. Achmad, 2015, Geological and operational summary: Well report series: Maryland coastal plain aquifer information system: Hy- Vienna, Virginia, US Department of the Interior, Min- drogeologic framework: Baltimore, Maryland, Maryland erals Management Service, Atlantic OCS Region, 49 p. Department of Natural Resources Open-File Report 12-02- Erbacher, J., and J. Thurow, 1997, Influence of oceanic an- 20, 121 p. oxic events on the evolution of mid-Cretaceous radiolaria Arthur, M., H. Jenkyns, H.-J. Brumsack, and S. Schlanger, in the North Atlantic and western Tethys: Marine 1990, Stratigraphy, geochemistry, and paleoceanography Micropaleontology, v. 30, no. 1–3, p. 139–158, doi: of organic carbon-rich Cretaceous sequences, in R. N. 10.1016/S0377-8398(96)00023-0. Ginsburg and B. Beaudoin, eds., Cretaceous resources, Erbacher, J., J. Thurow, and R. Littke, 1996, Evolution patterns events and rhythms: Dordrecht, the Netherlands, of radiolaria and organic matter variations: A new approach Kluwer, p. 75–119, doi:10.1007/978-94-015-6861-6_6. to identify sea-level changes in mid-Cretaceous pelagic – Baldwin, K., K. G. Miller, and G. S. Mountain, 2017, De- environments: Geology, v. 24, no. 6, p. 499 502, doi: lineating mid-Cretaceous seismic and well-log sequences 10.1130/0091-7613(1996)0242.3.CO;2. to assess carbon storage potential in the northern Balti- Flament, N., M. Gurnis, and R. D. Muller,¨ 2013, A review of more Canyon trough: The Geological Society of America observations and models of dynamic topography: Lith- – 129th Annual Meeting, Seattle, Washington, October osphere, v. 5, no. 2, p. 189 210, doi:10.1130/L245.1. 22–25, 2017. Friedrich, O., R. D. Norris, and J. Erbacher, 2012, Evolution — Blackburn, T. J., P. E. Olsen, S. A. Bowring, N. M. McLean, of middle to Late Cretaceous oceans A 55 my record of ’ D. V. Kent, J. Puffer, G. McHone, E. T. Rasbury, and Earth s temperature and carbon cycle: Geology, v. 40, no. 2, p. 107–110, doi:10.1130/G32701.1. M. Et-Touhami, 2013, Zircon U-Pb geochronology links Gradstein, F. M., J. G. Ogg, M. Schmitz, and G. Ogg, 2012, the end-Triassic extinction with the Central Atlantic The geologic time scale 2012: Amsterdam, Elsevier, Magmatic Province: Science, v. 340, no. 6135, p. 941–945, 1176 p. doi:10.1126/science.1234204. Greenlee, S. M., W. J. Devlin, K. G. Miller, G. S. Mountain, Brown, L., and W. Fisher, 1977, Seismic-stratigraphic inter- and P. B. Flemings, 1992, Integrated sequence stra- pretation of depositional systems: Examples from Brazilian tigraphy of Neogene deposits, New Jersey continental rift and pull-apart basins: Application of seismic reflection shelf and slope: Comparison with the Exxon model: configuration to stratigraphic interpretation, in C. E. Payton, Geological Society of America Bulletin, v. 104, no. 11, ed., Seismic stratigraphy: Applications to hydrocarbon ex- p. 1403–1411, doi:10.1130/0016-7606(1992)104<1403: ploration: AAPG Memoir 26, p. 213–248. ISSOND>2.3.CO;2. Browning, J. V., K. G. Miller, P. P. McLaughlin, M. A. Kominz, Greenlee, S. M., and T. C. Moore, 1988, Recognition and P. J. Sugarman, D. Monteverde, M. D. Feigenson, and interpretation of depositional sequences and calculation fi J. C. Hernandez,´ 2006, Quanti cation of the effects of of sea-level changes from stratigraphic data—Offshore eustasy, subsidence, and sediment supply on Miocene New Jersey and Alabama Tertiary, in C. K. Wilgus, sequences, mid-Atlantic margin of the United States: B. S. Hastings, C. G. S. C. Kendall, H. W. Posamentier, – Geological Society of America Bulletin, v. 118, no. 5 6, C. A. Ross, and J. C. Van Wagoner, eds., Sea-level – p. 567 588, doi:10.1130/B25551.1. changes: An integrated approach: Tulsa, Oklahoma, Browning, J. V., K. G. Miller, P. J. Sugarman, M. A. Kominz, SEPM Special Publication 42, p. 329–353, doi:10.2110 P. P. McLaughlin, A. A. Kulpecz, and M. D. Feigenson, /pec.88.01.0329. 2008, 100 Myr record of sequences, sedimentary facies Grow, J. A., and R. E. Sheridan, 1988, US Atlantic conti- and sea level change from Ocean Drilling Program on- nental margin; A typical Atlantic-type or passive conti- shore coreholes, US Mid-Atlantic Coastal Plain: Basin nental margin, in R. E. Sheridan and J. A. Grow, eds., Research, v. 20, no. 2, p. 227–248, doi:10.1111/j.1365- The Atlantic continental margin, U.S.: The geology of 2117.2008.00360.x. North America: Boulder, Colorado, Geological Society Catuneanu, O., 2006, Principles of sequence stratigraphy: of America, v. I-2, p. 1–8. Amsterdam, Elsevier, 375 p. Gurnis, M., 1992, Rapid continental subsidence following Dix, C. H., 1955, Seismic velocities from surface measure- the initiation and evolution of subduction: Science, ments: Geophysics, v. 20, no. 1, p. 68–86, doi:10.1190 v. 255, no. 5051, p. 1556–1558, doi:10.1126/science.255 /1.1438126. .5051.1556. Doyle, J., 1982, Palynology of continental Cretaceous sedi- Hansen, H., 1982, Hydrogeologic framework and potential ments, Crisfield geothermal test well, eastern Maryland, utilization of the brine aquifers of the Waste Gate For- pt. 2: Waste Gate Formation: Baltimore, Maryland, mation, a new unit of the Potomac Group underlying Maryland Department of Natural Resources, Geological the Delmarva Peninsula, pt. 1: Waste Gate Formation: Survey Open-File Report 82-02-01, p. 51–87. Baltimore, Maryland, Maryland Department of Natural Edson, G. M., 1987a, Shell Wilmington Canyon 586-1 well: Resources Geological Survey Open-File Report 82-02- Geological and operational summary: Well report series: 01, p. 1–50.

444 Cretaceous Fluvial-Deltaic Sequences of the Southern Baltimore Canyon Trough Haq, B. U., 2014, Cretaceous eustasy revisited: Global Kulpecz,A.A.,K.G.Miller,P.J.Sugarman,andJ.V.Browning, and Planetary Change, v. 113, p. 44–58, doi:10.1016 2008, Response of late Cretaceous migrating deltaic fa- /j.gloplacha.2013.12.007. cies systems to sea level, tectonics, and sediment supply Haq, B. U., J. Hardenbol, and P. R. Vail, 1987, Chronology of changes, New Jersey coastal plain, USA: Journal of Sedi- fluctuating sea levels since the Triassic: Science, v. 235, mentary Research, v. 78, no. 2, p. 112–129, doi:10.2110 no. 4793, p. 1156–1167, doi:10.1126/science.235.4793 /jsr.2008.009. .1156. Lehmann, J., 2000, Upper Albian ammonites from ODP Leg Hardenbol, J., J. Thierry, M. B. Farley, T. Jacquin, P.-C. De 171B off northern Florida: Palaeontology, v. 43, no. 1, Graciansky, and P. R. Vail, 1998, Mesozoic and Cenozoic p. 41–61, doi:10.1111/1475-4983.00118. sequence chronostratigraphic framework of European Libby-French, J., 1981, Lithostratigraphy of Shell 272-1 and basins, in P.-C. De Graciansky, J. Hardenbol, T. Jacquin, 273-1 wells: Implications as to depositional history of and P. R. Vail, eds., Mesozoic and Cenozoic sequence Baltimore Canyon Trough, mid-Atlantic OCS: AAPG stratigraphy of European basins: Tulsa, Oklahoma, Bulletin,v.65,no.8,p.1476–1484, doi:10.1306 SEPM Special Publication 60, p. 3–13, doi:10.2110 /03B59571-16D1-11D7-8645000102C1865D. /pec.98.02.0003. Libby-French, J., 1984, Stratigraphic framework and petro- Jansa, L. F., 1981, Mesozoic carbonate platforms and banks leum potential of northeastern Baltimore Canyon trough, of the eastern North American margin: Marine Geology, Mid-Atlantic outer continental shelf: AAPG Bulletin, – v. 44, no. 1–2, p. 97–117, doi:10.1016/0025-3227(81) v. 68, no. 1, p. 50 73. 90114-6. Liu, S., and D. Nummedal, 2004, Late Cretaceous subsidence Katz, M. E., J. D. Wright, K. G. Miller, B. S. Cramer, in Wyoming: Quantifying the dynamic component: Ge- – K. Fennel, and P. G. Falkowski, 2005, Biological over- ology, v. 32, no. 5, p. 397 400, doi:10.1130/G20318.1. print of the geological carbon cycle: Marine Geology, Manspeizer, W., and H. L. Cousminer, 1988, Late Triassic- v. 217, no. 3–4, p. 323–338, doi:10.1016/j.margeo Early Jurassic synrift basins of the US Atlantic margin, in R. E. Sheridan and J. A. Grow, eds., The Atlantic .2004.08.005. Kennedy, W., I. Walaszczyk, and W. Cobban, 2005, The continental margin: US: The geology of North America: Boulder, Colorado, The Geological Society of America, global boundary stratotype section and point for the base v. I-2, p. 197–216. of the Turonian stage of the Cretaceous: Pueblo, Col- McKenzie, D., 1978, Some remarks on the development of orado, USA: Episodes—Newsmagazine of the Interna- sedimentary basins: Earth and Planetary Science Letters, tional Union of Geological Sciences, v. 28, p. 93–104. v. 40, no. 1, p. 25–32, doi:10.1016/0012-821X(78) Klitgord, K. D., D. R. Hutchinson, and H. Schouten, 1988, 90071-7. U.S. Atlantic continental margin; Structural and tec- Miller, K. G., J. V. Browning, G. S. Mountain, M. A. Bassetti, tonic framework, in R.E.SheridanandJ.A.Grow,eds.,The D. Monteverde, M. E. Katz, J. Inwood, J. Lofi, and Atlantic continental margin, U.S.: The geology of North J.-N. Proust, 2013a, Sequence boundaries are impedance America: Boulder, Colorado, The Geological Society of contrasts: Core-seismic-log integration of Oligocene– – America, v. I-2, p. 19 55, doi:10.1130/DNAG-GNA-I2.19. Miocene sequences, New Jersey shallow shelf: Geosphere, Klitgord, K. D., C. Poag, C. Schneider, and L. North, 1994, v. 9, no. 5, p. 1257–1285, doi:10.1130/GES00858.1. Geophysical database of the East Coast of the United Miller, K. G., J. V. Browning, P. J. Sugarman, D. H. Monteverde, – States northern Atlantic margin crosssectionsand D.C.Andreasen,C.Lombardi,J.Thornburg,Y.Fan,and gridded database: Georges Bank Basin, Long Island R. E. Kopp, 2017, Lower To Mid-Cretaceous sequence Platform, and Baltimore Canyon Trough: Woods Hole, stratigraphy and characterization of CO2 storage poten- Massachusetts, US Geological Survey Open-File Report tial in the Mid-Atlantic US coastal plain: Journal of 94-637, 189 p., doi:10.3133/ofr94637. Sedimentary Research, v. 87, no. 6, p. 609–629, doi: Klitgord, K. D., and C. Schneider, 1994, Geophysical data- 10.2110/jsr.2017.33. base of the East Coast of the United States northern Miller,K.G.,M.A.Kominz,J.V.Browning,J.D.Wright, Atlantic margin: Velocity analyses: Woods Hole, Mas- G.S.Mountain,M.E.Katz,P.J.Sugarman,B.S.Cramer, sachusetts, US Geological Survey Open-File Report 94- N. Christie-Blick, and S. F. Pekar, 2005, The Phanerozoic 192, 74 p., doi:10.3133/ofr94192. record of global sea-level change: Science, v. 310, Klitgord, K. D., and H. Schouten, 1986, Plate kinematics of no. 5752, p. 1293–1298, doi:10.1126/science.1116412. the central Atlantic, in P. R. Vogt and B. E. Tucholke, Miller, K. G., C. J. Lombardi, J. V. Browning, W. J. Schmelz, eds., The western North Atlantic region: Boulder, Col- G. Gallegos, K. Baldwin, and G. S. Mountain, 2018, Back orado, Geological Society of America, p. 351–378, doi: to basics of sequence stratigraphy: Early Miocene and 10.1130/DNAG-GNA-M.351. Mid Cretaceous examples from the New Jersey paleo- Kominz,M.A.,J.Browning,K.Miller,P.Sugarman, shelf: Journal of Sedimentary Research, v. 88, no. 1, S. Mizintseva, and C. Scotese, 2008, Late Cretaceous p. 148–176, doi:10.2110/jsr.2017.73. to Miocene sea-level estimates from the New Jersey and Miller, K. G., G. S. Mountain, J. V. Browning, M. E. Katz, Delaware coastal plain coreholes: An error analysis: Basin D. Monteverde, P. J. Sugarman, H. Ando, M. A. Bassetti, Research,v.20,no.2,p.211–226, doi:10.1111/j.1365- C. J. Bjerrum, and D. Hodgson, 2013b, Testing sequence 2117.2008.00354.x. stratigraphic models by drilling Miocene foresets on

SCHMELZ ET AL. 445 the New Jersey shallow shelf: Geosphere, v. 9, no. 5, United States Atlantic margin: New York, Van Nostrand p. 1236–1256, doi:10.1130/GES00884.1. Reinhold, p. 293–341. Miller, K. G., G. S. Mountain, J. V. Browning, M. Kominz, Muller,R.D.,M.Seton,S.Zahirovic,S.E.Williams,¨ P.J.Sugarman,N.Christie-Blick,M.E.Katz,and K. J. Matthews, N. M. Wright, G. E. Shephard, J. D. Wright, 1998, Cenozoic global sea level, sequences, K. T. Maloney, N. Barnett-Moore, and M. Hosseinpour, and the New Jersey transect: Results from coastal plain 2016, Ocean basin evolution and global-scale plate and continental slope drilling: Reviews of Geophysics, reorganization events since Pangea breakup: An- v. 36, no. 4, p. 569–601, doi:10.1029/98RG01624. nual Review of Earth and Planetary Sciences, v. 44, Miller, K. G., P. J. Sugarman, J. V. Browning, M. A. Kominz, no. 1, p. 107–138, doi:10.1146/annurev-earth-060115- R. K. Olsson, M. D. Feigenson, and J. C. Hernandez,´ 012211. 2004, Upper Cretaceous sequences and sea-level his- Neal, J., and V. Abreu, 2009, Sequence stratigraphy hi- tory, New Jersey coastal plain: Geological Society of erarchy and the accommodation succession method: America Bulletin, v. 116, no. 3, p. 368–393, doi:10.1130 Geology, v. 37, no. 9, p. 779–782, doi:10.1130 /B25279.1. /G25722A.1. Mitchum, R., and P. Vail, 1977, Seismic stratigraphy and Nittrouer, C. A., J. A. Austin, M. E. Field, J. H. Kravitz, global changes of sea level: Part 7. Seismic stratigraphic J. P. Syvitski, and P. L. Wiberg, 2007, Continental margin interpretation procedure: Application of seismic re- sedimentation: From sediment transport to sequence flection configuration to stratigraphic interpretation, stratigraphy: Ghent, Belgium, International Associa- in C. E. Payton, ed., Seismic stratigraphy–Applications tion of Sedimentologists Special Publication 37, 560 p., to hydrocarbon exploration: AAPG Memoir 26, doi:10.1002/9781444304398. p. 135–143. Norris, R., and J. Firth, 2002, Mass wasting of Atlantic Mitchum, R., P. Vail, and J. Sangree, 1977a, Seismic stra- continental margins following the Chicxulub impact tigraphy and global changes of sea level: Part 6. Strati- event, in C. Koeberl and K. G. MacLeod, eds., Cata- graphic interpretation of seismic reflection patterns in strophic events and mass extinctions: Impacts and depositional sequences: Application of seismic reflection beyond: Boulder, Colorado, Geological Society of configuration to stratigraphic interpretation, in C. E. Payton, America Special Paper 356, p. 79–96, doi:10.1130 ed., Seismic stratigraphy–Applications to hydrocarbon /0-8137-2356-6.79. exploration: AAPG Memoir 26, p. 117–133. Olsson, R., and S. Wise, 1987, Upper Maestrichtian to middle Mitchum, R., P. Vail, and S. Thompson III, 1977b, Seismic Eocene stratigraphy of the New Jersey slope and coastal stratigraphy and global changes of sea level: Part 2. plain, in J. H. Blakeslee and E. Whalen, eds., Initial The depositional sequence as a basic unit for strati- reports of the Deep Sea Drilling Project 93: Washington, graphic analysis: Application of seismic reflection con- DC, US Government Printing Office, p. 1343–1365. figuration to stratigraphic interpretation, in C. E. Payton, Olsson, R. K., T. G. Gibson, H. J. Hansen, and J. P. Owens, ed., Seismic stratigraphy–Applications to hydrocarbon 1988, Geology of the northern Atlantic Coastal Plain: exploration: AAPG Memoir 26, p. 53–62. Long Island to Virginia, in R. E. Sheridan and J. A. Grow, Mitrovica,J.,C.Beaumont,andG.Jarvis,1989,Tilting eds., The Atlantic continental margin, U.S.: The geology of continental interiors by the dynamical effects of of North America: Boulder, Colorado, The Geologi- subduction:Tectonics,v.8,no.5,p.1079–1094, doi: cal Society of America, v. I-2, p. 87–105, doi:10.1130 10.1029/TC008i005p01079. /DNAG-GNA-I2.87. Moucha, R., A. M. Forte, J. X. Mitrovica, D. B. Rowley, Owens, J., and G. Gohn, 1985, Depositional history of the S. Quer´ e,´ N. A. Simmons, and S. P. Grand, 2008, Dy- Cretaceous series in the US Atlantic Coastal Plain: Stra- namic topography and long-term sea-level variations: tigraphy, paleoenvironments, and tectonic controls of There is no such thing as a stable continental platform: sedimentation, in C. W. Poag, ed., Geologic evolution of Earth and Planetary Science Letters, v. 271, no. 1–4, the United States Atlantic margin: New York, Van p. 101–108, doi:10.1016/j.epsl.2008.03.056. Nostrand Reinhold, p. 25–86. Mountain, G., and D. Monteverde, 2012, If you’ve got the Parsons, B., and J. G. Sclater, 1977, An analysis of the vari- time, we’ve got the depth: The importance of accurate ation of ocean floor bathymetry and heat flow with age: core-seismic correlation: American Geophysical Union Journal of Geophysical Research, v. 82, no. 5, p. 803–827, Fall Meeting Abstracts, San Francisco, December 3–7, doi:10.1029/JB082i005p00803. 2012, abstract PP51B-2111. Petersen, K. D., S. Nielsen, O. Clausen, R. Stephenson, and Mountain, G., J. Proust, D. McInroy, and C. Cotterill, and the T. Gerya, 2010, Small-scale mantle convection produces Expedition 313 Scientists, 2010, Proceedings of the In- stratigraphic sequences in sedimentary basins: Science, tegrated Ocean Drilling Program: Volume 313 Expedi- v. 329, no. 5993, p. 827–830, doi:10.1126/science tion Reports: New Jersey Shallow Shelf: Tokyo, .1190115. Integrated Ocean Drilling Program Management In- Poag, C. W., 1985a, Cenozoic and Upper Cretaceous sedi- ternational, doi:10.2204/iodp.proc.313.2010. mentary facies and depositional systems of the New Mountain, G. S., and B. E. Tucholke, 1985, Mesozoic and Jersey slope and rise, in C. W. Poag, ed., Geological evo- Cenozoic geology of the US Atlantic continental slope lution of the United States Atlantic margin: New York, and rise, in C. W. Poag, ed., Geologic evolution of the VanNostrandReinhold,p.343–365.

446 Cretaceous Fluvial-Deltaic Sequences of the Southern Baltimore Canyon Trough Poag, C. W., 1985b, Depositional history and stratigraphic of clastic depositional facies, in C. E. Payton, ed., Seismic reference section for central Baltimore Canyon trough, stratigraphy–Applications to hydrocarbon exploration: in C. W. Poag, ed., Geologic evolution of the United AAPG Memoir 26, p. 165–184. States Atlantic margin: New York, Van Nostrand Schlanger, S., and H. Jenkyns, 1976, Cretaceous oceanic Reinhold, p. 217–263. anoxic events: Causes and consequences: Geologie & Poag, C. W., 1987, The New Jersey transect: Stratigraphic Mijnbouw, v. 55, p. 179–184. framework and depositional history of a sediment-rich Schlische, R. W., M. O. Withjack, and P. E. Olsen, 2003, passive margin: Initial reports of the Deep Sea Drilling Relative timing of CAMP, rifting, continental breakup, Project 95: Washington, DC, US Government Print- and basin inversion: Tectonic significance, in W. Hames, fi – ing Of ce, p 763 817, doi:10.2973/dsdp.proc.95.132 J. G. Mchone, P. Renne, and C. Ruppel, eds., The central .1987. Atlantic magmatic province: Insights from fragments of Poag, C. W., and W. D. Sevon, 1989, A record of Appala- Pangea: Washington, DC, American Geophysical Union chian denudation in postrift Mesozoic and Cenozoic Geophysical Monograph Series 136, p. 33–59, doi: sedimentary deposits of the US middle Atlantic conti- 10.1029/136GM03. nentalmargin:Geomorphology,v.2,no.1–3, p. 119–157, Scholle, P. A., 1977, Geological studies on the COST No. B-2 doi:10.1016/0169-555X(89)90009-3. well, US mid-Atlantic outer continental shelf area, Cir- Poag,C.W.,andP.C.Valentine,1988,MesozoicandCeno- zoic stratigraphy of the United States Atlantic continen- cular: Reston, Virginia, US Geological Survey Circular tal shelf and slope, in R. E. Sheridan and J. A. Grow, 750, 71 p., doi:10.3133/cir750. eds., The Atlantic continental margin: US: The geology Seker, Z., 2012, Cretaceous well-log and sequence strati- of North America: Boulder, Colorado, The Geological graphic correlation of the outer continental shelf and ’ Society of America, v. I-2, p. 67–86, doi:10.1130 upper slope off of New Jersey, Master s thesis, Rutgers /DNAG-GNA-I2.67. University, New Brunswick, New Jersey, 155 p, doi: Posamentier, H., and D. James, 1993, An overview of 10.7282/T3X06624. sequence-stratigraphic concepts: Uses and abuses, in Seton, M., R. Muller,¨ S. Zahirovic, C. Gaina, T. Torsvik, H.Posamentier,C.Summerhayes,B.U.Haq,and G. Shephard, A. Talsma, M. Gurnis, M. Turner, and G. P. Allen, eds., Sequence stratigraphy and facies as- S. Maus, 2012, Global continental and ocean basin sociations: Ghent, Belgium, International Association of reconstructions since 200 Ma: Earth-Science Reviews, Sedimentologists Special Publication 18, p. 3–18, v. 113, no. 3–4, p. 212–270, doi:10.1016/j.earscirev doi:10.1002/9781444304015. .2012.03.002. Posamentier, H., M. Jervey, and P. Vail, 1988, Eustatic Sheridan, R. E., and J. A. Grow, 1988, The Atlantic conti- controls on clastic deposition I—Conceptual framework, nental margin, U.S.: The geology of North America, in C. K. Wilgus, B. S. Hastings, C. G. S. C. Kendall, Boulder, Colorado, Geological Society of America, v. I-2, H. W. Posamentier, C. A. Ross, and J. C. Van Wagoner, 610 p., doi:10.1130/DNAG-GNA-I2. eds., Sea level changes: An integrated approach: Tulsa, Sikora, P. J., and R. K. Olsson, 1991, A paleoslope model of Oklahoma, SEPM Special Publication 42, p. 109–124, late Albian to early Turonian foraminifera of the western doi:10.2110/pec.88.01.0109. Atlantic margin and North Atlantic basin: Marine Mi- Post, P., R. Klazynski, E. Klocek, T. DeCort, T. Riches, K. Li, cropaleontology, v. 18, no. 1–2, p. 25–72, doi:10.1016 E. Elliott, and R. Poling, 2012, Assessment of un- /0377-8398(91)90005-Q. discovered technically recoverable oil and gas resources Spasojevic, S., and M. Gurnis, 2012, Sea level and vertical of the Atlantic outer continental shelf 2011 as of Jan- motion of continents from dynamic earth models since uary 1, 2009: Washington, DC, Bureau of Ocean En- the Late Cretaceous: AAPG Bulletin, v. 96, no. 11, ergy Management, 39 p. p. 2037–2064, doi:10.1306/03261211121. Prather, B., 1991, Petroleum geology of the Upper Jurassic Spasojevic, S., L. Liu, and M. Gurnis, 2009, Adjoint models and Lower Cretaceous, Baltimore Canyon trough, western of mantle convection with seismic, plate motion, and North Atlantic Ocean: AAPG Bulletin, v. 75, no. 2, stratigraphic constraints: North America since the Late p. 258–277. Reynolds, D. J., M. S. Steckler, and B. J. Coakley, 1991, Cretaceous: Geochemistry Geophysics Geosystems, The role of the sediment load in sequence stratigraphy: v. 10, no. 5, 24 p., doi:10.1029/2008GC002345. The influence of flexural isostasy and compaction: Journal Steckler, M., and A. Watts, 1978, Subsidence of the Atlantic- ofGeophysicalResearch.SolidEarth,v.96no.B4, type continental margin off New York: Earth and Plan- – p. 6931–6949, doi:10.1029/90JB01914. etary Science Letters, v. 41, no. 1, p. 1 13, doi:10.1016 Rowley, D. B., A. M. Forte, R. Moucha, J. X. Mitrovica, /0012-821X(78)90036-5. N. A. Simmons, and S. P. Grand, 2013, Dynamic to- Steinkraus, W. E., 1981, Preliminary biostratigraphic report, pography change of the eastern United States since 3 OCS-A 00075 #1, Baltimore Canyon, Block 17: Mobil, million years ago: Science, v. 340, p. 1560–1563, doi: 2p. 10.1126/science.1229180. Steinkraus, W. E., and J. W. Bebout, 1985, Biostratigraphic Sangree, J., and J. Widmier, 1977, Seismic stratigraphy and report, OCS-A 0097 #1, Baltimore Canyon, Block 273: global changes of sea level, part 9: Seismic interpretation Shell, 3 p.

SCHMELZ ET AL. 447 Steinkraus, W. E., H. L. Cousminer, and R. E. Hall, 1985a, Vail, P. R., 1987, Seismic stratigraphy interpretation using Biostratigraphic report, OCS-A 0131 #1, Baltimore sequence stratigraphy: Part 1: Seismic stratigraphy in- Canyon, Block 495: Tenneco, 15 p. terpretation procedure, in E. W. Bally, ed., Atlas of seismic Steinkraus, W. E., C. Fry, and L. Bielak, 1985b, Bio- stratigraphy:AAPGStudiesinGeology1,p.1–10. stratigraphic report, OCS-A 0096 #1, Baltimore Canyon, Van Wagoner, J. C., R. Mitchum, K. Campion, and Block 272: Shell, 3 p. V. Rahmanian, 1990, Siliciclastic sequence stratig- Sugarman, P., D. Monteverde, R. Pristas, M. Girard, J. Boyle, raphy in well logs, cores, and outcrops: Concepts K. Miller, J. Browning, Y. Fan Reinfelder, P. Romero, for high-resolution correlation of time and facies: and A. Kulpecz, 2011, Characterization of the carbon dioxide storage potential beneath the New Jersey coastal AAPG Methods in Exploration Series 7, 55 p. plain: Preliminary characterization of CO sequestration Watts, A., G. Karner, and M. Steckler, 1982, Lithospheric 2 fl potential in New Jersey and the offshore coastal region: exure and the evolution of sedimentary basins: Trenton, New Jersey, New Jersey Geological Survey and Philosophical Transactions of the Royal Society of Columbus, Ohio, Battelle, p. 1–44. London A: Mathematical: Physical and Engineering Taner, M. T., and F. Koehler, 1969, Velocity spectra—Digital Sciences, v. 305, no. 1489, p. 249–281, doi:10.1098 computer derivation applications of velocity functions: /rsta.1982.0036. Geophysics, v. 34, no. 6, p. 859–881, doi:10.1190/1 Widess, M., 1973, How thin is a thin bed?: Geophysics, v. 38, .1440058. no. 6, p. 1176–1180, doi:10.1190/1.1440403. Triezenberg, P., P. Hart, and J. Childs, 2016, National Ar- Withjack, M. O., R. W. Schlische, and P. E. Olsen, 1998, chive of Marine Seismic Surveys: A USGS data website Diachronous rifting, drifting, and inversion on the passive for the curation and dissemination of marine seismic margin of central eastern North America: An analog reflection profile data within the US Exclusive Economic for other passive margins: AAPG Bulletin, v. 82, no. 5, Zone (EEZ) and Extended Continental Shelf: Reston, – Virginia, US Geological Survey, doi:10.5066/F7930R7P. p. 817 835. Turgeon, S. C., and R. A. Creaser, 2008, Cretaceous oceanic Zapecza, O. S., 1989, Hydrogeologic framework of the anoxic event 2 triggered by a massive magmatic episode: New Jersey coastal plain: Reston, Virginia, US Geo- Nature, v. 454, no. 7202, p. 323–326, doi:10.1038 logical Survey Professional Paper 1404-B, 61 p., doi: /nature07076. 10.3133/pp1404B.

448 Cretaceous Fluvial-Deltaic Sequences of the Southern Baltimore Canyon Trough