Evaluation and Modelling of Tertiary Source Rocks in the Central Arctic Ocean

ARTICLE IN PRESS

Marine and Petroleum Geology xxx (2009) 1–16

Contents lists available at ScienceDirect

Marine and Petroleum Geology

journal homepage: www.elsevier.com/locate/marpetgeo

Evaluation and modelling of Tertiary source rocks in the central Arctic Ocean

Ute Mann a,*, Jochen Knies b, Shyam Chand b, Wilfried Jokat c, Ruediger Stein c, Janine Zweigel a,1 a SINTEF Petroleum Research, N-7465 Trondheim, Norway b Geological Survey of Norway, N-7491 Trondheim, Norway c Alfred Wegener Institute for Polar and Marine Research, D-27568 Bremerhaven, Germany article info abstract

Article history: With the recently recovered organic-rich sediments of early Tertiary age from the Lomonosov Ridge by Received 1 April 2008 the Integrated Ocean Drilling Program (IODP) Expedition 302, the first data collection directly from Received in revised form source rocks of the central basins of the Arctic Ocean is now available. Using the results of seismic 16 December 2008 interpretations and published sedimentological and organic geochemical data from Expedition 302, the Accepted 5 January 2009 framework for the first quantitative assessment of source-rock quality and distribution of the Palaeogene Available online xxx sediments was modelled in the central Arctic Ocean. The modelling results suggest that an approximately 100-m-thick Early to Middle Eocene sedimentary sequence of good to very good source rocks Keywords: Source-rock modelling exists along a 75 km long transect across the Lomonosov Ridge. In-situ generation of hydrocarbons is w Source-rock potential unlikely because the overburden ( 200–250 m) and consequently the thermal maturity are too low. Tertiary Burial history and thermal modelling reveal that an additional overburden of at least 1000 m is necessary Central Arctic Ocean to start hydrocarbon generation along the ridge. However, source-rock modelling results show that good IODP Leg 302 source-rock potential may exist in correlative units in the adjacent Amundsen Basin. Simulated organic Lomonosov Ridge carbon contents of 1.5–5%, coupled with an overburden of w1000–1200 m, and heat flow anomalies (117 Amundsen Basin and 100 mW m2) due to the vicinity to the Gakkel Ridge spreading centre indicate that necessary conditions for hydrocarbon expulsion are already reached, and point to viability of a potential petroleum system. Our results support the hypothesis that deposition of a potentially good hydrocarbon source rock occurred across the entire Arctic Basin and adjacent margins during the early Tertiary. Ó 2009 Elsevier Ltd. All rights reserved.

1. Introduction unrecognized petroleum systems (Montgomery, 2005 and references therein). A potential source-rock unit for these systems might Hydrocarbon discoveries along the Arctic Alaskan margins be the organic-rich, lower Eocene section of the Canning Formation (Mackenzie Delta–Prudhoe Bay), the Canadian Arctic Islands (Mikkelsen Tongue) (Montgomery, 2005) which has organic carbon (Sverdrup–Ellesmere Basin), and on the Eurasian shelves (southern (TOC) contents typically between 1 and 2 wt% and maximum values Barents Sea, western Siberia) (Fig. 1) demonstrate that favourable up to 12.3 wt% (Keller et al., 1999). Tertiary source rocks have also conditions for hydrocarbon generation and entrapment are wide- been recognized in the Beaufort Mackenzie Basin in Arctic Canada, spread in the Arctic Ocean region (USGS World Petroleum Assess- where Snowdon et al. (2004) reported that the organic-rich ment, 2000). The primary source of these oil and gas accumulations Richards Formation and early Eocene coal-rich units have is thought to be source-rock units of Palaeozoic and Mesozoic age contributed to the Tertiary oil accumulations of the Beaufort Shelf. (e.g. Leith et al., 1992; Pinous et al., 2001; Peters et al., 2006; Vys- More recently, the recovery of organic-rich, lower-middle sotski et al., 2006). In contrast, Tertiary oils in the Beaufort Mack- Eocene sediments from the Lomonosov Ridge by the Integrated enzie basin off northwestern Canada appear to be derived from Ocean Drilling Program (IODP) Expedition 302 (Fig. 1)(Backman organic-rich, middle-upper Eocene deposits (Richards Sequence) et al., 2006), coupled with evidence from organic-rich Eocene (e.g. Brooks, 1986a,b). Recently, a new assessment of the hydro- deposits on the New Siberian Islands (Kos’ko and Trufanov, 2002), carbon resources along the Arctic Alaskan margin suggests that has given rise to speculations that widespread, organic-rich, Eocene and Miocene sequences have given rise to previously potential source rocks might be present across the entire Arctic Basin and its margins (Durham, 2007). These strata are characterised by the widespread occurrence of large quantities of the * Corresponding author. Tel.: þ47 73591243; fax: þ47 73591102. E-mail address: [email protected] (U. Mann). freshwater fern Azolla deposited during the onset of the middle 1 Present address: B. StatoilHydro ASA, N-7501 Stjørdal, Norway. Eocene (w50 Ma) (Brinkhuis et al., 2006; Houseknecht et al., 2007).

Please cite this article in press as: Mann, U., et al., Evaluation and modelling of Tertiary source rocks in the central Arctic Ocean, Marine and Petroleum Geology (2009), doi:10.1016/j.marpetgeo.2009.01.008 ARTICLE IN PRESS

2 U. Mann et al. / Marine and Petroleum Geology xxx (2009) 1–16

Fig. 1. Bathymetry of the Arctic Ocean (Jakobsson et al., 2003) showing locations of the IODP Expedition 302 boreholes 2A and 4A and seismic lines discussed in the text on the Lomonosov Ridge (LR) and in the Amundsen Basin (Jokat et al., 1992; Jokat and Micksch, 2004). Yellow asterisks indicate approximate locations of Azolla fern discoveries reported by Brinkhuis et al. (2006) and Houseknecht et al. (2007).AR¼ Alpha Ridge, MR ¼ Mendeleev Ridge, GR ¼ Gakkel Ridge, MD ¼ Mackenzie Delta. (For interpretation of the references to colour in this ﬁgure legend, the reader is referred to the web version of this article.)

The well-stratiﬁed organic-rich sediments probably resulted from Mod (Mann and Zweigel, 2008) provides quantitative analyses of an euxinic ‘‘Black Sea-type’’ (Stein et al., 2006) depositional envi- source-rock potential and prediction of source-rock quality in regions ronment that developed across the entire Arctic Ocean and may where well control is limited or lacking. The modelling was applied have triggered formation of highly proliﬁc source rocks. here to test if hydrocarbon source rocks formed in the central Arctic In this study, we apply new proxy data generated from IODP Ocean during the early Tertiary and if they could have been wide- Expedition 302 core material (Holes 2A, 4A, w200–400 mcd) (Back- spread throughout the Arctic. man et al., 2006) for 2D/3D source-rock modelling of the central Arctic Stein (2007) concluded that the organic-rich sediments depos- Ocean including the Lomonosov Ridge and the adjacent Amundsen ited on the Lomonosov Ridge during the early-middle Eocene have Basin (Fig. 1). The organic facies/source-rock modelling software OF- a (fair to) good source-rock potential. He determined, however, that

SU 1600 AWI-91090

1800 LR6 2000 LR5 LR4 2200 LR3 ? ? 2400 Two-way traveltime (msec.) Amundsen Basin Makarov Basin

Fig. 2. The seismic reflection profile AWI-91090 from the Lomonosov Ridge showing seismic units LR6–LR3 defined by Jokat et al. (1995).

U. Mann et al. / Marine and Petroleum Geology xxx (2009) 1–16 3

- 43 a Water Depth(m) Age: 56.2 Ma 0 43 86 129 171 214 257 300 343 386 water depth (m) N

M0004 W E

0 water depth (m) 42 - 46 84 64 126 174 283 168 393 (m) 210 503 252 612 722 294 Amundsen Basin 832 Makarov Basin 336 941 378 1051 79 72 65 58 51 44 37 30 23 16 9 2 (km)

- 46 Water Depth (m) Age: 44.4 Ma b 64 174 283 393 503 612 722 832 941 1051 water depth (m)

W M0004 E 400 480 water depth (m) 560 - 46 640 64 720 174 283 800 393 (m) 880 503 960 612 722 1040 Amundsen Basin 832 1120 Makarov Basin 941 1051 1200 79 72 65 58 51 44 37 30 23 16 9 2 (km)

Fig. 3. 2D/3D reconstruction of palaeowater depth for the Lomonosov Ridge transect at (a) w56.2 Ma and (b) 44.4 Ma. the possibility of in-situ generation of hydrocarbons was not additional objective of this study is to analyse the source-rock possible due to the immaturity of the sediments and suggested potential of the basins adjacent to the Lomonosov Ridge (e.g. the instead, that the presence of hydrocarbons in the sediments is most Amundsen Basin) (Fig. 1), and to investigate if burial and temper- likely due to more deeply buried equivalents of the Eocene sedi- ature in these deeper basinal areas are sufﬁcient to initiate ments deposited in adjacent basins (Stein, 2007). Hence, an hydrocarbon generation.

4 U. Mann et al. / Marine and Petroleum Geology xxx (2009) 1–16

2. Geological setting and stratigraphic framework 2006) recently confirmed that the Lomonosov Ridge comprises a sliver of continental crust that was rifted away from the Barents/ The Lomonosov Ridge divides the Arctic Ocean into two oceanic Kara Sea margin (Jokat et al., 1992, 1995; Jokat, 2005). Rift related basins, the Amerasia Basin and the Eurasia Basin (Fig. 1). The subsidence below sea level started during the early Tertiary present shapes and extents of these basins were created by at least (w57 Ma) (Moran et al., 2006). This timing corresponds to the oldest two rifting events, one during the Mesozoic creating the Canada magnetic anomaly identified in the Amundsen Basin, tentative and Makarov basins, and the second during the Cenozoic forming anomaly 25 (w56 Ma), implying the onset of rifting began prior to the Nansen and Amundsen basins (e.g. Heezen and Ewing, 1961; w56 Ma (Brozena et al., 2003). Water depths in the Nansen Basin Grantz et al., 1979; Jokat et al., 1992). The initial break-up of the range between 3000 and 4000 m with very flat topography in the Canada Basin is widely accepted to have occurred during the early central part of the basin. The Amundsen Basin has the same hori- Cretaceous (w135 Ma) (Grantz et al., 1979, 1990; Vogt et al., 1979), zontal dimensions as the Nansen Basin, but is slightly deeper and was likely completed by the early Late Cretaceous (w80 Ma). (4500 m). Estimated sediment thicknesses vary between 0 and The geological evolution of the Alpha/Mendeleev Ridge and the 4.5 km for the Nansen Basin and 1.7–2.0 km for the Amundsen Basin Makarov Basin is unclear. Taylor et al. (1981) suggested the (Jokat and Micksch, 2004). Crustal thicknesses of the basins are Makarov Basin was formed by seafloor spreading between 84 and estimated to range from 3 to 6 km by Weigelt and Jokat (2001). 49 Ma, whereas Grantz et al. (1990) argued it formed as part of the The modelling approach in this study is applied to early Canada Basin in the Early Cretaceous. Cenozoic sediments deposited on the Lomonosov Ridge and in the The Gakkel Ridge, a midocean ridge that represents the northern adjacent Amundsen Basin (Fig. 1). Borehole data (Holes 2A and 4A) continuation of the Mid-Atlantic ridge, divides the Eurasia Basin into recovered during IODP Expedition 302 (Backman et al., 2006) and the Nansen and Amundsen basins, which are bounded by the seismic reflection data (AWI-091090, AWI-20010300) (Jokat et al., Barents/Kara Sea shelves (south) and the Lomonosov Ridge (north) 1992; Jokat and Micksch, 2004) were used as input for model (Fig. 1). A clear pattern of seafloor magnetic anomalies demonstrates calibration (see details below). The stratigraphic framework for that the Eurasia Basin was formed by seafloor spreading beginning the Lomonosov Ridge was established from geophysical data during the early Palaeogene (Vogt et al., 1979; Kristoffersen, 1990; (Jokat et al., 1992; Jokat et al., 1995)(Fig. 2) and was recently Brozena et al., 2003). Results from IODP Expedition 302 (Moran et al., modified by IODP Expedition 302 biostratigraphic data (Backman

Sed. Rate (cmka-1) Sand Fraction (%) modelled data measured data modelled data measured data 0 2 4 6 8 10 0 5 1015 20 25 1486 44.4 Ma 1511 44.6 Ma

1508 1533 1531

1555 1553

1577 1575 Azolla 48.6 Ma

1599 1599

(m)

1621 1621

1643 1643

1665 PETM 1665 55 Ma

1687 1687 55.9 Ma

1709 1709 0 246810 0 51015 2025

0 246 8 10 (cmka-2) 0 51015 2025(%)

Fig. 4. Plots of measured and modelled lithological parameters (sedimentation rate (cm ka1); sand fraction (%)) versus depth at the location of IODP Expedition 302 boreholes (composite of Holes 2A and 4A). Measured data were taken from Backman et al. (2006, 2008).

U. Mann et al. / Marine and Petroleum Geology xxx (2009) 1–16 5

W M0004 E 400

480 Sed. Rate (cmka-2) 560 0 640 2 4 720 6 800 8 (m) 880 10 12 960 14 1040 Amundsen Basin 16 1120 Makarov Basin 18 20 1200 79 72 65 58 51 44 37 30 23 16 9 2 (km)

W M0004 E 400 480 Sand Frac. (%) 560 0 640 2.5 5 720 7.5 800 10 (m) 880 12.5 15 960 17.5 1040 Amundsen Basin 20 Makarov Basin 1120 22.5 25 1200 79 72 65 58 51 44 37 30 23 16 9 2 (km)

Fig. 5. Cross-section showing simulated variations in sedimentation rate (top) and sand fraction (bottom) between 56.2 and 44.4 Ma along the Lomonosov Ridge transect. Borehole position is indicated by drilling symbol. et al., 2008). An upper seismically reflective unit (w430 m) of 3. Modelling approach Cenozoic age rests in angular unconformity on lower, less reflective seismic units. Strata lying on the unconformity have an age of To estimate the volume of potential source rocks in the Palae- w56.2 Ma (Backman et al., 2008). Jokat et al. (1995) divided the ogene sediments in the central Arctic Ocean several modelling reflective seismic unit into four subunits (LR3–LR6). A disconfor- techniques have been applied. The techniques include SEMI mity separates the upper (LR5–LR6) from the lower seismic PaleoWater software, a tool to restore basin topography (Kjennerud subunits (LR3–LR4) (Fig. 2) and represents a hiatus spanning the and Sylta, 2001; Kjennerud and Gillmore, 2003), the forward period w44.4–18.2 Ma (Backman et al., 2008). Our target interval source-rock modelling software OF-Mod (Mann and Zweigel, for the modelling is the sediment sequence deposited between 2008), and finally the PetroMod 1D Express software for burial and both unconformities (i.e. subunits LR3–LR4) covering the period thermal history modelling (download at http://www.iesgmbh.eu/ from the Late Palaeocene (w56.2 Ma) to the middle Eocene index.php). Sedimentological and stratigraphic input data were (w44.4 Ma) (Backman et al., 2008). The TOC content in this adopted from Backman et al. (2006, 2008). Applied geochemical sequence varies considerably, ranging from 1 to 6 wt% (Stein et al., data including TOC content and Rock-Eval pyrolysis data are pub- 2006). lished in Stein et al. (2006) and Stein (2007). The nitrogen data are The Lomonosov Ridge seismic stratigraphy was projected into taken from Knies et al. (2008). the Amundsen Basin using downlapping reflectors and the age of The source-rock modelling software OF-Mod (Organic Facies the underlying seafloor (Jokat et al., 1995). Thereby, Jokat et al. Modelling) is a process-based modelling tool, which simulates (1995) correlated the seismic units AB3–AB6 in the Amundsen the development and variation of organic facies along 2D tran- Basin with subunits LR3 and LR4 on the Lomonosov Ridge. The sects or in 3D-grids (Mann and Zweigel, 2008). Consequently, it southward dip of these units (w56 and w43 Ma, chron 20) clearly allows a quantitative prediction of source-rock potential away identifies the Lomonosov Ridge as their source area. The sediments from well control. OF-Mod considers the main factors and their were deposited during the early drift phase of the ridge, when interactions which control the deposition of organic carbon in erosion occurred prior to its subsidence below sea level. Sedi- siliciclastic sediments (i.e. shales and sandstones). For further mentation that followed the ultimate submergence of the ridge details on the concept and the capabilities of the OF-Mod soft- occurred at more constant sedimentation rates. A second profile ware we refer to Mann and Zweigel (2008) and Knies and Mann acquired in 2001 (Jokat and Micksch, 2004) revealed the existence (2002). of the downlapping units in other parts of the Amundsen Basin, An important input parameter is the initial basin topography or, which indicates that they were not the result of a local sedimen- because OF-Mod is restricted to the marine/aquatic realm, the tation regime. palaeowater depth. This parameter is important because processes

6 U. Mann et al. / Marine and Petroleum Geology xxx (2009) 1–16

Fig. 7. Cross-plot between the Rock-Eval hydrogen index (HI) (in mg HC/g TOC) and

the maximum pyrolytic hydrocarbon generation temperatures (Tmax) from IODP Expedition 302 material (composites of boreholes 2A and 4A). Fields of kerogen types and vitrinite reﬂection values taken from Isaksen and Ledje (2001).

each time layer of the basin-fill model as a basis for the subsequent organic sedimentation modelling. Fig. 6. Cross-plots between (a) the amorphous and aquatic macerals (in %) identified from kerogen microscopy and the fraction of inorganic nitrogen in the total nitrogen fraction (% Ninorg) and (b) the sum of amorphous and terrigenous macerals (in %) and % 4. New proxy data for input in source-rock modelling Ninorg in IODP Expedition 302 boreholes (composite of Holes 2A and 4A). Original petrographical data are adapted from Stein et al. (2006) and Boucsein and Stein (2008). For process-based source-rock modelling it is necessary to quantify how much total organic carbon is derived from marine/ of organic matter accumulation are simulated at the time of aquatic sources and from terrestrial sources. Traditionally, TOC and deposition. Palaeowater depth restorations for the interpreted Rock-Eval data combined with kerogen microscopy and paly- seismic line AWI-91090, which stretches over 80 km across the nofacies results are used to estimate proportions of various organic Lomonosov Ridge have been performed and provide the bounding matter (OM) types (marine/algal vs. terrestrial) and the preserva- surfaces for the depositional model (Fig. 3). We applied the SEMI tion conditions (Stein, 1991; Tyson, 1995). Alternatively, various PaleoWater software (Kjennerud and Sylta, 2001), which addresses fractions of the total nitrogen content of a sediment (Ntot) can isostatic response (Airy isostasy), correction for compaction, represent different sources (aquatic vs. terrigenous) of OM and may thermal subsidence and eustatic sea-level variations in the course be quantitatively separated. Recent studies from the marginal, of back stripping procedure. We used a simplified lithology model modern Arctic Ocean demonstrate the potential of using relative that assumes 100% shale for all units and (de)compaction after amounts of inorganic nitrogen (Ninorg) and organic nitrogen (Norg) Sclater and Christie (1980). A constant thermal subsidence has been to track inputs of terrigenous (TOM) and aquatic/marine organic applied along the whole section. We assumed thermal subsidence matter (MOM) to marine sediments (Winkelman and Knies, 2005; to have started at 54 Ma according to Moore and the Expedition 302 Knies et al., 2007). Scientists (2006). Fig. 3 shows the restored palaeowater depth Applying this latter approach to the Palaeogene sediments from distribution for the Lomonosov Ridge transect. the central Arctic Ocean with published petrographical data In the next step, we used OF-Mod to fill inorganic sediment (macerals) (Boucsein and Stein, 2008), we identify a prominent between the bounding unconformity surfaces (Fig. 4). The OF-Mod negative correlation (R2 ¼ 0.8) between aquatic (marine/algal) model has a grid resolution of 397 50 200 cells splitting the macerals and % Ninorg (Fig. 6). It supports the conclusion drawn time interval from 56.2 to 44.4 Ma into 200 layers. This resulted in from surface sediments that % Ninorg (% Norg) may be useful to a horizontal spatial resolution of 200 m per cell and a temporal quantitatively separate the marine from the terrigenous OM frac- resolution of about 60 ka per time step (Fig. 5). The sea-level curve tion. However, some caveats exist in the data set. The amount of from Miller et al. (2005) as well as the age assignments, the sand amorphous organic matter (AOM) originally considered being fraction data, and sedimentation rates from Expedition 302 bore- predominantly of marine origin (Stein et al., 2006) is inconsistent holes 2A and 4A were used to calibrate the stratigraphic model with the linear regression between aquatic macerals and % Ninorg (Figs. 4 and 5)(Backman et al., 2006, 2008). Based on that, we used (Fig. 6). Stein et al. (2006) described the AOM, forming up to 55% of

U. Mann et al. / Marine and Petroleum Geology xxx (2009) 1–16 7

assignment to either MOM or TOM. Boucsein and Stein (2008) concluded based on detailed petrographical investigations of IODP Expedition 302 material that the AOM is composed of both TOM and MOM, respectively. Thus, using the proportions of Ninorg and Norg in the total N fraction, we may obtain quantitative information on the MOM and TOM (AOM) supply in the Arctic Ocean during the Paleogene. The % Ninorg values range between 30 and 100 (Fig. 6) implying a strong variability of OM supply during the course of the Palaeogene. This indicates a much higher contribution of TOM than formerly suggested (Stein et al., 2006). The latter is further expressed by the prevalence of kerogen type III (terrestrial OM) and type II (marine OM) inferred from cross-plotting HI values against the maximum pyrolytic hydrocarbon generation temperatures (Tmax)(Fig. 7). The plot shows that immature, freshly derived TOM is a dominant component in the whole sequence. To obtain quantitative information on palaeoproductivity changes in the Arctic Ocean during the Palaeogene, we estimated the palaeoproductivity (PP) in surface waters from % Norg-derived MOM data of the underlying sediments by applying the equations published by Mann and Zweigel (2008) and Knies and Mann (2002). The PP calculation was modified by considering the carbon flux through the water column and the burial efficiency in the sediment, as well as applying a preservation factor to account for the enhanced organic carbon preservation and the special environmental conditions that prevailed during early (PETM) and mid- (Azolla) Eocene (see Knies et al., 2008 for details).

4.1. The Palaeogene spatiotemporal evolution of the depositional environment from a modelling perspective

TOM (þAOM) and MOM flux rates and estimated marine palaeoproductivity of Paleogene strata on the Lomonosov Ridge are shown in Fig. 8. Average accumulation rates of organic carbon are 2 1 Fig. 8. Accumulation rate of marine (black) and terrestrial (grey) organic carbon (in g highest at the base and the top (0.14 g C cm ka ), while lower cm2 ka1) and estimated palaeoproductivity under oxic (PP1) and anoxic (PP2) values (w0.04 g C cm2 ka1) prevail during the middle parts of the conditions (in g C m2 a1) in IODP Expedition 302 material (composites of boreholes 1 sequence. The highest accumulation rates of MOM occurred in the 2A and 4A). PP was calculated by applying the equations of Mann and Zweigel (2008) w 2 1 and Knies and Mann (2002) and considering carbon flux through the water column early-middle Eocene ( 0.07 g C cm ka )(Fig. 8). TOM accumu- and burial efficiency. For PP2 a preservation factor of 1% was introduced to account for lation rates, in contrast, were enhanced during the late Palaeocene increased preservation during distinct time intervals, e.g. the PETM and the Azolla and late middle Eocene (w0.1 g C cm2 ka1)(Fig. 8), which is events. consistent with interpretations of bulk organic and kerogen microscopy data (Stein et al., 2006; Stein, 2007, 2008). total macerals in samples from the Paleocene–Eocene Thermal Estimated surface-water palaeoproductivity (PP) on the Lomo- Maximum (PETM), as unstructured and weakly fluorescent bitu- nosov Ridge was highest during periods of maximum supply of minite. This material, however, is associated with rather low Rock- MOM (e.g. Azolla event) (Brinkhuis et al., 2006; Knies et al., 2008; Eval hydrogen index (HI) values of 150–300 mg HC/g TOC as Boucsein and Stein, 2008), but displays generally strong fluctua- observed during the PETM (Stein et al., 2006) and is indicative of tions during the Palaeogene (Fig. 8). Considering the euxinic envi- poorly preserved AOM (Tyson, 1995). This is somewhat surprising ronment and increased OM preservation during the early (PETM) given the fact that generally good OM preservation conditions and middle Eocene Azolla interval, as suggested by Stein et al. prevailed in the central Arctic Ocean during the PETM (Sluijs et al., (2006), Sluijs et al. (2006), and Brinkhuis et al. (2006), calculated PP 2006). According to Tyson (1995), one possible explanation for this values under anoxic conditions should be slightly lower (PP2, discrepancy could be that abundant AOM is attributed to Fig. 8), since the low oxygen conditions in sediments would have allochthonous OM. The observation that AOM may be derived from favoured preservation of labile marine OM. The estimated Palae- huminic acids (Taylor et al., 1998) supports this view. By following ogene PP is comparable to PP in modern, euxinic environments, the assumption that Ninorg in marine sediments is allochthonous in such as enclosed and/or silled oceanic basins like the Baltic Sea or origin (Knies et al., 2007), the positive linear regression (R2 ¼ 0.8) Black Sea (Romankevich, 1984; Berger et al., 1989; Stein, 1991; between % Ninorg and the sum of terrigenous macerals and AOM Antoine et al., 1996). For development of these ‘‘Black Sea-type’’ (Fig. 6) might be another argument in addition to the low fluores- conditions, Sluijs et al. (2006) and Stein et al. (2006) suggested that cence and HI values for assigning the AOM to an allochthonous widespread salinity stratification would have been likely in, at that source rather than considering it to be produced as primary time, an isolated Arctic Ocean. These authors further suggested that marine/algal OM. Another option might be scavenging and higher increased primary production due to enhanced fluvial runoff might adsorbance of AOM on fine-grained material in the water column have amplified development of euxinic conditions. In contrast, thereby accelerating the transfer of microbially degraded OM, i.e. Knies et al. (2008) argued, that moderate PP values (90 g AOM, from surface waters to the deep sea. This would explain the Cm2 a1) in highly stratified waters were sufficient to sustain affinity of AOM to % Ninorg, however, it does not justify the anoxic bottom-water environments throughout the early-middle

8 U. Mann et al. / Marine and Petroleum Geology xxx (2009) 1–16

TOC (wt%) MOC (wt%) t-TOC (wt%) modelled data measured data modelled data measured data modelled data measured data 012345 012345 012345 1486 44.4 Ma 1511 44.6 Ma

1508 1533 1531

1555 1553

1577 Azolla 1575 48.6 Ma 1599 1599 (m) 1621 1621

1643 1643

1665 PETM 1665 55 Ma

1687 1687 55.9 Ma

1709 1709 012345 012345 012345

012345(%) 012345(%) 012345(%)

Fig. 9. Plots of measured and modelled total organic carbon (TOC), marine organic carbon (MOC), and terrestrial organic carbon (T-TOC) contents versus depth at the location of IODP Expedition 302 boreholes (composite of Holes 2A and 4A). Measured TOC data are published in Stein et al. (2006). Stratigraphic data are adopted from Backman et al. (2008).

Eocene as it is the case in the central part of the Black Sea. More- agreement with calculated PF in Holocene anoxic deep-water over, increased PP in surface waters due to enhanced fluvial environments (Bralower and Thierstein, 1984). Fig. 9 shows that the nutrient supply is rather unlikely, as evidence for atmospheric N2 modelled values of TOC, marine organic carbon (MOC), and fixation during this time interval implies absence of additional terrigenous organic carbon (T-TOC) at the borehole location (at nitrogen sources (Knies et al., 2008). Thus, lateral transport of TOM 58.6 km) show an overall good match with the measured values. may be facilitated by isopycnal diffusion (i.e. diffusion along Modelled hydrogen indices (HI) slightly exceed measured values a constant density surface) below the pycnocline. This process is (Fig. 10), but agree very well with HI0 which is the TOC-normalised known from different continental margin settings (Walsh and sum of both the hydrocarbon potential (S2) and the hydrocarbons Nittrouer, 1999; Hwang et al., 2004; McPhee-Shaw, 2006) and has generated already (S1) (Fig. 10). This agrees with the suggestion by been invoked previously to explain the impregnation of 14C Stein (2007) that very labile organic compounds generated during depleted, ‘‘old’’ dissolved and particulate organic carbon in central low temperature pyrolysis contribute to the S2 peak (and thus HI) gyres of the North Atlantic and North Pacific (Bauer and Druffel, in very immature samples. Indicators of low maturity 1998). It would explain the long-distance transport of TOM from (Tmax < 400 C, vitrinite reflection Ro w0.25%) during these inter- Siberian shelves to the Lomonosov Ridge without providing addi- vals support this view (Stein, 2007). tional nutrients to the photic zone. We therefore envisage a depo- Cross-sections of simulated TOC and HI values along the Lomo- sitional environment on the Lomonosov Ridge that was nosov Ridge show very little lateral variability due to the large lateral characterised by moderate primary production (i.e. PP2) in less extent of the undisturbed sediments (Fig. 11) deposited after the saline surface waters and by laterally transported material below ridge subsided below sea level (Jokat et al., 1992, 1995). The highest the pycnocline that may have contained fresh TOM supplied by TOC contents are identified in early-Middle Eocene deposits during enhanced river discharge. Moderate PP values (max. PP2 ¼ w60– the Azolla event (w48.6 Ma) and at 44.6 Ma (>5 wt% TOC), respec- 70 g C m2 a1) in surface water correspond to inferences that the tively. The lower values towards the Makarov Basin are likely due to water column was anoxic during most of the middle Eocene (Stein, higher sedimentation rates. The expected OM quality is expressed by 2007; Weller and Stein, 2008). the modelled HI values (Fig. 11). Maximum and minimum values Applying the calculated PP for anoxic environments (PP2) and nearly follow the simulated TOC contents. Generally, modelled HI TOM data as input parameters for our model, we are able to test the values vary between 200 and 350 mg HC/TOC within the early- robustness of the modelling results (Fig. 9). Except for short periods Middle Eocene sequence, while distinctly lower values (<200 mg during the early Eocene where oxic conditions prevailed, an anoxic HC/TOC) occur towards the base and top of the entire sequence depositional environment was suggested for the entire time (Fig. 11). Interestingly, the TOC maximum (w6 wt%) at 44.6 Ma interval. The same conclusion as that reached from bulk organic directly below the hiatus (44.4–18.2 Ma) (Backman et al., 2008)is and biomarker data (Sluijs et al., 2006, 2008; Stein et al., 2006; not accompanied by higher HI values (Fig. 11). In contrast, HI values Weller and Stein, 2008). A variable preservation factor (PF) of about of w100 mg HC/TOC indicate a significant contribution of TOM to 1% with a maximum value up to 2.5% was applied which is in good the sediments. With the applied subsidence model (Moore and the

U. Mann et al. / Marine and Petroleum Geology xxx (2009) 1–16 9

HI (mgHC/gC) HI HI´ (S1+S2)/TOC modelled data measured data measured data 0 100 200 300 400 0 100 200 300 400 500 1486 44.4 Ma 1511 44.6 Ma

1508 1533

1531

1555 1553

1577 1575 Azolla 48.6 Ma

1599 1598 (m)

1621 1621

1643 1643

1565 PETM 1565 55 Ma

1687 1687 55.9 Ma

1709 1709 0 100 200 300 400 500

0 100 200 300 400 500 (mgHC/gC)

Fig. 10. Plots of measured and modelled hydrogen indices (HI) and the corresponding quantity of pyrolyzable hydrocarbons (S2) per gram TOC (mg HC g1 TOC) versus depth at the location of IODP Expedition 302 boreholes (composite of Holes 2A and 4A). Measured HI and S2 values are published in Stein et al. (2006). For comparison purposes, measured HI0 values ((S1 þ S2)/TOC in mg HC/g TOC) are also shown. A reference line is set at 250 mg HC/g TOC.

Expedition 302 Scientists, 2006) and a modelled palaeowater depth (1986) and recalculated the source-rock potential from the of about 500 m, the most plausible explanation might be lateral modelled TOC and HI values (Fig. 12). Applying this approach to transport of TOM by isopycnal eddy diffusion from Siberian shelves. the borehole data, we confirmed the results of Stein (2007), who More recently, O’Regan et al. (2008) proposed an alternative model found that most of the Eocene sediments on the Lomonosov Ridge focusing on the mid-Eocene onset of tectonic uplift, and on vertical have a good source-rock potential. Indeed, we have identified an movement of the Lomonosov Ridge that culminated in subaerial approximately 100 m thick Lower to Middle Eocene sedimentary exposure during the w25 Ma long hiatus. If this exposure can be sequence of good to very good source-rock potential (Fig. 12). The verified, and if parts of the Lomonosov Ridge were already exposed source rocks can be followed along the entire 75-km-long seismic at 44.6 Ma, then they might have acted as a source of high amounts transect across the Lomonosov Ridge. This sequence is thicker but of TOM in the adjacent marine basins. Analogues to such a scenario of lower quality towards the Makarov Basin, whereas it is are well known (e.g. Saller et al., 2006 and references therein) and condensed and has a very good source potential towards the provide indications that high input of TOM by rivers results in TOC Amundsen Basin (Fig. 13). The thermal immaturity of the organic- values that are sufficiently high to indicate good source-rock rich sediments in the borehole is inferred from Tmax and vitrinite formation, as recent observations/simulations from the Mahakam reflectance data (Stein, 2007; Boucsein and Stein, 2008). Higher Delta (Saller et al., 2006) and Svalbard margin (Knies and Mann, maturities along the transect cannot be expected as overburden 2002) have suggested. thickness is generally less than 250 m (Fig. 13). Burial and thermal history modelling suggests that hydrocarbon generation and 5. Source-rock potential and hydrocarbon prospectivity in expulsion in this area would require an additional overburden of the central Arctic Ocean at least 1000 m thickness (Fig. 14). To test the hypothesis that hydrocarbons generated from more Because the Rock-Eval parameters S1 and S2 are not modelled mature sequences migrated into the organic-rich mid-Eocene in OF-Mod, we simplified the original classification by Peters deposits on the Lomonosov Ridge (Stein, 2007), we extended the

W M0004 E 400 480 TOC (wt%) 560 0 640 0.5 1 720 1.5 800 2 (m) 880 2.5 3 960 3.5 1040 4 Amundsen Basin 1120 Makarov Basin 4.5 5 1200 79 72 65 58 51 44 37 30 23 16 9 2 (km)

W M0004 E 400 480 HI (mgHC/gC) 560 0 640 50 720 100 150 800 200 (m) 880 250 960 300 350 1040 Amundsen Basin 400 Makarov Basin 1120 450 1200 500 79 72 65 58 51 44 37 30 23 16 9 2 (km)

Fig. 11. Cross-section showing simulated variations in TOC content (wt%) (top) and HI (mg HC/g TOC) (bottom) between 56.2 and 44.4 Ma along the Lomonosov Ridge transect. Borehole position is indicated by drilling symbol.

Source Rock Potential 44.4 Ma 1511 44.6 Ma

1533

1555

1577 Azolla

48.6 Ma

1599 (m) 1621 Source Rock Potential 500

400 1643 Very Good

300

PETM 1565 200 55 Ma Good HI mgHC/gC 100 1687 55.9 Ma Poor Fair 0 012345678910 1709 TOC % 1234 poor fair good very good

Fig. 12. (right) Source-rock potential classes based on HI and TOC values according to Peters (1986). (left) Modelled source-rock potential in the Lomonosov Ridge borehole. Stratigraphic data are taken from Backman et al. (2008). Hatchured areas indicate no recovery.

U. Mann et al. / Marine and Petroleum Geology xxx (2009) 1–16 11

0 50 100 150 200 250 overburden thickness (m) (km) 80 70 60 50 40 30 20 10 0

M0004

M0004 W source rock potential E 400 480 560 640 720 800 (m) 880 960 1040 Amundsen Basin 1120 Makarov Basin 1200 79 72 65 58 51 44 37 30 23 16 92 (km)

Fig. 13. Cross-section showing simulated source-rock potential in sediments deposited between 56.2 and 44.4 Ma along the Lomonosov Ridge transect (bottom) and corresponding overburden thickness (in metres) (top). Borehole position is indicated by drilling symbol. modelling approach into the adjacent Amundsen Basin. We well-stratiﬁed waters, TOM supply from river discharge, and ‘‘Black assumed that the environmental conditions for organic matter Sea-type’’ anoxia). The main argument for this assumption is the deposition during the middle Eocene in the Amundsen Basin were presence of the freshwater fern Azolla in organic-rich Middle like those inferred for the Lomonosov Ridge (i.e. moderate PP in Eocene (48.6 Ma) deposits, not only on the Lomonosov Ridge

Age (Ma) 56 50 40 30 20 10 0 CENOZOIC Pg Ng 0

200

Layer_6 Additional 400 modelled overburden

Temperature [Celsius] 20 40 60 80 100 0

200

400

Layer_6 Layer_5 Overburden 600 Temperature (C°) Neogene Hiatus 0 –15 Layer_3 Mid Eocene 800 15 –30 incl. Azolla Event 30 –45 45 –60 Layer_2 Late early Eocene

Depth [meter] 1000 60 –75 75 –90 Layer_1 Layer_5 90 –105 Early early Eocene incl. PETM 1200 105 –120 Layer_3 120 –135 135 –150 Layer_2 1421

Fig. 14. 1D thermal and burial history modelling for IODP Expedition 302 borehole (Lomonosov Ridge). Model shows that an additional 1000 m overburden and a constant heat ﬂow of 100 mW m2 are required to initiate hydrocarbon generation.

12 U. Mann et al. / Marine and Petroleum Geology xxx (2009) 1–16

a W E 2000 2200 TOC (wt%) 2400 0 0.5 2600 1 1.5 2800 2 3000 2.5

(m) 3 3200 3.5 4 3400 4.5 3600 5 Lomonosov Ridge 3800 Gakkel Ridge 4000 120 108 96 84 72 60 48 36 24 12 0 (km)

W E b 2000 HI 2200 (mgHC/gC) 2400 0 40 2600 80 120 2800 160 3000 200

(m) 240 3200 280 320 3400 360 3600 400 Lomonosov Ridge 3800 Gakkel Ridge 4000 120 108 96 84 72 60 48 36 24 12 0 (km)

Fig. 15. Cross-section showing (a) simulated variations in TOC content (wt%) and (b) HI (mg HC/g TOC) in the Amundsen Basin between 50 and 44.4 Ma. The stratigraphic interpretation of this sequence is based on the seismo-stratigraphic framework of Jokat et al. (1995) and Jokat and Micksch (2004).

(Brinkhuis et al., 2006), but also in coeral deposits in the Canada the entire Arctic Basin during the Early to Middle Eocene. The fact Basin, including the Alaskan Beaufort Shelf and the Chukchi Shelf that the Arctic Ocean was a tectonically closed basin until the Early (Houseknecht et al., 2007). This distribution implies that favourable Miocene (Jakobsson et al., 2007) and that lower and middle Eocene conditions for deposition of organic matter may have existed across organic-rich deposits are present in Alaska (Montgomery, 2005)

820700 940 1060 1180 1300 overburden thickness (m)

pseudo well position W E 2000 Source 2200 Rock Potential 2400 poor

2600 fair 2800 good 3000

(m) very 3200 good 3400 3600 Lomonosov Ridge 3800 Gakkel Ridge 4000 120 108 96 84 72 60 48 36 24 12 0 (km)

Fig. 16. Modelled source-rock potential (cf. Peters, 1986) (bottom) and overburden thickness (top) in the Amundsen Basin between 50 and 44.4 Ma. Black arrows mark the Azolla horizon (48.6 Ma). The position of a ‘‘pseudo’’ well shown in Fig. 17 is indicated.

U. Mann et al. / Marine and Petroleum Geology xxx (2009) 1–16 13

Lomonosov Ridge Amundsen Basin M0004 core position Pseudo well position 4040 44.4 Ma

1517 4080

1527 4120

1537 4160

1547 4200

1557 4240 (m)

1567 4280

1577 4320

Azolla Event ~ 48.6 Ma 1587 4360

1597 4400

1607 4440 50 Ma

source rock potential poor fair good very good

Fig. 17. Plots of source-rock potential versus depth on the Lomonosov Ridge (Site M0004) and in the Amundsen Basin (pseudo-well position at 23.8 km, see Fig. 16).

and in Beaufort Mackenzie Basin in Canada (Dixon et al., 1992; Basin, we applied a constant preservation factor of 4% (Bralower Snowdon et al., 2004) further support this view. and Thierstein, 1984). This is somewhat higher than that used at the Applying the seismo-stratigraphic model from the Amundsen Lomonosov Ridge sediments due to inferred deeper water and Basin (seismic line AWI-20010300; Fig. 1) (see Jokat et al., 1995; potentially more sluggish circulation. Jokat and Micksch, 2004) as framework for the OF-Mod modelling, The simulated TOC contents vary around 2.5 wt%, the highest we reconstructed the palaeowater depth with the same input values (7.1 wt%) occurring in sediments of a condensed section parameter set as outlined above. However, the modelled time between 36 km and 54 km (Fig. 15a). Modelled HI values are on interval of the Amundsen Basin model comprises only 44.4–50 Ma average slightly lower than on the Lomonosov Ridge, although and the transect covers only the first 120 km from the Lomonosov occasionally exceeding 350 mg HC g1 TOC (Fig. 15b). Remarkably, Ridge into the Amundsen Basin. The reason for this is that in the the source-rock potential of these lateral equivalents of the Amundsen Basin the sediment sequence rests on oceanic crust (cf. Lomonosov Ridge organic-rich deposits remains generally good, Jokat and Micksch, 2004). As a result of the spreading at the Gakkel with excellent conditions during the Azolla event (w48.6 Ma) (Figs. Ridge, this crust represents a diachronous surface with an age of ca. 16 and 17). The latter is consistent with observations in the 56 Ma towards the Lomonosov Ridge but younging (45 Ma) Amerasia Basin and adjacent margins (e.g. Houseknecht et al., towards the Gakkel Ridge spreading centre (Jokat and Micksch, 2007) implying the existence of a widespread, organic-rich source 2004). A palaeowater depth reconstruction for the lowermost rock across the entire Arctic Basin and its margins. The modelled reflector was therewith not possible since seafloor spreading accumulated thicknesses of lower and middle Eocene rocks having cannot be handled with the SEMI PaleoWater tool. very good and good source potential are noticeably greater in the Modelled sedimentation rates in the Amundsen Basin vary Amundsen Basin (up to 250 m) than on the Lomonosov Ridge (max. between 15 and 20 cm ka1 and increase up to 30 cm ka1 towards 110 m), even though the lower Eocene sediments (56–50 Ma) the Gakkel Ridge. Assuming the prevalence of ‘‘Black Sea-type’’ including the PETM (at w55 Ma) were not considered in the euxinic conditions between w50 and 44.4 Ma in the Amundsen source-rock modelling of the Amundsen Basin transect (Fig. 18).

14 U. Mann et al. / Marine and Petroleum Geology xxx (2009) 1–16

Age (Ma) 50 45 40 35 30 25 20 15 10 5 0 CENOZOIC Pg Ng 0

Layer_3 500 Overburden Neogene

1000

Layer_2 Mid Eocene incl. Azolla Event

Temperature [Celsius] 20 40 60 80 100 120

0 Layer_1 Early Eocene 200 incl. PETM 400 Layer_3 Temperature (C°) 600 0 –15 15 –30 800 30 –45 1000 45 –60 Layer_2 60 –75 Depth [meter] 1200 75 –90 90 –105 1400 105 –120 Layer_1 120 –135 135 –150 1700

Fig. 18. 1D thermal and burial history modelling at pseudo-well position in the Amundsen Basin. The location of the pseudo-well is indicated in Figs. 1 and 16. A slightly higher heat ﬂow varying between 117 and 100 mW m2 (Drachev et al., 2003) than at the Lomonosov Ridge was assumed due to the closer proximity of the pseudo-well to the Gakkel Ridge spreading centre.

Lomonosov Ridge E

44.4 Ma

56.2 Ma ? W Amundsen Basin ?

M0004

~ 44.4 Ma

~ 47 Ma

~ 50 Ma ~ 56 Ma

Accumulated thickness very good and good source rock potential (m) 0 50 100 150 200 250

Fig. 19. Accumulated thicknesses of rocks having very good and good hydrocarbon source potential on the Lomonosov Ridge and in the Amundsen Basin plotted against their respective seismic proﬁles. Note that the duration of the time interval with good or very good potential is shorter in the Amundsen Basin (w44.4–50 Ma) than in the Lomonosov Ridge (44.4–56.2 Ma).

U. Mann et al. / Marine and Petroleum Geology xxx (2009) 1–16 15

Burial history and thermal modelling data (Fig. 19) further suggest Brinkhuis, H., et al., 2006. Episodic fresh surface waters in the Eocene Arctic Ocean. that, due to the thicker overburden and the slightly higher Nature 441, 606–609. 2 Brooks, P.W., 1986a. Biological marker geochemistry of crude oils and condensates geothermal gradient (between 117 and 100 mW m ; Drachev from the Beaufort–Mackenzie Basin. Bulletin of Canadian Petroleum Geology et al., 2003), the Eocene sediments in the Amundsen Basin may 34, 490–505. have generated and expelled hydrocarbons, which possibly Brooks, P.W., 1986b. Unusual biological marker geochemistry of oils and possible source rocks, offshore Beaufort–Mackenzie Delta, Canada. Organic Geochem- migrated into the age equivalent deposits on the Lomonosov Ridge istry 10, 401–406. (Stein, 2007). Brozena, J.M., Childers, V.A., Lawner, L., Gahagan, L.M., Forsberg, R., Faleide, J.I., Eldholm, O., 2003. New aerogeophysical study of the Eurasia Basin and Lomonosov Ridge: implications for basin development. Geology 31, 825–828. 6. Conclusions Dixon, J., Dietrich, J., Snowdon, L.R., Morrell, G., McNeil, D.H., 1992. Geology and petroleum potential of upper Cretaceous and Tertiary Strata, Beaufort–Mack- Seismic, borehole and modelling data from the central Arctic enzie Area, Northwest Canada. AAPG Bulletin 76 (6), 927–947. Drachev, S.S., Kaul, N., Beliaev, V.N., 2003. Eurasia spreading basin to Laptev Shelf Ocean suggest the existence of organic-rich early and middle transition: structural pattern and heat flow. Geophysical Journal International Eocene sedimentary rocks that extend for at least 75 km across the 152 (3), 688–698. Lomonosov Ridge and have mainly good (partly very good) source- Durham, L.S., 2007. Source ideas boost Arctic promise. AAPG Explorer 28, p. 6, 8, 14. Grantz, A., Eittreim, S., Dinter, D.A., 1979. Geology and tectonic development of the rock potential. In-situ generation of hydrocarbons from this continental margin north of Alaska. Tectonophysics 59, 263–291. sequence is impossible, because it was never buried deeper than Grantz, A., May, S.D., Taylor, P.T., Lawver, L.A., 1990. Canada Basin. In: Grantz, A., w200 m. Thermal modelling data reveal that hydrocarbon expul- Johnson, G.L., Sweeney, W.J. (Eds.), The Arctic Region. Geological Society of sion would require an additional overburden of at least w1000 m. America, Boulder, pp. 379–402. Heezen, B.C., Ewing, M., 1961. The Mid-Ocean Ridge and its extension through the By applying the constraints for the depositional environment Arctic Basin. In: Raasch, G.O. (Ed.), Geology of the Arctic. University of Toronto derived from proxy data of the Lomonosov Ridge to the adjacent Press, Toronto, pp. 622–642. Amundsen Basin, we found evidence that thicker sequences of Houseknecht, D.W., Bird, K.J., Bujak, J., 2007. Petroleum systems of emerging and future importance in the Arctic Alaska petroleum province. In: AAPG Annual mainly good source-rock potential exist there, too. In the Amund- Convention and Exhibition, April 1–4, 2007, Long Beach, California, USA sen Basin, conditions for hydrocarbon generation are more (Abstract). favourable than on the Lomonosov Ridge because an overburden of Hwang, J., Druffel, E.R.M., Griffin, S., Smith Jr., K.L., Baldwin, R.J., Bauer, J.E., 2004. Temporal variability of d14C, d13C, and C/N in sinking particulate organic matter >1000 m is indicated by seismo-stratigraphic data and because at a deep time series station in the northeast Pacific Ocean. Global Biogeo- higher heat flow may be present due to closer proximity to the chemical Cycles 18, GB4015, doi:10.1029/2004GB002221. Gakkel Ridge spreading centre. Expulsion and lateral migration of Isaksen, G.H., Ledje, K.H.I., 2001. Source rock quality and hydrocarbon migration pathways within the greater Utsira High area, Viking Graben, Norwegian North hydrocarbons from the Amundsen Basin may explain the presence Sea. AAPG Bulletin 85 (5), 861–883. of hydrocarbons in equivalent deposits on the adjacent Lomonosov Jakobsson, M., Grantz, A., Kristoffersen, Y., Macnab, R., 2003. Physiographic prov- Ridge. The data also suggest that good potential source rocks may inces of the Arctic Ocean. Geological Society of America Bulletin 115, 1443– 1455. have been deposited across the entire Arctic Basin during the early Jakobsson, M., et al., 2007. The early Miocene onset of a ventilated circulation Tertiary. regime in the Arctic Ocean. Nature 447, 986–990. Jokat, W., Micksch, U., 2004. Sedimentary structure of the Nansen and Amundsen basins, Arctic Ocean. Geophysical Research Letters 31, L02603, doi:10.1029/ Acknowledgments 2003GL018352. Jokat, W., Uenzelmann-Neben, G., Kristoffersen, Y., Rasmussen, T.M., 1992. Lomo- The samples were provided by the Integrated Ocean Drilling nosov Ridge – a double-sided continental margin. Geology 20, 887–890. Program (IODP). We sincerely thank Hermann Weiss and Maarten Jokat, W., Weigelt, E., Kristoffersen, Y., Rasmussen, T., Scho¨ne, T., 1995. New insights into the evolution of the Lomonosov Ridge and the Eurasian Basin. Geophysical Felix for comments on an earlier draft of the manuscript. Thomas Journal International 122, 378–392. Moore is greatly acknowledged for improving the English language Jokat, W., 2005. The sedimentary structure of the Lomonosov Ridge between 88N and additional scientific comments.We would also like to thank and 80 N: consequences for tectonic and glacial processes. Geophysical Journal International 163, 698–726. Brian Popp and the School of Ocean and Earth Science and Tech- Keller, M.A., Bird, K.J., Evans, K.R., 1999. Petroleum source rock evaluation based on nology, SOEST, at the University of Hawaii for hosting U. Mann and sonic and resistivity logs. In: The Oil and Gas Resource Potential of the Arctic J. Knies during preparation of the manuscript. This work was sup- National Wildlife Refuge. U.S. Geological Survey Open-File Report 98-34, Menlo Park, California, 62 pp. ported by the Norwegian Research Council (Leiv Eiriksson Mobility Kjennerud, T., Gillmore, G., 2003. Integrated Palaeogene palaeobathymetry of the Program). northern North Sea. Petroleum Geoscience 9 (2), 125–132. Kjennerud, T., Sylta, Ø., 2001. Application of quantitative paleobathymetry in basin modelling, with reference to the northern North Sea. Petroleum Geoscience 7 References (4), 331–341. Knies, J., Brookes, S., Schubert, C.J., 2007. Re-assessing the nitrogen signal in Antoine, D., Andre´, J.-M., Morel, A., 1996. Oceanic primary production. 2. Estimation continental margin sediments: new insights from the high northern latitudes. at global scale from satellite (coastal zone color scanner) chlorophyll. Global Earth and Planetary Science Letters 253, 471–484. Biogeochemical Cycles 10, 57–69. Knies, J., Mann, U., 2002. Depositional environment and source rock potential of Backman, J., Jakobsson, M., Frank, M., Sangiorgi, F., Brinkhuis, H., Stickley, C., Miocene strata from the central Fram Strait: introduction of a new computing O’Regan, M., Lovlie, R., Pa¨like, H., Spofforth, D., Gattacecca, J., Moran, K., King, J., tool for simulating organic facies variations. Marine and Petroleum Geology 19, Heil, C., 2008, Age model and core-seismic integration for the Cenozoic Arctic 811–828. Coring Expedition sediments from the Lomonossov Ridge, Paleoceanography, Knies, J., Mann, U., Popp, B., Stein, R., Brumsack, H.-J., 2008. Surface water 23, PA1S03, doi:10.1029/2007PA001476. productivity and paleoceanographic implications in the Cenozoic Arctic Ocean. Backman, J., Moran, K., McInroy, D.B., Mayer, L.A., the Expedition 302 Scientists, Paleoceanography 23, PA1S16, doi:10.1029/2007PA001455. 2006. Arctic Coring Expedition (ACEX). In: Proc. Integrated Ocean Drilling Kos’ko, M.K., Trufanov, G.V., 2002. Middle Cretaceous to Eopleistocene sequences on Program, 302, College Station, Texas. the New Siberian Islands: an approach to interpret offshore seismic. Marine and Bauer, J.E., Druffel, E.R.M., 1998. Ocean margins as a significant source of organic Petroleum Geology 19, 901–919. matter to the deep open ocean. Nature 392, 482–485. Kristoffersen, Y., 1990. Eurasia Basin. In: Grantz, A., Johnson, G.L., Sweeney, W.J. Berger, W.H., Smetacek, V.S., Wefer, G., 1989. Ocean productivity and paleo- (Eds.), The Arctic Region. Geological Society of America, Boulder, pp. 365–378. productivity – an overview. In: Berger, W.H., Smetacek, V.S., Wefer, G. (Eds.), Leith, T.L., Weiss, H.M., Mørk, A., Århus, N., Elvebakk, G., Embry, A.F., Brooks, P.W., Productivity of the Ocean: Present and Past. John Wiley, Hoboken, NJ, pp. 1–35. Stewart, K.R., Pchelina, T.M., Bro, E.G., Verba, M.L., Danyushevskaya, A., Boucsein, B., Stein, R., 2008. Black shale formation in the late Paleocene/early Borisov, A.V., 1992. Mesozoic hydrocarbon source-rocks of the Arctic region. In: Eocene Arctic Ocean and paleoenvironmental conditions: new results from Vorren, T.O., Bergsager, E., Dahl-Stamnes, Ø.A., Holter, E., Johansen, B., Lie, E., a detailed organic petrological study. Marine and Petroleum Geology, Lund, T.B. (Eds.), Arctic Geology and Petroleum Potential. Norwegian Petroleum doi:10.1016/j.marpetgeo.2008.04.00. Society (NPF), Special Publication, vol. 2, pp. 1–25. Bralower, T.J., Thierstein, H.R., 1984. Low productivity and slow deep-water circu- Mann, U., Zweigel, J., 2008. Modelling source rock distribution and quality variations: lation in mid-Cretaceous oceans. Geology 12, 614–618. the organic faciesmodelling approach. In: deBoer,P.L., Postma,G., vanderZwan, C.J.,

16 U. Mann et al. / Marine and Petroleum Geology xxx (2009) 1–16

Burgess, P.M., Kukla, P. (Eds.), Analogue and Numerical Forward Modelling of Snowdon, L.R., et al., 2004. Organic geochemistry and organic petrology of Sedimentary Systems; from Understanding to Prediction. Special Publication of the a potential source rock of early Eocene age in the Beaufort–Mackenzie Basin. International Association of Sedimentologists, vol. 40, pp. 239–274. Organic Geochemistry 35, 1039–1052. McPhee-Shaw, E., 2006. Boundary–interior exchange: reviewing the idea that Stein, R., 1991. Accumulation of Organic Carbon in Marine Sediments. In: Lecture internal-wave mixing enhances lateral dispersal near continental margins. Notes in Earth Sciences, 34. Springer Verlag, Berlin, 217 pp. Deep Sea Research II 53, 42–59. Stein, R., 2007. Upper Cretaceous/Lower Tertiary black shales near the North Pole: Miller, K.G., et al., 2005. The Phanerozoic record of global sea-level change. Science organic-carbon origin and source rock potential. Marine and Petroleum 310, 1293–1298. Geology 24, 67–73. Montgomery, S.L., 2005. Petroleum geology and resource assessment: 1002 area, Stein, R., Boucsein, B., Meyer, H., 2006. Anoxia and high primary production in the Arctic Wildlife Refuge. AAPG Bulletin 89 (3), 291–310. Paleogene central Arctic Ocean: ﬁrst detailed records from Lomonosov Ridge. Moore, T.C., the Expedition 302 Scientists, 2006. Sedimentation and subsidence Geophysical Research Letters 33, L18606, doi:10.1029/2006GL026776. history of the Lomonosov Ridge. In: Backman, J., Moran, K., McInroy, D.B., Stein, R., 2008. Arctic ocean sediments: processes, proxies, and palaeoenvironment. Mayer, L.A., Expedition 302 Scientists, Arctic Coring Expedition (ACEX) (Eds.), Developments in Marine Geology, Vol. 2, Elsevier, Amsterdam, 587 pp. Proc. Integrated Ocean Drilling Program, 302, College Station, Texas. Taylor, G.H., Teichmu¨ ller, M., Davis, A., Diessel, C.F.K., Littke, R., Roberts, P., 1998. Moran, K., et al., 2006. The Cenozoic palaeoenvironment of the Arctic Ocean. Nature Organic Petrology: a New Handbook Incorporating Some Revised Parts of Sta- 441, 601–605. ch’s Textbook of Coal Petrology. Gebru¨ der Borntra¨ger, Berlin, Stuttgart, 704 pp. O’Regan, M., et al., 2008. Mid-Cenozoic tectonic and paleoenvironmental setting of Taylor, P.T., Kovacs, L.C., Vogt, P.R., Johnson, G.L., 1981. Detailed aeromagnetic the central Arctic Ocean. Paleoceanography 23, PA1S20, doi:10.1029/ investigation of the Arctic Basin, 2. Journal of Geophysical Research 86, 6223– 2007PA001559. 6333. Peters, K.E., 1986. Guidelines for interpreting petroleum source rock using pro- Tyson, R., 1995. Sedimentary Organic Matter: Organic Facies and Palynofacies. grammed pyrolysis. AAPG Bulletin 70, 318–329. Chapman and Hall, London, 615 pp. Peters, K.E., Magoon, L.B., Bird, K.J., Valin, Z.C., Keller, M.A., 2006. North Slope, USGS World Energy Assessment Team, 2000. U.S. Geological Survey World Petro- Alaska: source rock distribution, richness, thermal maturity, and petroleum leum Assessment 2000 – Description and Results, DDS-60. U.S. Geological charge. AAPG Bulletin 90 (2), 261–292. Survey Digital Data Series, Denver, Colorado. Pinous, O.V., Levchuk, M.A., Sahagian, D.L., 2001. Regional synthesis of the Vogt, P.R., Taylor, P.T., Kovacs, L.C., Johnson, G.L., 1979. Detailed aeromagnetic productive Neocomian complexes of West Siberia: sequence stratigraphic investigations of the Arctic Basin. Journal of Geophysical Research 84, 1071– framework. AAPG Bulletin 85 (10), 1713–1730. 1089. Romankevich, E.A., 1984. Geochemistry of Organic Matter in the Ocean. Springer Vyssotski, A.V., Vyssotski, V.N., Nezhdanov, A.A., 2006. Evolution of the West Verlag, Berlin, 334 pp. Siberian Basin. Marine and Petroleum Geology 23, 93–126. Saller, A., Lin, R., Dunham, J., 2006. Leaves in turbidite sands: the main source of oil and Walsh, J.P., Nittrouer, C.A., 1999. Observations of sediment ﬂux to the Eel continental gas in the deep water Kutei Basin, Indonesia. AAPG Bulletin 90 (10), 1585–1608. slope, northern California. Marine Geology 154, 55–68. Sclater, J.G., Christie, P.A.E., 1980. Continental stretching: an explanation of the post- Weigelt, E., Jokat, W., 2001. Peculiarities of roughness and crustal thickness of mid-Cretaceous subsidence of the central North Sea Basin. Journal of oceanic crust in the Eurasian Basin, Arctic Ocean. Geophysical Journal Inter- Geophysical Research 85 (B7), 3711–3739. national 145, 505–516. Sluijs, A., et al., 2006. Subtropical Arctic Ocean temperatures during the Paleocene/ Weller, P., Stein, R., 2008. Paleogene biomarker records from the central Arctic Eocene thermal maximum. Nature 441, 610–613. Ocean (Integrated Ocean Drilling Program Expedition 302): organic carbon Sluijs, A., Ro¨hl, U., Schouten, S., Brumsack, H.-J., Sangiorgi, F., Sinninghe Damste´, J.S., sources, anoxia, and sea surface temperature. Paleoceanography 23, PA1S17, Brinkhuis, H., 2008. Arctic late Paleocene–early Eocene paleoenvironments doi:10.1029/2007PA001472. with special emphasis on the Paleocene–Eocene thermal maximum (Lomono- Winkelman, D., Knies, J., 2005. Recent distribution and accumulation of organic sov Ridge, Integrated Ocean Drilling Program Expedition 302). Paleoceanog- carbon on the continental margin west off Spitsbergen. Geochemistry, raphy 23, PA1S11, doi:10.1029/2007PA001495. Geophysics, Geosystems 6, Q09012, doi:10.1029/2005GC000916.