• Abstract and Figures
• Public Full-text

INTERPRETATION OF CHANNELIZED ARCHITECTURE USING THREE

DIMENSIONAL PHOTO REALISTIC MODELS, PENNSYLVANIAN

DEEPWATER DEPOSITS AT BIG ROCK QUARRY, ARKANSAS

ABSTRACT

Mapping the geological details and interpreting 3three-Ddimensional geometryies in a highly heterogeneous outcrop such as the exposure at Big Rock Quarry has been a continuous challenge especially due to the fact that at this specific location location largehigh vertical cliffs the steepness of the cliff faces makes access to most of the rocks difficult for direct geological observations and sampling. Previous interpretations of facies architecture were derived from gamma-ray profiles, a core and observations and measurements made on two-dimensional photomosaics. In this study field observations integrated with a three-dimensional photorealistic model of the outcrop with assigned lithologies effectively helped in reconstruction of submarine channel architecture.This paper represents the first attempt of 3three-Ddimensional interpretation of the geometry and facies pattern of the Jackfork nested-channel complex deposited at the base-of-slope. Examination of the three-dimensional photo realistic model of the outcrop with assigned lithologies allowed extraction of accurate 3-D qualitative (lithology, contacts), as well as, quantitative (bed, channel width and thicknessdimensions) geometric information. This facilitated interpretation and reconstruction of the submarine channel complex architecture making possible correlations of strata exposed on the two distinct sides of the quarry. Three dimensional photorealistic mapping techniques have been recently developed as effective tools for detailed outcrop studies, but still their potential has not been fully exploited. This study makes effective use of the quantitative information incorporated in three-dimensional virtual outcrops and provides new tools for geologic mapping and interpretation. The three-dimensional model of sedimentary bodies allowed capturing the three dimensional spatial distribution of lithological units, which is fundamentally important

1 for understanding the internal architecture of erosional and depositional features, in this case channelized features. Most of the exposed vertically and laterally stacked channels channels are large, have aggradational with well defined axial regions composed of high relief basal erosional surfaces overlain by well developed intraformationalmatrix- supported conglomerate/ breccia which grades upward into amalgamated sandstones. or thin-bedded sandstone interlayered with shale. The thickness of the sandstone decreases toward the southeastern end of the quarry where more shale is present. The channel infill here consists of thin-bedded sandstones interlayered with shale which overlain the breccia. The upper part of the quarry is made up of smaller, lateral migrating channels. Significant channel width/thickness variation can be recognized at outcrop scale. The 348 identified channels are characterized by a relatively low aspect ratio (width/thickness ratio runs from 4:1 to 3239:1) with channel dimensions ranging from 3025 m to 265314 m wide and 2 m to 2424 m deep. Compared to previous attempts of reconstruction our 3-D virtual model is more realistic and because the model has accurate real dimensions it can be used to calibrate simulation of processes in deep water environments. The three-dimensional model of sedimentary bodies allowed capturing the three dimensional spatial distribution of lithological units, which is fundamentally important for understanding the internal architecture of erosional and depositional features, in this case channelized features. Compared to previous reconstructions our 3-D virtual model is more realistic and because the model has accurate real dimensions it can be used to calibrate simulation of processes in deep water environments.

INTRODUCTION

2 Understanding the depositional processes and lateral geometry of deep-water base-of-the slope systems is important for predicting the extent and internal architecture of these reservoirs (Bouma et al., 1995; Slatt et al., 2000). Adequately documentation of reservoir properties and flow behavior at reservoir scale Deep-sea sediments have received considerable interest both for research purposes and economic reasons due to large hydrocarbon accumulations associated with turbidite deposits (Bouma, 2000). Ideally, characterization of hydrocarbon reservoirs and flow behavior requiress information about heterogeneity at a submetersub meter scale in three dimensions. However, it is difficult to include realistic distributions of shale and sandstone bodies in reservoir models because of the wide spacing of wells and limited vertical resolution of seismic surveys. One solution is to characterize shale-sandstone distribution using data from large, continuous three-dimensional outcrops outcrop exposures (Coleman et al., 2000; Slatt, 2000).). Accurate mapping of the architecture of these reservoirs requires interpretation of erosional and depositional geometries preserved in outcrops. Channelized facies of the Pennsylvanian Jackfork sandstone exposed at Big Rock Quarry in Arkansas have long been controversial. Originally thought to be fluvial in origin (Taff, 1902) they are now widely agreed to be submarine, although there is still debate about what kind of processes generated them (debris flow, slumps, high vs. low density turbidity flows) and what submarine settings they should be assigned to (upper canyon slope vs. base-of-slope). Also the geometry of these deposits was difficult to interpret due to high variability of bed thicknesses and lateral discontinuities in three dimensions. Previous Excellent cliff faces exposed at Big Rock Quarry were interpreted as proximal deep-water fans within slope channels canyons (Jordan et al., 1993) or at the base-of- slope (Bouma and Cook, 1994), but geometry of these deposits was difficult to be interpreted due to highly variability of bed thicknesses and lateral discontinuities in three dimensions. The upper Jackfork strata that crop out in the main face of the quarry have previously been the focus of several studies into facies architecture and individual bed

3 geometries. Ffacies interpretation was derived from gamma-ray profiles (Jordan et al., 1993), a core drilled about 15 m behind the outcrop face (Link and Stone, 1986) and from direct geological observations where the outcrop was easily accessible. Measurements on large photographic prints (Cook, 1993; Bouma and Cook, 1994) facilitated statistical characterization of width and thickness dimensions, as well as, reconstruction of facies architecture. The three loggamma profiless taken at this site illustrated potential problems in well log correlations of laterally discontinuous strata. When compareding the information from well logs and cores it was obvious that some of the thinner beds could not be identified on the gamma-ray profiles since they are often beneathlow the resolution of the logging tools. Cores provide the necessary resolution for distinguishing these features, but unfortunately there is only one core taken at this site. Also two- dimensional photomosaic interpretations are not effective tools in correlation of strata with a complicated three-dimensional geometry.

Understanding of turbiditic systems has improved with the use of high-resolution shallow 2-D and 3-D seismic data, but the detailed internal architecture of these reservoirs below seismic resolution remains uncertain. Seismic modeling of Big Rock Quarry outcrop indicated that internal sand-body architecture and geometry may not be clearly imageable with conventional seismic profiling (Coleman et al., 2000)) mostly due to the cemented nature of the sediments, providing little or no acoustic impedance contrast between sedimentary intervals. . Due to its horse shoe shape Big Rock Quarry has a horse shoe shape and iis a good candidate for three-dimensional studies imaging and interpretations. This study uses 3-D photorealistic methodology to understand the three-dimensional architecture of the channelized features. Facies architecture and dimensions of channel sandstones are highly anisotropic (width/length/depth). As a consequence it is critical to know how outcrops are oriented within depositional strike/dip to make meaningful quantitative analysis of sand body dimensions. Conventional outcrop analog are hampered by the 2-D nature of data sets and the inability to overcome or moreover to take advantage of parallax. Compared to previous two-dimensional outcrop photomosaics, the three-

4 dimensional virtual model adds more valuable information for outcrop interpretation (Bhattacharya et al., 2002; Aiken et al., 2004). In this study we use innovative 3-D methods to generate accurate dimensional data that represent true dip/strike orientation. These dimensions are compared to those derived from more traditional 2-D methods that have been used to build reservoir analog data bases. The study of bed thickness distributions of various facies can help to identify important reservoir facies within the fill of stacked channels at the base of slope where massive, thick sandstone is interbedded with thinner shales.Using a combination of real- time kinematic global positioning system (RTK-GPS), and laser scanners we were able to capture 3-D terrain data of the outcrop in global coordinates with centimeter accuracy. Oblique close-in photography acquired from the ground was integrated with terrain data and converted into a 3-D digital photorealistic model of the outcrop. Examination of the virtual model of the outcrop allowed extraction of accurate 3-D qualitative (lithology, contacts), as well as, quantitative (bed width and thickness) geometric information. This facilitated interpretation and reconstruction of bed architecture making possible further correlations of strata exposed on distinct sides of the quarry. Our interpretation is based on mapping of three-dimesional facies distributions and collection of new paleocurrent data. Compared to previous two-dimensional outcrop photomosaics, the three- dimensional virtual model adds more valuable information for outcrop interpretation (Bhattacharya et al., 2002).

GEOLOGIC SETTING

Big Rock Quarry is located in the southeastern part of the Ouachita Mountains along the north bank of the Arkansas River in North Little Rock, Arkansas (Fig. 1). The cliff faces of the quarry expose a twohree-dimensional view of the lower part of the upper Jackfork Group (Jordan et al., 1993). The exposure is oriented at different angleshas a horse-shoe shape and is up to 60 m high and almost 1000 1250 m long. In study area Jackfork Group is divided in to the lower Jackfork (Irons Fork Mountain Fm.) and upper Jackfork (Brushy Knob Fm.). The Jackfork Group was dated as Pennsylvanian (Morrowan) (Fig. 2) on the base of correlative units on the shelf and is

5 Pennsylvanian (Morrowan) in age (Fig. 2). and it represents the low stand system tract, time equivalent of a major unconformity on the shelf (Coleman, 2000) (Fig. 2). Sediments that crop out at this quarry were interpreted as fan channel deposits (Stone and McFarland, 1981), stacked channelized packages of an inner (upper) fan valley (Moiola and Shanmungan, 1983), channel-fill and levee deposits in a submarine canyon or upper part of a submarine fan channel system (Link and Stone, 1986; Link and Roberts, 1986) and slope canyon fill generated by retrogressive slope failure (Jordan et al., 1993; Slatt et al., 1997). Bouma and Cook, 1994 and Bouma et al., 1995 considered these sediments as part of a submarine channel complex likely deposited at the base of slope based on the limited development of levees and overflow deposits and on the stacked pattern of the channels. The presence in the southeastern part of the quarry of small, scoured depressions eroded in sandstone bodies and filled with mudstone and sandstone-clast debris flows or drapes of laminated mudstone suggest that scour and successive slope failures have cut out many originally more continuous units (Jordan et al., 1993).

6 Fig.1 Location map. Big Rock Quarry Outcrop Belt

T Coleman et al. (2000)his estimated that this channel complex is at least 9.6 km wide and 16 to 24 km long. and Iit appears to pinch out about 4 km north of the quarry. At least 14 channels are exposed here, some with relief exceeding 3.6 m. Most of the channels have flow indicators oriented west-southwest (Coleman et al., 2000). Shanmungan and Moiola, 1997 suggested that Jackfork sandstones were not deposited by high-density turbidity currents and are predominantly of sandy debris flow origin because traction-generated sedimentary structures are not present and the matrix content in sandstone is high. However this view received much critique (Slatt et al., 1997; Lowe, 1997; Coleman ColemanJr., 1997; Bouma et al., 1997; D’Agostino and Jordan, 1997). The major fill of the base-of-slope channels consist of somewhat clayey sandstone layers because the high-density currents that transported this fill may have not been able to move their very fine grained material in suspension toward their upper part and tail (Bouma et al., 20020).

Facies architecture of the Jackfork channel complex at Big Rock Quarry

The channel complex infill consists mostly of amalgamated sand-rich individual channels. These massive sandstones comprise a significant proportion of the total channel fill and their overall thickness decreases in an easterly direction where more shale is present. This and the fact that the basal upper Jackfork sandstone is thicker at Big Rock

7 than in exposures in the northwest and southeast suggest proximity to the axis of the sandy submarine canyon fill (Jordan et al., 1993). Jackfork channel complex consist of 38 erosive-based channels that stack up in a nested pattern. The lower two thirds of the quarry are made up of large, sandstone-rich aggradational channels while the upper part consists mostly of small, laterally migrating channels. The larger, aggradational channels have well defined axial regions and display massive fill with tabular geometry in the central part of the exposure while toward the southeastern end the channels display a layered fill with convergent geometry. Mud clasts dispersed in a sandy matrix are common in the channel axis. Sometimes the basal erosional surface is directly overlain by highly amalgamated sandstones. Smaller channels dominate the upper part of the quarry. Some of them display lateral accretion surfaces. Sand lenses at the channel margin incline in the direction of channel migration; most of the time northward. Lateral accretion packages are characterized by interbedded high-concentration turbidites deposited by suspension (massive sandstones) and mud-clast conglomerates deposited by traction as bed load (Abreu et al., 2003). The formation of the sinuous channels at the top of the aggradational channels can be explained by a reduction in the volume of the turbidity currents that preceded the overall abandonment of this part of the turbidite system. In the upper part of the quarry there are also present some intervals with thin bedded sandstone and shale (mud-clast conglomerates) which might represent remnants of laterally equivalent levees of the channels. The levees are absent at the base of the channel complex.

Main ffacies associations of the Jackfork channel complex at Big Rock Quarry

Jackfork channel complex consist of 34 erosive-based channels that stack up in a nested pattern. The lower two thirds of the quarry are made up of sandstone-rich aggradational channels while the upper one third consists of laterally migrating channels. The formation of the sinuous channels at the top of the aggradational channels can be explained by a reduction in the volume of the turbidity currents that preceded the overall abandonment of this part of the turbidite system. Levees are absent at the base of the channel complex, but remnants of bedded levee-upper overbank deposits can be found up

8 in the section. However the lateral extension of the levees is limited. They only formed when the canyon morphology was almost filled. The channel complex infill consists mostly of amalgamated sand-rich individual channels. These massive sandstones comprise a significant proportion of the total channel fill and their overall thickness decreases in an easterly direction where more shale is present. This and the fact that the basal upper Jackfork sandstone is thicker at Big Rock than in exposures in the northwest and southeast suggest proximity to the axis of the sandy submarine canyon fill (Jordan et al., 1993). Jackfork channel succession display layered or massive fill with flat or lenticular geometry; vertical fining and thinning is common. Most of the channels are erosional (conduit for sediment transport into the basin) with scoured contacts (erosion between successive channels is common). Sandstone amalgamation, sediment deformation and rip-up clasts are common. Each individual channel shows a deep basal cut with evidence of by-pass processes (mud-clast breccia).The debris flow deposit grades upward into amalgamated, thick- bedded sandstones or thin bedded, fine–grained sandstone interbedded with shale which in turn are overlain by a drape of shale deposited during the waning stages of channel filling (Fig. 4). The sedimentary section at Big Rock consists of four lithofacies: (1) matrix-supported breccia, (2) massive, thick bedded, fine-grained sandstone, (3) massive to parallel- laminated, fine-grained sandstone,thin-bedded, fine-grained sandstone interbeded with siltstone, shale intraclast breccia with a sandy matrix, shale intraclast breccia with a shale matrix, and (4) finely laminated shale (Link and Stone, 1986; Cook, 1993).

1. Matrix-supported breccia

This facies association is the product of cohesive debris flow and consist of two main groups of facies depending on the nature of the matrix either shale or sand. The breccia with a sandy matrix occurs in the upper part of the outcrop. Its formation might be related to levee collapse since thin-bedded turbidite are incorporated into the sandy matrix.

9 The breccia with a muddy matrix results mostly from the incorporation of slope to pelagic mudstones eroded during transport. The mud clasts are subangular and range in size from few mmmillimeters to centimeters (Fig. 3A). The base of the shaly-matrix breccia is an erosional surface cutting into the underlying channel. This facies constitute non-reservoir units because of their high shale content.

2. Massive, thick-bedded, fine-grained sandstone

Channels located in the central part of the quarry which is considered to be the axis of the canyon (Jordan et al., 1993) are typically filled by massive, amalgamated, thick to very thick (6 - 106 m) tabular to irregular-bedded, fine-grained, and massive sandstones deposited from high-concentration turbidity currents (Fig. 3B). Very few sedimentary structures are preserved except for dish and pillar dewatering structures; flute cast are rare, load structures are present (Fig. 4). The sandstone bodies have flat upper surfaces and undulating lower contacts which are interpreted as channel scours. These sandstones represent Bouma Ta divisions and are the fill of mixed erosive- depositional to depositional channels. Massive Ta divisions (Fig.4C) are the most common Bouma divisions in the section and they are also the thickest.

Individual turbidite beds rarely exceed 50 cm in thickness, but they commonly occur in amalgamated units that have an average thickness of 2m and frequently attain thickness of 6-8m. The sandstone bodies have flat upper surfaces and undulating lower contacts which are interpreted as channel scours.

3. Thin-bedded, fine-grained sandstone interbeded with siltstone

The channel complex margin association is composed of thin bedded, massive to planar stratified, fine–grained sandstone deposited from low-density turbidity currents separated by discontinuous centimeter- scale parallel laminated siltstone and shale (Fig. 3C. ). They represent Bouma Tb, c, d divisions.

10 Sandstone bed thickness ranges from cm to 1m. A large spill-over lobe m thick marks the end of the channel fill. It consists of stacked tabular sandstone beds with plane parallel stratification.

4. Shale (mudstone)

This facies is interpreted either as the pelagic background sedimentation or as a drape of shale deposited during the wanning stage of channel filling. Trace fossils present in shale (few of them are traces burrowed in the sand) indicate a deep water setting. Horizontal pascichnial (Helmintoida) and agrichnial (Urohelmintoida, Megagrapton, Neonereites) trails of the Nereites ichnofacies occur on the bedding planes.

METHODOLOGY

3Three-Dimensional pPhotor realistic mMapping

Three -dimensional photo realistic mapping techniques have been recently developed (Nielsen et al., 1999; Xu et al., 1999; 2000; Thurmond et al., 2000; Xu, 2000) as effective tools for detailed outcrop studies, but their potential has not been fully exploited. This study makes effective use of the quantitative information incorporated in three-dimensional virtual outcrops and provides new tools for geologicquantitative mapping of sedimentary facies and interpretation.

Modeling of three-dimensional outcrops from two-dimensional images has been a challenging problem for geologists. Three dimensional mapping techniques for detailed studies of geologic outcrops have been recently developed (Xu et al., 1999; 2000; Xu, 2000; Xu et al., 2000) so that more quantitative information is available to be interpreted for a better quantitative outcrop mapping.

11 Using a combination of real-time kinematic global positioning system (Leica RTK-GPS 530) and high-density laser scanners (Riegl, LMS-Z360), the surface morphology of the outcrop is captured and can be expressed in a global reference system with centimeter accuracy. Oblique close range photography acquired from the ground can than be integrated with the terrain data and converted into a three-dimensional digital photo realistic model of the outcrop (Fig. 35). CyberMapping software (Xu, 2000) allowed for real time three-dimensional visualization and acquisition of scanner-mapped features. This software is also used to accurately drape photography onto centimeter digital terrain models generated by scanning the outcrop. We show that channels located in the central part of the quarry are exposed again on the NW face close to the river. The quarry walls are spaced 150-300m apart. Bed boundaries were easily followed on the three-dimensional view of the outcrop (Fig. 7, Fig. 8). This way strata exposed in the central part of the quarry were found to appear again to the NW. In order to correlate beds from one wall to the other we paid attention to the lateral continuity/discontinuity of the strata, to the geometry of the channels and their relative position on the outcrop and also we took in consideration the fact that the quarry is tilted about 10˚ toward northwest.

Statistical Aanalysis of cChannel dDimensions and aAspect Rratios

Channel width and depth parameters can easily be obtained from outcrops with sufficient exposure, allowing aspect ratios and cross-sectional areas to be calculated. A statistical analysis was performed Statistical analyses have been performed using parameters like bed thickness and length for each facies; width/depth (thickness) ratio of channel fill was estimated in order to evaluate the potential improvements of 3three-Ddimensional over classical 2two- Ddimensional quantitative analysis methods iIn this study study we usinge dimensional

12 and architectural data (width vs. thickness measurements) of individual channels andto estimate channel aspect ratios (width:depthwidth: depth) and to evaluate the potential improvements of three-dimensional over classical two-dimensional analysis methods were estimated. Measurements of bedchannel lengwidtth and thickness were done on the 3-D photo realistic model of the outcrop built in Gocad using a parallel view, also known as an isometric view. The main use of this type of view is to avoid viewing parallax distortions of length, area, volume and angle that are unavoidable in most 2-D photomosaics. It is very useful for comparing the dimensions of objects that are at different distances from the viewer. Since the quarry face is oriented at different angles and is not a real really cross- section through the channels the beds width/t and thickness wasere projected in on a vertical plane perpendicular to the maiean paleocurrent direction (as measurestimated from the outcrop) and measured accordingly. This way all widths have been corrected to widths perpendicular to paleocurrent, from apparent widths measured at the outcrop. Since successive channels cut into each other removing the upper part of previously filled channels the maximum measupreservred thickness of each individual channel has been considered.

Quantitative facies analysis

Thicknesses for each identified lithofacies have been measured inside individual channels. Since the rock layers have various geometries (lenticular, tabular, undulating bases) and consequently they do not have a constant thickness across the channel width the maximum thickness for each layer has been considered. Facies percentages of total channel thickness and net to gross ratio for each type of channel have been estimated.

DEPOSITIONAL INTERPRETATIONRESULTS 3-D Correlations of channels DISCUSSIONS

13 Understanding the depositional processes and lateral geometry of deep-water system base-of-the slope environment is important for predicting the extent and internal architecture of these reservoirs (Bouma et al., 1995). To adequately document reservoir properties at reservoir scale requires the study of large, continuous 3-D outcrops (Slatt, 2000). A good understanding of the architecture of these reservoirs requires interpretation of erosional and depositional geometries preserved in outcrops.In this study field observations integrated with a 3-D photorealistic model of the outcrop with assigned lithologies effectively helped in reconstruction of submarine channel architecture. Reconstruction of body geometry was accurate in areas where lithologies were highly distinct and where surfaces/bodies could be correlated on the distinct sides of the quarry. Bounding surfaces were mapped on the three-dimensional photorealistic model, as well as, the internal stratigraphy between surfaces. Quantitative 3-D bed continuity information at reservoir scale as well as the internal architecture of channel fill strata from the 3-D photorealistic outcrop produced better tools for more accurately interpreting the ancient record. Jackfork channel sequences display layered or massive channel fill with flat or lenticular geometry. Most of the channels are erosional (conduit for sediment transport into the basin) with scoured contacts (erosion between successive channels is common removing part of the channel fill). The channel fill usually contains debris flow at the base which grades upward into thin bedded, fine–grained sandstone interbedded with shale overlain by a drape of shale deposited during the wanning stage of channel filling (Fig. 4). In the central part of the quarry, massive fill is comprised of thick-bedded (10m) erosional-based, massive-appearing to amalgamated, fine-grained sandstones. Very few sedimentary structures are preserved except for dish and pillar dewatering structures; flute cast are rare, load structures are present. The massive character in outcrop of the thicker sandstone interval is evident in the photo of Fig.4C. The sandstone bodies are typically flat on the upper surfaces, but undulated on the lower. These undulatory surfaces are at the contact with underlying mudstone so they are assigned to soft sediment deformation and differential loading of soft sediment by sand.

14 Paleoflow direction was inferred from direct measurements of erosional as well as, depositional sedimentary structures identified in strata exposed at the base of the outcrop. Longitudinal scours and few grooves at the base of sandstone beds, as well as ripple morphology and cross-lamination are indicators of paleocurrent direction (Fig. 45, 6). However in most cases it was only possible to tell the direction of the flow and not its sense. Most measured paleocurrents suggest a southwest flow direction; few of them have southeast markers (Fig. 65). Theise is results are in agreement with previous studies (Jordan et al., 1993, Stone and McFarland, 1982) and with the regional paleoflow directions ofin the Ouachita basin. The total length of the exposure is about 1250 m with the quarry walls being at least 120 m up to 270 m apart. The southeast part of the quarry has a length of about 800 m; the central part is about 150 m long and the northwest part is about 300 m. Since most of the paleocurrent measurements indicate that the principal direction of transport is southwest channels have flow indicators oriented west-southwest the outcrop belt displays a nearly orthogonal cut across the downcurrent direction to the north, southeast and a more oblique, a dip-oriented profile into the basin to the northwest and an oblique depositional strike section with respect to paleoflow to the southeast (Fig. 6). The northwest end of the quarry close to the river displays again a nearly cross-sectional view. Bed boundaries were easily followed on the three-dimensional photo real model of the outcrop (Fig. 7, Fig. 8). This way strata exposed in the central part of the quarry were found to appear again on the NW quarry face. In order to correlate beds from one wall to the other we paid attention to the lateral continuity/discontinuity of the strata, to the geometry of the channels and their relative position on the outcrop and also we took in consideration the fact that the quarry is tilted about 10˚ toward northwest. The lower boundaries of the channels steep northward; they incline with about 3º at the southeastern end and reach 10º in the central part of the quarry. Therefore it was possible to think about correlation of these channelized sandstone bodies located at different positions along the axis of a submarine channel system.We interpreted a channel (Fig. 9) oriented N67E/S67W which is exposed on both sides of the quarry spaced apart by about 200 m. The part of the channel exposed on the northeastern face reveals a strike oriented view while the one on the northwestern face is

15 more dip oriented. The part of the channel that crops out on the main face is thinner; debris flow at the base is about 3.5 m in both, but the sandstone fill is about 2.5 m thinner in the channel on the NE face. It is assumed that the beds become thicker down current due to non-uniformity of the flow, reduction in confinement at the base of slope and a downstream decrease in bed shear stress.

Channel dimensions and aspect ratios

We show that channels located in the northernmost part of the quarry are exposed again in the NW part. The quarry walls are spaced apart by 150-300m. Bed boundaries were easily followed on the three-dimensional view of the outcrop. This way strata exposed in the central part of the quarry were found to appear again to the NW. In order to correlate beds from one wall to the other we paid attention to the lateral continuity/discontinuity of the strata, to the geometry of the channels and their relative position on the outcrop and also we took in consideration the fact that the quarry is tilted 0-10˚ toward north. The mean paleocurrent orientation as determined from direct measurements at the outcrop is N34E/S34W. However, we considered the median N27E/S27W as the main paleocurrent direction since it is more robust. Therefore the orientation of the main outcrop belt (~N9W) is approximately at 3647˚ to the transport direction of the turbidite which means that the estimated widths from the main face of the quarry on the 3-D virtual model are overestimated compared to the real widths measured in the true cross-section of the channels. The digitized channel boundaries from the photo real virtual model were projected on a plane oriented N563W/S563E to correspond to the realestimated strikecross-section of the channels and the width measured accordingly. This way all widths have been corrected to widths perpendicular to paleocurrent, from apparent widths measured at the outcrop. The other face of the outcrop close to the river is oriented N168E/S168W, almost parallel to the mean paleoflowalong dip. 34 channels have been identified. Significant channel width/thickness variation can be recognized at outcrop scale (Fig. 910). The mMinimum channel thickness observed at the outcrop is 2.119 m and the maximum is 2324.716 m with a mean of 88.3

16 m. The lower two-thirds of the quarry are composed mainly of larger, sandstone-rich, aggradational channels, at the base of the quarry have widths of about 145 m and thickness of about 10 m. They can reach widths of 300 m, but when corrected they have a maximum width of about 200 m. while cChannels within the upper part of the quarry are smaller, laterally migrating channels which vary in thickness between 2 m and 86.3 m (mean ~54.3 m) and have widths of at least 3250 m with an average of 59 m. When corrected with respect to the mean paleocurrent direction they are at least 16 m in width and a maximum of 75 m with a mean 40 m. They display multiple internal erosion surfaces and lateral accretion is common suggesting some channel sinuosity. Sand lenses at channel margins incline in the direction of channel migration - northward and are perpendicular to the flow in the channel (Fig. 23, ‘lateral accretion packages’-LAPs of Abreu et al., 2003). Non-amalgamated, mixed traction-suspension LAPs are characterized by interbedded high-concentration turbidites deposited by suspension (massive sandstones) and mud (rip-up)-clast conglomerates deposited by traction as bedload. The formation of the sinuous channels at the top of the aggradational channels can be explained by a reduction in the volume of the turbidity currents that preceded the overall abandonment of this sector of the turbidite system. Estimated cross-channel width of the channel shaped unit (perpendicular to the main paleocurrent direction, N34E-SW) and maximum measured thickness of individual channels were plotted on a log-log graph (Fig.101) and the aspect ratio calculated for each channel. As can be noticed the cChannel width commonly is one order of magnitude higher than channel depth. Channel bodies have rather flat channel fills (width/thickness ratio 4 3:1 to 3932:1) with a mean aspect ratio of 17.415:16.. Sinuous channels have a minimum apparent ratio of 9:1, a maximum of 17:1 and a mean value of 14:1.Aggradational channels have a minimum apparent ratio of 4:1, a maximum of 32:1 and a mean value of 16:1. When measured across the estimated cross- channel section all the ratios have lower values compared to those measured from the outcrop. Aggradational channels have a minimum width to thickness ratio of 3:1, a maximum of 20:1 and an average value of 11:1. Sinuous channels have a minimum ratio of 7:1, a maximum of 12:1 and a mean value of 9:1.

17 Quantitative Facies Analysis Facies thickness and facies percentages of total channel thickness, as well as net to gross ratios have been estimated for each channel. There are mainly three types of channels (Fig. 12) exposed at the outcrop: (1) Large, sandstone-rich, aggradational channels located in the axis of the canyon typically filled by amalgamated, thick to very thick tabular to irregular-bedded, fine-grained, and massive sandstones deposited from high-concentration turbidity currents (Fig.12A). (2) Large aggradational channels specific to channel complex margin association show a deep basal cut with evidence of by-pass processes. Mud-clast breccia overlies the erosive surface and grades upward into thin- bedded, fine-grained sandstone interlayered with shale overlain by massive-to-planar stratified sandstone deposited from high density flows over-spilling the filled channel (Fig.12B). and (3) Small, laterally migrating channels in the upper part of the quarry characterized by interbedded massive sandstones deposited from high-concentration flows and mud-clast breccia deposited by traction as bed load (Fig. 12C). The sandstone comprises a significant proportion (in average 83.28%) of the total channel thicknesses (Fig. 14), while shale and debris flow deposits make about 17.24% with a range from 3.66% to 59.2% (minimum net/gross = 40.8). If we consider only the large, aggradational channels that make up the lower two thirds of the quarry the sandstone will be at least 40.8% of the total channel thickness with an average of 82.78%. Shale and debris flow deposits make up about 18% of the channel thickness. The smaller channels at the top of the quarry have in average 84.86% sandstone ranging from 68.19% to 93.84%. The lateral accretion packages have not been considered when taken these measurements, only thicknesses in the axis of the main channel have been measured. Shale and debris flow make up about 15.54% of the total channel thickness with a range from 6.16% to 31.81%. When occurring in amalgamated units sandstone has a mean thickness of about 60 cm covering a range from 13 cm to 5 m (Fig. 13 A). The thickness of the amalgamated units reaches in average 4 m with a maximum thickness of 16 m. Thin-bedded sandstone layers have an average of 34 cm being at least 9 cm thick up to 2 meters (Fig. 13 B).

18 The interpreted channel (Fig.11) is oriented N67E/S67W and is exposed on both sides of the quarry which are spaced apart by at least 200m. The part of the channel exposed on the northeastern face is more strike oriented while the channel on the northwestern face is more dip oriented (Fig. 12).The part of the channel exposed on the northwestern face is thicker than its correspondent on the main face (debris flow at the base is about 3.5 m in both, but the sandstone fill is about 2.5m thicker in the channel on the NW face). Due to non-uniformity of the flow, reduction in confinement at the base of slope and assuming a downstream decrease in bed shear stress the beds become thicker down current.Shale intraclast breccia has an average thickness of about 1 m (Fig. 13C). Submeter thickness usually occurs in the lateral accretion packages of the sinuous channels or in some of the preserved levees while thicker beds occur at the base of aggradational channels. Thin layers of shale have an average thickness of 8 cm covering a range from 2 cm to 30 cm (Fig. 13D).

DISCUSSIONS Channel dimensions - Uncertainty 2-D vs. 3-D

Significant variations exist among channel dimensions (width and thickness) and aspect ratios in turbidite systems as they are measured from modern and ancient records (Clark and Pickering, 1996) especially because of the difference in scale. Our data (channel width and thickness) when compared to other modern and ancient channels using data published by Clark and Pickering, 1996 fall in the lower aspect ratio area of the data scatter (Fig. 11). Their data come from modern, ancient and subsurface data sets with channels, canyons and channel-shaped elements of deep-water depositional system plotted all together which explains the wide range covered by the data. Jackfork stacked channel sequences deposited at the base-of-slope have low aspect ratios (4:1 to 3329:1). and wWhen compared with outcrop data from similar deep- water settings such as the outcrop example, from Fan#3 Tanqua Karoo basin, South Africa our results shows similarsomehow lower values for the width/thickness ratio. The

19 2 km long outcrop along the Ongeluks River is interpreted to be a sand-rich channel complex deposited most probably at the base of slope (Bouma et al., 1995). The channel complex is characterized by high sandstone/shale ratio and rather flat channel fills (width/thickness ratio 30:1 to 50:1). This might be explained by the different methods used to measure the channel dimensions (3-D vs. 2-D measurements). The total length of the exposure is about 1250 m with the quarry walls being at least 120 m up to 270 m apart. The main face of the quarry has a length of about 808 m with a projected length along y axis (north south direction) of 732 m. Bouma and Cook, 1994 estimate from 2-D photomosaics a total length of the main face of the quarry of 838.2 m (2750 feet) which is overestimated by 100 m. Since the maximum channel width measured on the digital outcrop is 265 m the layers that make up the channel infill should have at a maximum this length. At least some of the layers measured by Bouma and Cook, 1994 are overestimated since the maximum measured length for sand, thin beds and mud is well over 300 m. It is mentioned in their paper that the actual layer length is not the distance between both ends of one layer, but the distance along the y-axis of the projected length. Since the projected length is shorter then the actual length of the outcrop the length of the layers in outcrop would be even bigger.

CONCLUSIONS

The primary purpose of studying this outcrop was to determine its viability as an analog input for deep-water base-of-the-slope hydrocarbon reservoir models. Due to the fact that geometry and facies pattern of nested-channel complex at the base-of-slope may provide models of heterogeneity distribution this paper described the geometry and the internal architecture of the channels within the channel complex of the Jackfork Group. This study represents the first attempt of three-dimensional interpretation of the deposits exposed at Big Rock based on correlation of channels from both sides of the quarry, a difficult exercise without a three-dimensional dataset. Despite relative limited exposures relative to channel dimensions and intense tectonism in the region the three-

20 dimensional photo realistic model provides superior quantitative and spatial data suitable for three-dimensional 3-D facies architecture studies. Field observations integrated with the three-dimensional photo realistic methodology effectively helped in understanding the 3-D architecture of channelized deposits in a deepwater setting, base-of-the-slope environment. Reconstruction of body geometry was accurate in areas where lithologies were highly distinct and where surfaces/bodies could be correlated on the two sides of the quarry. Bounding surfaces were mapped on the three-dimensional photo realistic model, as well as, the internal architecture of channel fill stratastratigraphy between surfaces. Quantitative 3-D bed continuity information at outcrop scale as well as the internal architecture of channel fill strata from the 3-D photorealistic outcrop produced better tools for more accurately interpreting the ancient record. Three-dimensional photorealistic methodology effectively helped in understanding the 3-D architecture of channelized deposits in a deepwater setting, base- of-the-slope environment. Three-dimensional photorealistic models of outcrops are useful tools for visualization, analysis and interpretation.

Channel dimensions and morphology, stacking pattern ,and a high sandstone/shale ratio and the inferred presence of common upward-thinning cycles suggest deposition on the inner fanat the base of slope. Most of the channels have erosional contacts, low aspect ratio (width: thickness = 15:1) and high net to gross (more than 80%).

Three-dimensional photo realistic models of outcrops are useful tools for visualization, analysis and interpretation. High-resolution outcrop data sets help in reducing uncertainty in the geologic interpretation of sparse and lower resolution well-log and subsurface data. Thinning- and fining-upwards cycles are believed to be produced by filling and abandonment of channels.

Massive Ta divisions are the most common Bouma divisions in the section and they are also the thickest. The deposits at BRQ were deposited in a sand-rich environment dominated by rapid deposition of thick, massive Ta divisions.

21 Acknowledgements

This study benefited from a GCSSEPM Ed Picou Fellowship Grant for Graduate Studies in the Earth Sciences. I am especially grateful to GCA SEPM for financial support.Xueming Xu for building the three-dimensional photo real model of the outcrop at Big Rock.

REFERENCES

Abreu, V., Sullivan, M., Pirmez, C., and Mohrig, D., 2003, Lateral Accretion Packages (LAPs):An: An Important Reservoir Element in Deep Water Sinuous Channels, Marine and Petroleum Geology 20, 631-648

Aiken, C. L.V., Xu, X., Thurmond, J., Abdelsalam, M., Olariu, M. I., Olariu, C., and Thurmond, A., 2004, 3-D Laser Scanning and Virtual Photorealistic Outcrops: Acquisition, Visualization and Analysis, AAPG short course

Bhattacharya, J.P., Aiken, C.L.V., Corbeanu, R.M., McMechan, G.A., Xu, X., Zeng, X., and White C.D., 2002, 3-D Outcrop Reservoir Characterization, (abs.), AAPG Annual Meeting, Houston.

Bouma, A. H., and T. W. Cook, 1994, Architecture of a submarine channel complex, Jackfork Formation, Arkansas: Louisiana Basin research Institute Bulletin, v.4, p. 11-22

Bouma, A. H., G. H. Lee, O. van Antwerpen, and T. W. Cook, 1995, Channel Complex Architecture of Fine-Grained Submarine Fans at the Base-of-Slope: Transactions – Gulf Coast association of Geological Societies, v. 45, p.65-70

Bouma, A. H., DeVries, M. B., and Stone, C.G., 1997, Reinterpretation of depositional processes in a classic flysch sequence (Pennsylvanian Jackfork Group), Ouachita Mountains, Arkansas and Oklahoma: discussion. AAPG Bulletin, 81,470-472

Bouma, A.H., 2000, Fine-grained, mud-rich turbidite systems: model and comparison with coarse-grained, sand-rich systems, in A. H. Bouma and C. G. Stone, eds., Fine-grained turbidite systems, AAPG Memoir 72/SEPM Special Publication 68, p. 9–20.

22 Clark, J. D. and K. T. Pickering, 1996, Submarine Channels: Processes and Architecture: Vallis Press, London, 231 pp.

Coleman Jr., J. L., 2000, Carboniferous Submarine Basin development of the Ouachita Mountains of Arkansas and Oklahoma, in A.H. Bouma and C.G. Stone, eds. Fine-grained turbidite systems, AAPG Memoir 72/SEPM Special publication 68, p.21-32

Coleman Jr., J. L., Sheppard, F., C., and Jones, T., K., 2000, Seismic Resolution of Submarine Channel Architecture as Indicated by Outcrop Analogs, in A. H. Bouma and C. G. Stone, eds., Fine-grained turbidite systems, AAPG Memoir 72/SEPM Special Publication 68, p. 119–126.

Coleman Jr., J. L., 1997, Reinterpretation of depositional processes in a classic flysch sequence (Pennsylvanian Jackfork Group), Ouachita Mountains, Arkansas and Oklahoma: discussion. AAPG Bulletin, 81, 466-469

Cook, T. W., 1993, Facies Architecture of deep-water channel deposits, Brushy Canyon Formation, West Texas, and Jackfork Group, Arkansas: Unpublished Master’s Thesis, Louisiana State University, 108 p.

D’Agostino, A. E. and Jordan D. V., 1997, Reinterpretation of depositional processes in a classic flysch sequence (Pennsylvanian Jackfork Group), Ouachita Mountains, Arkansas and Oklahoma: discussion. AAPG Bulletin, 81,473-475

Jordan, D. W., D. R. Lowe , R. M. Slatt, A. D’Agostino, M. H. Scheihing, R. H. Gillespie, and C. G. Stone, 1993, Scales of Geological Heterogeneity of Pennsylvanian Jackfork Group, Ouachita Mountains, Arkansas: Applications to Field Development and Exploration for Deep-Water Sandstones, Guidebook Dallas Geological Society, Field Trip Guidebook no. 3., 142 p.

Langley, R., 1998, RTK GPS, GPS World, 9, 70-76

Link, M.H., and Roberts, M.T., 1986, Pennsylvanian paleogeography for the Ozarks, Arkoma, and Ouachita basins in east-central Arkansas, in Stone, C. G., and B.R. Haley, eds., Sedimentary and Igneous Rocks of the Ouachita Mountains of Arkansas: Arkansas Geological Commission, Guidebook, pt.2, p. 37-40.

Link, M.H., and Stone, C.G., 1986, Jackfork Sandstone at the abandoned Big Rock Quarry, North Little Rock, Arkansas, in Stone, C. G., and B.R. Haley, 1986, Sedimentary and Igneous Rocks of the Ouachita Mountains of Arkansas: Guidebook GSA Annual Meeting, San Antonio, p. 1-8.

Lowe, D. R., 1997, Reinterpretation of depositional processes in a classic flysch sequence (Pennsylvanian Jackfork Group), Ouachita Mountains, Arkansas and Oklahoma: discussion. AAPG Bulletin, 81, 460-465

23 Moiola, R. J., and Shanmungan, G., 1983, Submarine fan sedimentation, Ouachita Mountains, Arkansas and Oklahoma, AAPG Bulletin, vol.67, no.3, pp.517, Mar 1983

Nielsen, K.C., Aiken, C.L.V., and Xu, X., 1999, Digital mapping methods: accurate digital data capture and analysis for the field geologist, GSA Continuing Education Manual, 119p.

Olariu, C., 2002, Study of Cretaceous Delta Front Deposits, Integrating Outcrop, GPR and 3-D Photorealistic Data, Panther Tongue Sandstone, Utah, Master Thesis, UTD

Shanmungan, G. and Moiola, R. J., 1997, Reinterpretation of depositional processes in a classic flysch sequence (Pennsylvanian Jackfork Group), Ouachita Mountains, Arkansas and Oklahoma. AAPG Bulletin, 79, 672-695

Slatt, R. M., Al-Siyabi, H. A., Williams E.T., Stone C. G., Weimer P., Davis R.J., and Jordan D.W., 1997, Oil and Gas Journal, vol. 95, no.33, pp. 67-72, 18 Aug 1997

Slatt, R. M., Weimer, P., and Stone, C.G., 1997, Reinterpretation of depositional processes in a classic flysch sequence (Pennsylvanian Jackfork Group), Ouachita Mountains, Arkansas and Oklahoma: discussion. AAPG Bulletin, 81,449-459

Slatt, R. M.,2000, Why outcrop characterization of turbidite systems, in A. H. Bouma and C. G. Stone, eds., Fine-grained turbidite systems, AAPG Memoir 72/SEPM Special Publication 68, p. 181–186.

Slatt, R. M., Stone, C.G., and Weimer, P., 2000, Characterization of Slope and Basin Facies Tracts, Jackfork Group, Arkansas, with Applications to Deepwater (Turbidite) Reservoir Management, GCSSEPM

Stone, C. G., and J. D. McFarland, III, 19821981, Stop 1 – Abandoned Big Rock Quarry in upper part of Jackfork sandstone, in C. G. Stone and J. D. McFarland, III, eds., Field Guide to the Paleozoic rocks of the Ouachita Mountain and Arkansas Valley provinces, Arkansas: Arkansas Geological Commission guidebook 81-1, p. 7-13

Thurmond, J.B., Drzewiecki, P.A., Markello, 2001, Three dimensional mapping of Permian mud-mounds, Last Chance Canyon, New Mexico, Abstracts with Programs: GSA Annual Meeting 2001.

Thurmond J. B., Løseth, T. M., Rivenæs, J. C., Martinsen, O. J., Aiken, C. L.V., and Xu, X., 2000, Using Outcrop Data in the 21st Century – New methods and applications, with example from the Ainsa Turbidite System, Ainsa, Spain, AAPG Annual Meeting Program, 9, A7

24 Xu, X., Aiken, C. L. V., and Nielsen K. C., 1999, Real Time and the Virtual Outcrop Improve Geological Field Mapping, EOS, Transactions, American Geophysical Union, Vol. 80, N0.29, July 20, p. 317, 322-324.

Xu, X., 2000, Three-dimensional Virtual Geology: Photorealistic outcrops, and their acquisition, visualization and analysis, Ph.D. thesisdissertation, UTD

Xu, X., Aiken, C. L. V., Bhattacharya J. P., Corbeanu R. M., Nielsen, K. C., McMechan G. E., and Abdelsalam, M. G., 2000, Creating virtual 3-D outcrop, The Leading Edge, Vol. 19, No. 2, p. 197-202.

Xu, X, Bhattacharya, J.P., Davies, R.K., Aiken, C.L.V., 2002, Digital geologic mapping of the Ferron Sandstone, Muddy Creek, Utah with GPS and reflectorless laser rangefinders. GPS Solutions, v. 5, no. 1, p.15-24

Xu, X., Aiken, C. L. V., Thurmond, J., Neubert, B., 2004, The Future of 3-D Virtual Models, Geoworld, p.38-41

The 3D topographic model was collected with a .5m point spacing, which was deemed the best compromise between data acquisition speed and accuracy. With modern scanners, capable of collecting tens to hundreds of thousands of points per second, the acquisition time could have been reduced, and the resolution of the initial dataset could have been increased. However, the .5m model was sufficiently detailed for the purposes of this projec The resulting model (Figure 4) is a three-dimensional replica of the main outcrop of the Ainsa II, with a pixel resolution of approximately 3-4 cm and sub- decimeter accuracy. RTK GPS receiver, which has a nominal accuracy of 2 cm

REFERENCES de la paperul lui john Adams, E.W, Schröder, S., Grotzinger, J.P. & McCormick, D.S. 2004. Digital reconstruction and stratigraphic evolution of a microbal-dominated, isolated carbonate platform (terminal Proterozoic, Nama Group, Namibia). Journal of Sedimentary Research, 74, 479-497. Arbués, P., Muñoz, J.A., Poblet, J., Puigdefàbregas, C. & McClay, K. 1998. Significance of submarine truncation surfaces in the sedimentary infill of the Ainsa basin (Eocene of south-central Pyrenees, Spain). Abstracts of the 15 th International Sedimentological Congress, Alacant, Spain. Publicaciones de la Universidad de Alicante, 145-146. Arbués, P., Gjelberg, J., Puig, M., Sánchez-Villanueva, S., Puigdefàbregas, C., Muñoz, J.A. & Marzo, M. 2000. Anatomy and evolution of an intra-slope turbidite system in a foredeep basin; the Ainsa system (Eocene, Spanish Pyrenees) (abs.): AAPG Annual Meeting Program, 9, A7

25 Clark, J.D. 1995. Detailed section across the Ainsa II channel complex, south central Pyrenees, Spain. In: Atlas of Deep-Water Environments (ed. By K.T. Pickering, R.N. Hiscott, N.H. Kenyon, F. Ricci Lucchi and R.D.A. Smith), Chapman & Hall, London, 139-144. Fernández, O., Muñoz, J.A., Arbués, P., Falivene, O. & Marzo, M. 2004. Three- dimensional reconstruction of geological surfaces: An example of growth strata and turbidite systems from the Ainsa basin (Pyrenees, Spain). AAPG Bulletin, 88, 1049-1068. Muñoz, J.A., McClay, K. & Poblet, J. 1994. Synchronous extensión and contraction in frontal thrust sheets of the Spanish Pyrenees. Geology, 22, 921-924. Muñoz, J.A., Arbués, P. & Serra-Kiel, J. 1998. The Ainsa basin and the Sobrabe oblique thrust system: Sedimentological and tectonic processes controlling slope and platform sequences deposited synchrnously with a submarine emergent thrust system. In: Melendez Hevia, A. & Soria A.R. (eds.) Field trip guidebook of the 15 th International Sedimentological Congress, 213-223. Mutti, E., Remacha, E., Sgavetti, M., Rosell, J., Valloni, R. and Zamorano, M. 1985. Stratigraphy and facies characteristics of the Eocene Hecho Group turbidite systems, South Central Pyrenees. In: Excursion Guidebook of the 6 th European Regional Meeting of Sedimentologists, Llerida (ed. by M.D. Mila and J. Rosell). International Association of Sedimentologists, 521-576. Posamentier, H.W. & Kolla, V. 2003. Seismic geomorphology and stratigraphy of depositional elements in deep-water settings. Journal of Sedimentary Research, 73, 367- 388- Wynn, R.B., Kenyon, N.H., Masson, D.G., Stow, D.A.V. & Weaver, P.P.E. 2002. Characterization and recognition of deep-water channel-lobe transition zones. AAPG Bulletin, 86, 1441-1462.

Thinning- and fining-upwards cycles are believed to be produced by filling and abandonment of channels. and the inferred presence of common upward-thinning cycles

Recognition and mapping of bounding surfaces in outcrop sections are important components of facies analysis.

In order to study both upward (in time) and basinward (in space) evolutions of geometries and facies the stratigraphic units have to be correlated over the whole outcrop. Trace fossils indicating a deep water setting are abundant in shale (pre- depositional traces), few of them have been found in sand (post-depositional). These undulatory surfaces are at the contact with underlying mudstone so they are assigned to soft sediment deformation and differential loading of soft sediment by sand.

26 The upper Jackfork strata that crop out in the main face of the quarry were interpreted as proximal deep-water fans within slope channels canyons (Jordan et al., 1993) or nested channels at the base-of-slope (Bouma and Cook, 1994).

Using a combination of real-time kinematic global positioning system (RTK- GPS), and laser scanners we were able to capture 3-D terrain data of the outcrop in global coordinates with centimeter accuracy. Oblique close-in photography acquired from the ground was integrated with terrain data and converted into a 3-D digital photorealistic model of the outcrop. Examination of the virtual outcrop model allowed extraction of accurate 3-D qualitative (lithology, contacts), as well as, quantitative (bed width and thickness) geometric information. This facilitated interpretation and reconstruction of bed architecture making possible further correlations of strata exposed on distinct sides of the quarry. The massive character in outcrop of the thicker sandstone interval is evident in the photo of. The strata exposed in the central part of the quarry were deposited in a sand-rich environment dominated by rapid deposition of thick, massive Ta divisions. Modeling of three-dimensional outcrops from two-dimensional images has been a challenging problem for geologists. Three dimensional mapping techniques for detailed studies of geologic outcrops have been recently developed (Xu et al., 1999; 2000; Xu, 2000) so that more quantitative information is available for interpretation. In the central part of the quarry, which is considered to be the axis of the canyon (Jordan et al., 1993) thick-bedded (10m) erosional-based, massive-appearing to amalgamated, fine-grained sandstones are dominant.

Our interpretation is based on mapping of three-dimensional facies distributions and architecture of the channelized features and collection of paleocurrent data.

These might have been eroded by gravity flows during transport or collapsed from the channel margin.

27 The channels within the chanel complex are characterized The channels within the channel complex are characterized by: 1) lateral change from channel to levee to overbank deposits, 2) relative vertical fining and thinning of the sediment fill, 3) many erosional surfaces, rip-up clasts, and plant fragments, 4) erosional remnants within amalgamated sandstones, and 5) soft sediment deformation of the sediment below the channel surface. These channels are relatively shallow and range in lateral extent. Overall, it seems that they exhibit good lateral and vertical sandstone connectivity. The channels belonging to the channel-levee-overbank depositional environments cut through thin-bedded turbidite facies, which represent levee-overbank deposits of another channel. Sandstone amalgamation, slumping sediment deformation, rip-up clasts, and plant fragments are common.In this study field observations integrated with a three- dimensional photorealistic model of the outcrop with assigned lithologies effectively helped in the reconstruction of submarine channel complex architecture.

The aim of this paper is to describe the geometry and the internal architecture of the channels within the channel complex of the Jackfork Group. Geometry and facies pattern of nested-channel complex at the base-of-slope may provide models of heterogeneity distribution. The levee development is limited. They only formed when the canyon morphology was almost filled and are of limited lateral extension. Levee aggradation and construction only occur if the flow is turbulent enough to maintain fine-grained material in suspension.In the proximal fan setting levees will be poorly developed because the flow are highly concentrated and turbulence is limited (Mulder&Alexander,2001).

At the onset of the Carboniferous the Oachita basin was approximately 450 km north to south and 550 km east to west (Coleman, 1990; Coleman et al., 1994). The basin was primarily fed with sediment which was derived from the north and east shelves. Jackfork Group is up to 2100m and was accumulated at an average rate of 400m/mil years. for the most part is interpreted as… Jackfork Group deepwater clastic represents deposition during a period of tectonic quiescence over a period of 5 million years.

1) erosional remnants within amalgamated sandstones, and These channels exhibit good lateral and vertical sandstone connectivity.

Introduction Level of turbidite outcrop studies This paper represent the first attempt of 3-D interpretation of the geometry and facies pattern of the Jackfork nested-channel complex deposited at the base-of-slope. 2D vs. 3D interpretation –why is important…heterogeneities

28 What this paper bring new: - first attempt of 3D interpretation of Big Rock Quarry; first 3-D facies architecture of a channelized (toe of the slope) turbidity deposits using photorealistic method - first attempt of 3D interpretation of Big Rock Quarry ; Channelized deposits importance for turbidite systems, heterogeneities; channels vs slumps –2-3 phrases – can be more at discussions Stage of channelized outcrop studies 2D vs 3D – publications Big rock has beed studied in 2D but not in 3D. Because BRQ has the shape of a horse shoe it is suitable for 3D studies (interpretations). This study uses 3-D photorealistic methodology to understand 3-D architecture of channelized deposits. Within paper we will discuss – 3-D continuity of channelized deposits; channels vs. scours. Big Rock Geology & Previous Studies

Methodology : Digital Mapping Techniques + interpretation methodology (digitize major discomformities (channels); following the same bed around the quarry and correlate beds above and below; 3-D acquisition + interpretation ; Data collection ; StatisticsStatistiques Results 2-D photomosaics ; 3-D interpretation ; Statistical measurements Discussions - are channels or scours –which data show that are scours or channels, what if are both? Compare with other toe of the slope outcrop or other data studies Conclusion - restate the problems from introduction List of figures 1 Location 2 Stratigrphical location (time-column) + typical lithology vertical section 3 Photorealistic model – data acquisition + processing 4 interpretation-working steps 5 Facies photo; paleocurrent measurements 6 2-D photomosaics (east & wwest part) –bedding diagram, interpretation

29 7 3-D interpretationinterpretations 8 3-D bed measurements – comparison with 2-D 9 ???strike and dip at the base of the interpreted channels? 10 3-D block diagram with interpretation and dimensions – compared with other studies (outcrops, seismic, seafloor)

The 3-D geometry of the quarry allowed for bed correlation between facing walls despite the lateral discontinuity of the strata. Integration of traditional tools for examining outcrops (facies analysis and interpretation) and new techniques (3-D detailed geologic mapping) helped in paleo- environment reconstruction. This interpretation is based on mapping of 3-D facies distributions and collection of new paleocurrent data

Quantum detectors are faster and more sensitive than thermal detectors. However, they need to be cooled down in order to be efficient in the longwave range, which makes them difficult to use as field instruments. In addition, their detectivity is spectrally dependent and vanishes above a given wavelength threshold. Thermal detectors, however, can be operated at room temperature and their detectivity is independent of the wavelength. The presence in the southeastern part of the quarry of small, scoured depressions eroded in sandstone bodies and filled with mudstone and sandstone-clast debris flows or drapes of laminated mudstone suggest that scour and successive slope failures have cut out many originally more continuous units (Jordan et al., 1993). Deposition of dense sand over weak plastic clay cause differential loading of the underlying sediments. Examination of the virtual model of the outcrop allowed extraction of 3-D accurate geometric information. Reconstruction of body geometry was accurate in areas where lithologies were highly distinct and where surfaces/ bodies could be correlated on distinct sides of the quarry. Compared to previous two-dimensional outcrop photomosaics, the three-dimensional virtual model adds more valuable information for outcrop interpretation (Bhattacharya et al., 2002). Spill-phase deposits capping channels - sandstone Deposition of dense sand over weak plastic clay cause differential loading of the underlying sediments. In other types of turbidite depositional system, asymmetric cycles have been interpreted as products of retrogressive flow sliding (Pickering, 1979) or of sea-level

30 changes (Mutti, 1985). Whatever the interpretation, asymmetric cycles of bed-thickness are generally picked out by eye from pictorial logs or from graphs of thickness against height in the section (e.g. Mutti, 1974; Ricci Lucchi, 1975).

FIGURE CAPTION

Fig. 1 Location map of the study area. Aerial photo shows the Big Rock Quarry (BRQ) outcrop belt. Big Rock Quarry The quarry is located along the north bank of the Arkansas riverRiver in North Little Rock, Arkansas. Map view of the quarry shows the outcrop belt.

Fig. 2 Correlation chart for Mississippian and Pennsylvanian formations between deep- water Ouachita-Trough and shelfal areas to the north (from Jordan et al, 1993).

Fig.3 Three-dimensional photorealistic mapping and interpretation of the deep-water succession exposed at the Big Rock Quarry outcrop. A. The outcrop belt displays a nearly orthogonal cut across the downcurrent direction to the southeast and a more oblique, dip-oriented profile to the north. B. 3-D digital terrain model generated by the laser scanning of the outcrop C. Surface built in Gocad from terrain data D. Oblique close-in photography acquired from the ground is integrated with terrain data and converted into a 3-D digital photorealistic model of the outcrop E. Bedding diagram superimposed on 3-D virtual model of the outcrop F. Submarine channel architecture. Bounding surfaces are digitized in Gocad directly from the virtual 3-D photorealistic model Fig.43 Facies and facies associations of the Jackfork channel complex at Big Rock Quarry

31 A. Erosive contact at the base of a channel with debris flow (DF) at the base overlain by massive sandstone (MS). Erosion between succesive channels is common removing part of the channel fill. Jackfork channel sequences display layered (thin-bedded, fine-grained sandstone interlayered with shale) or massive (thick-bedded, massive appearing to amalgamated fine-grained sandstone) channel fill with flat or lenticular geometry Matrix supported breccia deposited by cohesive debris flows (DF); its base (thick line) corresponds to a sharp erosional surface which cuts into the underlying channel. This type of facies is found in the basal part of the individual channel fills, just above the erosional surface and below massive sandstone (MS) and is interpreted as material left behind by flows transiting through the channel conduits. Depending on the nature of the matrix there are two types of breccia exposed in the outcrop a sandy–matrix and a shaly- matrix breccia. The mud clasts are subangular and range in size from few mm to cm (see lens cap for scaling; diameter = 50 mm).

B. Amalgamated sandstones: Thick to very thick tabular to irregular beds, fine-grained, structureless sandstones constitute the major part of the channel fill of channels exposed in the central part of the quarry which is considered to be the axis of the canyon. These massive sandstones show only water escape pillar and dish structures. They were deposited from high-density sediment gravity flows and represent Bouma Ta divisions. Detail of shale intraclast breccia shown in A: Matrix-supported breccia is the product of cohesive debris flow and consists of two main groups of facies depending on the nature of the matrix (sandy/shaly) C. Thin-bedded turbidite: Thin to medium lenticular to tabular beds of fine-grained sandstone deposited from low-concentration turbulent flow interbedded with subordinate suspension shale drapes. Amalgamated sandstones: Thick to very thick tabular to irregular beds, fine-grained, structureless sandstones constitute the major part of the channel fill of channels exposed in the central part of the quarry which is considered to be the axis of the canyon. These massive sandstones show only water escape pillar and dish structures. They were deposited from high-density sediment gravity flows and represent Bouma Ta divisions.

32 D. Thin to medium lenticular to tabular beds of fine-grained sandstone deposited from low-concentration turbulent flow interbedded with subordinate suspension shale drapes E. Interlaminated sandstone and siltstone: laminae of very fine, parallel, wavy and riplleripple-laminated sandstone and parallel laminated siltstone Predominantly traction depositiodepositedn from dilute, low concentration turbulent flows.

Fig.54 Sedimentary structures identified in strata exposed at the base of the outcrop A. ripple cross-laminated sandstone N72E/S72W B. groove marks on the base of sandstone bed N54E/S54W C. cross-bedded sandstone D. climbing ripples N18E/S18W E. longitudinal scours on the base of shale N45E/S45W F. ripple cross-laminated shale N18E/S18W G. (H.) flute cast on the base of sandstone N16W/S16E (N18E/S18W)

Fig.5 Three-dimensional photorealistic mapping and interpretation of the deep-water succession exposed in the outcrop at the Big Rock Quarry. A. The outcrop belt displays an oblique cut across the downcurrent direction to the southeast and a nearly dip-oriented profile to the north. B. 3-D digital terrain model generated by the laser scanning of the outcrop C. Surface built in Gocad from terrain data D. Oblique close-in photography acquired from the ground is integrated with terrain data and converted into a 3-D digital photorealistic model of the outcrop E. Bedding diagram superimposed on the 3-D virtual model of the outcrop F. Submarine channel architecture. Bounding surfaces are digitized in Gocad directly from the virtual 3-D photo real model (blue = debris flow; red= thin to medium sandstone interbedded with shale; yellow = fine-grained, structureless, amalgamated sandstones).

Fig. 6 Horse-shoe shape of the outcrop at Big Rock with paleocurrent orientations. Paleoflow direction was inferred from direct measurements of sedimentary structures identified in strata exposed at the base of the outcrop. Longitudinal scours and few

33 grooves at the base of sandstone beds, as well as ripple morphology and crosslamination are useful indicators of paleocurrent direction. However in most cases it was only possible to tell the direction of the flow and not its sense. Most paleocurrents suggest a southwest flow direction; few of them have southeast markers. Since most of the channels have flow indicators oriented west-southwest the outcrop belt displays a nearly orthogonaln oblique cut across the downcurrent directionview to the southeast and a more oblique,nearly dip-oriented profile into the basin to the north.

Fig.7 Three-dimensional interpretation of channelized features present on the main face of the quarry A. 3-D photo real model of the outcrop made up of 12 images - one picture is missing (y is north) B. Bedding diagram superimposed on the 3-D virtual outcrop (red - channel boundaries) C. Facies architecture of channel complex. Different channel types are highlighted in yellow – large, aggradational channels with well defined axial regions composed of high relief basal erosional surfaces overlain by matrix-supported breccia which grades upward into amalgamated sandstones, orange - small, laterally migrating channels and red - aggradational channels made out of thin-bedded sandstones interlayered with shale which overlain the basal breccia. D. Bedding diagram showing the stacked channel pattern. Dimensional and architectural data (width vs. thickness measurements) for each facies were also estimated. 12 µm ± The surface connecting the base of the channel on both sizes of the quarry is oriented strike/dip=N67.19E/9.13W

Strike daca consideram toate triunghiutile care compun suprafata de la baza canalului Min max mean median stdev -89.9907 89.9861 59.1968 67.1917 25.6358 Strike positive in cadranul NE =72111 triunghiuri 0.0113 89.9861 64.0223 67.1962 11.1987

34 Strike negative in cadranul NW= 3363 tringh -89.9907 -0.0024 -44.2745 -45.2640 29.2298 Dip = 75474 triung 15.9315 89.9632 78.1357 80.8709 9.4009

N33=strike/dip=N47.63E/8.27E SS =65891 tri -89.9943 89.9993 43.5317 47.6364 21.0066 spp =63306 tri 0.0336 89.9993 47.2313 47.6390 8.5884 snn =2585 tri -89.9943 -0.0132 -47.0709 -50.2560 29.9631 DD =65891 tri -89.7743 -31.5688 -80.4877 -81.7306 5.3268

Some packages of sandstone are comprised of amalgamated sandstones in the center and pass laterally into a wedge of shale suggesting that each succeeding current deposited a sand layer that is overlain by shale The axis of the current is sufficiently strong to remove the soft cover of shale in the center of density current that results in an amalgamated contact. Ancient erosionally-based channel fills record at least two changes in flow properties the first to initiate the erosion and the second to initiate the fill (Kneller, 2003).

Fig.8 Three-dimensional interpretation of channelized features present on the central and northwest face of the quarry close to the river A. 3-D photo real model of the outcrop made up of 6 images - one picture is missing (y is north) B. Bedding diagram superimposed on the 3-D virtual outcrop (red - channel boundaries, yellow - channel fill) C. Bedding diagram showing the stacked channel pattern.

35 D. Facies architecture of channel complex. Large, aggradational channels are highlighted in yellow – they have well defined axial regions composed of high relief basal erosional surfaces overlain by matrix-supported breccia which grades upward into amalgamated sandstones.

Fig.9 Three-dimensional interpretation of channelized features at Big Rock A. Three-dimensional photo real model of the central part of the outcrop - oblique view of the quarry looking northeast (y is north and x is east). The horse-shoe shape of the quarry allowed for correlation between facing walls. Interpreted channel is shown in yellow while its projections on the planes parallel/perpendicular to the median paleocurrent orientation (N27E - N117E) are purple and magenta. B. Bounding surfaces are easily followed on the 3-D photo real model of the outcrop making possible correlation of strata exposed on the distinct sides of the quarry. C. Geometry pattern of a nested channel complex at the base-of-slope. Surfaces are built in Gocad from the scanned (x, y, z) data points to correspond to each individual channel. D. A surface is built to connect the parts of the channel that crop out on the two sides of the quarry. The channel is oriented NNE-SSW.

Fig. 10 Histogram frequency of channel dimensions. Channel width and depth measurements have been taken from submarine channels identified on the 3-D photo real model. A. Histogram frequency of apparent channel width as measured at the outcrop B. Histogram frequency of channel widths as measured in a plane perpendicular to the mean paleocurrent orientation (N34E) C. Histogram frequency of maximum preserved channel thickness D. Histogram frequency of channel thickness (difference in height measured between upper and lower channel boundaries). All four histograms display a bimodal distribution showing larger, aggradational channels to the left and smaller, lateral migrating channels to the right.

36 Fig.11 Channel aspect ratio (width vs. thickness). Estimated cross-channel width of the channels (apparent (red stars), as measured from the 3-D outcrop and corrected (black stars) width, perpendicular to the main paleocurrent direction, N34E) and maximum measured thickness of individual channels were plotted on a log-log graph and the aspect ratio calculated for each channel. Channel width commonly is one order of magnitude higher than channel depth. Channel bodies have rather flat channel fills (width/thickness ratio 4:1 to 16:1) with a mean aspect ratio of 9:1. Dark dots are channel dimensions from modern and ancient turbidite systems as published by Clark and Pickering, 1996.

Fig.12 Typical lithologic columns for the three types of channel infill exposed in the outcrop and associated depositional processes. A. Large, sandstone-rich, aggradational channels located in the axis of the canyon are typically filled by amalgamated, thick to very thick tabular to irregular-bedded, fine- grained, and massive sandstones deposited from high-concentration turbidity currents. B. The channel complex margin association is composed of thin bedded, massive to planar stratified, fine–grained sandstone deposited from low-density turbidity currents separated by centimeter-scale parallel laminated siltstone and shale. The channels show a deep basal cut with evidence of by-pass processes with mud-clast breccia overlying the erosive surface. At the top of the channel massive-to-planar stratified sandstone deposited from high density flows over-spills the filled channel. C. Small, laterally migrating channels in the upper part of the quarry are characterized by interbedded massive sandstones deposited from high-concentration flows and mud-clast breccia deposited by traction as bed load.

Fig. 13 Histogram frequency of bed dimensions. Facies thickness measurements have been taken from submarine channels identified on the 3-D photo real model. A. Histogram frequency of thickness of amalgamated sandstone and statistics. B. Histogram frequency of thickness of thin beds and statistics. C. Histogram frequency of thickness of debris flow deposits and statistics. D. Histogram frequency of thickness of shale layers and statistics.

37 Fig. 14 Facies percentages of total channel thickness. The four identified lithofacies are color coded and plotted as percentages of total channel thickness. The channels that are circled correspond to the small ones at the top of the quarry, while the others are large, aggradational channels that make up the lower two thirds of the quarry. Sandstone comprises a high proportion (more than 80%) of the channel fill.

The degree of connectivity of this stacking pattern (system) Most of the intra-channel element aspect ratios lie between, whereas the channel bodies have aspect ratio less than .

Deep-sea sediments have received considerable interest both for research purposes and economic reasons due to large hydrocarbon accumulations associated with turbidite deposits (Bouma, 2000). Ideally, characterization of hydrocarbon reservoirs and flow behavior requires information about heterogeneity at a submeter scale in three dimensions. To adequately document reservoir properties at reservoir scale requires the study of large, continuous three-dimensional outcrops (Slatt, 2000).

apparently moved episodically rather than progressively, as indicated by acoustic impedance contrasts within the channel complexes (individual channel bodies can commonly be differentiated) and by their lithological heterogeneity in the wells. Kolla et al. (2001), in their discussion of continuous vs. discontinuous (discrete) channel migration, inferred discrete channel migration to be associated with individual large flow events.We suggest that this discrete jumping of channel positions occurs on longer timescales and may be associated with temporary channel abandonment involving deposition of (1) finer grained drapes on point bars (see for example, Elliott, 2000), (2) debris-flow channel plugs, or (3) thinning- and fining-upward sequences.

38 Reestablishment of the channel in the new position involves erosion that may partially or entirely remove these deposits. We concur with Kolla et al. (2001) that discontinuous migration likely introduces more baffles into the reservoir than is likely to be the case in fluvial High-frequency alternations between aggradation and erosion are possibly related to changes in flow parameters on a timescale of a few thousand years (fifth order, Badalini et al., 2000). A relative fall in sea level may give rise to an increase in flow size and flow density, which will increase the efficiency of flows and reduce the slope gradient. Note however that any increase in grain size will tend to partially offset the increase in flow efficiency. Similarly, during a relative rise in sea level, flows may become less efficient due to decreases in the size and density of the flows. Again, any decreases in grain sizewill tend to reduce the effect but an overall increase in slope gradient is likely. On longer timescales, as long as accommodation remains or is created by longer term changes, such cycles may occur repeatedly, producing stacked sequences of aggradational channel (Figure 13). Short-term fluctuations in flow parameters may result in alternations between erosion/bypass and aggradation in an overall long-term aggradational system. Samuel et al, 2003AAPG

39

Load more

  39

Copy Link