ABSTRACT

Deep-water clastics have been researched, explored, and produced from for several decades, but our understanding is still rudimentary. Terms as "turbidites" and "submarine fans" are well known and, like many other geological terms, are used both correctly and incorrectly. Models have been published during the last twenty years and they can be helpful either to direct our thinking or to prevent us from going astray. On the other hand, both the terminologies and models can be detrimental to the user, due to their tendency to create a stifling effect. Although an abundance of excellent work has been carried out, we have to ask ourselves several questions: how much do we really understand; do we feel comfortable with our deductions; and are we able to identify some of our shortcomings?

With the lease sales moving into continental slope environments and with the well known difficulties of obtaining good correlations in turbidite areas, more pressure is generated by higher management to produce workable exploration concepts, seismic characterization, and well log characteristics.

We feel that a major step in the realization of our shortcomings was taken during the September 1982 COMFAN (Committee on Fans) meeting in Harmarville, Pennsylvania, hosted by Gulf Research and Development Company. An international group of specialists on modern fans and ancient turbidite sequences convened and focussed their attention on (1) review of the current status of research on modern and ancient turbidites in fan and nonfan settings, (2) identification of the main reasons why comparisons between modern fans and ancient turbidite sequences are still in their infancy, (3) identification of the primary areas of confusion (i.e., nonuniform uses of terminology, process concepts, and data types), (4) presentation of the most recent, generally unpublished results of turbidite research by active specialists in the field, (5) attempts to recognize and define specific criteria such as facies types, morphologic features, morphometric zonations, sediment distribution and depositional processes that can be applied to both fan and nonfan turbidite environments (Normark, Mutti, and Bouma, 1983-84a), and (6) critical review and modification, if necessary, of the proposed Deep Sea Drilling Project program on the Mississippi Fan.

Little is understood about the influence on the distribution and characteristics of sands by the type of continental margin, size and shape of the sedimentary basin, sand/clay ratio of the sediment source, single over multiple point (line) source, eustatic sea level changes, and seafloor gradients that may strongly affect the transport processes and, thus, the sorting of the sands.

Published models are generally based on detailed studies of one or, at most, a few modern sub-marine fans or ancient turbidite systems. As a result, the observer has to be extremely cautious in applying one of those models to another area. In addition, models derived from ancient turbidite systems tend to show more differences than similarities with modern deposits and vice versa.

The COMFAN participants agreed to publish the examples discussed during the meeting in a special issue of Geo-Marine Letters (Normark et al., 1983-84b). All contributions followed a set format to make the papers more compatible. A wall chart was designed to (1) portray at a scale of 1:5,000,000 the variation in shape and size of several submarine fans, (2) present morphometric maps of modern submarine fans to accentuate the type of knowledge available and the directions of data acquisition taken, and (3) present in table format a large amount of data from many fans.

The lack of a common data set to compare modern submarine fans and ancient turbidite sequences is the result of differences in both the scale of observation and the data acquisition methods (Normark et al., 1979). It is highly probable that no common data set for comparison exists. Data collection from ancient turbidite units provides details primarily on layer thickness, distribution in lateral and vertical sense over short intervals, composition, grain size, color, sedimentary structures on bedding planes, trace fossils, and paleocurrent directions. Detailed stratigraphic sections, ranging in thickness from a few feet to hundreds of feet, can be measured and individual layers can sometimes be followed across the width of the outcrop. Biostratigraphic control does not have the same resolution as the depositional events that are represented by single layers or packets of layers. Marker beds generally are rare to provide time correlations basinwide. Erosion, masking of layers by debris or vegetation, and structural complications can prevent, or make it difficult to carry out detailed correlation attempts and, thus, reconstruction through time of the original basin configuration. Paleocurrent measurements seldom define the source area adequately.

Studies of modern submarine fans can provide a good insight into basin size, shape, and gradients. The surface morphology can be defined to a certain degree, and the source area(s) can often be identified accurately. A seismic stratigraphic framework can be developed and seismic facies identified and sometimes correlated. The total thickness of the fan and of subportions can be established. However, most detail can only be obtained from the water-sediment interface and from the upper sediment column by surface and near-bottom towed high-resolution equipment (side scan sonar, multibeam, high resolution reflection systems). Detail of information decreases with penetration, and basically these high resolution techniques do not provide the details observed

End_Page 345------------------------

in outcrops. Deep drilling with continuous coring, such as the Deep Sea Drilling Project and the new Ocean Drilling Project, are invaluable ways of narrowing this gap.

Observations on the surface and near surface areas represent the upper skin of a modern fan and, therefore, may not be representative at all of the principal depositional systems, Channels, for example, can be a few miles wide and tens to hundreds of feet thick. If the channel fill is characterized by a fining-upward sequence, observations in both modern and ancient systems will be severely handicapped by the depth-restricted data set.

Because only a few of the major points of discussion brought forward during the COMFAN meeting were published in Geo-Marine Letters, the most applicable ones are listed below.

1. There is no unanimity in the application of the term fan. This word normally suggests a fan-shaped deposit that results from long term active turbidite-type deposition, related to a point source and taking place in a large, unconfined basin of low relief and with gentle gradients. However, examples of modern fans often involve the infilling of narrow, irregular depressions that prevent the development of a typical fan shape. When viewing the wall map, one observes that the shapes are strongly influenced by basin configuration, and that an unrestricted basin is more often the exception rather than the common occurrence. These basins shape influences must have been equally common throughout geologic time. Most ancient turbidite systems now exposed in outcrops were deposited in active-margin settings and the chance that the receiving basins were rather narrow and oblong is high. Those fans that developed in passive-margin settings had a better chance to build a more fan - shaped deposit.
2. Channels and lobes can often be defined on modern fans. For ancient turbidite systems, we can only describe the sedimentary facies, and therefore it is advised to describe ancient deposits without specific morphologic connotations.
3. Entire turbidite sandstone bodies can seldom be seen in outcrop. As a result, the original morphology and three-dimensional shape of an ancient turbidite environment can only be inferred.
4. Vertical facies-sequence analysis has been widely used to define the physiographic subdivision of an ancient turbidite system. Internal changes in bed thickness, sandstone/shale ratio, texture, sedimentary structures, and other characteristics define directional relationships and finally serve to reconstruct the morphology of the ancient fan. Although researchers demonstrated the workability of the concepts from areas where detailed studies led to the development of a depositional model (Mutti and Ricci Lucchi, 1972), many followers misuse the model.
5. Thinning- and/or fining-upward sequences and thickening- and/or coarsening-upward sequences have been widely and often successfully used to recognize channel-fill and lobe deposits, respectively. A comparison of modern distributary channels and channel-mouth bars in fluvially-dominated deltas was used as the basis. The actual validity of such trends as environmental indicators has not been tested yet by sufficiently deep cores from modern fans. The meaning of accretion and progradation processes is therefore completely interpretive. Throughout the literature much confusion exists about the criteria and even the terminology that should be applied to establish those sequences. Unfortunately, this has resulted in highly inconsistent usage.
6. It was generally agreed upon that too few investigators discuss the tectonic setting of the ancient or modern fan location, and consequently, the potential tectonic influence on source material, basin shape, and other controlling factors is seldom mentioned.
7. Investigators of ancient turbidite sequences may mention a point or a line source for the sediments without defining what such a source means. The term line source, especially, can result in confusion if it is not defined, provided it even exists.
8. Little study has been attempted as to the mechanisms that set the material from a shallow water source area into motion. Terms like oversteepening, heavy storms and wave action, earthquakes, and pore-water pressure are too freely used and avoid the real question.
9. Few investigators have a proper feeling for a channel, its dimensions and its fill. The dimensions of a channel depth and fill may range from a few tens to hundreds of feet. In actuality, the channel depth at any given time is in the order of tens of feet, whereas the total accumulated fill can be hundreds of feet thick.
10. We know very litle about the fill of a channel. Is it always a fining-upward sequence with possibly a coarse grained lag deposit overlain by a passive fill, or can it be an active depositional fill? In both cases we deal with channel deposits. However, the first case refers to a conduit, the second to an aggradational channel.
11. It became apparent that small and large sized sediment failures are more common than heretofore realized. Reasons and mechanisms are not yet being seriously studied.
12. Specifically, the specialists working on modern fans agreed that eustatic sea level variations could have a major influence on forming submarine fans with the major activity taking place near low sea level stands.

From these points, it becomes clear that we may be in a position to formulate specific questions which will direct our attention while resolving those inquiries. Concentrated discussions about the general topic and basically restricting it to one area will advance our knowledge. Because of the high costs involved in data gathering offshore, such as different types of seismic acquisition (high resolution and CDP), deep coring and well logging, the industry should pull together to set up a COST-well program utilizing Deep Sea Drilling Project coring techniques.