A COMPUTER PROGRAM FOR GRAIN-SIZE DATA1

The computer program presented here seeks to improve estimation of statistical parameters for grain-size data by use of interpolated values. Interpolation is made by fitting a series of overlapping parabolas to the data, and follows the method of Snyder (1961). The values are used in moment formulas to compute standard statistical measures. Skewness and kurtosis are reduced by the interpolation data, and extreme positive values of kurtosis tend to be greatly reduced. The program also picks major modes, the median, and sediment type.     

Authigenic pyrite and gypsum in South West African continental slope sediments

An unusual association of authigenic pyrite and authigenic gypsum has been found in silty clays recovered from the South West African continental slope. Nannofossil content suggests that the sediments are Upper Miocene-Lower Pliocene in age. Pyrite occurs as (1) granular masses, (2) ‘worm’tubes, (3) foraminiferal infillings, and (4) framboids. Gypsum occurs as euhedral single or twinned crystals of selenite. ‘Worm’tubes and foraminiferal infillings of pyrite are partially or completely enclosed in some gypsum crystals. Electron microprobe analyses show a relatively high concentration of manganese in both the granular masses and tubes. Present-day waters off this coast are dominated by the upwelling of cold, nutrient-enriched waters (the Benguela Current). These rich waters support an enormous population of plankton. Death and decomposition of these plankton consume oxygen, thereby creating a belt of anaerobic sediments close to shore. An Upper Miocene-Lower Pliocene regression (Dingle & Scrutton, 1974) lowered sea level and shifted an older analogue of this upwelling zone seaward, eventually establishing an anaerobic environment in places on the present continental slope. Anaerobic bacteria thrived in these conditions. They reduced SO₄ dissolved in sea water, initiating the formation of H₂S. The H₂S reacted with iron minerals present in the sediment to form FeS. Addition of elemental sulphur produced pyrite. This strongly reducing, low pH, environment became saturated with calcium obtained by the dissolution of planktonic calcareous organisms. Gypsum was precipitated once the product of the concentrations of dissolved calcium and SO₄ exceeded the gypsum solubility product.     

Petrology of recent caliche pisolites, spherulites, and speleothem deposits from central Texas

Recent caliche, including nodules, pisolites, crusts, internal sediment, speleothem deposits, and spherulites, has formed within the dolomitic Cretaceous Edwards Formation of central Texas. As weathering altered the host strata, rhombic crystals of calcite were precipitated concomitantly with dissolution of the dolomite, thereby forming nodules. The highly altered dolomite (i.e. pulverulite) was then removed and spar, internal sediment, and travertine accumulated in the internodular voids. Nodular masses of calcite and dolomite are the most prominent constituent of the caliche. Some of the nodules have a well developed concentric structure as well as other characteristics similar to hypersaline pisolites. Features which appear to be useful in distinguishing caliche from hypersaline pisolites are: regional geological setting, association with other caliche and palaeosoil deposits, types of fossils present, and the presence of rhombic calcite and/or bladed sparry calcite with triangular shaped cross-section. A brick-like calcite texture and relict aragonite rays characterize hypersaline pisolites. Incipient neomorphism of the nodules and pisolites has resulted in the development of a radial pattern of spar within these structures. Geopetal deposits of internal sediment, including terra rossa soil, inhibited spar growth in the upward direction; consequently, spar is much better developed on the undersides of pisolites. Crusts and travertine flowstone (speleothem) deposits are intimately associated with the nodular masses and internal sediment. The brecciated thin crusts and travertine flowstone are end products of the same processes. The crusts formed during times of periodic desiccation of the growing surface while the flowstone formed when water was relatively abundant. Spherulitic bodies within the caliche, commonly 1–2 mm in diameter, display a radial texture and yet are composed of single crystals of calcite. The structures are the product of neomorphosed Microcodium or Microcodium-like globular bodies.     

The application of high-resolution sequence stratigraphy to fluvial systems: a case study from the Upper Carboniferous Breathitt Group, eastern Kentucky, USA

The Pennsylvanian Pikeville, Hyden and Four Corners formations of the Breathitt Group in eastern Kentucky, USA, contain six major facies associations along with a number of subassociations. These facies associations are offshore siltstone, rhythmically bedded mouthbar heteroliths, predominantly fine-grained floodplain deposits, minor channel fills, major distributary channels and major, stacked fluvial bodies. The stacked fluvial bodies are incised into a variety of open marine and delta plain deposits, have widths of several kilometres and exhibit a range of sandy fill types. These fluvial complexes are interpreted as incised valley fills. Parasequences and parasequence sets are not identifiable. Nonetheless, it is possible to identify systems tracts on the basis of sequential position, facies associations and systematic changes in architectural style and sediment body geometries. The studied portion of the Breathitt Group comprises stacked 4th-order sequences, which occur in lowstand, transgressive and highstand sequence sets related to the development of a lower frequency base level cycle. In the lowstand sequence set, incision associated with successive 4th-order sequence boundaries has commonly removed all the HST and TST of the underlying sequences, such that succeeding 4th-order incised valley fills are amalgamated. Within the transgressive sequence set, incision is at a minimum and incised valley fills tend to stack discretely with the maximum amount of fine-grained TST and HST between them. The highstand sequence set is transitional between the lowstand and transgressive sequence sets in terms of the amount of transgressive and highstand deposits preserved. Incised valley fills tend to stack discretely.     

Injection clastic dykes in the Lower Oxford Clay (Jurassic) of central England: relationship to compaction and concretion formation

An injection dyke of fine-grained sandstone derived from the Kellaways Sand Formation intrudes overlying organic-rich shales and shell beds of the Lower Oxford Clay. The dyke shows cross-cutting relationships with early carbonate concretions, and fills uncompacted kosmoceratid ammonite shells both within the concretions and surrounding shales. Internally the dyke displays flow-like features, and the walls show lobate flow structures. Clasts of uncompacted Lower Oxford Clay and fragments of pyrite-rich concretions occur within the sandstone intrusions. The sandstone of the dyke was cemented by calcite identical to that precipitated in septarian cracks in the concretions. This cementation took place prior to final compaction of the Oxford Clay. The dyke has a sub-parallel relationship to the nearby Tinwell-Marholm fault suggesting that the dyke may be related to local tectonic events during the Middle Jurassic.     

WHICH PRINCIPAL COMPONENTS TO UTILIZE FOR PRINCIPAL COMPONENT REGRESSION

Principal components(PCs)for principal component regression(PCR)have historically been selectedfrom the top down for a reliable predictive model.That is,the PCs are arranged in a list starting withthe most informative(PC associated with the largest singular value)and proceeding to the leastinformative(PC associated with the smallest singular value).PCs are then chosen starting at the top ofthis list.This paper discusses an alternative procedure of treating PC selection as an optimization prob-lem.Specifically,without any regard to the ordering,the optimal subset of PCs for an acceptablepredictive model is desired.Five data sets are analyzed using the conventional and alternative approaches.Two data sets are spectroscopic in nature,two data sets deal with quantitative structure-activityrelationships(QSARs)and one data set is concerned with modeling.All five data sets confirm thatselection of a subset without consideration to order secures the best results with PCR.One data set isalso compared using partial least squares 1.     

DATA ANALYSIS WITH MINIMAL ASSUMPTIONS

The computer has made it possible to scrutinize data rapidly by means of graphics. This should be doneprior to the application of any model to the data, since the model must be validated before using it asa means of analyzing the data. The procedure is illustrated in terms of two examples of real experimentaldata.     

Flow and sediment dynamics in a low sinuosity, braided river: Calamus River, Nebraska Sandhills

The interaction between channel geometry, flow, sediment transport and deposition associated with a midstream island was studied in a braided to meandering reach of the Calamus River, Nebraska Sandhills. Hydraulic and sediment transport measurements were made over a large discharge range using equipment operated from catwalk bridges. The relatively low sinuosity channel on the right-hand side of the island carries over 70% of the water discharge at high flow stages and 50–60% at low flow stages. As a result, mean velocity, depth, bed shear stress and sediment transport rate tend to be greater here than in the more strongly curved left-hand channel. The loci of maximum flow velocity, depth and bed shear stress are near the centre of the channel upstream of the island, but then split and move towards the outer banks of both channels downstream. Variations in these loci depend on the flow stage. Topographically induced across-stream flows are generally stronger than the weak, curvature-induced secondary circulations. Water surface topography is controlled mainly by centrifugal accelerations and local changes in downstream flow velocity. The averaged water surface slope of the study reach varies very little with discharge, having values between 0·00075 and 0·00090. As bed shear stress generally varies in a similar way to mean velocity, friction coefficients vary little, normally being in the range 0·07–0·13. These values are similar to those in straight channels with sandy dune-covered beds. Bedload is moved mainly as dunes at all flow stages. Grain size is mainly medium sand with coarse sand moved in thalwegs adjacent to the cut banks, and with fine sand at the downstream end of the island. These patterns of flow velocity, depth, water surface topography, bed shear stress, bedload transport rate and mean grain size can be accurately predicted using theoretical models of flow, bed topography and sediment transport rate in single river bends, applied separately to the left and right channels. During high flow stages deposition occurs persistently near the downstream end of the island, and cut banks are eroded. Otherwise, erosion and deposition occurs only locally within the channel as discharge varies. Abandonment and filling of a strongly curved channel segment may occur by migration of an upstream bar into the channel entrance at a high flow stage.     

Halite saltern in the Canning Basin, Western Australia: a sedimentological analysis of drill core from the Ordovician-Silurian Mallowa Salt

The Late Ordovician-Early Silurian Mallowa Salt of the Carribuddy Group, Canning Basin, north-west Australia, is the largest halite deposit known in Australia, attaining thicknesses of 800 m or more within an area of approximately 200 000 km². Study of 675 m of drill core from BHP-Utah Minerals’ Brooke No. 1 well in the Willara Sub-basin indicates that the Mallowa Salt accumulated within a saltern (dominantly subaqueous evaporite water body) that was subject to recurrent freshening, desiccation and exposure. Textures and bromine signatures imply a shallow water to ephemeral hypersaline environment typified by increasing salinity and shallowing into evaporitic mudflat conditions toward the top of halite-mudstone cycles (Type 2) and the less common dolomite/anhydrite-halite-mudstone cycles (Type 1). The borate mineral priceite occurs in the capping mudstones of some cycles, reinforcing the idea of an increasing continental influence toward the top of mudstone-capped halite cycles. The rock salt in both Type 1 and Type 2 cycles typically comprises a mosaic of large, randomly orientated, interlocking halite crystals that formed during early diagenesis. It only partially preserves a primary sedimentary fabric of vertically elongate crystals, some with remnant aligned chevrons. Intraformational hiati, halite karst tubes and solution pits attest to episodic dissolution. Stacked Type 2 cycles dominate; occasional major recharges of less saline, perhaps marine, waters in the same area produced Type I cycles. The envisaged saltern conditions were comparable in many ways to those prevailing during the deposition of halite cycles of the Permian Salado Formation in New Mexico and the Permian San Andres Formation of the Palo Duro Basin area in Texas. However, in the Canning Basin the cycles are characterized by a much lower proportion of anhydrite, implying perhaps a greater degree of continental restriction to the basin. The moderately high level of bromine in the Mallowa Salt (156·5 ± 43·5 ppm Br for primary halite, 146·1 ± 54·7 ppm Br for secondary halite) accords with evolved continental brines, although highly evaporative minerals such as polyhalite and magnesite are absent. The bromine levels suggest little or no dissolution/reprecipitation of primary halite and yet, paradoxically, there is little preservation of the primary depositional fabric. The preservation of early halite cements and replacement textures supports the idea of an early shutdown of brine flow paths, probably at burial depths of no more than a few metres, and the resultant preservation of primary bromine values in the secondary halite.