Downhole Geophysical Logs Research Articles

Effect of spatial variability of downhole geophysical logs on machine learning exercises

Machine Learning (ML) based estimation of petrophysical properties from downhole geophysical logs has attracted a significant attention from across the geoscience community in recent years. The main focus of the studies in the area is to identify the most accurate ML algorithm for petrophysical property prediction using logging data often from a very limited number of boreholes (e.g. only one or two). Despite having vertical variability in one borehole, the horizontal variability cannot be captured using limited number of boreholes thus the effect of spatial variability on ML based models' performance needs a specific attention. We have therefore developed an innovative workflow that assesses the impact of spatial variation in training geophysical logs data on the accuracy of ML models in predicting formation properties e.g. compressional wave velocity. To build this systematic workflow, a large dataset containing geophysical logs from 50 boreholes in a coal mine in Bowen Basin, QLD, Australia was selected for the analysis. Boreholes mainly intersected geological formations from Rangal coal measure rocks. The data included density, compressional wave velocity (VP), gamma ray, resistivity and depth. The workflow consists of i) a clustering analysis using the K-mean algorithm to determine if the data could be partitioned into smaller, more homogeneous clusters based on density, gamma ray and measured depth, ii) putting aside five boreholes for validation (these boreholes are not seen by the models, named test boreholes) and iii) selectin training datasets randomly from the remaining boreholes grouping them into 1, 5, 10, 20, 30 and 40 boreholes. The data selection from each group was repeated randomly five times to capture the potential randomness; and iv) the Least Squares Support Vector Regression model of each group was trained with 10, 20, 30, …, 90 and 100% of the data of each group forming 300 predictive models. The accuracy of developed models to predict VP from other logs was evaluated by calculating R2 value comparing the estimated and the actual VP values for five test boreholes. The results from the proposed workflow interestingly suggest that the model should prioritize incorporating data from various boreholes rather than collecting more data points from a limited number of boreholes. This observation is mainly linked to high lateral variability in the study area. Consequently, utilizing a smaller portion of data from a larger number of boreholes yields a simpler yet more accurate predictive model, which also reduces the computational cost.

In-situ physical and elastic properties of Archaean basement granitoids in the Koyna seismogenic zone, western India from 3 km downhole geophysical well logs: Implications for water percolation at depth

The Koyna region in western India is characterized by recurrent artificial water reservoir triggered seismicity since the impoundment of the Shivaji Sagar lake in 1962. Physical and mechanical properties of the crystalline basement rock beneath the Koyna seismogenic zone are determined from the downhole geophysical logs in a 3 km research borehole. The borehole, KFD1, is located in the proximity of the Donichawadi fissure zone (DFZ), the surface manifestation of the fault associated with the 1967 M 6.3 Koyna earthquake. KFD1 penetrated 1247 m thick Deccan traps and continued 1767 m into the underlying granite-gneiss basement. Geophysical logs provide a rare opportunity to characterize the basement granitoids, which host the recurrent seismicity. Several major zones of anomalous physical and mechanical properties were delineated below a depth of 2 km from the analysis of downhole geophysical logs. High Poisson's ratio and high neutron porosity across anomalous zones imply the presence of water saturated rock. The analyses of Nuclear Magnetic Resonance log reveals the presences of clay bound and free fluid across these zones. Sharp decrease in dynamic Young's modulus with increasing Poisson's ratio imply the presence of fractures saturated with water. These hydraulically conductive fractures are potential locations for fault reactivation. Additionally, strike of major and minor fractures associated with the anomalous zones are compared with the strike of Donichawadi fault. High angle between strike of the major shear fractures in the anomalous zone and strike of the Donichawadi fault imply that borehole KFD1 may have either intersected subsidiaries of the main fault at multiple depths or a new hidden fault system.

Constraining the lithostratigraphic architecture of a buried bedrock valley using surface electrical resistivity and seismic refraction tomography

Buried bedrock valleys are common erosional features in northern mid-latitude environments that form through glaciofluvial or paleoalluvial processes and are typically infilled by Quaternary-aged sediments. The erosional extent and geometry of the valley including a weathered interface, along with sediment infill that can contain complex sequences of unconsolidated aquifer and aquitard sediments, mean these features may act as preferential pathways to deeper bedrock aquifers. Noninvasive geophysical tools can provide rapid, high-resolution subsurface characterization of these features. This study evaluates the application of electrical resistivity and seismic refraction tomography along two transects centred over a buried bedrock valley in Elora, Ontario, Canada. Geophysical measurements were combined with existing continuous core records and an electrofacies model based on downhole geophysical logs to constrain the morphology and infilled lithostratigraphic architecture of the valley. Bedrock competency associated with lithology may act as a control on depth and width of valley incision during erosion, with resistivity measurements of the bedrock revealing a potential association between interpreted mechanical properties and variations in the resolved valley morphology. Seismic velocity corroborated these contrasting valley widths but could not assess bedrock competency variability below the bedrock interface. This study reveals the sequence of events depositing sediments in the valley, yielding a revised architectural mapping that improves on previous regional scale lithostratigraphic interpretations. Results will be of use to groundwater practitioners requiring detailed conceptualization of this buried bedrock valley and its role on preferential zones of groundwater flow. Similar approaches can be used for delineation of these common and hydrogeologically significant features.

A multiphysics field-scale investigation of gas pre-drainage in sorptive sediments

Gas pre-drainage through underground-to-inseam (UIS) horizontal boreholes is one of the common practices for the management of environmental and safety concerns in underground coal mining. In the design of UIS boreholes, local experience is invaluable; however, an experienced-based design is not necessarily optimal from both economic and technical perspective. In this study, we present an informed, data-driven design and optimization methodology that employs a multiphysics coal seam gas flow model for simulating gas pre-drainage on different scenarios of UIS borehole pattern. The implementation of the methodology is presented through a case study from an Australian underground coal mine. The methodology uses an in-house three-dimensional finite element (FEM) model (referred to as NETCoal) that has been developed based on the non-equilibrium thermodynamics and continuum mechanics. NETCoal considers the interacting processes of matrix gas diffusion, fracture gas flow and rock stress evolution in a coupled manner. This allows the permeability tensor of each grid element to be updated in response to evolving stresses on fractures induced by matrix shrinkage due to gas desorption (sorptive-mechanical effect) and gas pressure depletion (hydro-mechanical effect). The proposed methodology for the UIS borehole design requires a three-dimensional geological/structural grid model to be constructed for the area under consideration using available field data. Material and flow properties of coal (e.g., Isotherm, flow, mechanical properties etc.) are then calculated particularly using downhole geophysical logs (DGL) and physical equations/empirical correlations at borehole locations and distributed onto other grid cells using the kriging method to build a property model. NETCoal imports these geological and property models in a corner point gridding format and generates its own finite element mesh that suits the UIS borehole pattern under consideration. A Warren-Root type shape factor is used for the calibration of the numerical results of gas content with actual borehole flow data. Once calibrated, the model simulates gas drainage for different scenarios to determine an optimal borehole pattern which is the one satisfying the threshold limit value (TLV) and lead time constraints while requiring least drilling length. In addition to introducing an innovative practical methodology for the design of UIS pre-drainage boreholes, this study provides insights into the effect of coal seam and geometry parameters on gas pre-drainage through a sensitivity analysis.

A novel workflow based on physics-informed machine learning to determine the permeability profile of fractured coal seams using downhole geophysical logs

Permeability is arguably the most important parameter in gas pre-drainage and production practices in coal mining and coal seam gas industries. In fractured coal seams, permeability can significantly vary both laterally and vertically. Knowing the vertical variation of permeability in relatively thick coal seams is important in optimizing inseam pre-drainage design in underground coal mining and improving reservoir management and completion design efficiency in the coal seam gas industry. Existing empirical/statistical coal permeability models often yield poor results in estimating permeability profiles where there is significant variation in the characteristics of coal fracture systems. This study presents a novel workflow for evaluating the permeability profile of naturally fractured thick coal seams using downhole geophysical logging data. The workflow combines physics-based simulation, laboratory experiments, and a data-driven machine learning approach for estimating the permeability profile. As part of this workflow, several coal specimens from the study coal seam are first tested under different stresses to measure their permeability, density, and ultrasonic responses. Numerical simulation of the Navier-Stokes fluid flow and elastic wave propagation was then performed on a range of constructed coal block geometries with various fracture densities. In the next step, a physics-informed, neural network-based model (PINN) was trained using the data obtained from both laboratory experiments and numerical simulation. The model inputs included density, as well as compressional and shear wave velocities of the coal seam and its output is the trend of permeability variation across the coal seam interval. In-situ permeability measurement at a certain depth (e.g., from well pressure testing) was then used to convert the permeability trend to the entire permeability profile of the coal seam along its interval. The performance of the proposed workflow was finally evaluated using a suite of downhole geophysical logging data from a borehole intersecting two coal seams in an eastern Australian basin. This study demonstrates that the proposed workflow is relatively simple to implement and yet accurate for evaluating the permeability profile of a coal seam across its interval.

Delineation of Damage Zones From 3 km Downhole Geophysical Logs in the Koyna Seismogenic Zone, Western India

AbstractDelineation of subsurface faults and damage zones is a major goal of scientific drilling projects in seismically active areas. Geophysical logs acquired in a 3‐km deep scientific borehole KFD1 in the Koyna seismogenic zone, a site of recurrent reservoir triggered seismicity over the past 55 years, provide an unprecedented opportunity to investigate the rock properties and delineate the fault zones. KFD1 passed through 1,247‐m thick Deccan traps and continued for 1,767 m into the underlying granitic basement rocks that host the seismic activity in the depth range 2–10 km. We have studied the physical properties and acoustic behavior of basement granitoids from the analysis of geophysical logs from 1,500 to 3,000 m. Salient results are as follows. (1) Seven anomalous zones are identified below 2,100‐m depth based on electrical resistivity, caliper, density, neutron porosity, self‐potential, and sonic data. (2) The anomalous zones are characterized by significant shear wave velocity anisotropy (up to 25%), as revealed by cross‐dipole sonic data. (3) Dispersion analysis of dipole flexural modes confirms that the anisotropy is primarily stress induced; fast polarized shear wave azimuth (FSA) therefore indicates the orientation of maximum horizontal compressive stress SHmax. (4) Comparison of FSA with independent estimates of SHmax orientations obtained from drilling induced tensile fractures and strike of inclined fractures in the anisotropic zones shows that FSA is controlled mainly by the stress regime. Therefore, the stress rotations inferred from anisotropy analyses in the anomalous zones indicate their association with subsurface fault damage zones.

Geophysical, geological, and hydrogeological characterization of a tributary buried bedrock valley in southern Ontario

Buried bedrock valleys infilled with Quaternary-aged sediment have the potential to become productive aquifers owing to prevalent sand and gravel deposits often associated with these topographic lows. In areas where groundwater is drawn from the underlying bedrock aquifer, buried bedrock channels may significantly affect the spatial distribution of recharge and localized contaminant pathways. Therefore, understanding the form, distribution, and the nature of Quaternary infill sediments within these buried bedrock river valleys, and their relationship to hydraulically transmissive bedrock features is an important aspect of groundwater resource management. Here, we evaluate the effectiveness of electrical resistivity and seismic refraction collected over a partially urbanized 150 ha area with variable vegetation, roads, and structures, to map the spatial distribution of sediments and delineation of a channel segment associated with a regional bedrock valley. Electrical resistivity and seismic refraction was performed along 13 (covering ∼11.6 km) and seven transects (covering ∼0.9 km), respectively, to map and characterize the bedrock surface morphology beneath a variable thickness of unconsolidated deposits. Three continuously cored holes and downhole geophysical logs, supplemented with four nearby water well records captured the in-channel as well as adjacent Quaternary stratigraphy (∼15–40 m). Cores recorded multiple glacial till deposits and ice-marginal processes associated with ice advances and retreats. Hydraulic transmissivity of the bedrock around the valley feature was evaluated using a FLUTe hydraulic transmissivity profiling technique. This study demonstrates the potential of combining several surface geophysical methods with sedimentological analysis of continuous cores and hydraulic data for characterizing tributary bedrock channel morphology and Quaternary infill at a scale relevant to localized studies of municipal production well recharge zones and contaminant transport and fate.

Robust Library Building for Autonomous Classification of Downhole Geophysical Logs Using Gaussian Processes

Machine learning techniques such as Gaussian Processes can be used to identify stratigraphically important features in geophysical logs. The marker shales in the banded iron formation hosted iron ore deposits of the Hamersley Ranges, Western Australia, form distinctive signatures in the natural gamma logs. The identification of these marker shales is important for stratigraphic identification of unit boundaries for the geological modelling of the deposit. Machine learning techniques each have different unique properties that will impact the results. For Gaussian Processes (GPs), the output values are inclined towards the mean value, particularly when there is not sufficient information in the library. The impact that these inclinations have on the classification can vary depending on the parameter values selected by the user. Therefore, when applying machine learning techniques, care must be taken to fit the technique to the problem correctly. This study focuses on optimising the settings and choices for training a GPs system to identify a specific marker shale. We show that the final results converge even when different, but equally valid starting libraries are used for the training. To analyse the impact on feature identification, GP models were trained so that the output was inclined towards a positive, neutral or negative output. For this type of classification, the best results were when the pull was towards a negative output. We also show that the GP output can be adjusted by using a standard deviation coefficient that changes the balance between certainty and accuracy in the results.

Integrated Geological and Geophysical Interpretation for the Koodaideri Detrital Iron Deposits, Fortescue Valley, Western Australia

This paper presents a review of integrated interpretation of geophysical surveys with geological data for interpretation and exploration targeting for detrital iron deposits at the Koodaideri Project, Western Australia. Significant previous exploration has been conducted in the area and this has identified a number of detrital iron deposits. The aim of this project has been to integrate all available geoscientific data in order to assess remaining prospectivity of the area and provide a framework for future exploration and evaluation projects. Previous exploration at Koodaideri has used a variety of techniques including drilling, downhole geophysical logs, sparse refraction seismic, airborne and ground gravity, airborne magnetics and time-domain airborne electromagnetics. Spatial coverage of the individual exploration datasets is irregular, and the first stage of this project has focussed on an area of approximately 42 km × 12 km within which there is reasonably good coverage of all data types. The relatively high data density has allowed the relationships between the various data types to be assessed and effective exploration parameters to be defined. The larger detrital iron deposits at Koodaideri occur within palaeochannels or depressions within the basement, which is mainly comprised of units of the Wittenoom Formation. The detrital iron deposits are considered to have been sourced from erosion of bedded iron deposits of the Brockman Iron Formation which outcrops on the high ground both upstream and immediately to the southwest of the area of interest. The known deposits generally occur beneath cover of variable thickness of up to 50 m. The detrital deposits themselves may have thicknesses in excess of 100 m in major palaeochannels and sinkholes. The detrital deposits have a higher density than other cover units due to their high iron content (>50% Fe). However, gravity alone is not an effective exploration technique because the gravity signature is complicated by significant variability in the depth to higher density Wittenoom Formation bedrock. A more recent development is use of a modified seismic refraction method (Sparse Refraction Seismic) to constrain the basement topography, and then to model and remove the basement response from the observed gravity data to identify areas of anomalous excess mass. This approach has allowed cost-effective semi-regional exploration and has been successful in identifying all known major detrital iron deposits. This study extended the excess mass approach by constructing revised basement models from the sparse refraction seismic and drilling and from interpretation of the SkyTEM airborne electromagnetic and drilling. The results show that the basement interface can be interpreted from either the seismic or airborne electromagnetic datasets, although the airborne electromagnetic interpretation is complicated by highly saline groundwater in the northeastern quadrant of the area and by conductive shale units within the Wittenoom Formation bedrock. Almost all of the known detrital and channel iron deposits are spatially associated with an overlying pisolite unit, which can be identified from the magnetic data via its characteristic magnetic texture. These studies have shown that the derived excess mass is spatially associated with the known detrital and channel iron mineralisation. Significantly, almost all of the known deposits were also successfully identified from the simple geological model, in the absence of drill hole constraints. A number of untested areas of possible mineralisation have been identified, as well as potential extensions or alternate trends to known mineralisation. The modelling scenarios tested confirmed that the minimum elements for exploration targeting at Koodaideri are a geological model incorporating basement topography, interpreted magnetic domains and geologically constrained inversion of the gravity data.

Integrated Use of Geological, Geophysical, Radiocarbon and Stable Isotopes Data for Tracing the Conduit Flow Paths in a Small Karstic Aquifer: Poocher Swamp Freshwater Lens, South Australia

<p class="1Body">An integrated approach combining lithological logs, downhole geophysics, electromagnetic survey and the distribution of radiocarbon (<sup>14</sup>C) and the stable isotopes of water molecules (<sup>18</sup>O) were used to identify the conduit flow paths of a small freshwater lens. Lost circulation zones, where drilling fluid flows into geological formation instead returning up the annulus recorded during water well drilling, were considered as the major fracture zones. The presence of high porosity zones within boreholes were identified using caliper, gamma and neutron logs. These methods were used to identify the depth intervals at which cavities and the existence of conduit porosity within the boreholes. Transient electromagnetic (TEM) method was used to investigate resistivity anomalies in the profiles along nine pre-determined lines across the freshwater lens. Resistivity anomalies were related to borehole information and other surface features such as sinkholes. Low resistivity zones of the TEM tomography sections had excellent correlation to fracture zones identified during well drilling, and downhole geophysical logs. Similarly, high resistivity zones in the profiles correlate well with the zones of cemented or recrystallized limestone identified from the lithological logs. The interpreted resistivity anomaly accurately depicts depth to watertable at 15-18 m, presence of main fracture zone at 20-25 m, presence of possible conduits flow paths and the cemented or recrystallized limestone below 35 m depth. The <sup>14</sup>C, <sup>13</sup>C and <sup>18</sup>O signatures of the groundwater confirm the presence of conduits and potential pathways of preferential flows. This investigation illustrates the effectiveness using an integrated approach to trace the conduit flow paths in karst aquifers. The information gained from the study is currently being used for the management of the freshwater lens.</p>

Three-dimensional hydrofacies assemblages in ice-contact/proximal sediments forming a heterogeneous ‘hybrid’ hydrostratigraphic unit in central Illinois, USA

Three-dimensional (3-D) hydrostratigraphic modelling of glacial sediment assemblages was undertaken as part of a groundwater study in central Illinois, USA. Sediments comprising these assemblages, informally referred to as the Glasford deglacial unit, form discontinuous sand-gravel layers including small aquifer zones, and fine-grained interstratified layers that may impede groundwater movement. This unit is stratigraphically above a regional aquitard overlying the important Mahomet aquifer. The study improves understanding of the internal stratigraphic architecture and hydrostratigraphic character of the unit. Data include descriptions of continuous cores, profiles of near-surface and downhole geophysical logs, and sediment descriptions from water well logs. Discrete bounding surfaces constructed using gOcad represent the main lithofacies assemblages forming a 3-D framework. The framework was further partitioned into a 3-D cellular grid for mapping the spatial distribution of fine- and coarse-grained facies. Hydraulic conductivity (K G) estimates were used to convert these lithofacies into hydrofacies. Medium- to coarse-grained hydrofacies (K G = 1.25 × 10−5 m/s) represent 46 % of the total volume, the remainder being fine-grained hydrofacies (K G = 3.01 × 10−8 m/s). The spatial pattern of these hydrofacies is highly heterogeneous, thus, designating the Glasford deglacial unit as an aquifer or aquitard would be conceptually misleading. The term “hybrid hydrostratigraphic unit” is introduced to better represent conceptually this type of unit in hydrostratigraphic models.

Lithostratigraphy from downhole logs in Hole AND-1B, Antarctica

The ANDRILL (Antarctic Drilling Project) McMurdo Ice Shelf (MIS) project drilled 1285 m of sediment in Hole AND-1B, representing the past 12 m.y. of glacial history. Downhole geophysical logs were acquired to a depth of 1018 mbsf (meters below seafloor), and are complementary to data acquired from the core. The natural gamma radiation (NGR) and magnetic susceptibility logs are particularly useful for understanding lithological and paleoenvironmental change at ANDRILL McMurdo Ice Shelf Hole AND-1B. NGR logs cover the entire interval from the seafloor to 1018 mbsf, and magnetic susceptibility and other logs covered the open hole intervals between 692 and 1018 and 237–342 mbsf. In the upper part of AND-1B, clear alternations between low and high NGR values distinguish between diatomite (lacking minerals containing naturally radioactive K, U, and Th) and diamictite (containing K-bearing clays, K-feldspar, mica, and heavy minerals). In the lower open hole logged section, NGR and magnetic susceptibility can also distinguish claystones (rich in K-bearing clay minerals, relatively low in magnetite) and diamictites (relatively high in magnetite). Sandstones can be distinguished by their high resistivity values in AND-1B. On the basis of these three downhole logs, diamictite, claystones, and sandstones can be predicted correctly for 74% of the 692–1018 mbsf interval. The logs were then used to predict facies for the 6% of this interval that was unrecovered by coring. Given the understanding of the physical property characteristics of different facies, it is also possible to identify subtle changes in lithology from the physical properties and help refine parts of the lithostratigraphy, for example, the varying terrigenous content of diatomites and the transitions from subice diamictite to open-water diatomite.

Physical properties of sediment from the Mount Elbert Gas Hydrate Stratigraphic Test Well, Alaska North Slope

This study characterizes cored and logged sedimentary strata from the February 2007 BP Exploration Alaska, Department of Energy, U.S. Geological Survey (BPXA-DOE-USGS) Mount Elbert Gas Hydrate Stratigraphic Test Well on the Alaska North Slope (ANS). The physical-properties program analyzed core samples recovered from the well, and in conjunction with downhole geophysical logs, produced an extensive dataset including grain size, water content, porosity, grain density, bulk density, permeability, X-ray diffraction (XRD) mineralogy, nuclear magnetic resonance (NMR), and petrography. This study documents the physical property interrelationships in the well and demonstrates their correlation with the occurrence of gas hydrate. Gas hydrate (GH) occurs in three unconsolidated, coarse silt to fine sand intervals within the Paleocene and Eocene beds of the Sagavanirktok Formation: Unit D-GH (614.4 m–627.9 m); unit C-GH1 (649.8 m–660.8 m); and unit C-GH2 (663.2 m–666.3 m). These intervals are overlain by fine to coarse silt intervals with greater clay content. A deeper interval (unit B) is similar lithologically to the gas-hydrate-bearing strata; however, it is water-saturated and contains no hydrate. In this system it appears that high sediment permeability ( k) is critical to the formation of concentrated hydrate deposits. Intervals D-GH and C-GH1 have average “plug” intrinsic permeability to nitrogen values of 1700 mD and 675 mD, respectively. These values are in strong contrast with those of the overlying, gas-hydrate-free sediments, which have k values of 5.7 mD and 49 mD, respectively, and thus would have provided effective seals to trap free gas. The relation between permeability and porosity critically influences the occurrence of GH. For example, an average increase of 4% in porosity increases permeability by an order of magnitude, but the presence of a second fluid (e.g., methane from dissociating gas hydrate) in the reservoir reduces permeability by more than an order of magnitude.

Attempts to model the Lower Cretaceous pelagic sedimentary cover of Shatsky Rise (NW Pacific Ocean) from downhole geophysical records

Ocean Drilling Program (ODP) Leg 198 to Shatsky Rise cored Lower Cretaceous sediments at three sites (Sites 1207, 1213, and 1214). Unfortunately, core recovery was poor at all sites, most likely due to the contrast in physical properties between poorly cemented, very soft carbonate ooze/chalk and well-cemented, very hard biosiliceous chert. Downhole geophysical logs from two sites (1207 and 1213) provide continuous, in situ records of physical and chemical properties that can be used to elucidate the sedimentary history of Shatsky Rise in the early Cretaceous. The downhole logs show that at both sites two stratigraphic intervals (dated as late Barremian–Aptian and early Albian) are characterised by low porosity, high resistivity sediments. The timing of these intervals is synchronous at both sites and it has been previously proposed that the ‘resistivity highs’ are the result of increased biosiliceous sedimentation and subsequent silica cementation. In order to test this hypothesis, we use matrix algebra to invert the logging data from Site 1207 to produce two compositional models — the first reconstructs lithology, the second reconstructs mineralogy. The results of the lithology model suggest that the abundance of chert is highest during the resistivity highs. The results of the mineralogical model are more complex because of the probable presence of additional minerals such as barite.

Pitfalls in the geological characterization of alluvial deposits: site investigation for reactive barrier installation at Aznalcóllar, Spain

The alluvial deposits of the Agrio River in SW Spain have been studied using terrace mapping, boreholes, trenching and vertical electrical sounding to select an adequate place for a Permeable Reactive Barrier. Geological and hydrogeological data available prior to the barrier construction suggested a simple geological model based on three terraces, the most modern being deeper than the oldest. The barrier was accordingly trenched through the youngest terrace. However, excavation of the barrier andsubsequent subsoil data demonstrated that the internal structure of the Agrio alluvial deposits does not follow such a simple model. A revised model, less favourable to the existing barrier design, revealed the oldest terrace to be deeper than the youngest, and to form a palaeochannel oblique to the surface terraces and river trends. The study methods used are standard, but proved to have insufficient resolution to detect key features of the alluvial geology. It is concluded that such characterization methods, though widely used, are not appropriate where alluvial terraces may display complex internal structures not reflected in the modern surface geomorphology. In hindsight, the study area should have been larger, so as to encompass at least the width of the alluvial plain and to extend for a hundred metres or so up and downstream from the proposed barrier location. Lithological logs of the boreholes should have been carefully described from both cuttings and downhole geophysical logs, which would have allowed more accurate delineation ofthe stratigraphy of the alluvial deposit. Subsequentgeophysical methods should have been calibrated to this stratigraphy to characterize the internal structure and basal contact of the alluvial deposit.

Hydraulic transience and the role of bedding fractures in a bedrock aquitard, southeastern Wisconsin, USA

Hydraulic heads in a dolomitic shale bedrock aquitard in Wisconsin, USA, are apparently not in equilibrium with drawdown in the underlying aquifer system caused by pumping for municipal supply over the last century. Measurements of head with depth, downhole geophysical logs, and estimates of hydraulic conductivity indicate very low vertical hydraulic diffusivity, and show that high horizontal permeability within the aquitard due to bedding‐plane fracture zones can allow lateral groundwater flow. Unlike the hydrogeological conceptual models used in many investigations, flow in aquitards cannot always be inferred to be primarily one‐dimensional and vertical. Failure to account for transient conditions and lateral flow in similar settings of intensive groundwater pumping could lead to error in estimates of aquitard leakage and underlying aquifer properties.

Evaluation of multifrequency airborne EM for opal prospect definition at Lightning Ridge, NSW

The NSW Department of Mineral Resources has been evaluating the use of geophysical techniques as a tool to better understand controls on opal formation in the Lightning Ridge district of New South Wales. A detailed study over the Gurley Ridge, approximately 25km southwest of the Lightning Ridge township, commenced in June 2001. This Cretaceous sedimentary ridge hosts the Coocoran Opal Field and includes the productive Allahs and Natelies Dream prospects. Current geological understanding predicts opal occurrence at the interface of clay and sandstone horizons within the sedimentary package. The study included a compilation of historic drill records, detailed surface geological mapping, lithofacies mapping, and acquisition of detailed airborne electromagnetic data. An electrical model of the subsurface generated from previous ground work was used to guide selection of the transmit?receive frequencies employed in the airborne survey. The acquired multifrequency airborne data were inverted and conductivity depth images for each flight and tieline produced. The CDI sections revealed the existence of an ubiquitous conductor located between 10m and 20m below the surface, covering the majority of the ridge area. A ground follow-up program was instigated to investigate the identified conductive zone. Eight high-resolution, detailed transient EM ground lines were acquired coincident with various conductive signatures identified in the airborne data. A small drill program was conducted to ground truth both the airborne and TEM subsurface conductive features. Drill results, downhole geophysical logs, and the detailed ground TEM data have been used to validate and refine inversion of the airborne data.

Multi-scalar structure at DSDP/ODP Site 504, Costa Rica Rift, III: faulting and fluid circulation. Constraints from integration of FMS images, geophysical logs and core data

Abstract Downhole geophysical logs and high-resolution electrical images (FMS) from DSDP/ODP Hole 504B are analysed in combination with core data to obtain an integrated description of oceanic faults met in the hole. About 34 500 fractures were mapped from FMS images over 1672 m of basement. The fracture distribution from FMS confirms the presence of a main fault zone between 800 and 1100 mbsf (metres below sea floor), elsewhere detected from seismic data as well as magnetic, acoustic, and electrical resistivity measurements. The fracture density profile reveals the presence of two other highly fractured zones, (1) between 400 and 575 mbsf and (2) close to the bottom of Hole 504B (1700 to 2100 mbsf). Consequently, we infer that Site 504 was submitted first to an extensional stress regime near the ridge axis, with circulation of high-temperature fluids and pervasive alteration of the basalts. This initial phase is associated with the main fault met in Hole 504B. Similar but less developed deformation was generated off-axis, with lower-temperature parageneses, such as that cored between 400 and 575 mbsf. The present compressional to strike-slip stress regime is expressed in subhorizontal fracturing detected in discrete zones, such as within the main fault zone and the lower fracture zone (1700 to 2100 mbsf) in Hole 504B.

Aptian‐Albian sea level history from Guyots in the western Pacific

Relative sea level fluctuations are an important control on patterns of sedimentation on continental margins and provide a valuable tool for regional correlations. One of the main objectives of combined Ocean Drilling Program Legs 143 and 144 was drilling the thick carbonate caps of a suite of seamounts, called guyots, scattered over the northwestern Pacific. The array of drowned Cretaceous banks includes four carbonate banks of Aptian‐Albian age. These particular carbonate banks display emergent surfaces if regional sea level falls faster than the rate of guyot subsidence, or intervals of condensed parasequences and well‐cemented peritidal crypto‐algal flats if the rate of sea level fall is slightly less than guyot subsidence. Rapid rises of sea level following these sequence boundaries are recorded as drowning of the emergent horizons or as pronounced deepening of facies. The cored lithologies and downhole geophysical and geochemical logs were used to identify depositional sequences and surfaces of exceptional shallowing or deepening. A combination of biostratigraphic datums, carbon and strontium isotope curves, relative magnitude of surfaces of emergence, relative thicknesses of depositional sequences, sea level events, and counts of upward shallowing cycles or parasequences were used to correlate sequences among the four sites. After compensating for thermal subsidence rates at each guyot, an identical pattern of major Aptian‐Albian eustatic sea level events is evident throughout this large portion of the Pacific Ocean. There are approximately 12 Aptian and 12 Albian significant sequence boundaries, of which a third were associated with major episodes of emergence. When these events are compared with Aptian‐Albian relative sea level changes observed in European shelf successions, the major sequence boundaries and transgressive surges can be easily correlated, and it appears that both regions also display the same number of minor events. Therefore we can apply the relative timing of these events from the thermal subsidence rates and parasequence counts of the Pacific banks to construct an improved scaling of the associated ammonite zones and biostratigraphic datums in the Aptian‐Albian interval.An electronic supplement of this material may be obtained on adiskette or via Anonymous FTP from KOSMOS.AGU.ORG (LOGINto AGU's FTP account using ANONYMOUS as the username andGUEST as the password. Go to the right directory by typing APEND.Diskette may be ordered from American Geophysical Union, 2000Florida Ave., N.W., Washington, D.C. 20009, $15.00. Payment mustaccompany order.

Physical properties of 110 MA oceanic crust at Site OSN‐1: Implications for emplacement of a borehole seismometer

An extensive suite of downhole geophysical logs was acquired through a 70 m thick section of basalt drilled by the Ocean Drilling Program (ODP) into 110 Ma oceanic crust 225 km SW of Hawaii. These data include standard sonic, electrical, and nuclear logs, as well as electrical and acoustic images of the wellbore, and enable a vertically continuous characterization of the physical properties of shallow, old oceanic crust. Fracture porosity and alteration increase with depth in this hole, and considered together with results obtained elsewhere in both young and old (5 to 110 Ma) crust, suggest that physical properties gradients in the uppermost few hundred meters of oceanic basement are controlled by crustal morphology and alteration and are ubiquitously variable. As this hole was drilled primarily for pilot experiments at the first site of the Ocean Seismic Network, evaluation of the log data demonstrates the importance of careful in‐situ data collection in order to emplace a broadband monitoring device below the seafloor for long‐term seismic measurements.