The Value of Rock Physics in Generating Soft Data

Accurately inferring realistic subsurface structures according to different types of geophysical observations and interpretations remains a significant challenge as the morphology controls the subsurface flow and transport behavior. Hard data and soft data are most commonly used in subsurface characterization. Among them, hard data are observations obtained from subsurface measurements. Soft data are remotely sensed geophysical information and interpretations. The deep learning technique is able to automatically learn the nonlinear relationships and extract implicit features of prior samples. In this work, we propose a deep-learning-based multiple-condition fusion network (MCF-Net) to accurately characterize subsurface structures according to both hard and soft data. Two encoders are used to extract the deep implicit features of the hard and soft data in multiple-level representations. To effectively integrate the multiple-level representations of multiple data, multiple-condition fusion blocks are introduced to extract the representing characteristics. These blocks are also equipped for the reconstruction of heterogeneous structures. MCF-Net can accurately estimate subsurface structures using a variety of available observations. Experimental results prove the utility of MCF-Net in many fields of Earth science.

The development of data fusion systems capable of incorporating soft human-generated data into the fusion process is an emerging trend in the fusion community, motivated mainly by asymmetric warfare situations where the observational opportunities for traditional hard sensors are restricted. This paper describes an extension of our prototype soft/hard data fusion system, based on the random set theory, from centralized into a fully distributed computational framework. A fully distributed data fusion algorithm relies only on information exchange between local sensor nodes and hence promises enhanced scalability, reliability, and robustness in contrast to the conventional centralized fusion approach. We propose a novel approach for distributed estimation of average soft data using the consensus propagation algorithm. The distributed estimation of aggregated hard data is accomplished through an average consensus filter. Based on the proposed approach, we describe a single-target tracking system capable of processing soft and hard data. The preliminary experiments demonstrate the efficiency of the consensus propagation based approach for distributed aggregation of soft data, as well as the advantages of incorporating soft data into the distributed data fusion process.

International maritime crime is becoming increasingly sophisticated, often associated with wider criminal networks. Detecting maritime threats by means of fusing data purely related to physical movement (i.e., those generated by physical sensors, or hard data) is not sufficient. This has led to research and development efforts aimed at combining hard data with other types of data (especially human-generated or soft data). Existing work often assumes that input soft data is available in a structured format, or is focused on extracting certain relevant entities or concepts to accompany or annotate hard data. Much less attention has been given to extracting the rich knowledge about the situations of interest implicitly embedded in the large amount of soft data existing in unstructured formats (such as intelligence reports and news articles). In order to exploit the potentially useful and rich information from such sources, it is necessary to extract not only the relevant entities and concepts, but also their semantic relations, together with the uncertainty associated with the extracted knowledge (i.e., in the form of probabilistic knowledge graphs). This will increase the accuracy of, and confidence in, the extracted knowledge and facilitate subsequent reasoning and learning. To this end, we propose Maritime DeepDive, an initial prototype for the automated construction of probabilistic knowledge graphs from natural language data for the maritime domain. In this paper, we report on the current implementation of Maritime DeepDive, together with preliminary results on extracting probabilistic events from maritime piracy incidents. This pipeline was evaluated on a manually crafted gold standard, yielding promising results.

A geostatistical methodology is proposed for integrating elevation estimates derived from digital elevation models (DEMs) and elevation measurements of higher accuracy, e.g., elevation spot heights. The sparse elevation measurements (hard data) and the abundant DEM-reported elevations (soft data) are employed for modeling the unknown higher accuracy (reference) elevation surface in a way that properly reflects the relative reliability of the two sources of information. Stochastic conditional simulation is performed for generating alternative, equiprobable images (numerical models) of the unknown reference elevation surface using both hard and soft data. These numerical models reproduce the hard elevation data at their measurement locations, and a set of auto and crosscovariance models quantifying spatial correlation between data of the two sources of information at various spatial scales. From this set of alternative representations of the reference elevation, the probability that the unknown reference value is greater than that reported at each node in the DEM is determined. Joint uncertainty associated with spatial features observed in the DEM, e.g. the probability for an entire ridge existing, is also modeled from this set of alternative images. A case study illustrating the proposed conflation procedure is presented for a portion of a USGS one-degree DEM. It is suggested that maps of local probabilities for over or underestimation of the unknown reference elevation values from those reported in the DEM, and joint probability values attached to different spatial features, be provided to DEM users in addition to traditionally reported summary statistics used to quantify DEM accuracy. Such a metadata element would be a valuable tool for subsequent decision-making processes that are based on the DEM elevation surface, or for targeting areas where more accurate elevation measurements are required.

The mapping of saline soils is the first task before any reclamation effort. Reclamation is based on the knowledge of soil salinity in space and how it evolves with time. Soil salinity is traditionally determined by soil sampling and laboratory analysis. Recently, it became possible to complement these hard data with soft secondary data made available using field sensors like electrode probes. In this study, we had two data sets. The first includes measurements of field salinity (EC a ) at 413 locations and 19 time instants. The second, which is a subset of the first (13 to 20 locations), contains, in addition to EC a , salinity determined in the laboratory (EC2.5). Based on a procedure of cross-validation, we compared the prediction performance in the space-time domain of 3 methods: kriging using either only hard data (HK) or hard and mid interval soft data (HMIK), and Bayesian maximum entropy (BME) using probabilistic soft data. We found that BME was less biased, more accurate and giving estimates, which were better correlated with the observed values than the two kriging techniques. In addition, BME allowed one to delineate with better detail saline from non-saline areas.

Treating an artificial catchment as ungauged: Increasing the plausibility of an uncalibrated, process-based SVAT scheme by using additional soft and hard data

<h3>Background</h3> Patient experience is identified by NHS England as a key component in improving care quality. Although several indicators have been developed to measure patient experience and provide institutional benchmarks, comparatively little is known about the range of sources of information about patient experience within healthcare, or about how staff interpret and use experiential data to enhance knowledge and improve care. <h3>Aims</h3> To understand more clearly the breadth of patient experience data which is currently available and how it is being responded to by healthcare professionals to improve patient care. <h3>Methods</h3> We conducted ethnographic observations and interviews with a purposive sample of healthcare staff in Intensive Care Units and Acute Medical Units in three NHS hospital sites, involving over 100 hours of observations and 45 interviews. <h3>Results</h3> We identified a spectrum of types of information about patient experience, including written feedback from surveys and complaints; 'bedside' conversations or patients returning to speak about their experiences; and information received in a sensory way such as a hug, or gifts of chocolates or flowers. Some types of data were codified and used as intelligence within organisations to monitor and improve services, but much of the information about patient experience remained 'below the line': it was recognised by staff and used to shape their practice informally, but never entered the formal economy of data for improvement. <h3>Conclusions</h3> We suggest that, rather than trying to convert this 'soft' information into hard data that can be counted at an organisational level, there is value in recognising the different ways in which different types of information about the patient experience can be harnessed to drive improvement. Soft data can promote grass-roots improvement through reinforcing good practice and facilitating reflection.

Prediction method of physical parameters based on linearized rock physics inversion

Summary A novel method for Total Organic Carbon (TOC) content prediction was proposed utilizing pre-stack elastic impedance inversion. First of all, a new rock physics model was established considering mineral compositions, pore type and fluid content according to the elastic property experiment of shale gas core sample. S wave velocity curve was estimated by the rock physics modelling and the elastic parameters were calculated by P wave velocity, S wave velocity and density. Then, the pre-stack super gather was pre-processed for obtaining high quality angle gather. We derived P wave velocity, S wave velocity and density through pre-stack elastic impedance inversion in framework of Bayes theory under the constraint of prior information such as well logging data, geology knowledge. Finally, under the direction of rock physics analysis, the relationships between TOC content and elastic parameters were established. We find that P wave velocity to S wave velocity ratio had a good relationship with TOC content and we get a high accurate prediction of TOC content. Real data example test showed that TOC content predicted result agreed with geology property and well log interpretation result. The new approach can offer a reliable geophysical evidence for TOC evalutaion.

An evaluation of empirical and rock physics models to estimate shear wave velocity in a potential shale gas reservoir using wireline logs

This is the first in a series of papers demonstrating the application of rock physics as the linchpin in inverting geophysical measurements to predict engineering and geologic rock and fluid properties. This paper establishes the importance of having accurate rock physics models for inversion. It highlights many of the effects that need to be incorporated into an analytical rock physics model, which is our preferred embodiment. The papers following this one (Baechle et al., 2008; Chen et al., 2008a; Chen et al., 2008b; Deffenbaugh et al., 2008; Xu et al., 2008) discuss in more detail aspects of the rock physics model in carbonates, its use in 4‐D applications, and its validation with computational rock physics techniques. A crucial step in successfully inverting geophysical data for rock and fluid properties that directly describe hydrocarbon volumes and producing rates is the development of robust, accurate, and practical rock physics models. Our model is an extension of the Xu‐White (1995) model which incorporates many other known physical and geological effects. We have determined that application of the model achieves better predictions and deeper understanding of the subsurface than empirical models.

Summary Carbonate sediments are prone to rapid and pervasive diagenetic alterations that change the pore structure within carbonate rocks. Rock physics plays an important role in linking elastic properties of rock to petrophysical properties, so a rock physics model which might be capable to model pore structure of carbonate rocks can be very beneficial for characterization of these reservoirs. In this study well logs and seismic data of an oil field located in Iran are used for characterization of a carbonate reservoir. The Arab formation, which consists of dolomite, calcite and anhydrite is studied in order to examine efficiency of an inclusion rock physics model in S-wave velocity estimation and litho-fluids discrimination. The rock physics model is optimized by calibrating the estimated properties with the measured log data, and then the model is used for generating a rock physic template. Then, the litho-fluids distinguished based on this template and the results of simultaneous pre-stack seismic inversion are classified by using Bayes’ rule. The results show that presented rock physics model helps to estimate S-wave velocity more accurately and improves litho-fluids classification of Arab formation in the studied field.

Data integration is very important for the success of quantitative interpretation project. Well logs, core measurements, reservoir fluid properties, existing geological model and pre-stack/post-stack seismic data need a close-up analysis before integrating for quantitative analysis. Any of the poor quality data set may limit the analysis/interpretation. Limitations of seismic data due to noise, frequency content and imaging issues are well known. In this paper we have discussed well log data issues with core measurements and fluid properties. In general, well log data is always considered very accurate. However, the precision required for quantitative seismic interpretation indicates that if possible, well log data should also be corrected for necessary corrections like poor borehole condition, well trajectory, mud-filtrate invasion and tools problem. Therefore, poor quality log data should be identified and edited using appropriate techniques like empirical relations and rock physics modeling. In this study, rock physics modeling has been used for identification, correction and to generate missing logs of any wells in the Sarawak basin, Malaysia. The main focus of log editing/modeling was on the elastic logs i.e. P-sonic, shear sonic and density. Thin laminated sands and shaly sands impose major challenges for rock physics modeling and log editing. Special emphasis has been given for modeling of thin laminated sands and shaly sands in the study area. Different fluid substitution methods and upscaling techniques have been tested and it has been found that an iterative editing and well seismic calibration is a must for correct rock physics model for laminated and shaly sands in Sarawak basin. The rock physics based well log condition workflow will be elaborated further in the paper. A rigorous quality check including iterations for petrophysical interpretation, depth dependent rock physic model parameters are very crucial to explain the recorded well logs and fluid substituted responses through integration with core measurements and seismic data. It has be established that pre-stack seismic gathers and synthetic CDP gather matching should be analyzed for insitu case before embarking on modeling other fluid responses in the reservoir. The rock physics model has been used to show the impact of well log conditioning for AVO modeling and reliable quantitative interpretation. Based on analysis of several wells in Sarawak Basin, it can be mentioned that data QC and conditioning is very important to understand amplitudes and to boost the confidence especially for thinly laminated sand-shale sequences.

Carbonate rocks have a wide variety of pore shapes and different types of grains, which greatly affect the elastic properties and characteristics of the reservoir. This causes certain difficulties in petroelastic modeling. One of the problems is the scale of the input data, which is then used to build the rock physics model. The paper presents the results of studying three core samples of carbonate rocks of the Upper Devonian and Lower Carboniferous age, which are located in the South Tatar arch (Volga-Ural oil and gas basin (Russia)). To evaluate the structural characteristics of the pore space, the effective medium theory is used. The input data are the results of laboratory studies that include measurements of the velocities of longitudinal and transverse waves, porosity, and thin section and computed tomography analysis. When using the computed tomography, the core samples are analyzed at different resolution (12–37 µm/voxel). The tomography studies of pore space at different scales provide rather different values of porosity and pore aspect ratio. The tomography-based porosity estimations also differ from the experimentally measured porosity (up to 10%). The pore space characteristics provided by different datasets are used to build a rock physics model for the studied rocks that helps to estimate the elastic wave velocities with three different methods of effective medium theory (self-consistent approximation, differential effective medium (DEM), and the Kuster–Toksöz method). A comparison of the velocity estimations with their experimental analogs for dry rocks may indicate the presence of microcracks whose size is beyond the tomography resolution. Improved rock physics models incorporating both pores and microcracks are then used to predict the elastic wave velocities of fluid-saturated rock in a wide porosity range. It is demonstrated that the predicted values significantly differ (up to 30%) from those provided by the rock physics (RP) models constructed without the support of the tomography results. Moreover, other types of models are considered in which the difference in experimental and theoretical velocities is attributed to changes in the host matrix properties as compared to the calcite polycrystal, which are caused by various reasons.

This paper proposes data envelopment analysis (DEA) as a suitable data analysis tool to overcome facility management (FM) benchmarking difficulties: FM performance benchmarking analysis is often unsophisticated, relying heavily on simple statistical representation, linking hard cost data with soft customer satisfaction data is often problematic. A case study is presented to show that DEA can provide FM personnel with an objective view on performance improvements. An objective of the case study is to investigate the relative efficiency of nine facilities with the same goals and to determine the most efficient facility. The case is limited to nine buildings in FM on four inputs and nine output criteria. The paper concludes by demonstrating that DEA-generated improvement targets can be applied when formulating FM outsourcing policies, strategies and improvements. Facility manager can apply DEA-generated improvement targets in formulating FM outsourcing policies, specifications development, FM strategy and planning. FM benchmarking with DEA can enhance continuous improvement in service efficiency and cost saving. This will help reduce utility cost as well as pollution. This paper fills the gap in the research of FM benchmarking by applying DEA which studies both soft and hard data simultaneously. It also contributes to a future research of a trade-off sensitivity test between FM cost, services performance and reliability.