IAMG 2019 Topics
The conference will cover the usual topics in geo-mathematics, geostatistics, and geomodeling but in particular will bring to fore geomodeling issues at the intersection of food, water, and energy. The challenge of meeting the increased demand for food, water, and energy and the resultant stress on our geo-sphere is broadly accepted as one of the major scientific challenges facing mankind. The feedback processes intrinsic to this tri-partite cycle are best-studied using sophisticated mathematical and statistical modeling tools. One of the broad themes of the proposed conference will be to highlight research being performed to address these issues.
List of Sessions
Click on the titles marked to see a descriptions.
- Classical GeostatisticsRicardo OleaThe term classical geostatistics has been recently introduced because of the need of making a difference with multiple-point geostatistics. Classical geostatistics characterizes spatial correlation in terms of two-point moments, such as the covariance or the semivariogram, by using the same data conditioning the modeling. In contrast, multiple-point geostatistics determines the spatial correlation using a second dataset in the form of a training image. The session covers any aspect related to Classical Geostatistics, be it new modeling methods, advancements in programming implementations, or novel applications using existing approaches.
- Classical StatisticsKarel Hron, K. Gerald van den BoogaartThe Geosciences provide various methodological challenges for the statistical analysis of geodata. Some of these led to spezialized topics ourdays filling specialized session e.g. on geostatistics or compositional data analysis. But once upon a time these subjects were new and needed a forum to be presented. We are this forum. Our session should be a home to all statistical contributions showing advanced methods for statistical challenges in the geosciences, which do not fall in one of the specialized categories of the other Sessions. Your data is special but not compositional, or your data is stochastically dependent but not spatially, or your data requires an advanced Bayesian model or an innovative kind of robust analysis, or maybe you just show how a simple classical method allowed you to really understand how that mountain range was formed. Then this Session is the home for your contribution, because we are interested in what is new, in what nobody has thought of before, in what will fascinate the audience, regardless of what exactly it is.
- Compositional Data AnalysisJosep Martin-FernándezCompositional analysis deals with vectors whose components show the relative importance of parts of a whole. This type of data (CoDa) appears in many geological applications commonly expressed as percentages, ppm, or the like. Since the log-ratio approach to CoDa was introduced back in the eighties, steady progress has been made in methodological approaches, visualization techniques, data modelling and applications. Submissions that cover these topics for CoDa in geosciences are welcomed, with particular emphasis on the problems related to food, water and energy, considering both generation and the associated environmental impacts.
- Fractal and Multifractal ModelingFrits AgterbergIn recent years there has been a significant increase in nonlinear modeling studies applied to geophysical, geochemical and other data obtained for geoscientific process modeling, mineral exploration and environmental assessment. Various forms of fractal/multifractal modeling including the application of universal multifractals to geoscientific survey data have been proven to provide useful new types of information. One example is singularity theory that allows the delineation of local anomalies which are superimposed on broader regional map patterns commonly constructed by better known contouring methods. Papers in this session will be concerned with new theoretical developments as well as practical applications of fractal/multifractal theory and singularity analysis.
- GeohydrologyJaime Gómez-HernándezMathematical and/or modeling of flow and transport in the subsurface. Stochastic hydrogeology. Inverse modeling. Incorporating secondary information for aquifer characterization. Big data and machine learning for predictive modeling. Going beyond parameter characterization. Conceptual model uncertainty. Successful case studies of stochastic modeling in geohydrology. New data acquisition techniques. Upscaling of flow and transport properties. Modeling reactive transport. Open-source codes for aquifer modeling. Artificial intelligence for aquifer protection and remediation.
- GeoinformaticsJeffrey YarusGeoinformatics is the science of how to use data, information and knowledge to improve geological characterization of the subsurface and the delivery of associated services. With recent step-changes in awareness of data analytical tools and the big data digital tsunami, significant changes in our understanding of the subsurface, quantitatively describe it, and deliver useful products to academia, the public, and industry is being realized. The ability to consume massive amounts of data, synthesized through fully or assisted automated earth models, scaled from plate to pore, delivering insights ranging from paleoclimate to rock physics at formation, pore, and nano scales, is the cutting edge of geoinformatics today. The session covers the various aspects of geoinformatics and its impact on better understanding of the subsurface. This can include cloud-based computing and architecture, management of geoscience data, (semi) automated workflows and models, and the socialization of these processes and results.
- Geometry and Topology in GeosciencesGuillaume CaumonTopological and geometrical considerations are gaining interest in geoscientific modeling and data analysis research. In this session, we welcome contributions which use geometrical or topological concepts to: - Analyze complex natural objects from observations and measurements (e.g., facies, fractures, karsts, ore deposits, seismic events, etc.). - Analyze, process or visualize results of geoscientific simulations (e.g., geostatistical simulations, flow and transport in porous media, seismic waveforms, ocean or atmosphere dynamics, etc.) - Define model parameters for geological descriptive models (e.g., surface-based or level set methods), forward physical modeling on unstructured meshes.
- Geophysical Data Processing, Interpretation and Machine LearningWeichang LiThis session will cover recent progress in geophysical data processing and interpretation, including but not limited to data denoising, multiple suppression, interferometry, velocity model building, migration/imaging and inversion, as well as fiber optic DAS data processing. In addition, recently there has been a surge of research activity and interest in applying machine learning in geoscience in general and geophysical applications in particular. This session will also show case recent progress and the state-of-the-art in machine learning techniques applied to geophysical data processing and interpretation.
- Coupled Modeling of Food, Water and Energy SystemsMadalyn BlondesAs global energy needs increase, it is important to understand the critical interplay between energy production, water use, and food. For this session, we are soliciting abstracts that take a mathematically and statistically rigorous approach to this theme. Topics may include but are not limited to: 1) waste water “produced water” treatment and beneficial reuse from oil and gas production, 2) competing water use needs between energy production and agriculture, and 3) environmental impacts of energy production and agriculture on drinking water resources. Submissions using compositional data analysis (CoDa), novel geospatial techniques, and big data approaches are particularly encouraged.
- Machine Learning and Optimization MethodsQiuming Cheng, Guoxiong ChenRecent years has found great interests of machine learning (e.g., logistical regression, random forest, support vector machine, self-organizing map, convolution neuronal network) in geoscience applications as diverse as lithofacies classification, extreme event detection, long-term forecasting, mineral resource prediction and so on. As geosciences enter the era of big data, machine learning methods offer immense potential to contribute to many geoscience problems including but not limited to classification, modeling, inversion and forecasting. Meanwhile, geoscience applications introduce novel challenges for machine learning methodologies especially for optimization option. This session aims to provide a forum for the exchange of innovative ideas, methods, softwares and case studies in the field of applying the state-of-art machine learning methods in support for addressing diverse geoscience problems.
- Marine Geosciences: Coasts and GatewaysJan Harff (University of Szczecin, Institute of Marine and Coastal Sciences, Poland), Wenyan Zhang (Helmholtz-Zentrum, Institute of Coastal Research Geesthacht, Germany )Today, such as in the historical past, and even increasingly in the future the societal development needs a deep understanding of the interaction not only between the oceans and the continents in the coastal zones, but also between the marine basins including marginal seas. The gateways between the marine basins form bottle necks for the natural exchange of different water masses, through the millennia they form also “lanes” for migration and maritime trade ways. Coastal zones and marine gateways are shaped by hydrographic processes, eustatic sea-level changes, but also by tectonics and isostatic movements of the Earth’s crust. Oceanographers try to understand mass and energy-flux between continent and oceans as well as between different ocean basins. Archaeologists are interested in the reconstruction of the prehistoric maritime human communities’ migration paths around the globe. Civil engineers need to protect coastal environment and to construct navigational channels for maritime worldwide connections, and economists are obliged to design the present and the future of maritime trade routes. All these multiple tasks need a transdisciplinary co-operation between marine geologists, oceanographers, coastal engineers, nautical scientists and economists in order to reconstruct paleogeographic scenarios and to project future scenarios superposing natural processes with human induced climate change. For realistic paleo- and future scenarios functional models have to be developed and linked with existing archives of measured data and maps that have to be made accessible for the parameterization of models describing the behaviour of the Earth’s crust, oceanographic and climate processes as well as nearcoast morphodynamics. Not only specialists, but all scientists who want to present and discuss their ideas, concepts and results in elaborating and applications of models in marine geosciences in general are welcome to join the session. It is anticipated to foster international geoscientific organizations to pave the roads for data- and model-exchange beyond already existing data communication. The session shall provide a stage to discuss a new field international collaboration, open data and model sharing in the big data world, and transdisciplinary research.
- Medical geologyJennifer McKinleyMedical geology involves the interaction between the natural geological environment and human and animal health. Since rocks are the main source of chemical elements found in soils and stream waters, knowledge and valid interpretation of the spatial distribution of these chemical elements in the environment is key to understanding the influence of environmental factors on the geographical distribution of health issues. Interdisciplinary collaboration is essential to investigate the use of different approaches to address the issues related to medical geology. This session welcomes submissions from geochemists, mathematicians, statisticians, geoscientists and public health researchers to present innovative research in this area.
- Mining modelingJeff BoisvertNumerical modeling in the mining industry has been largely focused on building a single geomodel. Applications and advancements towards better estimation based models is important. In addition to these traditional investigations, we are also soliciting abstracts that quantify mineral deposit uncertainty; such as, decision making based on an increased understanding of geological or grade uncertainty, through various techniques such as simulation and machine learning. There is potential overlap with the machine learning session as machine learning algorithms become more popular for mineral deposit modeling. Interesting case studies and new theoretical developments are welcome.
- Unconventional oil and gas resources modelingMasa Prodanovic (The University of Texas at Austin)Unconventional resources such as mudrocks (esp. shale), tight carbonates and tight sandstones have recently revitalized oil and gas energy production in United States and worldwide. The IAMG 2019 host Penn State University, in particular, is next door to Marcellus shale development. While mudrocks make up more than fifty percent of rocks in geologic record, they are extremely difficult to study due to very fine grain constituents that create nanometer scale pore systems. Interplay of clay, silt and organic material further produces layered and highly heterogeneous and anisotropic nature, low permeability fabric (that is, depending on organic content either reservoir source rock or seal), compositional complexity, and unique geomechanical behaviors ranging from ductile to extremely fissile. In this session we invite presentations on quantitative characterization and modeling of tight unconventional reservoirs on scales ranging from individual pores to reservoir. The topics include, but are not limited to characterization of tight matrix porosity types representative of each facies, the relationship between pore structure, reservoir quality and transport properties, characterization and modeling of natural and hydraulic fractures, deformation and fracture network formation, wellbore stability, reservoir optimization and resource management. In the context of the 2019 conference focus, we especially welcome presentations related to water management in unconventional resources.
- Pattern Recognition Contributions to Inverse Methods in GeosciencesBehnam JafarpourInverse problems in Geosciences often involve reconstructing high-dimensional distributed model parameters (e.g., material property images) from limited nonlinear and indirect measurements. In many cases, these problems are ill-posed and include unknown geologic properties that exhibit complex spatial patterns, which are challenging to represent and preserve during inversion. Further complicating the problem is the uncertainty in the conceptual geologic models that are proposed to constrain the solutions. This session focuses on novel contributions that leverage recent progress in pattern recognition and machine learning to solve challenging inverse problems in Geosciences that involve reconstruction of complex and high-dimensional spatial patterns. Example topics of interest include approaches for handling geologic uncertainty, maintaining geologic plausibility, effective parameterization and regularization methods, dimensionality reduction, as well as computationally efficient implementations, including surrogate models that are derived from predictive analytics and machine learning algorithms.
- Spatiotemporal GeostatisticsDionissios Hristopulos (Technical University of Crete)The development of spatiotemporal models is at the forefront of current research in geostatistics. Flexible and realistic models that can capture the complex behavior of dynamic processes are crucial for improved prediction of spatiotemporal processes in the geo-sphere and for reliable evaluation of the uncertainties involved. Significant modeling challenges stem from issues such as the increasing size, heterogeneity, multivariate dependence, multiple correlation scales and non-Gaussian probability distribution of modern spatiotemporal data. This session seeks contributions that will advance spatiotemporal geostatistics by proposing novel concepts and methodologies, computational algorithms, and innovative applications to address such issues. A non-exhaustive list of topics includes the development of novel space-time covariance functions (e.g., non-separable models, spherical geometry, multivariate dependence), covariance-free approaches (e.g., models based on stochastic partial differential equations and explicit precision operators), innovative simulation methods, applications to interesting or challenging spatiotemporal datasets, as well as approaches for non-Gaussian space-time data and multiscale models. Contributions that combine geostatistics with current developments in applied mathematics (e.g., uncertainty quantification and multifidelity frameworks), as well as in machine learning (e.g., hierarchical models, sparse and multiscale Gaussian process regression) are also welcome.
- Analysis, Simulation, and Optimization of Subsurface SystemsMary WheelerMathematical modeling of complex physical phenomena for predictive understanding is challenging because these problems involve multiple, interacting physical and chemical processes. Multiphysics models need to be developed to simultaneously simulate such processes and their interactions. This session addresses three major challenges: (1) How to mathematically couple individual processes to adequately simulate them and their interactions at various spatial and temporal scales? (2) How to numerically solve highly nonlinear multiphysics problems accurately and efficiently? (3) How to validate multiphysics mathematical models and to verify their corresponding numerical solutions? Validation and verification, model reduction and data assimilation play an important role between mathematical modeling and its practical applications. The goal of this session is to discuss breakthroughs in core technologies needed by mathematical models based on laboratory experiments and field data sets, coupled with both computational methods for modeling these physical phenomena and statistical analyses of controlling factors.
- Fracture Characterization and ModelingRajesh Goteti (Aramco Research Center)Fracture characterization in the subsurface is of critical importance in many disciplines and industries including hydrogeology, civil, mining and tunneling engineering, nuclear waste storage/disposal, and activities related to oil and gas production. The biggest challenge in fracture characterization arises from limited sampling of fractures near well bores or in tunnels and from technologies that support inference of fractures using various geophysical techniques. In order to address this challenge, subsurface geologists and engineers rely heavily on mathematical and statistical models (e.g., discrete fracture network) to populate 3D fracture network geometries away from the observation points. In addition to the 3D geometrical characterization, modeling of fracture networks is essential to assess their impacts on flow, rock mass stability, hydro-fracturing and on subsurface stress estimations. In this session, we invite contributions from a wide range of disciplines (O&G, Civil, Mining and Environmental) that highlight novel advances in mathematical and computational techniques in the characterization and process-based modeling of fractured rock masses.
- Coupling Geomechanics and Flow Systems in Subsurface ApplicationsSanghyun LeeNumerical models for coupled PDEs involving mechanics and flow in porous media can be challenging to solve, and have many applications including groundwater, petroleum, environmental, and biomedical problems. Numerical models involving the coupled PDEs can be challenging to solve due to strong nonlinearity and high heterogeneity as well as different length and time scales between individual physics. This session aims at highlighting the current research status, model capabilities, and numerical methods for solving the coupled geomechanical fluid-rock(soil) interactions. Considered processes may also include flow and deformation of fractured geological formations and geochemical effects, which are relevant in applications such as subsurface CO2 storage, hydraulic fracturing, landslides, subsidence, and sink holes.
- Atmospheric and Earth System ScienceSteven GreybushThis session solicits abstracts related to atmospheric and earth system science, including the areas of meteorology, oceanography, climate science, cryology, and related fields. Submissions that bridge several disciplines, or use computational or methodological tools relevant across the geosciences, are particularly welcome.