Topics&Sessions

For the last 200 years, geologists have developed the science and methodology to construct maps. A geological map reflects both observations and conceptual knowledge about an area of study. It not only tells a geological story but also conveys critical information for natural resources exploration and production. The advent of computers at the end of the 20th century naturally led to the development of 3D geological mapping methods (also known as Geomodeling), which are able to describe 3D geometry of geological objects without any ambiguity. This session aims at exchanging new developments, theories and concepts about 3D geomodeling. Case studies are acceptable but should include methodological novelty. Topics of interest include: - 3D structural model building from dense or sparse data - Efficient updating of geomodels - Knowledge-driven and data driven geomodeling - Event-based geomodeling - Assessing and reducing 3D structural uncertainty - Validation strategies of 3D geomodels - 3D restoration and palinspastic reconstructions - Querying and interacting with geomodels - Ontology and interoperability issues - From geomodels to grids: discretization issues

Eighty percent of political decisions are based on spatial data, which requires precisely described and harmonized data, which base on a conceptual data model. However, this aspect is still neglected in favour of technical solutions and the development of logical data models as direct and often isolated solutions. Trans-organisational as well as trans-national/intra-national date exchange is prevented. In many cases, the description and harmonisation of data subsequently remains at a very basic level, which causes lots of effort with only small benefit. Therefore, the development and public implementation of widely accepted (minimum) data models that can be used for manual and automated data exchange between organisations, administrations, countries, software packages and even other data models will be key for supporting the future significance of Geodata. Consequently, the communication of Geomodels already starts with the acquisition and influences the entire production workflow of digital of Geodata. The need for integrated data models, which are trans-dimensionally valid and cover the entire workflow, is challenging, but will be, nevertheless, the most important success factor for the communication and consumer adaption of digital Geomodels. Thus, the seamless communication of digital Geomodels becomes rapidly more important, but has also to be seen in the area of conflict between the Open Geodata initiative and the political and legal aspects. This session addresses questions concerning (i) the fashion of developing data models designed for the multi-usage in various disciplines, (ii)the state-of-the-art communication and distribution of Geomodels to all kind of users and for machine-readable interaction between independent systems, (iii) the role and responsibility of the suppliers, (iv)technical solutions and tools supporting the communication, (v) the economic benefit of a harmonized data as well as (vi) the role of political and/or legal restrictions. Contributions focussing on the practical application and best practices are welcome.

The session is exclusively dedicated to the presentation of commercial 3D geomodeling and mining software by their vendors. It comprises oral software presentations and demonstrations as well as an exhibition of companies’ presentations with roll-ups, monitors, posters, etc. The session will take place in Haus Formgebung, Bernhard-von-Cotta-Straße 4, Freiberg, on Sep 8 and 9, 2015, only. For detailed information please visit http://www.software-iamg2015.de/. On the occasion of IAMG2015, this session is open to the general public.

The relationships between rock properties including mineralogy, assays and metallurgical indices, and metallurgical process performance in mining have become very relevant challenges for optimizing the revenue in mining projects. In this session, we aim to convene the latest modeling techniques for metallurgical variables, case studies with advanced applications of spatial predictive modeling and developments in incorporating geological understanding with process modeling for geometallurgical challenges. The session will emphasize the use of novel geostatistical and other modeling techniques to generate geometallurgical domains, to characterize the spatial distribution of relevant variables within domains with uncertainty, and to post-process them to construct predictive models of process performance, risk analysis and for optimizing the drilling, mining, blending and processing of ores.

In mining, extraction of natural resources from the earth comprises several stages including prospecting, exploration, development, exploitation and reclamation. Of these stages, exploration, development and exploitation are of critical importance in terms of the economic viability of the overall operation. Considering the required mathematical modelling related to these aforementioned stages, linear and non-linear geostatistical algorithms have been used to construct geological and ore body models with a reasonable accuracy. The mining industry has also frequently made use of secondary data in which the non-invasive geophysical data is incorporated into the sparse exploration borehole data, through the multivariate geostatistical algorithms. Given the uncertainties associated with these stages, stochastic approach, where the conditional simulation algorithms are implemented, is used to generate equi-probable realisations of the ore bodies to account for the uncertainties in grade variation, lateral variability of the ore body units, etc. Considering the exploitation stage, there are various mine planning-related problems that need to be addressed to ensure the economic viability of the project. Some of the mines planning problems include mine design, determination of optimum cut-off grade policy, long- and short-term production scheduling, equipment selection and dispatching. Operations Research has been used to provide optimum solutions to these problems. There are however still various challenges in relation to ore body modelling and mine planning activities depending on the mineral commodity that is mined. The aim of this session is therefore to show several examples where the geostatistical and operation research algorithms are implemented in mining data sets. The session also welcomes contributions focussing on new theory/algorithms and practical applications.

Geochemistry mapping is a data visualization method to determine in a systematic manner the abundance and spatial distribution of chemical elements across the Earth’s surface. GeoMapping, along with attached sections, histograms, tables of chemical and mineralogical analyses, and other geochemical materials, help in the interpretation of the causes of anomalous concentrations of elements compared with their background content in the surrounding areas. Geochemical maps may facilitate the location of promising areas when exploring for deposits of endogenic and exogenic minerals. They have also a substantial role in point out pollution phenomena in the environment with special interest in medicine and agriculture, since illnesses of humans and animals are associated. Importantly, the statistical analysis of chemical elements concentrations concerns the analysis of compositional data (CoDa). Nowadays CoDa analysis is based on the principle of working on coordinates, that is, express concentrations in terms of its log-ratio coordinates and apply the typical methods. The session focus the attention on the use of the log-ratio techniques in the investigation of spatial compositional phenomena. The session focus the attention on the use of the log-ratio techniques in the investigation of spatial geochemical phenomena. Contributions are welcome in any of the open problems of log ratio regional geochemistry: among other, how to produce univariate analyses, precision issues, problems with values below detection limit and irregular data, concepts of background and anomaly, multivariate exploration techniques, reconciliation of multi-analytical results, fractionation studies, or sofwtare advances. Selected contributions will be invited to submit a paper to a special issue of the journal Applied Geochemistry.

Many quantitative models have been used for forecasting oil recovery. They can help in research of hydrocarbon accumulation, reserves estimation, reservoir characterization, extraction data modeling, petrophysical modeling, data inversion and history matching, etc. The goal of this session is to show the utility of mathematics and quantitative models in oil recovery. Submissions with new mathematical models, case studies and application on real data are welcome.

Multiphase fluid flow problems (water-gas(-oil) …) are usually modelled at a macro (i.e. field) scale by systems of strongly nonlinear time-dependent partial differential equations (PDE), often convection-dominated, and thus their numerical simulation provides a formidable task. The same applies to transport problems, one-sided or fully coupled (e.g. density driven flow) to the fluid flow. These transport problems may range from linear diffusion-dispersion-convection-problems for tracers to large to very large systems of again strongly nonlinear time-dependent PDEs, to describe multicomponent reactions, both in the fluid and at the porous skeleton. In the last decades great progress has been made in improving the accuracy and efficiency in the numerical simulation of such processes. This relies on new (high order) locally mass conservative discretization techniques (mixed finite elements, finite volumes, discontinuous Galerkin methods, …) possible combined with dynamic grid adaptation and improvements in the solution of the emerging large sets of nonlinear equations. Put into a parallel framework taking advantage of current computer architecture, these developments allow the approximation of more and more complex problems. Thus the numerical approximation is less and less the bottleneck in tackling advanced problems, but the question of mathematical modelling becomes more and more critical. For a porous medium (i.e. the underground) being intrinsically heterogeneous and having multiple spatial scales, rigorous ways of a transition from micro (i.e. pore) scale, where continuum mechanics based models can be formulated, to a macro scale is indispensable. Recently it was e.g. possible to deal also with problems with a changing pore geometry induced by surface reactions (e.g. clogging or karst processes) leading to macro scale models of a new (micro-macro) type. This session welcomes contributions addressing some of the aspects mentioned, also indicating concrete cases of application (e.g. CO2 sequestration).

Medical geology is an emerging interdisciplinary scientific field studying the relationship between natural geological factors and their effects on human and animal health. It also aims to improve our understanding of the ways in which the geological environment has an impact on the geographical distribution of health problems. This session welcomes talks on modelling the spatial distribution of geology related health issues, such as the relationship between groundwater arsenic or radon and cancer, or health impact of mining activities.

Understanding the functioning of marine environmental systems depend increasingly on the quantitative interpretation of empirical research data and its integration by numerical process models. Of special interest for the society are processes of the coastal ocean where geosphere, hydrosphere, atmosphere, ecosphere and anthroposphere interfere. Models describing the dynamics of transfer and transformation of matter and energy are particularly important for future projections needed for integrated management of the coastal zone. Those models are to be elaborated based on hypotheses deducted from in situ observations during the modern period or/and by paleo-reconstructions based on multi-proxy data extracted from the geological record. Validated models can be used for the generation of future scenarios of marine and coastal environments in front of assumptions about changing climate and socio-economic conditions. The advancement of comprehensive multi-scale numerical model is an extremely complex task and needs cooperation of scientists at this wide field of marine and coastal sciences, from geology and biology through oceanography, climatology to mathematics and computer sciences. Various modeling approaches and their applications such as inverse model, process-based model and geo-statistical model are welcome to be presented and discussed at the session proposed here. A field excursion to the southern Baltic Sea coast in conjunction with the technical sessions will introduce the participants to the morphogenetic processes occurring in a coast facing sea level rise and coastal erosion. The state of the art of coastal engineering and management will be explained using selected sites at the German and the Polish coast.

Sediments are products of a complex web of processes acting at the surface of the Earth, ranging from initial weathering of the source rocks to sedimentary sorting during final deposition. Sedimentary provenance analysis aims at deciphering these processes and their interactions in a deductive approach. This includes reconstruction of source-area geology and drainage patterns and their evolution through geologic time. Similar concepts and techniques are also used in many other studies of provenance allocation, where dirt, dust, minerals, sediments, rocks or any other materials found after some transport and deposition processes must be assigned to their source. These include forensic studies, volcanic tephra allocation and volcanostratigraphic correlation, pollution origin determination, archaeometric studies, etc. Initially being a largely descriptive and qualitative approach, the development of new analytical and tailor-made mathematical-statistical techniques has paved the way towards a comprehensive and quantitative understanding of the provenance of sediment or other materials. This session is dedicated to contributions covering all facets of quantitative provenance analysis (QPA) with a special focus on modern quantitative techniques pushing the limits of traditional approaches. This includes automated bulk petrographic-mineralogical techniques (e.g., Mineral Liberation Analyzer), the integration of multiple data sets (e.g., petrography, geochemistry, granulometry), linear and complex sediment or detritus evolution models, multi-method single-grain analysis (i.e., two or more geochemical and/or radiometric techniques applied to a specific detrital mineral phase), and statistical techniques to decompose complex mixtures of single-grain observations. We invite contributions addressing these and other aspects of QPA, from the conceptual and statistical-numerical point of view to practical field examples from modern and ancient settings.

Many processes lead to structures in rocks and other materials, which vice versa provide information about these processes. However, most structures are complex and, therefore, quantification is needed for precise characterisation and in order to allow comparison with the results of experiments and modelling, and of technical design. Contributions are invited from all fields of geosciences, which deal with all types of structures from brittle to ductile and from micro- to mega-scale. Quantification methods may be but not necessarily have to be related to fractal geometry. We would like to specifically receive contributions dealing with quantification in 3D.

Inverse problems arise everywhere in the geosciences: Geophysicists, e.g., try to reconstruct the subsurface distribution of material parameters such as density, elasticity, magnetization, or electric conductivity, just to name a few, by measuring the appropriate physical fields. These distributions appear as coefficients in a number of differential equations governing the underlying physics, for instance, the heat equation, Maxwell's equations, Navier-Stokes equation etc. Their solutions can be partially observed. The resulting parameter estimation problem, i.e., the reconstruction of the coefficients of a differential equation from measured incomplete solutions, is a typical inverse problem. Similar problems arise in geodesy, meteorology, oceanography, hydrology etc. Inverse problems are still challenging because they are notoriously ill-posed. Intensive areas of research are currently the efficient solution of the associated forward problems, the numerical computation of sensitivities, the selection of an appropriate regularization technique, and the handling of the uncertainty in the data. This session provides a forum, where modelers confronted with inverse problems in the geosciences can interact with applied mathematicians and numerical analysts.

Directional statistics is the branch of statistics that is concerned with handling observations that are directions (unit vectors) in 2 or 3 dimensions, axes/poles (unsigned unit vectors), or rotations. Typical applications include analyses of directions of palaeocurrents or of ocean waves, estimation of Euler poles of tectonic plates, and analyses of focal mechanisms of earthquakes. This session welcomes contributions on all aspects of directional statistics in the Earth and Environmental Sciences.

Machine-learning and knowledge discovery techniques that are designed to handle complex nonlinear relationships, high-dimensional or noisy data have been gaining recognition across the geosciences. This session explores the utility and potential of these novel techniques for geoscience applications including, for example, geophysical data analysis, environmental remote sensing, reservoir modeling and natural hazards. In addition to contributions on recent advances in developing and applying machine-learning and knowledge discovery algorithms to the geosciences, studies that compare these to or establish links with geostatistical approaches are particularly welcomed.

Abstracts and Papers for plenary invited talks i.e. the winners of the Felix Chayes Prize for Excellence in Research in Mathematical Petrology, the Vistelius Award winner, the Matheron Lecturer and the Plenary Keynote Speaker(s) should submit to this virtual session.

Differential equation modelling of geodynamic systems covering scales from regional crustal to global Earth

Differential equation models of sedimentary basin infills and of hydrocarbon reservoir building/behaviour

This session focuses on quantitative models that involve the numerical solution of partial differential equations that describe a process of geoscientific interest (e.g., multiphase flow in hydrocarbon reservoirs and soils, transport of pollutants in groundwater). Given the complex structure of the media in which such processes take place as well as the uncertainty in the initial and boundary conditions, this session will focus on models that involve a stochastic component which accounts for the spatial/temporal variability and various uncertainties. Contributions include data assimilation models that incorporate available data in the solution of the underlying differential equations, such as ensemble Kalman filters and their generalizations. In addition, this session contains also novel approaches that combine the numerical solution of systems of partial differential equations with stochastic models and representations of the system's parameters.

Image analysis and machine learning methods for data from airborne or satellite sensors, or from seismic campaigns.

Quantitative models for groundwater reservoir characterization or simulation

Geomodels (computer volumetric visualization models, like Gocad) for underground natural resources: water, mineral, oil, gas, etc. This session welcomes advanced contributions on 3D geomodeling with a special focus on mineral deposits, resources evaluation, targeting, oil and gas, and geothermal energy. Papers may concern new methodological insights for modeling complex geological structures related to natural resources, but also advanced time models used to reconstitute the geological scenario having lead to deposits. Contributions on multidisciplinary and subsurface data reconciliation integrating several geophysical and geochemical exploration methods together with the 3D modeling are also welcome. Real case studies showing the role of models for targeting, exploration, resources evaluation and understanding deposits are particularly encouraged.

A general session on Geostatistics.

Geostatistical methods in mineral resource characterization, two-point (variogram-based) or multi-point (training image-based)

Methods for delineating on a map the potential presence of mineral deposits

Geostatistical. and space/time statistical evaluations of data sets of geophysical variables and geophysical mdoels.

Quantitative evaluation of hydrocarbon resources, excluding models involving differential equations.

Applications of geoinformatics, databases and e-Geoscience standards

Statistical approaches to mineral resources estimation, locally (scale one deposit) or globally (regional to planetary scale), but without geostatistical methods.

Several methods of analysis of compositional data and densities, not necessarily with a logratio approach, without any strong geostatistical bias.

We present a unique infrastructure for 3D immersion, the Freiberg CAVE.