141 documents found in 318ms
# 1
Jatnieks, Janis • Sips, Mike • De Lucia, Marco • Dransch, Doris
Abstract: Geochemical models are used to seek answers about composition and evolution of groundwater, spill remediation, viability of geothermal resources and other important geoscientific applications. To understand these processes, it is useful to evaluate geochemical model response to different input parameter combinations. Running the model with varying input parameters creates a large amount of output data. It is a challenge to screen this data from the model to identify the significant relationships between input parameters and output variables. For addressing this problem we developed a Visual Analytics approach in an ongoing collaboration between Geoinformatics and Hydrogeology sections of GFZ German Research Centre for Geosciences. We implement our approach as an interactive data exploration tool called the GCex. GCex is a Visual Analytics approach and prototype that supports interactive exploration of geochemical models. It encodes many-to-many input/output relationships by the simple yet effective approach called Stacked Parameter Relation (SPR). GCex assists in the setup of simulations, model runs, data collection and result exploration, greatly enhancing the user experience in tasks such uncertainty and sensitivity analysis, inverse modeling and risk assessment. While in principle model-agnostic, the prototype currently supports and is tied to the popular geochemical code PHREEQC. Modification to support other models would not be complicated. GCex prototype was originally written by Janis Jatnieks at GFZ-Potsdam. It relies on Rphree (R-PHREEQC geochemical simulation model interface) written by Marco De Lucia at GFZ-Potsdam. A compatible version of Rphee is bundled with this installation.
https://gitext.gfz-potsdam.de/sec15pub/GCex/tags/1.0
# 2
Jatnieks, Janis • De Lucia, Marco • Sips, Mike • Dransch, Doris
Abstract: Surrogate playground is an automated machine learning approach written for rapidly screening a large number of different models to serve as surrogates for a slow running simulator. This code was written for a reactive transport application where a fluid flow model (hydrodynamics) is coupled to a geochemistry simulator (reactions in time and space) to simulate scenarios such as underground storage of CO2 or hydrogen storage for excess energy from wind farms. The challenge for such applications is that the geochemistry simulator is typically slow compared to fluid dynamics and constitutes the main bottleneck for producing highly detailed simulations of such application scenarios. This approach attempts to find machine learning models that can replace the slow running simulator when trained on input-output data from the geochemistry simulator. The code may be of more general interest as this prototype can be used to screen many different machine learning models for any regression problem in general. To illustrate this it also includes a demonstration example using the Boston housing standard data-set.
# 3
Warsitzka, Michael • Závada, Prokop • Pohlenz, Andre • Rosenau, Matthias
Abstract: This dataset provides friction data from ring-shear tests (RST) for a quartz sand used in analogue experiments at the Institute of Geophysics of the Czech Academy of Science (IGCAS) (Kratinová et al., 2006; Zavada et al., 2009; Lehmann et al., 2017; Krýza et al., 2019). It is characterized by means of internal friction coefficients µ and cohesion C. According to our analysis the materials show a Mohr-Coulomb behaviour characterized by a linear failure envelope. Peak friction coefficients µP of the tested material is ~0.75, dynamic friction coeffi-cients µD is ~0.60 and reactivation friction coefficients µR is ~0.64. Cohesions of the material range between 90 and 130 Pa. The material shows a minor rate-weakening of <1% per ten-fold change in shear velocity v.
# 4
Dietze, Elisabeth • Dietze, Michael
Abstract: EMMA – End Member Modelling Analysis of grain-size data is a technique to unmix multimodal grain-size data sets, i.e., to decompose the data into the underlying grain-size distributions (loadings) and their contributions to each sample (scores). The R package EMMAgeo contains a series of functions to perform EMMA based on eigenspace decomposition. The data are rescaled and transformed to receive results in meaningful units, i.e., volume percentage. EMMA can be performed in a deterministic and two robust ways, the latter taking into account incomplete knowledge about model parameters. The model outputs can be interpreted in terms of sediment sources, transport pathways and transport regimes (loadings) as well as their relative importance throughout the sample space (scores).
# 5
Deng, Bin • Schönebeck, Jan • Warsitzka, Michael • Rosenau, Matthias
Abstract: This dataset provides friction data from ring-shear tests (RST) on natural and artificial granular materials used for analogue modelling in the experimental laboratory of the Chengdu University of Technology (CDUT, China). Six samples, four types of quartz sands and two types of glass beads, have been characterized by means of friction coefficients µ and cohesions C. The material samples have been analysed at the Helmholtz Laboratory for Tectonic Modelling (HelTec) at the GFZ German Research Centre for Geosciences in Potsdam in the framework of the EPOS (European Plate Observing System) Transnational Access (TNA) call of the Thematic Core Service (TCS) Multi-scale Laboratories (MSL) in 2017 as a remote service for the CDUT. According to our analysis the materials show a Mohr-Coulomb behaviour characterized by a linear failure envelope. Peak friction coefficients µP of the quartz sand samples range between 0.62 and 0.79 and µP of the glass beads between 0.61 and 0.64. Except for one quartz sand sample, peak cohesions CP of all materials are smaller than or around zero meaning that these materials are cohesionsless. All materials show a minor rate-weakening of 1-2 % per ten-fold change in shear velocity v.
# 6
Willingshofer, Ernst • Sokoutis, Dimitrios • Kleinhans, Maarten • Beekmann, Fred • Schönebeck, Jan-Michael • (et. al.)
Abstract: This dataset provides friction data from ring-shear test (RST) on a plastic (polyester) sand material that has been used in flume experiments (Marra et al., 2014; Kleinhans et al., 2017) and is now used in the Tectonic Laboratory (TecLab) at Utrecht University (NL) as an analogue for brittle layers in the crust or lithosphere. Detailed information about the data, methodology and a list of files and formats is given in the data description and list of files that are included in the zip folder and also available via the DOI landing page. The material has been characterized by means of internal friction coefficient and cohesion as a remote service by GFZ Potsdam for TecLab (Utrecht University). According to our analysis the material behaves as a Mohr-Coulomb material characterized by a linear failure envelope and peak, dynamic and reactivation friction coefficients of 0.76, 0.60, and 0.66, respectively. Cohesions are in the order of few tens of Pa. A minor rate-weakening of 3% per ten-fold rate change is evident.
# 7
Warsitzka, Michael • Ge, Zhiyuan • Schönebeck, Jan-Michael • Pohlenz, Andre • Kukowski, Nina
Abstract: This dataset provides friction data from ring-shear tests (RST) for two types of foam glass beads and a mixture of foam glass beads with quartz sand (“G12”; Rosenau et al., 2019). These materials have been used in analogue experiments in Helmholtz Laboratory for Tectonic Modelling (HelTec) at the GFZ German Research Centre for Geosciences in Potsdam and in the Analogue laboratory of the Institute of Geosciences of the Friedrich Schiller University of Jena (FSU Jena). The materials have been characterized by means of internal friction coefficients µ and cohesion C. According to our analysis the materials show a Mohr-Coulomb behaviour characterized by a linear failure envelope. Peak friction coefficients µP of all tested materials range between 0.70 and 0.75, dynamic friction coefficients µD between 0.52 and 0.55 and reactivation friction coefficients µR between 0.60 and 0.62. Peak cohesions CP of all materials are negative indicating that they are cohesionless. All materials show a minor rate-weakening of ~1% per ten-fold change in shear velocity v. Further information about materical characteristics, measurement procedures, sample preparation, the RST (Ring-shear test) and VST (Velocity stepping test) procedure, as well as the analysed method is proviced in the data description file. The list of files explains the file and folder structure of the data set.
# 8
Purinton, Benjamin • Bookhagen, Bodo
Abstract: Grain-size distributions and their associated percentiles are a key geomorphic metric of gravel-bed rivers. Traditional measurement methods include manual counting or photo sieving, but these are typically achievable only at the patch (1 square meter) scale. With the advent of unmanned aerial vehicle systems and increasingly high-resolution cameras, we can now generate orthoimagery over large areas at resolutions of <1 cm. These scales, along with the complexity of many natural environments in high-mountain rivers, necessitate different approaches for photo sieving. Here, a new open-source algorithm is presented: PebbleCounts. As opposed to other image segmentation methods that use a watershed approach and automatically segment entire images, PebbleCounts relies on k-means clustering in the spatial and spectral (color) domain and rapid manual selection of well-outlined grains. This results in improved estimates for complex river-bed imagery without the need for post-processing.
# 9
Dreiling, Jennifer • Tilmann, Frederik
Abstract: BayHunter is an open source Python tool to perform an McMC transdimensional Bayesian inversion of receiver functions and/ or surface wave dispersion. It is inverting for the velocity-depth structure, the number of layers and noise parameters (noise correlation and amplitude). The forward modeling codes are provided within the package, but are easily replaceable with own codes. It is also possible to add (completely different) data sets. The BayWatch module can be used to live-stream the inversion while it is running: this makes it easy to see how each chain is exploring the parameter space, how the data fits and models change and in which direction the inversion progresses.
# 10
Blanchet, Cécile L.
Abstract: The database presented here contains radiogenic neodymium and strontium isotope ratios measured on both terrestrial and marine sediments. It was compiled to help assessing sediment provenance and transport processes for various time intervals. This can be achieved by either mapping sediment isotopic signature and/or fingerprinting source areas using statistical tools (see supplemental references). The database has been built by incorporating data from the literature and the SedDB database and harmonizing the metadata, especially units and geographical coordinates. The original data were processed in three steps. Firstly, a specific attention has been devoted to provide geographical coordinates to each sample in order to be able to map the data. When available, the original geographical coordinates from the reference (generally DMS coordinates, with different precision standard) were transferred into the decimal degrees system. When coordinates were not provided, an approximate location was derived from available information in the original publication. Secondly, all samples were assigned a set of standardized criteria that help splitting the dataset in specific categories. We defined categories associated with the sample location ("Region", "Sub-region", "Location", which relate to location at continental to city/river scale) or with the sample types (terrestrial samples – “aerosols”, “soil sediments”, “river sediments” - or marine samples –“marine sediment” or “trap sample”). Thirdly, samples were discriminated according to their deposition age, which allowed to compute average values for specific time intervals (see attached table "Age_determination_Sediment_Cores.csv"). The dataset will be updated bi-annually and might be extended to reach a global geographical extent and/or add other type of samples. This dataset contains two csv tables: "Dataset_Nd_Sr_isotopes.csv" and "Age_determination_Sediment_Cores.csv". "Dataset_Nd_Sr_isotopes.csv" contains the assembled dataset of marine and terrestrial Nd and/or Sr concentration and isotopes, together with sorting criteria and geographical locations. "Age_determination_Sediment_Cores.csv" contains all background information concerning the determination of the isotopic signature of specific time intervals (depth interval, number of samples, mean and standard deviation). Column headers are explained in respective metadata comma-separated files. A human readable data description is provided in portable document format, as well. Finally, R code for mapping the data and running statistical analyses is also available for this dataset (see supplemental references).
spinning wheel Loading next page