Displays

Forests and woodlands offer many benefits to people. They can provide timber and food, store carbon to help deal with the effects of climate change, decrease flooding and soil erosion, and provide recreation for people and habitat for a multitude of species we care to conserve. Scottish forests cover roughly 19% of the country. The Scottish government has the ambition to add several thousand hectares a year over the next decades, to support the rural economy, the environment, and communities. It is important that a substantial proportion of the expansion is made up by native trees and shrub species due to better habitat for wildlife.

These challenges were explored with a case study of virtual forest landscape from Cairngorms National Park (CNP) which was used to test preferences for scenarios of future woodland expansion. Spatial Multi-criteria Analysis (sMCA) has been applied to decide where to plant new forests and woodlands, recognizing a range of land-use objectives while acknowledging concerns about possible conflicts with other uses of the land. The tools used in the development and implementation of the 3D model were PC and Mobile based, and enable the incorporation of interactive functionality for manipulating features. Model inputs comprise 5m DTM, 25cm Aerial Imagery, 3D Tree Species, GIS layers of Current Forest and Woodland Expansion inside CNP. Afforestation animation has been attached in Google My Maps. This is through setting different keyframes by storyboard camera path animation around the area of CNP. Stereo panorama has been applied to selection of woodland expansion scenarios (e.g. Broadleaved potential corridors, Conifer potential corridors), which is viewed with mobile technology and Virtual Reality (VR) equipment.

The 3D model with simulation of woodland expansion was used at the event of 2019 Royal Highland Show and European Forest Institute Annual Conference 2019. Audience feedback suggested the enhancement of user interaction through VR has potential implications for the planning of future woodland to increase the effectiveness of their use and contribution to wider sustainable ecosystems.

How to cite: Wang, C., Gimona, A., Compagnucci, A. B., and Jiang, Y.: Use of Digital and 3D Visualisation Technology in Planning for Woodland Expansion, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-1243, https://doi.org/10.5194/egusphere-egu2020-1243, 2020.

How to cite: Wendelin, E., Bastani, M., Persson, L., Curtis, P., Sopher, D., and Daniels, J.: Three-dimensional geological mapping and modelling at the Geological survey of Sweden, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-5132, https://doi.org/10.5194/egusphere-egu2020-5132, 2020.

With more and more data being gathered from environmental sensor networks, the importance of automated quality-control (QC) routines to provide usable data in near-real time is becoming increasingly apparent. Machine-learning (ML) algorithms exhibit a high potential to this respect as they are able to exploit the spatio-temporal relation of multiple sensors to identify anomalies while allowing for non-linear functional relations in the data. In this study, we evaluate the potential of ML for automated QC on two spatio-temporal datasets at different spatial scales: One is a dataset of atmospheric variables at 53 stations across Northern Germany. The second dataset contains timeseries of soil moisture and temperature at 40 sensors at a small-scale measurement plot.

Furthermore, we investigate strategies to tackle three challenges that are commonly present when applying ML for QC: 1) As sensors might drop out, the ML models have to be designed to be robust against missing values in the input data. We address this by comparing different data imputation methods, coupled with a binary representation of whether a value is missing or not. 2) Quality flags that mark erroneous data points to serve as ground truth for model training might not be available. And 3) There is no guarantee that the system under study is stationary, which might render the outputs of a trained model useless in the future. To address 2) and 3), we frame the problem both as a supervised and unsupervised learning problem. Here, the use of unsupervised ML-models can be beneficial as they do not require ground truth data and can thus be retrained more easily should the system be subject to significant changes. In this presentation, we discuss the performance, advantages and drawbacks of the proposed strategies to tackle the aforementioned challenges. Thus, we provide a starting point for researchers in the largely untouched field of ML application for automated quality control of environmental sensor data.

How to cite: Schmidt, L., Mollenhauer, H., Rebmann, C., Schäfer, D., Claussnitzer, A., Schartner, T., and Bumberger, J.: On the potential and challenges of using machine-learning for automated quality control of environmental sensor data, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-20777, https://doi.org/10.5194/egusphere-egu2020-20777, 2020.

How to cite: Pravia-Sarabia, E., Gómez-Navarro, J. J., Montávez, J. P., and Jiménez-Guerrero, P.: A new tracking algorithm for cyclones with tropical characteristics in the Mediterranean basin, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-9590, https://doi.org/10.5194/egusphere-egu2020-9590, 2020.

Due to global climate changes, an intensification of extreme events such as floods is expected in many regions, affecting an increasing number of people. An assessment of the flood frequencies is then a public concern. For several years now, numerous studies are undertaken on geological paleoclimate records and especially on lake sediments to understand the fluctuations of the flood activities in contrasting climatic contexts and over long time periods. Flood events produce turbidity currents in lake basins that will usually lead to a normal graded detrital layer that differs remarkably from the continuous sedimentation. Currently, in an overwhelming majority of studies, once identified, the layers with the same characteristics (e.g. texture, geochemical composition, grain-size) are usually counted by naked-eye observation. Unfortunately, this method is time-consuming, has a low spatial resolution potential and can lead to accuracy bias and misidentifications. To resolve these shortcomings, high-resolution analytical methods could be proposed, as X-ray computed tomography or hyperspectral imaging. When coupled with algorithms, hyperspectral imaging allows automatic identifications of these events.

Here, we propose a new method of flood layer identification and counting, based on the combination of two high-resolution techniques (hyperspectral imaging and high-resolution XRF core scanning). This approach was applied to one sediment core retrieved from the Lake Le Bourget (French Alps) in 2017. We use two hyperspectral sensors from the visible/near-infrared (VNIR, pixel size: 60 µm) and the short wave infrared (SWIR, pixel size: 200 µm) spectral ranges and several machine learning methods (decision tree and random forest, neural networks, and discriminant analysis) to extract instantaneous events sedimentary signal from continuous sedimentation. The study shows that the VNIR sensor is the optimal one to create robust classification models with an artificial neural network (prediction accuracy of 0.99). This first step allows the estimation of a classification map and then the reconstruction of a chronicle of the frequency and the thicknesses of the instantaneous event layers estimated.  

High-resolution XRF core scanning (XRF-CS) analyses were performed on the same core with a 200 µm step. Titanium (Ti) and Manganese (Mn) were selected as a high-resolution grain size indicator and a redox-sensitive element that shows abrupt inputs of oxygenated water-related to floods, respectively. Both elements have thus been added to the model in order to refine the chronicle derived from hyperspectral sensors. The combination of both hyperspectral and XRF-CS signal indicator allows to decipher floods from instantaneous deposits (e.g slump). This combined chronicle is in good agreement with the expected frequency obtained from the naked-eye chronicle realized on the same core (r² = 0.8). In this study, we present for the first time, an innovative approach based on machine learning which allows to propose fast automatized flood frequencies chronicles. This work was assessed by traditional deposits observations, but it can be easily applied to very micrometric deposits, undistinguishable to the naked eye. Finally, this model can be implemented with other indicators. It then represents a promising tool not only for flood reconstructions but also for other paleoenvironmental issues.

How to cite: Jacq, K., Rapuc, W., Develle, A.-L., Sabatier, P., Fanget, B., Coquin, D., Debret, M., Wilhelm, B., and Arnaud, F.: Combining hyperspectral and XRF analyses to reconstruct high-resolution past flood frequency from lake sediments, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-20333, https://doi.org/10.5194/egusphere-egu2020-20333, 2020.

            Fuzzy modelling of multisource geoscience data and its implications to mineral prospectivity mapping is drawing wide attention of the mineral exploration sector. Mineral deposits are basically end products resulted from optimal combination of certain metallogenetically favourable earth processes which leave their imprints in the associated geological entities at a range of scales that can be interpreted from their direct or indirect manifestation in several geospatial datasets. Therefore, in geologically potential yet under-explored or greenfield areas with no or very few discovered mineral deposits, the qualitative knowledge on the spatial relationship between mineralisation of interest and geoscience data could be an important guide to delineate exploration targets. In such a case, fuzzy set theory aided with Geographic Information System (GIS) is preferred as an effective mechanism for the transformation of subjective knowledge into quantitative information that further helps in modelling of earth science data.

            The Archaean to Paleo-proterozoic Sonakhan Greenstone Belt (SGB) located in the north-eastern fringe of the Baster Craton in central India is considered as a potential geological terrane for mesothermal gold mineralisation based on its geological and geochemical similarities with other mineralised greenstone belts. In this case study, a part of SGB has been taken as a target area that exposes sequence of metamorphosed mafic to ultramafics rocks and associated metasedimentary units in a gneissic country and younger granites. Since the study area represents a less explored terrane in terms of mineralisation, the objective of this research is to generate gold prospectivity maps using fuzzy logic modelling. A total of 17 multiclass evidential maps were generated using four independent geoscience datasets viz. geological, geochemical, geophysical, and remote sensing. The sources of data include existing databases from the Geological Survey of India (GSI) and published work along with the newly produced exploratory data in this research. Fuzzy membership values (0-1) were assigned to each class of evidential maps based on subjective judgement. The fuzzyfied evidential maps were then combined using fuzzy operators (AND, OR, SUM, PRODUCT, and GAMMA) through a series of logical steps i.e. the fuzzy inference network. Two different fuzzy inference networks were created using several combinations of fuzzy operations and accordingly, two prospectivity maps resulted which were classified as very high, high, moderate, low, very low favourable zones. To further enhance the result, the two maps were intersected to produce the final gold prospectivity map in support of targeting gold exploration in the region. A part of the study area, that is the Baghmara gold block, that was already identified as a gold enriched block based on traditional exploration works, coincides with the very high to high favourable zones predicted in the final map and this ensures the reliability of the gold prospectivity map and the efficiency of the adopted fuzzy logic approach in delineating promising targets for exploration.

How to cite: Behera, S. and Panigrahi, M. K.: Application of various fuzzy inference networks to integrate mineral exploration datasets: Implication for gold prospectivity mapping in Sonakhan greenstone belt, India, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-7576, https://doi.org/10.5194/egusphere-egu2020-7576, 2020.

This study reveals the use of high resolution images collected by small Unmanned Aerial Vehicle (UAV) and Digital Image Processing (DIP) from Structure from Motion (SfM) technique applied to the prospection and geometric characterization of fossil tanks in the Guanambi region, located at Bahia state, Brazil. Geologically, the region is located in the Guanambi Batolith, composed of granites, migmatites and orthognaisses. In the research region for example, there was the Lagoa das Abelhas fossiliferous tank, which was previously excavated and in which bone fragments of various pleistocene mammal taxa, such as those of the order Xenarthra, were found, represented by sloths, glyptodonts and armadillos. Considering that there are no records of an effective scientific method to identify these features, the main objective of this work is to map the distribution of fossiliferous tanks excavated as well as those with prospective potential, and to estimate the geometries that they present through the use of the high resolution DIP. The Phantom 4 Advanced equipped with RGB 1’’ CMOS effective 20 M sensor were the UAV model type and camera used for conducting the flight plan. The Pix4D Capture was the tablet/smartphone application used for conducting the flight operation and image collection in an area with ​​80 ha. After this step, the images were submitted as DIP routines using the SfM technique from the Agisoft Metashape software, version 1.5.1. The DIP is divided into stages like point cloud calculation, 3D models generation from mesh and texture procedures, digital elevation model (DEM) and orthomosaic. With the integration of images (DEM and orthomosaic) it was possible to identify and delineate a total of 14 targets through geometric information such as surface area, length, width, depth and internal format. The configuration in relation to soil type, vegetation and rock outcrops was the same around the Lagoa das Abelhas fossil tank. After that, the team came back to fieldwork and found fossil fragments of three out of fourteen targets. Thus, this study could show the potential of using UAV to cover large areas directed to the prospecting part of fossiliferous tanks with good flight autonomy, low cost and fast data analysis. Some of the 11 targets can be prospected because they have a high prospective potential due to their similarity to past prospects which became sites for future paleontological prospection.

How to cite: Magalhães, J., Pereira, G., Leão, A., and Scherer, C.: UAV-based digital image processing applied to the fossiliferous tanks prospection: insights at Guanambi region, northeast Brazil, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-21002, https://doi.org/10.5194/egusphere-egu2020-21002, 2020.

Earthquakes have disastrously impacted communities by destructing the buildings and infrastructure and creating substantial setbacks in the socio-economic development of a region in addition to the huge human loss. They are inevitable and considered extremely difficult to predict. Earthquake prediction research is being carried out for more than 100 years with no well acknowledged model achieved till date. However, the analysis of past seismic stress history of an active fault may help in understanding the stress build up and the local breaking points of the faults. Yet, analysing and interpreting the abundant seismological dataset is most time consuming and is a herculean task.

The possibilities to solve big data, complex problems with Deep Learning are undeniable, however, it’s usage in Seismology is still in its early stage. The implementation of Deep Learning algorithms has the potential to decipher the complex patterns and hidden information in past stress history that is nearly impossible for scientists. The careful implementation of various Deep Learning algorithms in the exponentially growing seismic data can significantly improve the Early Warning System. In the present study, we train a time efficient machine/deep learning algorithm to self-learn and decode the intricate stress accumulation and release pattern, to estimate the probability of local breakdowns of the fault.

The study region for the present research is Indonesia, which under the influence of the Eurasian, Indo-Australian, Philippine and Pacific plates, immensely suffers due to high seismic activity. The principal contributor in the seismicity of the region is Java-Sunda Trench which lies in the Pacific Ring of Fire (PROF). Owing to the high frequency of earthquakes striking every year from different epicentres, the region provides a huge database. The earthquakes triggering in the region from 1970-2018 is downloaded from the International Seismological Centre website (http://www.isc.ac.uk). These earthquake data comprised of ~270,000 events with the information of Latitude, Longitude, Time of the event and focal depth. To respect the bias which is unavoidable due to the change of the quality of the sensors and the data over the decades, the data is divided into subsets. We considered both small and large magnitude earthquakes along the subduction line to generate a localized time series of stress release to understand the seismic history of the region. By using different neural network models such as one dimensional Convolutional Neural Network (CNN), Recurrent Neural Network (RNN), an optimized Deep Learning algorithm is trained to understand the intricate pattern associated with the seismic stress release in region. This specialized model is expected to empower seismologist by providing a time saving, automated process for the identification of the zone of failures.

How to cite: Srivastava, N., Zhou, K., Steinheimer, J., Faber, J., and Stoecker, H.: Pattern recognition of Seismic Activity in Indonesia through Deep Learning, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-1538, https://doi.org/10.5194/egusphere-egu2020-1538, 2020.

The aim of this work is to study the subsurface structures in the west Beni Suef area of the Western Desert in Egypt and to determine their effects on surface geologic structures. A detailed land magnetic survey has been carried out for the total component of the geomagnetic field using two proton magnetometers. The necessary corrections concerning daily variation, the regional gradient and time variations have been applied. Then, the total magnetic intensity anomaly map (TMI) has been constructed and transformed to the reduced to the pole magnetic map (RTP). The reduction-to-pole magnetic and Bouguer anomaly maps were used to obtain regional extensions of this subsurface structure. Regional–residual separation is carried out using the power spectrum. Also, Edge detection techniques are applied to delineate the structure and hidden anomalies. Data analysis was performed using trend analysis, Euler deconvolution, the results indicate that the area is affected by tectonic forces in the N-S, NW-SE, NE-SW and E-W trends, which are correlated with the directions of surface geologic lineaments. In addition, depths to the basement rocks have been estimated using spectral analysis technique. The computed depths have been used to construct the basement relief map which resulted from gravity and magnetic data. They show that the depth to the basement rocks ranges from 2.3 km to 4.7 km.

KEYWORDS
Land magnetic, Gravity, Euler deconvolution, Edge detection and Spectral analysis.

How to cite: Khalil, A., El Emam, A., Abdel Hafeez, T., Saleh, H., and Mohamed, W.: Integrated Geophysical study on the subsurface structural characterization of West Beni Suef area, Western Desert, Egypt, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-148, https://doi.org/10.5194/egusphere-egu2020-148, 2020.

We investigated scaling properties of measurements of personal exposure to solar ultraviolet radiation (pUVR) using the 2nd order detrended fluctuation analysis (DFA2) and the wavelet transform spectral analysis (WTS). Studies of pUVR are important to identify populations at-risk of excess and insufficient exposure given the negative and positive health impacts, respectively, of time spent in the sun. These very high frequency recordings are collected by electronic UVR dosimeters. We analyzed sun exposure patterns of school children in South Africa and construction workers and work site supervisors in New Zealand, and we found scaling behavior in all our data. The observed scaling changed from uncorrelated to long-range correlated with increasing duration of sun exposure. We found peaks in the WTS spectra that mark characteristic times in pUVR behavior, which may be connected to both human outside activity and natural (solar) daily cycles. We further hypothesized that the WT slope would be influenced by the duration of time that a person spends in continuum outside and addressed this hypothesis by using an experimental study approach. To that end we performed combined DFA2-WTS analysis on a subset of individual records taken on the same day under very similar outdoor conditions and used the theoretical superposition rule provided by systematic assessments of effects of trends and nonstationarities on DFA2 as a methodological mean to trace and subsequently model human behavioral patterns in pUVR time series.

How to cite: du Preez, D. J., Blesic, S., Wright, C. Y., Stratimirovic, D., Ajtic, J., Allen, M., and Bencherif, H.: Characterization of human behavior in records of personal solar ultraviolet exposure records, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-303, https://doi.org/10.5194/egusphere-egu2020-303, 2020.

Rhourde Chegga field is located in the north of Hassi Messaoud giant field, Algeria. The main hydrocarbon-bearing reservoir in Rhourde Chegga field is the lower Triassic Argilo-Gréseux reservoir. The Triassic sand is deposited as fluvial channels and overbank sands with a thickness ranging from 15 to 20 m, lying unconformably on the Paleozoic formations. Lateral and vertical distribution of the sand bodies makes their mapping very difficult and, sometimes, even impossible with conventional seismic interpretation. 

To better define drilling targets within the Triassic sand in the Rhourde Chegga field, 3D stratigraphic seismic attribute analysis was performed along the reservoir level, using PSTM and mid angle stack seismic data. By combining various attributes (RMS amplitude, half energy, variance, etc.), the channelized feature has been clearly imaged and delineated on the horizon slices and the volume extraction. The relationship between the combined seismic attributes and reservoir properties at well locations showed a good correlation.

Based on this study, about ten produced wells have been successfully drilled, confirming the efficiency of seismic attribute analysis to predicted channel body geometry.

Keywords: Channel, Attributes, Amplitude, Fluvial reservoir.

How to cite: Medjdouba, N., benaissa, Z., and boudella, A.: Seismic attribute mapping of a fluvial reservoir in Rhourde Chegga field (Hassi Messaoud, Algeria), EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-4260, https://doi.org/10.5194/egusphere-egu2020-4260, 2020.

Geochemical anomalies are an important indicator in prospecting. In particular, geochemical anomalies of Cu play a very important role in geological prospecting of minerals. Geochemical anomalies of Cu are mainly related to mafic-ultramafic rocks and porphyry bodies, which are also associated with ore-forming elements of the Co-Zn-Cr-Ni-Cu combination. The conventional technique of geochemical prospecting involves superimposition of element symbols (Au, Fe, Cu, Al, Ca, etc.) on the geological map of an area by analysing geochemical anomalies using geochemical data. However, this technique is not suitable for regions where geochemical anomaly data are limited. The development of hyperspectral remote sensing has enabled the mapping of spectral features related to characteristic absorption bands of elements in minerals at high spatial resolution, providing a means for precise and detailed reconstructions of geochemical anomalies facies (surface). Compared to conventional techniques for identifying elements, reflectance spectroscopy offers a rapid, inexpensive, and non-destructive tool for determining the mineralogy of rock and soil samples. Hyperspectral remote sensing also provides data for prospecting in areas without sufficient geochemical data, and thus is of vital significance in prospecting for ores in such regions. However, approaches for remotely sensing elements are still lacking, particularly for element content. In this study, a level analysis of Cu content via spectral indices in the northwestern Junggar region, Xinjiang, was conducted. Based on four levels (0–100 ppm, 100–1000 ppm, 1000–10000 ppm, and >10000 ppm) of Cu content and corresponding spectral reflectance, simple and useful spectral indices for estimating Cu content at different levels were explored. The best wavelength domains for a given type of index were determined from four types of spectral indices by screening all combinations using correlation analysis. The coefficient of determination (R2) for Cu was calculated for all indices derived from the spectra of rock samples and was found to range from 0.02–0.75. With sensitive wavelengths and a significant correlation coefficient (R2 = 0.63, P < 0.005), the Normalized Difference (ND)-type index was the most sensitive to Cu content exceeding 10000 ppm. Although the ND-type index has a few limitations, it is a useful, simple, and robust indicator for determining Cu at high concentrations. With the advent of new platforms and satellites in the future, such relationships with other elements are required to enable the widespread use of this index in broad-scale surveys of mineral elements in the field.

How to cite: Shanshan, W. and Kefa, Z.: Study on the Geochemical Anomaly of Copper Element Based on Hyperspectral Indices, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-6433, https://doi.org/10.5194/egusphere-egu2020-6433, 2020.

An accurate estimation of electric power production (EPP) is a crucial first step to design a floating photovoltaics (PV) project. This study estimates the EPP of a floating PV system and validates the results by comparing with the actual EPP observed at the Hapcheon Dam, Korea. Typical meteorological year data and system design parameters were entered into System Advisor Model (SAM) software to estimate the hourly and monthly EPPs. Three-year average observed EPPs (2012, 2013, and 2015) were used as reference values for the validation. The results showed the seasonal EPPs were the highest in spring and the lowest in winter. The monthly estimated EPPs were lower than the monthly observed EPPs. These results are ascribed to the fact the SAM was unable to consider the natural cooling effect of the water environment on the PV module. The error results showed it was possible to estimate the monthly EPPs with an error of less than 15% simply by simulation. However, it may possible to estimate the monthly EPPs with an error of approximately 9% when considering empirical results: The floating PV efficiency was approximately 1.1 times (110%) the overland PV efficiency. This indicates that the approach of using empirical results can provide reliable monthly estimation of EPP in feasibility assessment stage of floating PV projects.

How to cite: Suh, J., Kim, S., and Choi, Y.: Estimation and validation of electric power output from a fixed-type floating photovoltaic system, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-6503, https://doi.org/10.5194/egusphere-egu2020-6503, 2020.

Combining multi-source data can improve the accuracy and the spatial resolution of the three-dimensional (3-D) displacements field. How to effectively integrate multi-source data to obtain high-precision and high spatial resolution 3-D displacements field is worthy of further study. The stochastic model and fusion model of integrating multi-source data affect the accuracy of data fusion. In this paper, based on the least squares method, the effects of different stochastic models and data fusion models on the 3-D displacements field’s accuracy are studied. The optimal method for estimating large-scale 3-D displacements field from integrated InSAR, leveling and GPS measurements is obtained. Then we realize the integrating InSAR, leveling and GPS measurements to obtain the high-precision 3-D displacements velocity field in Tianjin(China) from 2016 to 2018. The results are validated with GPS measurements at 6 independent stations, with the root mean squares (RMS) residuals of the discrepancies being 2.39mm/yr、2.54mm/yr and 2.83mm/yr in eastern, northern and vertical directions, respectively. By comparing different stochastic models, the 3-D displacements field obtained from multi-source data is optimized by the variance component estimation-least squares method, which is better than weighted least squares (WLS) method. By comparing different data fusion models, the accuracy of the horizontal displacements velocity is better than that of interpolated GPS results. The horizontal displacements component has a great influence on the vertical displacements velocity accuracy in the process of acquiring the 3-D displacements velocity by integrating InSAR, GPS and leveling measurements. This study provides a reference method for integrating multi-source data to obtain 3-D displacements field. This method effectively utilizes the advantages of GPS, InSAR and leveling measurements, and extends the limitations of single technical in describing surface-time scale applications. The 3-D displacements information with a large spatial scale and high spatial resolution provide a reliable data basis for studying the crustal movement and its dynamic mechanism.

How to cite: Guo, N.: Influence of Different Data Fusion Methods on the Accuracy of Three-Dimensional Displacements Field , EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-6601, https://doi.org/10.5194/egusphere-egu2020-6601, 2020.

The geological survey of Sweden (SGU) has carried out several detailed airborne TEM (Transient Electromagnetic) surveys in recent years. The data collected in these surveys were inverted to provide models of the resistivity of the subsurface, down to a few hundred meters depth. These resistivity models together with the data from existing boreholes and ground observations offer an excellent basis for further 3D geological modeling. 

The airborne TEM data presented in this study were collected between 2013 and 2016, covering large areas of the islands of Öland and Gotland, in Sweden. Both islands face problems with water supply due to limited groundwater resources. The aim of the surveys was to identify new groundwater resources, specify the depth to saline groundwater and to improve the understanding of the geology of the islands. On Öland, the Paleozoic sedimentary succession reaches thicknesses of approximately 250 m and is composed of Lower Cambrian sandstone, Middle Cambrian siltstone, and claystone followed by the Alum Shales of Upper Cambrian and Lower Ordovician age. Above this lies an up to 40 m thick Lower Ordovician limestone succession, which forms the bedrock at the surface across much of the island. The entire sedimentary sequence rests on Precambrian crystalline rocks. On the Island of Gotland, Silurian bedrock represents the upper part of a 250-800 m thick Paleozoic sequence overlying the crystalline basement. The Silurian bedrock is dominated by interbedded layers of limestone and marlstone, where the interface between limestone and marlstone is often the primary hydraulic conductor.

After acquisition, these data were processed and inverted (1D inversions with lateral constraints), to provide a series of large airborne datasets, providing a resistivity image down to depths of about 250 m in some areas. The considerable resistivity contrast between lithologies, e.g. limestone and marlstone on Gotland, provided an excellent opportunity to resolve boundaries between the different rock types. Borehole information, geological maps, ground geophysical data and the inversion results were incorporated in a 3D geological modelling software. On comparison of the airborne models, ground geophysical data and borehole information it was clear that the airborne resistivity models correlated well with the other available data. Hence, the resistivity models were used as the basis for constructing the 3D hydrogeological and geological models over significant parts of the islands. In this study we present the 3D geological models over the islands of Öland and Gotland which were constructed from the integrated interpretation of all the available data. The models are composed of voxels, each representing a certain lithology classified using a statistical approach. The classification is based on the resistivity range, distance to the neighboring wells/boreholes and the geological observations at the surface. The 3D voxel models will be/have been utilized in hydrological modelling, societal planning, and groundwater management.

How to cite: Bastani, M., Persson, L., Dahlqvist, P., Wendelin, E., and Daniels, J.: 3D geological models from combined interpretation of airborne-TEM and geological data- Two examples from Sweden, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-7204, https://doi.org/10.5194/egusphere-egu2020-7204, 2020.

A 3D geological model was constructed for the Gråbo site to investigate its suitability for artificial groundwater infiltration, to provide drinking water. The modelling work was performed by the Geological Survey of Sweden (SGU) as part of ongoing groundwater investigations. The site is located close to the city of Gothenburg, in western Sweden. A relatively thick succession of coarse-grained glaciofluvial sediment is located at the site, which overlies a typically finer grained and more clay rich sequence. Previously, the site has been the target of several investigations, the most extensive of these was performed in 2006, where a range of geophysical (seismic refraction, ground penetrating radar and resistivity) and borehole measurements were conducted. Based on previous studies the upper course-grained layer has the best potential for infiltration. However, although these investigations improved the understanding of the site, significant uncertainty remained as to the geometry of the upper course grained layer away from borehole locations.

In order to improve the understanding of the site, additional data was collected in 2018 using a tTEM (towed transient electromagnetic) system developed by Aarhus university. The system is comprised of a transmitter and receiver coil, which are towed behind an ATV (all terrain vehicle). Using the tTEM data a 3D resistivity model of the subsurface was generated down to a depth of between 50 and 70 m at the Gråbo site. On comparison with the available borehole data, it was clear that the course-grained layer could be mapped with relatively high accuracy as a region of high resistivity. The tTEM data was combined with the pre-existing geophysical and borehole data to construct both a voxel and layer-based model of the site. These 3D models have subsequently been used as part of ongoing efforts to evaluate the suitability of the site for infiltration (for example, to decide the location of additional investigation boreholes and to provide input to hydrogeological modelling). In this study we present the tTEM data and the 3D geological model. Finally, we exemplify how the 3D model has been used in subsequent investigations and decision making.

How to cite: Sopher, D., Wendelin, E., Lång, L.-O., Öhman, J., and Lindhe, A.: 3D geological model of the Gråbo site from ground TEM measurements, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-7516, https://doi.org/10.5194/egusphere-egu2020-7516, 2020.

We introduce a new approach for 3D joint inversion of potential fields and its derivatives under the condition of constraining data and information. The interactive 3D gravity and magnetic application IGMAS (Interactive Gravity and Magnetic Application System) has been around for more than 30 years, initially developed on a mainframe and then transferred to the first DOS PCs, before it was adapted to Linux in the ’90s and finally implemented as a cross-platform Java application with GUI called IGMAS+. The software has proven to be very fast, accurate and easy to use once a model has been established. Since 2019 IGMAS+ has been maintained and developed in the Helmholtz Centre Potsdam – GFZ German Research Centre by the staff of Section 4.5 – Basin Modelling and ID2 – eScience Centre.

The analytical solution of the volume integral for the gravity and magnetic effect of a homogeneous body is based on the reduction of the three-folded integral to an integral over the bounding polyhedrons (in IGMAS polyhedrons are built by triangles). Later the algorithm has been extended to cover all elements of the gravity tensor as well. Optimized storage enables very fast inversion of densities and changes to the model geometry and this flexibility makes geometry changes easy. The geometry is updated and the gravity is recalculated immediately after each change. Because of the triangular model structure, IGMAS can handle complex structures (multi Z surfaces) like the overhangs of salt domes very well. Geophysical investigations may cover huge areas of several thousand square kilometers but also models of Applied Geophysics at a meter scale. Due to the curvature of the Earth, the use of spherical geometries and calculations is necessary.

The model technique is user-friendly because it is highly interactive, operates ideally in real-time whilst conserving topology and can be used for both flat (regional) and spherical models (global) in 3D. Modeling is constrained by seismic and structural input from independent data sources and is essential toward true integration of 3D thermal modeling or even Full Waveform Inversion. We are close to the demand for treating all geophysical methods in a single model of the subsurface and aim of fulfilling most of the constraints: measurements and geological plausibility.

We demonstrate the flexibility of the software by modeling: (1) the southern segment of the Central Andes which is designed to assess the relationship between the characteristics of the overriding plate and the deformation and dynamics of the subduction system; (2) the South Caribbean margin which defines the two flat-slab subductions of the Nazca Plate and the Caribbean Plate, with variable mantle density distribution implemented by voxels; (3) the North Patagonian Massif Plateau in Argentina which provides insight into the main height differences between the plateau and the surroundings; and (4) an Alpine model which interrogates the strength of the lithosphere at different locations through the Alps and their forelands.

How to cite: Schmidt, S., Anikiev, D., Götze, H.-J., Gomez Garcia, À., Gomez Dacal, M. L., Meeßen, C., Plonka, C., Rodriguez Piceda, C., Spooner, C., and Scheck-Wenderoth, M.: IGMAS+ – a tool for interdisciplinary 3D potential field modelling of complex geological structures., EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-8383, https://doi.org/10.5194/egusphere-egu2020-8383, 2020.

In this study, we proposed a method to utilize a multi-sensor Unmanned Aerial System (UAS) for exploration of hydrothermal alteration zones. This study selected an area (10m × 20m) composed mainly of the andesite and located on the coast, with wide outcrops and well-developed structural and mineralization elements. Multi-sensor (visible, multispectral, thermal, magnetic) data were acquired in the study area using UAS, and were studied using machine learning techniques. For utilizing the machine learning techniques, we applied the stratified random method to sample 1000 training data in the hydrothermal zone and 1000 training data in the non-hydrothermal zone identified through the field survey. The 2000 training data sets created for supervised learning were first classified into 1500 for training and 500 for testing. Then, 1500 for training were classified into 1200 for training and 300 for validation. The training and validation data for machine learning were generated in five sets to enable cross-validation. Five types of machine learning techniques were applied to the training data sets: k-Nearest Neighbors (k-NN), Decision Tree (DT), Random Forest (RF), Support Vector Machine (SVM), and Deep Neural Network (DNN). As a result of integrated analysis of multi-sensor data using five types of machine learning techniques, RF and SVM techniques showed high classification accuracy of about 90%. Moreover, performing integrated analysis using multi-sensor data showed relatively higher classification accuracy in all five machine learning techniques than analyzing magnetic sensing data or single optical sensing data only.

How to cite: Choi, Y., Baek, J., Suh, J., and Kim, S.-M.: Application of multi-sensor unmanned aerial system for identification of hydrothermal alteration zones, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-12546, https://doi.org/10.5194/egusphere-egu2020-12546, 2020.

Nowadays, powerful hand-held devices, like smartphones, tablets and smartwatches, are ordinary things, which many people take anywhere they go. One of the major advantages of this technology is the ability to locate its user by means of GNSS or cellular positioning. Paired with popular, free mobile mapping applications, it greatly simplifies the problem of finding oneself in the unknown place, calculating the best route to one’s destination by various means of transport or tracking one’s movement. For this reason, outdoor navigation is a well-established and widespread technology. The problem arises, when positioning and wayfinding are needed in a GNSS-denied environment, e.g. a building or a mine. In a complex, large or multi-floor constructions modern techniques for easing the navigation through them are rarely applied. Recent years brought numerous new, promising approaches and algorithms for solving a problem of indoor positioning and navigation, but many of them can’t be easily implemented on a typical smartphone or conveniently used. This includes Simultaneous Localization and Mapping (SLAM) and algorithms based on Augmented Reality (AR). It seems that the most feasible and cost-efficient methods are those based on Wi-Fi Access Point (AP), low-cost Bluetooth Low Energy (BLE) or Ultra-Wideband (UWB) beacons. This research aims to describe the process of developing such an Indoor Positioning and Navigation System in one of the buildings, located on the campus of the Wroclaw University of Science and Technology, and identify the main challenges that have to be overcome during this process. Feasibility of available GIS software solutions for this application is analyzed. Directions for future research and development are discussed.

How to cite: Trybała, P.: Development of an Indoor Positioning and Navigation System using Wi-Fi network and BLE beacons for the Smart Campus: A case study, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-13858, https://doi.org/10.5194/egusphere-egu2020-13858, 2020.

The city of Gijón is located on the Cantabrian Coast (NW, Spain), and its subsurface is formed mainly by sand linked to an old estuarine mouth barrier (beach and dunes), sand bay and marshes. Under these sediments, there is a layer of clays related to the weathering of a Jurassic rock basement. This research addresses the setting of the estuary sediments in both the submerged area, located north of the city, and under the built-up area.

The seafloor morphology was investigated by means of a bathymetric survey with multi-beam echo sounder. A geophysical survey using high-resolution reflection seismic profiles allowed studying the thickness of the unconsolidated deposits that fill the bay of Gijón. Likewise, the distribution of coastal sediments under the city was reviewed from boreholes collected within a GIS-based geotechnical database.

The bathymetric reconstruction led to the identification of a paleo-valley supposedly excavated by the main river of the city, with N-S orientation that evolves to NNE-SSW towards the north. It shows a sandy bottom with a very low slope, a length of about 4 km and a width that ranges between 400 and 800 m. In this channel, the unconsolidated deposits reach a maximum thickness of around 15 m while at S, in the urban subsurface, the thickness exceeds 20 m locally. With these data, it was also possible to investigate the geometry of the bedrock under the sedimentary filling.

How to cite: Pando, L., Flor-Blanco, G., Rey Díaz de Rada, J., and García-Rodríguez, A.: Morphology and sedimentary filling of an ancient estuarine valley in an urban environment (Gijón, NW Spain), EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-14447, https://doi.org/10.5194/egusphere-egu2020-14447, 2020.

Austrian loess and loess loam deposits represent an important source of raw materials for the heavy clay industry for centuries. Building material quality of loess and loess loam deposits and their suitability for different applications is significantly influenced by their heterogeneous properties. These depend on the geology of the source area, climatic conditions, geomorphological location, stratigraphic position, intensity of weathering and redeposition potential. The description of occurrences, properties and availability of these raw materials is therefore an important prerequisite to meet the industrial quality requirements. A large number of different sub-datasets exist at the Geological Survey of Austria, which comprise grain-size analysis, bulk rock composition, clay mineralogy, and geochemistry data of loess and loess loam. Within our project, these individual data sets underwent a thorough examination and have been merged into a coherent database to enable the joint regional and statistical analysis of the data. By applying a log-ratio approach the compositional nature of the analysis data has been taken into account for multivariate statistical methods. 
Within our study we focused on the classic Austrian loess regions in the Northern Alpine foreland areas of Upper and Lower Austria and in the Vienna Basin. By transferring the results of the statistical analysis to a Geographic Information System (GIS) these served as the fundamental basis for our categorization of the loess and loess loam occurrences. Taking into account previously published approaches based on soil profile classifications as well as trends and patterns derived from the analysis data, we finally were able to delineate different districts of brick raw materials deposits. These will be made publically accessible to the industry and interested parties as part of the web application of the Austrian Interactive Raw Material Information System IRIS-Online.

How to cite: Reitner, H., Benold, C., Filzmoser, P., Heinrich, M., Hobiger, G., Mert, C., Rabeder, J., Reitner, J. M., and Wimmer-Frey, I.: Compositional data analysis of sedimentological, mineralogical and geochemical data for the evaluation of Austrian loess and loess loam deposits, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-20444, https://doi.org/10.5194/egusphere-egu2020-20444, 2020.

With the development of remote sensing technology, the copyright protection of remote sensing images has become an urgent problem to be solved. In this paper, a blind watermarking scheme based on invariant features is applied. In the embedding process, the stable image features are firstly extracted from the original host using block DCT, and the embedding positions are constructed adaptively according to feature processing theory. Then, the watermark is embedded into the low-frequency coefficients by modifying the DC coefficients. For watermark extraction, according to the invariant image features in each region, the watermark location and the watermark information can be extracted without the original host. Experimental results show that the proposed watermarking is not only invisible and robust against common image processing, such as noise addition, image filtering, and JPEG compression, but also robust against cropping attack.

How to cite: Ren, N. and Zhu, C.: Watermark-based Copyright Protection Using Invariant Features for Remote Sensing Images, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-21392, https://doi.org/10.5194/egusphere-egu2020-21392, 2020.

Evaluating digital soil mapping approaches to predict topsoil exchangeable calcium and magnesium in a sugarcane field of Australia

Maryem Arshad¹, Dongxue Zhao¹, Tibet Khongnawang¹ and John Triantafilis^1*

¹School of Biological, Earth and Environmental Sciences, Faculty of Science, UNSW Sydney, Kensington, NSW, 2052, Australia

Corresponding

John Triantafilis, School of Biological, Earth and Environmental Sciences, Faculty of Science, UNSW Sydney, Kensington, NSW, 2052, Australia

Email: j.triantafilis@unsw.edu.au

Abstract

Knowledge about spatial distribution of exchangeable (exch.) calcium (Ca) and magnesium (Mg) is needed to maintain sugarcane biomass in north Queensland, Australia. To create digital soil maps (DSM), herein, we evaluated three approaches, including; geostatistical (i.e. ordinary kriging [OK]), statistical and hybrid. We first determined the number of samples (10 – 120) required to compute variogram by calculating nugget to sill ratio (NSR) and sum of squared error (SSE). We then used this variogram with OK to predict topsoil (0 – 0.3 m) exch. Ca and Mg. For comparison, four statistical models, including; one linear regression (LR) and three machine learning (ML) models (i.e. Cubist, support vector machine [SVM] and random forest [RF]) were used. Doing so, usefulness of two digital data, including; gamma-ray (g-ray) and soil apparent electrical conductivity (EC_a), either individual or combined, was tested. Regression residuals (RR) were then added to find out improvement in prediction performance (i.e. Lin’s) and in hybrid approach. Influence of varying sample size (10 – 120) was also determined on all three DSM approaches. Comparisons were then drawn with a traditional soil type map and by calculating the mean square prediction error (MSPE). Finally, Digital soil maps (DSM) of exch. Ca and Mg were developed. Results showed that 50 samples were enough to compute a good variogram for exch. Ca (NSR = 11%, SSE = 0.39) and Mg (NSR = 33%, SSE = 0.005). Considering OK, exch. Ca and Mg were predicted with moderate agreement (Lin’s = 0.65 – 0.80). Comparing statistical models and to predict exch. Ca, RF (0.64) and SVM (0.63) outperformed Cubist and LR (0.60) while to predict exch. Mg, SVM (0.79), RF and Cubist (0.74) outperformed LR (0.62). Combined and individual g-ray data performed best and equally well. Hybrid models i.e. RK and CubistRR improved prediction of exch. Ca (0.76) and Mg (0.81) using individual g-ray and EC_a data, respectively. Considering sample size, OK and statistical models required 80 samples while hybrid models required only 30 samples to satisfactorily (Lin’s ≥ 0.70) predict exch. Ca and Mg. Comparisons based on MSPE showed that to predict exch. Ca, hybrid (RK = 1.44) was the best approach followed by geostatistical (OK = 1.94), statistical (Cubist = 2.15) and then traditional soil map (2.64). Same was the case for exch. Mg. DSM of predicted exch. Ca and Mg were consistent with contour plots of measured data. However, some poor predictions were apparent across field edges or areas where small scale variation in digital or soil data was prevalent.  

How to cite: Arshad, M.: Evaluating digital soil mapping approaches to predict topsoil exchangeable calcium and magnesium in a sugarcane field of Australia, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-292, https://doi.org/10.5194/egusphere-egu2020-292, 2020.