EnvirVis18

Permanent URI for this collection

https://diglib.eg.org/handle/10.2312/2632396

Browse

Now showing 1 - 11 of 11

Change Point Detection for Ocean Eddy Analysis
(The Eurographics Association, 2018) Banesh, Divya; Wendelberger, Joanne; Petersen, Mark; Ahrens, James; Hamann, Bernd; Karsten Rink and Dirk Zeckzer and Roxana Bujack and Stefan Jänicke
The detection and analysis of mesoscale ocean eddies is a complex task, made more difficult when simulated or observational ocean data are massive. We present the statistical approach of change point detection as a means to help scientists efficiently extract relevant scientific information. We demonstrate the value of change point detection for the characterization of eddy behavior in simulated ocean data. Our results show that change point detection helps with the identification of significant parameter values used in an algorithm or determination of time points that correspond to eddy activity of interest.
Developing a Concept to Visualize Object-based Weather Forecasting Ensembles
(The Eurographics Association, 2018) Feige, Kathrin; Posada, Rafael; Blahak, Ulrich; Karsten Rink and Dirk Zeckzer and Roxana Bujack and Stefan Jänicke
Operational weather forecasters face the challenge of having to process and interpret a large amount of available information. Therefore, condensation of extensive information is required. Research and development of forecasting techniques will on the one hand improve the forecast quality and on the other hand lead to an increased amount of data. A new extensive and valuable data set will emerge from the SINFONY project at Deutscher Wetterdienst (DWD). It aims at a seamless forecast of upcoming convective events from actual time up to some hours by combining observation-based nowcasting techniques and numerical weather prediction (NWP) ensembles into a single system. In this context, a group of products will comprise features ("cell objects") that were extracted from three-dimensional radar measurements and NWP ensemble simulations. A user-oriented intuitive visualization of the new meteorological data is crucial for weather warning and forecasting. Before including new data into forecast operation, extensive tests and evaluations have to be performed. It therefore requires a careful iterative development process with continuous evaluation by the users. To facilitate this process, an initial visualization mock-up is created, which will be used to prototype and refine visualization and data product concepts. The browser-based nature of the tool allows to quickly share an interactive design with the users which, in turn, will help to have in-depth discussions and to collect visualization requirements, before the final concept is implemented into the meteorological workstation. This paper presents the first use-case for this approach: The development of a concept to visualize object-based severe convective events based on matching observed and simulated features.
EnvirVis 2018: Frontmatter
(The Eurographics Association, 2018) Rink, Karsten; Zeckzer, Dirk; Bujack, Roxana; Jänicke, Stefan; Karsten Rink and Dirk Zeckzer and Roxana Bujack and Stefan Jänicke
How To Look at Data: Environmental Practitioners' Lens Through Two Case Studies
(The Eurographics Association, 2018) Ling, Meng; Johnson, Jeffrey A.; Feng, Zhiquan; Chen, Jian; Karsten Rink and Dirk Zeckzer and Roxana Bujack and Stefan Jänicke
Exploratory data visualization, an idea proposed by Tukey in 1977, is meant to output various types of visualization in order to make the data more understandable. While visualization researchers have generated many novel solutions to critical and complex environmental problems, to everyday environmental consultants, some practical considerations have to be made in the visualization analysis to help stakeholders generate and test hypotheses that would not be possible otherwise. We present two environmental case studies of using visualization to communicate key findings: constructing stratigraphic units (layered) and generating groundwater contaminant plumes (volumetric). These real-world cases show that many times visualization alone may not give us correct answers; often what works is the combination of visualization, domain experts' knowledge, and interpretation of the visualization solutions. The lack of any of them may lead to faulty conclusions. The first case study illustrates together how domain experts, visualization, and contamination conditions assisted in interpreting limited and ambiguous lithologic data. The second case study emphasizes conceptual and technical understanding and discusses some common factors affecting 3D interpolation, which again suggests that we must incorporate domain experts' knowledge as well as analytics into visualization for defensible decision making.
Interactive Visual Exploration of Teleconnections in Atmospheric Datasets
(The Eurographics Association, 2018) Antonov, Anatoliy; Lohmann, Gerrit; Ionita, Monica; Dima, Mihai; Linsen, Lars; Karsten Rink and Dirk Zeckzer and Roxana Bujack and Stefan Jänicke
Traditional analyses of geoscientific data and their features require a lot of manual scripting to organize various tools and software libraries. We present a tool developed to cover the typical workflow of the task of analyzing dependencies between regions of the climate system. We propose an interactive visual analysis tool that uses a series of automated analysis steps combined with an interactive visual exploration of patterns in the data. We base our analysis on the statistical approach of strongest negative correlations for discovery of teleconnections. The interactive visual analysis process uses selecting, highlighting, and filtering in four coordinated views. The views represent teleconnectivity information within a teleconnectivity map and a teleconnectivity links list as well as correlation information within a correlation map and a projection view of the correlation space. We apply the tool to different datasets to demonstrate its capabilities for the analysis and comparison of correlation patterns.
Predict Saturated Thickness using TensorBoard Visualization
(The Eurographics Association, 2018) Nguyen, Vinh The; Dang, Tommy; Jin, Fang; Karsten Rink and Dirk Zeckzer and Roxana Bujack and Stefan Jänicke
Water plays a critical role in our living and manufacturing activities. The continuously growing exploitation of water over the aquifer poses a risk for over-extraction and pollution, leading to many negative effects on land irrigation. Therefore, predicting aquifer water level accurately is urgently important, which can help us prepare water demands ahead of time. In this study, we employ the Long-Short Term Memory (LSTM) model to predict the saturated thickness of an aquifer in the Southern High Plains Aquifer System in Texas, and exploit TensorBoard as a guide for model configurations. The Root Mean Squared Error of this study shows that the LSTM model can provide a good prediction capability using multiple data sources, and provides a good visualization tool to help us understand and evaluate the model configuration.
TreeeX: Exploring the Diversity of Tree Species
(The Eurographics Association, 2018) Jänicke, Stefan; Karsten Rink and Dirk Zeckzer and Roxana Bujack and Stefan Jänicke
The GlobalTreeSearch database provides a mapping between 60,065 tree species known to science and the countries where these trees grow. TreeeX is a visual exploration system that supports multifaceted analyses of the GlobalTreeSearch data. Focusing on the entire earth or a group of countries, investigating research questions on biodiversity are visually supported by interactive choropleth maps that color countries according to frequency, diversity, or uniqueness of prevalent tree species. Focusing on a single country, similarities and differences to other countries can be analyzed in detail. Several examples outline the system's capability of delivering insights concerning the geographical diversity of tree species.
Visual Analysis of Urban Traffic Data based on High-Resolution and High-Dimensional Environmental Sensor Data
(The Eurographics Association, 2018) Häußler, Johannes; Stein, Manuel; Seebacher, Daniel; Janetzko, Halldor; Schreck, Tobias; Keim, Daniel; Karsten Rink and Dirk Zeckzer and Roxana Bujack and Stefan Jänicke
Urbanization is an increasing global trend resulting in a strong increase in public and individual transportation needs. Accordingly, a major challenge for traffic and urban planners is the design of sustainable mobility concepts to maintain and increase the long-term health of humans by reducing environmental pollution. Recent developments in sensor technology allow the precise tracking of vehicle sensor information, allowing a closer and more in-depth analysis of traffic data. We propose a visual analytics system for the exploration of environmental factors in these high-resolution and high-dimensional mobility sensor data. Additionally, we introduce an interactive visual logging approach to enable experts to cope with complex interactive analysis processes and the problem of the reproducibility of results. The usefulness of our approach is demonstrated via two expert studies with two domain experts from the field of environment-related projects and urban traffic planning.
Visualizing Electrical Power Systems as Flow Fields
(The Eurographics Association, 2018) Molnar, Samantha; Gruchalla, Kenny; Karsten Rink and Dirk Zeckzer and Roxana Bujack and Stefan Jänicke
We describe a method for visualizing data flows on large networks. We transform data flow on fixed networks into a vector field, which can be directly visualized using scientific flow visualization techniques. We evaluated the method on power flowing through two transmission power networks: a small, regional, IEEE test system (RTS-96) and a large national-scale system (the Eastern Interconnection). For the larger and more complex transmission system, the method illustrates features of the power flow that are not accessible when visualizing the power transmission with traditional network visualization techniques.
Web-based 3D Meteo Visualization: 3D Rendering Farms from a New Perspective
(The Eurographics Association, 2018) Koutek, Michal; Neut, Ian van der; Karsten Rink and Dirk Zeckzer and Roxana Bujack and Stefan Jänicke
We present an approach to highly responsive and interactive web-based 3D meteorological visualization. We describe important technical aspects of the current proof-of-concept implementation that leverages a 3D rendering farm. The work has been motivated by high demands of operational weather forecasters on ease of use of such a tool, quick system responsiveness and low operational end-to-end latency, instantaneous navigation in forecast-time (max. 5 seconds to navigate over 50 timesteps), and other challenging requirements considering the meteo data volumes and complexity of 3D visualization processing. We had limited computer resources but were allowed to spend any time needed in the back-end processing to maximize the responsiveness of the front-end (web-portal). Our goal was not to implement a perfect system, but to design a suitable architecture and implement a responsive system filled with 2D and 3D visualization products of operational NWP (numerical weather prediction) model runs for testing by forecasters and discovering good use-cases for 3D in operations. We succeed in bringing conventional 2D and 3D visualizations together in one web-portal. Customized visualizations layouts can be created to depict a certain atmospheric phenomenon. The webportal provides display of all 3D products (pre-rendered) also as VR (virtual reality) images through the web-browser. The user can instantaneously rotate the camera-view on 3D products and rotate the vertical cross-sections in the web-browser. The system when used on a 3D high-end workstation with VR display provides fully interactive VR data exploration of any given 3D visualization product shown in the 3D preview in the web-portal. Although from today’'s perspective (modern browsers, Web-GL and cloud 3D computing/visualization technologies) our approach of rigorously pre-rendering 3D products might seem unnecessary, the contrary is true. The pre-rendered 3D views off-line in back-end (pre-caching) make sure the user will not suffer a poor web-portal responsiveness on-line.
What if we use the "What if" Approach for Eco-Feedback? Designing an Electricity Consumption Analysis for Layman Users
(The Eurographics Association, 2018) Wambecke, Jérémy; Bonneau, Georges-Pierre; Blanch, Renaud; Vergne, Romain; Karsten Rink and Dirk Zeckzer and Roxana Bujack and Stefan Jänicke
Many households share the objective of reducing electricity consumption for either economic or ecological motivations. Ecofeedback technologies support this objective by providing users with a visualization of their consumption. However as pointed out by several studies, users encounter difficulties in finding concrete actions to reduce their consumption. To overcome this limitation, we introduce in this paper Activelec, a system based on the visualization and interaction with user's behavior rather than raw consumption data. The user's behavior is modeled as the set of actions modifying the state of appliances over time. A key novelty of our solution is its focus on the What if approach applied to eco-feedback. Users can analyze and experiment scenarios by selecting and modifying their usage of electrical appliances over time and visualize the impact on the consumption. We conduct two user studies that evaluate the usability of Activelec and the relevance of the What if approach for electricity consumption. Our results show that users understand the interaction paradigm and can easily find relevant modifications in their usage of appliances. Moreover participants judge these changes of behavior would require little effort to be adopted. Therefore our work opens new perspectives on using the What if approach in the context of eco-feedback technology.

Browse

Browsing EnvirVis18 by Title

Results Per Page

Sort Options