- Open Access
Vhub: a knowledge management system to facilitate online collaborative volcano modeling and research
© Palma et al.; licensee Springer. 2014
- Received: 15 August 2013
- Accepted: 7 February 2014
- Published: 14 February 2014
Knowledge of volcanic systems and the hazards they produce is rapidly advancing as internet resources become more readily accessible, new and more sensitive field techniques are developed, and ever greater amounts of data are collected. Such rapid advances drive the need for an online collaborative knowledge management system that enables the sharing of volcanological information, and modeling and analysis tools. Vhub (http://vhub.org) is a community cyberinfrastructure platform designed for collaboration in volcanology research, education, outreach, and discovery that complements existing volcano databases and other cyberinfrastructure projects. Vhub is unique in its functionality as a nucleus for the creation of collaborative groups focused on issues such as code development, field research, education, and hazard mitigation. In addition, Vhub serves as a clearinghouse and virtual platform for computational tools relating to volcanic processes and data analysis, as well as documentation to aid in the use and understanding of these tools. By providing a means for scientists to easily disseminate data, models, and ideas, Vhub aims to promote collaboration amongst scientists and to provide resources for science education while advancing the state of understanding of volcanoes and the hazards they produce. This paper introduces the Vhub cyberinfrastructure and provides an overview of select hub features and resources.
- Knowledge management
- Volcano modeling
- Volcanic hazards
Volcanology has been undergoing a fundamental transformation over the past 10–15 years brought about by the internet’s facilitation of access to large volumes of information, greater community acceptance of computational modeling and statistical analysis, and increased attention to volcanic hazards. Despite these trends, the development of scientific research and applications that require numerical computation is hampered by the relative isolation of individual researchers and research groups, and by a gap between research-level development of computational models and their deployment at observatories with real-world hazard mitigation responsibilities. The emergence of innumerable web-accessible resources has many positive aspects, but sthese resources often have low visibility, insufficient documentation, and/or require computational resources or expertise that would-be end users do not possess. Each of these issues represents a barrier to the development and application or adoption of computational tools. There is an urgent need for a better approach to confront these issues that fall under the general topic of collaborative knowledge management. Such an approach should take advantage of the rapid growth of resources, provide new mechanisms for global collaboration in research, and disseminate resulting tools to researchers, observatory practitioners, educators and stakeholders, while reducing barriers associated with local infrastructure, and supporting continued growth of more focused resources that are locally managed at sites around the world.
In the 1990s, the nanoscience research community recognized similar barriers to progress within their own field and so began the development of a cyberinfrastructure “hub” to serve as a central online resource promoting connections amongst researchers and facilitating accessibility to models and information. The result of their efforts, nanoHUB (http://nanohub.org), has had profound effects on the rate of progress of nanoscience, making research tool development and use easily accessible, and enabling the use of advanced research tools for educational purposes. To support these efforts, software developers at Purdue University created the HUBzero® Platform for Scientific Collaboration (McLennan and Kennell 2010), the cyberinfrastructure that provides nanoHUB with all the functionality to develop and run online simulation tools, create private groups for collaboration, manage digital resources including videos, presentations and wiki-style content, communicate through a basic email-like messaging system, and many other capabilities.
Comparison of usage statistics for nanoHUB during 2005 and 2013, and Vhub for 2013
Registered users (1)
Unregistered interactive users (2)
Unregistered download users (3)
Total users (4)
Simulation users (6)
Simulation runs (jobs)
Vhub provides infrastructure for knowledge management, defined as “the creation, archiving and sharing of valued information, expertise and insight within and across communities of people and organizations with similar interests and needs” (Rosenberg 2001). In this context, Vhub development efforts have been consistent with the four general objectives of knowledge management projects (Davenport et al. 1998): 1) create knowledge repositories, 2) improve knowledge access, 3) establish an environment conductive to more effective knowledge creation, transfer and use, and 4) manage knowledge as an asset. Note that we use the word knowledge to mean not just information (data), but also tools to improve understanding and application of that understanding. Vhub is designed to function as a platform for sharing a variety of forms of knowledge, including a unique emphasis on computational tools pertaining to volcanology, and as a platform that facilitates collaboration. Vhub itself is not a database; databases can be shared on the platform, but Vhub aims principally to provide capabilities that complement rather than duplicate existing efforts centered specifically on compiling and managing data about volcanoes, such as the Global Volcanism Program (http://www.volcano.si.edu; Siebert et al. 2010), the Volcano Global Risk Identification & Analysis Project (http://www.bgs.ac.uk/vogripa/; Crosweller et al. 2012). Instead, Vhub is a portal to simulations, lectures, existing databases, educational resources, and other content of interest.
Vhub takes advantage of the HUBzero® open-source software platform. This software incorporates several well-known open-source packages, including an Apache web server, the Joomla content management system, PHP web scripting, and a MySQL database for storing content and usage statistics (Klimeck et al. 2008). The software is specifically designed to support scientific collaboration, research, education, and discovery, the four pillars of Vhub’s organization (Figure 1). HUBzero can handle a variety of data types including documents created with Microsoft Office©, PDFs, ASCII files, images and video clips, analysis tools, and datasets. Additionally, all uploaded content is integrated with collaboration features such as forums, wikis, blogs and Q&A. Integrating all of these packages, the Vhub web site serves as the interface between the user and the platform.
Description of the different resource types available within Vhub
A collection of lectures and/or other types of educational materials that make a course on a specific topic.
Repository of data from volcanic experiments, modeling, field trips, GIS, hazard mapping or other related work.
Individual educational materials aimed at teaching a specific subject such as lectures and homework assignments. For example, it can be a narrated presentation or a written document. They may also include a short video or web link to other type of media. Generally they are part of a set of complementary resources on a specific topic.
A software application that allows users to run a specific type of calculation directly on VHub resources within your browser. They include numerical simulations of volcanic flows, statistical tools and algorithms to process and plot petrological data.
A downloadable software application that users can run on their own computer, subject to platform/software requirements.
A lecture, seminar or other type of presentation in various types of media. Useful for sharing material presented at conferences or workshops.
A document relevant to the community that has been published in some manner. It includes white papers, theses, pre-prints, and magazine articles.
A collection of lectures, seminars, and materials that were presented at a workshop.
Any resource that might not easily fit into any of the previous categories.
Topic pages and articles (2)
Wiki articles for explaining terminology, methodologies, resources or any topic of interest to the community.
Vhub blogs (3)
A public blog that any member can contribute to. It may be used, for example, to report volcanic activity or to describe meetings, workshops or courses. Members can upload their comments.
Personal blogs (4)
Every member can manage a personal blog in which only the member can add entries. They can be accessed by anyone.
A personalizable user panel, or ‘dashboard’, is included to allow users to organize resources and user-specific information. Connecting via this interface, users view and interact with the site without needing to be aware of the complex system of software, data-storage devices, supercomputing hardware and other middleware that host and manage the content being accessed. Vhub members can thus focus only on the resources of interest to them while working towards their individual research, monitoring, or hazard mitigation goals. The dashboard can be accessed through the “My Account” link located at the top of the page. In addition to accessing a customizable dashboard and communication resources, establishing a membership account with Vhub will automatically allot users 1 Gb of disk space on hub servers, enabling them to run online simulation tools via the hub infrastructure.
Vhub provides capabilities for online collaboration in a variety of ways, including (1) within teams conducting basic research, (2) between participants and external advisors on complex hazard assessment projects, (3) between mentors/teachers and students, and (4) amongst community stakeholders. All such collaborations can have personalized privacy levels.
Collaboration can be restricted to a small audience through the creation of ‘groups’. Each group has a dedicated url, a name, and a logo that are set when the group is created. Any registered user can create a group, set up the content of the group, and specify who can join. The user that creates the group automatically becomes a group manager. Managers can change the description and content of group pages, send messages to all or select members, cancel memberships, promote members to managers, and assign member roles.
Group privacy settings are controlled by group manager(s), including the option to determine what content is visible to the general public, to Vhub account holders, and to group members. If designated ‘visible’, groups can be found in searches as well as by browsing (https://vhub.org/groups/browse). If set to ‘hidden’, the group is only discoverable by group members. Unique access restrictions can be set to group content as well.
Four options are available to restrict the access to a group: “Anyone”, “Restricted”, “Invite Only” and “Closed”. This allows options for membership to be requested by users and approved automatically, requested by users and approved by a group manager, initiated by group managers and accepted by users with no opportunity for users to request access, or closed such that no new members are accepted to the group.
Description of the features available for Vhub groups
This section provides a general description of the group. It may also inform about membership restrictions and have announcements to the members. Subpages within this section may be created, for example to have a version of the overview in a different language, or to organize specific information about the group. The overview can only be edited by the managers of the group.
List of all the members of the group. It includes the status (whether the member is a managers of the group or not) and the role of the member if it has been assigned. Managers can send messages to one or all members.
Wiki pages are single pages that any member can create and comment on. It supports wiki formatting and allows the insertion of figures and documents. The content of these pages depends on the scope and organization of the group.
Presentations, data, documents, etc., that are uploaded by members of the group. When the resource is uploaded the author must select the group that the resource belongs to in the Group ownership section of the uploading process. Since all resources on Vhub have a unique url, members can add a direct link to the resource in any of the other group features.
In this section a member can start or follow up on a discussion in any topic relevant to the group. Once a discussion has been started, other members can add comments, questions, answers, or further information on the topic. The managers can create categories to maintain the discussing organized. In other websites this capability is known as Forum.
This section contains all messages sent by the managers of the group to all group members.
This feature allows members to add communications of interest to other members of the group. Other members can add comments to the original posts. Similar to other blog services on the web, the posts are organized by date and the most recent and popular blogs are highlighted.
A group wish is generally a suggestion for improvement that focuses on the content and work within the group. Any member can send a wish, comment and like or dislike existing wishes. A wish can then be accepted, rejected or granted.
To add and discover upcoming events or activities related to collaboration and communication within the group.
Vhub’s Wiki pages and Blogs are additional means of communication that members of the platform can take advantage of (Table 2). While Vhub blogs are public and designed to contain information relevant to the general community, individual members are also provided dedicated blog space for personal communication; privacy settings for individual posts can be set when they are created or edited. Blogs can be used, for example, to report descriptions of volcanic activity or field trips, or to share personal points of view on specific subjects. A Wiki is a web-based hypertext system which supports community-oriented authoring that allows collaborative content creation (Shih et al. 2008). Vhub Wiki Pages and Articles may receive contributions from interested members either through direct editing of the wiki or through suggestions. Topic pages can be used, for example, as tutorial articles describing how laboratory techniques or computational tools work, to discuss the definition and proper use of volcanological terms, or to describe and discuss volcanological concepts (an example focusing on volcano remote sensing capabilities is https://vhub.org/topics/SatelliteMonitoring).
Vhub hosts a Resource Warehouse organized into nine categories based on the type of content submitted: Courses, Data Sets/Collections, Education Materials, Offline Tools, Online Tools, Presentations, Publications, Workshops, and Miscellaneous. See Table 2 for a short description of each resource type. These contents are copyrighted by their authors and contributors, who give Vhub license to display and distribute the material according to the agreement detailed on the following webpage: https://vhub.org/legal/license. All resources on Vhub include a title, author list, abstract, and archival citation information. They are assigned a dedicated url that can be used for citation in journal publications (e.g. Rose 2011). Files uploaded to any individual resource are also assigned an exclusive url, which can be linked to from its abstract or from other resources. While there are no restrictions on the type of files that may be uploaded to the hub, in order to promote accessibility contributors are encouraged to avoid uploading files in proprietary formats that require commercial software for viewing. For example, it is recommended to save presentations and Microsoft Word documents as PDF documents before uploading to vhub.org. As part of the process of creating a resource, the contributor selects ‘tags’ to categorize the content. Tags are simply key words assigned to a resource by the user making the contribution. Tags are searchable and thus aid in resource discoverability.
Each time a new resource is created, the contributor must set the access level of its content. ‘Public’ resources grant full access, including access to the abstract and to any supporting documents uploaded with the resource, to anyone visiting the web page of that resource. When the resource is set to ‘protected’, the title and author list are visible to everyone but the abstract, references, citation information and any associated documents are only accessible by members of the specific group with which the resource is aligned. Like protected resources, ‘private’ resources must be aligned with a group. However in the case of private resources, only group members can discover and access the resource; in this case the resource and all associated content is hidden from all other users. Through these classifications, contributors have full control over whether a given resource is made available to the public or only to specified collaborators.
Since Vhub is a community-driven knowledge environment, we expect the community to contribute to this knowledge system and to learn from it at the same time. This increases the impact of collaboration and opens up new possibilities for users with few resources to gain access to collaboration pathways and shared resources.
The concept of knowledge transfer within Vhub includes development and dissemination of tools that promote the analysis and understanding of volcanic processes. To that end, one important feature of Vhub is its capacity to execute simulation, modeling, and data analysis programs within the user’s web browser while the program being executed runs on remote computing resources, negating the need for the user to install software on their local machine. These ‘Online simulation tools’ include a graphical user interface (GUI) and documentation to facilitate and encourage volcanologists, particularly those who do not regularly use modeling tools, not only to take advantage of the computational resources available to them but to use each tool appropriately. The development and use of computational tools can benefit from the many collaboration features available on Vhub including discussion panels, internal messaging, wish lists, wiki pages, and Q&A sections. These tools help to bring together program developers, experts in the field, tool users, and educators. To facilitate appropriate use of modeling tools, Vhub encourages developers and experts to describe what the models can and cannot do, their advantages over other techniques, and their limitations, as well as to provide advice on how to read and use the model results.
Questions asked from within an online tool are linked to that tool and can be accessed through the “Questions” tab on the resource page (e.g. https://vhub.org/resources/pingu/questions). The developers of the tool get an email every time a question about their tool is posted. Unless the question is closed, any member can answer a question and contribute to solve problems using a tool. Finding or posting questions (and their answers) for specific tools is easy for regular users and developers, and facilitates feedback and collaboration. All questions are listed under “Questions and Answers” in the Explore submenu.
In order to facilitate implementation of numerical simulations and/or processing algorithms by users, Vhub hosts a library of educational and training material prepared by tool developers and expert users. This material may exist in video format, as a topic wiki page, or as a downloadable pdf or other type document. Course and/or workshop materials also provide background information on the theory and practical applications of simulation tools. The developers of the tools, as well as the authors of any other resource material, can track the number of users that have accessed their contribution and obtain feedback from the community.
Additionally, tool users have the ability to ask Vhub community members to view and comment on simulation results. Researchers and students can participate in discussions and establish new collaboration paths or suggest improvements to existing tools. Findings can be shared with the community within blog or topic pages. Educators can use resources to facilitate student learning in a classroom environment (e.g. Courtland et al. 2012). From interactions like these, innumerable collaborative research opportunities emerge, including opportunities for improving both the work of those who contribute to Vhub’s knowledge repository and the users that benefit from that work.
Some of the computational tools currently available to Vhub users include: Energy Cone (rapid delineation of the impact zone by pyroclastic density currents; Sheridan and Macías 1995 Palma 2013a), Tephra2 (tephra dispersion forecast tool; Bonadonna et al. 2005 Bonadonna et al. 2013), Bent (atmospheric plume analysis; Bursik 2001; Bursik 2013), Hazmap (tephra dispersion forecast tool; Macedonio et al. 2005; Costa 2013), Titan2D (mass flow simulation tool; Pitman et al. 2003), PINGU (tool for plotting geochemical and petrological data; Cortés and Palma 2013) and BET_VH (Bayesian Event Tree analysis for Volcanic Hazards; Marzocchi et al. 2010; Tonini et al. 2012). The list of online simulation tools available on Vhub is expected to expand considerably as more volcanologists contribute tools and in so doing realize much more rapid dissemination of their work than through traditional routes (such as publishing a paper about a model and then providing the software upon request). Vhub is also becoming a central library for offline tools, which in some cases can be downloaded directly from the Vhub site (e.g., the Plumeria eruption column model; Mastin 2011) or in others provide links to the author(s)’s website where source code is available for download.
Some online simulation tools create files that are stored in the user’s account within Vhub, and they are kept after the user ends the tool session. These files can be accessed and managed, at any time, through the Workspace tool or through WebDav (see https://vhub.org/kb/tips/webdav). Below and in the Appendix we illustrate usage of the online tools Workspace, Tephra2, and Titan2D, and the offline tool OMIplot.
The workspace and model development
The Workspace is a Linux desktop environment that can be used to upload and download files stored in the user’s account, usually created by other online simulation tools (files can also be accessed through webdav, see https://vhub.org/kb/tips/webdav). The Workspace is also the perfect environment to edit, compile and test a tool in development. Developing online simulation tools within Vhub is achieved through a content-management-system for tool publication which provides a project area with wikis, a timeline that documents changes to each revision, and a subversion source-code repository (McLennan and Kennel 2010). Developers can chose whether they want to make source code private or accessible to other Vhub members, and can specify an open-source or other license. Once a tool is ready, it is committed by the developer and installed by Vhub’s server managers, at which point the developers are asked to test the working version of the tool before it is fully deployed on the hub. More information on how to create a tool for Vhub can be found at https://vhub.org/support/knowledgebase/development.
Modeling tephra fallout with Tephra2
Tephra2 is an advection–diffusion model used to forecast the accumulation of ejected volcanic material on the ground (Bonadonna et al. 2005). It can be used for civil protection purposes, such as giving public warnings and planning mitigation measures (Connor et al. 2001; Volentik et al. 2009), as a research tool for studying eruptions past and present (Johnston et al. 2012; Biass and Bonadonna 2011; Longchamp et al. 2011; Connor and Connor 2006; Volentik et al. 2010), or as an educational tool for volcanology, numerical modeling, or quantitative literacy (Courtland et al. 2012).
The Tephra2 online simulation tool applies equations for particle diffusion, transport, and sedimentation to a set of user-supplied eruption parameters. After ejection from the vent, particles undergo column diffusion before being released into an atmosphere composed of a series of vertically stacked layers within which wind speed and direction are constant but between which they are allowed to vary. While falling, particles experience diffusion due to atmospheric turbulence and advection due to the wind before falling onto the land surface.
Atmospheric data required by the model can either be downloaded from the NCEP-DEO Reanalysis website directly (http://www.esrl.noaa.gov; Kanamitsu et al. 2002), obtained and formatted for use by the Windre online simulation tool (https://vhub.org/resources/windre), or typed into a text file that includes altitude (m a.s.l.), direction (0–360 degrees from north) and magnitude (m s-1) of the local wind field. The easting, northing, and elevation of the ground surface at which to make calculations are listed in a separate ‘grid’ file. The remaining parameters (e.g., maximum height of eruption column, total mass of tephra ejected, pyroclast density, size distribution of particles) can either be defined in a third text file or typed into the appropriate GUI field (Figure 2). Each file is uploaded to the Tephra2 graphical user interface. Execution time scales with the number of points in the grid file. Example atmospheric profiles, grids, and configuration files for three volcanoes are provided with the tool. Users may either alter the input parameters for these example eruptions or provide information for the eruptive scenario of their choice.
Simulating dry granular flows over natural terrain with Titan2D
Titan2D simulates dry granular avalanches over digital representations of natural terrain (Pitman et al. 2003; Patra et al. 2005). The program is based upon a depth-averaged model for an incompressible Coulomb continuum, a “shallow water” granular flow, based on the work of Savage and Hutter (1989), Iverson (1997), Iverson and Denlinger (2001), Denlinger and Iverson (2001) and Mangeney-Castelnau et al. (2002). This assumption is grounded in the fact that the flow thickness is small compared to the entire area over which a long-run-out flow travels and deposits material. A principal feature of the code is the incorporation of topographical data into the simulations and grid structure, combining numerical flow simulations with digital elevation data of natural terrain. A pre-processing routine allows digital elevation data to be imported. These data define a two-dimensional spatial box in which the simulation will occur. The raw data provide elevations at specified locations. Using these data, interpolating between data points where necessary, a rectangular Cartesian mesh is generated. Adaptive gridding allows for the concentration of computing power on a region of interest. Mesh refinement captures complex flow features at its leading edge, as well as locations where the topography changes rapidly. The conservation equations for mass and momentum are solved with a Coulomb-type friction term for the interactions between grains and between the granular material and the basal surface. Finally, the grid data are used, together with simulation output, in post-computation visualization (Yu et al. 2009; Charbonnier and Gertisser 20112012).
Remote sensing of SO2 emissions with OMI data and OMIplot
OMIplot (Carn 2011) is a software package designed to ingest, visualize and analyze SO2 (i.e. sulfur dioxide) and other trace gas data collected with OMI: a hyperspectral UV and visible spectrometer aboard NASA’s Aura spacecraft. Aura is part of the A-Train satellite constellation and operates in a sun-synchronous, near polar orbit at an altitude of ~705 km (Levelt et al. 2006a). Each OMI orbit has a swath width of 2600 km and pixel size of 13 × 24 km at nadir (Levelt et al. 2006b), allowing continuous global measurements of ozone and various other trace gases including SO2, nitrogen dioxide (NO2) and bromine monoxide (BrO) (Carn et al. 2007). In clear sky conditions, OMI can provide daily remote measurements of volcanic SO2 emissions if they exceed the sensor’s detection limit (Carn et al. 2008). Level 2 OMI SO2 (OMSO2) data files in HDF-EOS5 format are publicly available at no cost and downloadable from the NASA Goddard Earth Sciences Data and Information Services Center website (http://disc.sci.gsfc.nasa.gov/Aura/data-holdings/OMI/omso2_v003.shtml).
This paper provides a brief overview of many of the capabilities of Vhub, such as collaboration within groups, the deployment of modeling tools, and examples of online and offline tools. Vhub is a knowledge-management-system aimed at facilitating collaborative efforts to better understand volcanoes and to more closely link observatory practitioners with external resources as they emerge from the research community, and non-observatory researchers with the monitoring data and knowledge generated within volcano observatories. Vhub provides a mechanism for volcanologists to share knowledge and tools, and to rapidly connect users with the resources they need, whether they be students, academics, or volcano observatory personnel.
At its core, Vhub is a community driven effort that is constantly evolving in terms of the resources available within the hub, the ways that those resources are used, and the underlying cyberinfrastructure (software and hardware) itself. For example, significant current effort is focused on re-engineering core modeling tools for mass flows and tephra dispersion in order to make them more flexible for changing concepts and user needs, and in implementation of probabilistic models. Vhub goes beyond the traditional, static web presentation to provide an interactive platform capable of promoting explosive developments in the field of volcanology and volcano risk mitigation.
Details of modeling and analysis tool examples
This Appendix provides details on the examples of Vhub modeling and analysis tools that are summarized in the main paper, namely Tephra2, Titan2D, and OMIplot based upon eruptive activity at Merapi volcano in 2006 and 2010. Merapi is a basaltic-andesite stratovolcano located 28 km to the north of the city of Yogyakarta, Indonesia. It is one of the most active volcanoes in the world, with eruptions typically recurring every 4–6 years during the twentieth century (Surono et al. 2012). During March-June 2006, Merapi exhibited the extrusion and partial collapse of a lava dome, generating pyroclastic flows on the southwest and south flanks of the volcano. The activity peaked on June 14 with two dome-collapse derived block-and-ash flows that travelled 5 and 7 km on the south flank, respectively, destroying the village of Kaliadem located 4.5 km from the summit (Charbonnier and Gertisser 2009).
In late October and early November of 2010, Merapi volcano produced its most explosive eruption in over a century. The event was characterized by a series of tephra fallout events and pyroclastic flows. The climactic event occurred on 4–5 November, producing an eruption column that reached an altitude of 17 km and pyroclastic flows that travelled approximately 16 km from the volcano (Surono et al. 2012).
As an example exercise, we can model tephra fallout using parameters available for the 2010 eruption of Merapi volcano. The Tephra2 model can be used to investigate the likely pattern of tephra fallout from the sustained eruption column produced by Merapi volcano on November 4th. Note that this example is meant to outline the utility of the Tephra2 simulation tool and is not meant to provide a validated tephra fallout map for the eruption, an exercise that would require additional field data and post processing.
The volume of tephra deposits from the activity of Nov. 4th-5th is estimated as 107 m3 (Surono et al. 2012). Assuming a mean density of 1950 kg m-3 (Lube et al. 2011), this corresponds to a total erupted mass of 1.95 × 1010 kg. The maximum plume height of the cataclysmic eruption was 17000 m a.s.l. The grain sizes considered by the model are set to range from −3 to 6 phi with a median of 1.6 and a standard deviation of 1.7 phi. These values are based on analysis of seven samples of the 2010 tephra fall deposit (Charbonnier, unpublished data). Tephra2 will use these values to approximate the grain size distribution of the erupted material assuming a Gaussian distribution. A diffusion coefficient is used to take into account the spreading of the plume, the interaction of the plume with the atmosphere, and the variability of the atmospheric conditions; its value can range anywhere from 100 to greater than 10,000 m2s-1. A value of 2000 m2 s-1 is used in this example. Ideally, this parameter would be determined either from the dispersion of tephra on the ground or comparison with similar eruptions. The eddy diffusivity of the atmosphere is a constant, 0.4, and the number of steps into which to discretize the eruption column is set to 100. The fall time threshold allows the model to distinguish between dense (lithic) and light (pumice) particles, however for simplicity the density of both particle classes is set to the median density of 1950 kg m-3. The fall time threshold can then be set to an arbitrarily high value. The plume model is set to 0 to represent a well-mixed plume and the plume ratio is set to 0.33 so that no particles will be released from the lower third of the column. Volcano coordinates are entered as Easting and Northing (438752 m E, 916425 m N). The input grid is 200,000 m on each side with a spacing of 1000 m between grid points. Although topography in the area is quite variable, the elevation of the landing surface is set to a constant 600 m a.s.l. Using a grid of constant elevation insures that numerical calculations have closed form solutions. The input atmospheric profile (see Figure 3) is determined from Reanalysis data (Kanamitsu et al. 2002) for Merapi volcano for Nov. 4th 2010. Once this information is loaded into the GUI, the user presses the ‘simulate’ button and tephra accumulation, in mass per unit area, is calculated for each grid point. Numerical output (not included here) includes both the mass loading of particles on the ground (kg m-2) and the spatial distribution of the various particle sizes considered by the model. An isomass map is automatically generated, allowing users to quickly visualize results (Figure 3). For the Merapi example, isomass contours indicate that tephra particles settled to the west/northwest of the volcano, in general agreement with eyewitness accounts of plume dispersion. The atmospheric profile and configuration file used in the calculation are also available for download from the graphical user interface. If specified, the output can be saved onto the user’s account within Vhub, which can be later accessed using the Workspace online tool or WebDav. Further details of how to execute the Tephra2 online simulation tool, as well as details of the mathematical calculation, are included in the Tephra2 user manual. Additional material available for download include video tutorials explaining how to use the tool, Perl scripts to aid in the creation of input files and interpretation of output files, and a graduate level Tephra2 geocomputation assignment. Each of these is available from the ‘Supporting Docs’ section of the Tephra2 tool (https://vhub.org/resources/tephra2/supportingdocs).
As an example we simulate the generation and runout distance of a dense volcanic granular flow of ~106 m3 in volume on the southern flank of Merapi volcano (Figure 5). The simulation presented in this example intends to reproduce one of the short-to-medium-runout block-and-ash flows (SM-BAFs) that were emplaced during the June 2006 dome-collapse events at Merapi (Charbonnier and Gertisser 2009, 2012). Due to its average values of runout distance, area covered, and volume of the associated deposits within the whole range of SM-BAF deposits (Charbonnier and Gertisser 2009, 2011), the characteristics of lobe 5 (see Figure one in Charbonnier and Gertisser 2012) were taken as a reference for evaluation of the simulation outputs.
Provide a 3D grid containing topographic information (e.g., a DEM for the simulation area) and entering this information into the ‘GIS’ tab of the GUI. Using the online DEM converter tool for Titan (Palma 2013b), the user can directly convert his/her own DEM from a Golden Software Grid file, an Arc/Info ASCII file, or GeoTIFF file to a Titan2D compatible GRASS format. DEMs with a spatial resolution of 30 m or better are desirable. The effect of DEM interpolation and sampling schemes on Titan2d model outputs is discussed in Stefanescu et al. (2012).
Defining dimensions of an initial “pile” of material (‘Piles’ tab) or a flux source (‘Flux Sources’ tab), which adds mass over a specified time period and area at a specified rate, including shape, footprint, height, volume, position and initial velocity (if required).
A variable that represents the angle of internal friction of the granular pile (‘Material map’ tab), which must be higher than the chosen bed friction angle. Analysis of input sensitivity and experience has shown that flows are relatively insensitive to the internal friction angle (Dalbey and Patra 2008).
A variable that represents the angle of bed friction between the granular pile and the substrate (‘Material map’ tab). This value is somewhat volume dependent (large volume flows require lower basal friction angles) and can be initially estimated from the Heim coefficient (height-dropped over runout, H/L) measured from natural deposits for a particular volcano, or taken from an H/L vs. volume plot. Using H/L vs. volume curves works well for simple channels, and provides a range of reasonable basal friction angles. For more complex terrain, ‘calibrating’ Titan2D by replicating natural deposits through trial and error, or employing material maps of varying friction may be required (Stinton et al. 2004; Charbonnier and Gertisser 2009, 2012). Simulated flow inundation is sensitive to the basal friction angle, and a range of inputs should be used to capture natural variability and uncertainty. A GIS-based classified material map (‘true’ option in the ‘Material map’ tab) which matches the area covered by the DEM can be used to define zones in the region where pronounced changes in the topographic surface (e.g., different substrates, slope breaks, and channel confinements and morphologies) result in a change in the bed friction angle.
Stopping criteria to halt the simulation (normally a limit on “maximum real time” or the maximum number of computational time steps) and other mass flow scaling parameters is entered in the ‘General’ tab of the GUI.
Once the GUI tabs have been filled for the DEM and mass flow parameters, the ‘Job Submission’ and ‘Job Monitor’ tabs allow the user to execute and monitor the Titan2d simulation within the Vhub environment. All Titan2D input and output data are saved and stored in the Vhub user’s personal ‘save’ directory specified in the ‘Load/Save’ tab of the GUI. Some of Titan2D’s outputs are flow depth and momentum over the whole computational domain (or DEM) at specified times, which can be used to compute field observable variables at different times during the flow including run-up height, inundation area, and velocity. Using the Titan2D viewer on Vhub (Palma 2012), the user can visualize the Titan2D standard outputs and/or output summary, plot an output parameter against time and/or display a map of the initial pile height, maximum height recorded during the simulation and/or a simulation variable at a particular computational time step. A ‘download’ option is available to download the current results to a new web browser window on the user’s desktop and create animations and/or movies.
For the Merapi example, the horizontal resolution of the DEM used is 15 m. The initial pile had a dome shape with a maximum depth of 10 m, with the base having major and minor axes of 100 and 70 m, respectively. A flux source of 50 s duration with a mean extrusion rate of 2.5 kg m-2 s-1 was added to the initial pile. The internal and bed friction angles were 30 and 25, respectively. Starting at rest, the granular material collapses and accelerates downslope until it reaches its maximum speed after 45 seconds (~45 m s-1).
According to this simulation, after descending the steepest slopes of the southeastern flank of the volcano, the Titan2D simulated flow hits the western side of Kendil hill after ~50 s, bends ~45° to enter the open slopes of the proximal Gendol valley and stops after ~120 s and 3.4 km run out distance (Figure 5).
Here we demonstrate the use of OMIplot to analyze OMI data collected during the November 2010 eruption of Merapi volcano (Java, Indonesia; Surono et al. 2012). Although OMIplot has been demonstrated to be a reliable tool for quantitative estimates of volcanic SO2 emissions (e.g. Thomas et al. 2011), in this example the emphasis is on the capabilities of OMIplot rather than the quantitative results.
Once launched in IDL, OMIplot offers several options, including ingest of a single OMI orbit, an entire day of OMI data (14–15 orbits) or automated processing of multiple days (dependent on data availability on the user’s system). Moreover, OMIplot can also process near-real time (NRT) OMSO2 data available within ~3 hours of the Aura overpass from NASA’s Land Atmosphere Near Real-Time Capability for EOS (LANCE) system (http://earthdata.nasa.gov/data/nrt-data/data-products/omi). In this example a single OMSO2 orbit from 8 November 2010 has been loaded.
The user enters the latitude and longitude ranges to plot, in this example a latitude −12:0 and longitude 104:115 are used. At this stage, the SO2 data may be filtered, either spatially or based on parameters such as the OMI-derived cloud fraction, reflectivity, and/or cloud pressure. The user must then select the desired retrieval algorithm, the Band Residual Difference (BRD) or Linear Fit (LF), for plotting. Other options related to the geometry of the plot are also available, such as plotting gridded (interpolated) data or actual OMI footprints, changing the map projection from the default (Mercator) and adjusting the map limits.
The user can choose to plot the locations of all or select active volcanoes in the area of interest (based on the Smithsonian Global Volcanism Program database, VOTW4.0, http://www.volcano.si.edu), as well as the locations of copper smelters (known to be strong point sources of SO2). Selecting a SO2 VCD scale from 0 to 2 Dobson Units (DU), examples of the resulting plots of the Lower Troposphere (TRL), Mid-troposphere (TRM) and Upper troposphere/lower stratosphere (STL) SO2 products are shown in Figure 6. The estimated SO2 mass is 26.63 kt at TRL with a maximum of 62.87 DU, 11.28 kt at TRM (max 22.16 DU), and 7.84 kt at STL (max 25.61 DU).
After plotting the software outputs SO2 mass data for the chosen geographic region, including total SO2 mass and maximum SO2 VCD. Options are available to analyze the SO2 cloud mass and emission rate, as well as to query individual pixels and measure distances. Plots can be exported in several formats, including KMZ, which is compatible with many of the most popular visualization programs and permits synoptic integration of the OMI data with the other simulation tools already available on Vhub.
Vhub is supported by the U.S. National Science Foundation, through NSF Grant DRL 0940831. The Vhub development team includes, in addition to the authors: M.I. Bursik, E.S. Calder, S.M. Gallo, P. Johnson, M. Jones, J.V. Bajo, S. Melander, C. Renschler, A. Patra, J. Cortés, D. Moore-Russo, B. Pitman (University at Buffalo); C. Connor, L. Connor (University of South Florida); S. Carn, and W. Rose (Michigan Technological University). Numerous others in the volcanology community have contributed to the resources at Vhub.
- Biass S, Bonadonna C: A quantitative uncertainty assessment of eruptive parameters derived from tephra deposits: the example of two large eruptions of Cotopaxi volcano, Ecuador. Bull Volcanol 2011, 73: 73–90. doi:10.1007/s00445–010–0404–5 doi:10.1007/s00445-010-0404-5 10.1007/s00445-010-0404-5View ArticleGoogle Scholar
- Bonadonna C, Connor CB, Houghton BF, Connor LJ, Byrne M, Laing A, Hincks T: Probabilistic modeling of tephra dispersion: hazard assessment of a multiphase rhyolitic eruption at Tarawera, New Zealand. J Geophys Res 2005, 110: B03203. http://dx.doi.org/10.1029/2003JB002896Google Scholar
- Bonadonna C, Connor LJ, Connor CB, Courtland LM: Tephra2. 2013. . Accessed 20 May 2013 https://vhub.org/resources/tephra2 . Accessed 20 May 2013Google Scholar
- Bursik MI: Effect of winds on the rise height of volcanic plumes. Geo Res Lett 2001, 28(18):3621–3624. 10.1029/2001GL013393View ArticleGoogle Scholar
- Bursik MI: Bent - Atmospheric Plume Analysis. 2013. . Accessed 20 May 2013 https://vhub.org/resources/bent . Accessed 20 May 2013Google Scholar
- Carn SA: OMIplot Software. 2011. . Accessed 16 Apr 2013 http://vhub.org/resources/682 . Accessed 16 Apr 2013Google Scholar
- Carn SA, Krotkov NA, Yang K, Hoff RM, Prata AJ, Krueger AJ, Loughlin SC, Levelt PF: Extended observations of volcanic SO2 and sulfate aerosol in the stratosphere. Atmos Chem Phys Discus 2007, 7: 2857–2871. 10.5194/acpd-7-2857-2007View ArticleGoogle Scholar
- Carn SA, Krueger AJ, Krotkov NA, Arellano S, Yang K: Daily monitoring of Ecuadorian volcanic degassing from space. J Volcanol Geotherm Res 2008, 176: 141–150. 10.1016/j.jvolgeores.2008.01.029View ArticleGoogle Scholar
- Charbonnier SJ, Gertisser R: Numerical simulations of block-and-ash flows using the TITAN2D flow model: examples from the 2006 eruption of Merapi Volcano, Java, Indonesia. Bull Volcanol 2009, 71: 953–959. 10.1007/s00445-009-0299-1View ArticleGoogle Scholar
- Charbonnier S, Gertisser R: Deposit architecture and dynamics of the 2006 block-and-ash flows of Merapi Volcano, Java, Indonesia. Sedimentology 2011, 58: 1573–1612. 10.1111/j.1365-3091.2011.01226.xView ArticleGoogle Scholar
- Charbonnier SJ, Gertisser R: Evaluation of geophysical mass flow models using the 2006 block-and-ash flows of Merapi Volcano, Java, Indonesia: Towards a short-term hazard assessment tool. J Volcanol Geotherm Res 2012, 231–232: 87–108.View ArticleGoogle Scholar
- Connor LJ, Connor CB: Inversion is the key to dispersion: understanding eruption dynamics by inverting tephra fallout. In Statistics in volcanology, Sp Pub IAVCIE No 1. Edited by: Mader HM, Coles SG, Connor CB, Connor LJ. London: Geological Society; 2006.Google Scholar
- Connor CB, Hill BE, Winfrey B, Franklin NM, LaFemina PC: Estimation of volcanic hazards from tephra fallout. Nat Haz Rev 2001, 2: 33–42.View ArticleGoogle Scholar
- Cortés JA, Palma JL: Petrological INput – Graphical oUtput. 2013. . Accessed 16 Apr 2013 https://vhub.org/resources/pingu . Accessed 16 Apr 2013Google Scholar
- Costa A: Hazmap. 2013. . Accessed 20 May 2013 https://vhub.org/resources/hazmap . Accessed 20 May 2013Google Scholar
- Courtland LM, Connor C, Connor L, Bonadonna C: Introducing students to numerical modeling of volcanic hazards: the example of Tephra2 on VHub.org. Numeracy 2012, 5(2):6. doi:10.5038/1936–46220.127.116.11 doi:10.5038/1936-4618.104.22.168View ArticleGoogle Scholar
- Crosweller HS, Arora B, Brown SK, Cottrell E, Diligne NI, Ortiz Guerrero N, Hobbs L, Kiyosugi K, Loughlin SC, Lowndes J, Nayembil M, Siebert L, Sparks RSJ, Takarada S, Venzke E: Global database on large magnitude explosive volcanic eruptions (LaMEVE). J Appl Volcanol 2012, 1: 4. doi:10.1186/2191–5040–1-4 doi:10.1186/2191-5040-1-4 10.1186/2191-5040-1-4View ArticleGoogle Scholar
- Dalbey K, Patra A: Input uncertainty propagation methods and hazard mapping of geophysical mass flows. J Geophys Res 2008, 113: 1–16.View ArticleGoogle Scholar
- Davenport TH, De Long DW, Beers MC: Successful knowledge management projects. Sloan Management Rev 1998, 39: 43–57.Google Scholar
- Denlinger RP, Iverson RM: Flow of variably fluidized granular masses across three-dimensional terrain: 2. Numerical predictions and experimental tests. J Geophys Res 2001, 106: 553–566. 10.1029/2000JB900330View ArticleGoogle Scholar
- Elmagarmid AK, Samuel A, Ouzzani M: Community-Cyberinfrastructure-Enabled discovery in science and engineering. Comp Sci Eng 2008, 10(5):46–53.View ArticleGoogle Scholar
- Iverson RM: The physics of debris flows. Rev Geophys 1997, 35: 245–296. 10.1029/97RG00426View ArticleGoogle Scholar
- Iverson RM, Denlinger RP: Flow of variably fluidized granular masses across three dimensional terrain 1. Coulomb mixture theory. J Geophys Res 2001, 106: 537–552. 10.1029/2000JB900329View ArticleGoogle Scholar
- Johnston EN, Phillips JC, Bonadonna C, Watson IM: Reconstructing the tephra dispersal pattern from the Bronze Age eruption of Santorini using an advection–diffusion model. Bull Volcanol 2012, 74: 1485–1507. doi:10.1007/s00445–012–0609-x doi:10.1007/s00445-012-0609-x 10.1007/s00445-012-0609-xView ArticleGoogle Scholar
- Kanamitsu M, Ebisuzaki W, Woollen J, Yang S-K, Hnilo JJ, Fiorino M, Potter GL: NCEP–DOE AMIP-II Reanalysis (R-2). Bull Am Meteorol Soc 2002, 83: 1631–1643. 10.1175/BAMS-83-11-1631View ArticleGoogle Scholar
- Klimeck G, McLennan M, Brophy SP, Adams GB III, Lundstrom MS: nanoHUB.org: Advancing Education and Research in Nanotechnology. Comp Sci Eng 2008, 10: 17–23.View ArticleGoogle Scholar
- Levelt PF, van den Oord GHJ, Dobber MR, Mälkki A, Visser H, de Vries J, Stammes P, Lundell JOV, Saari H: The ozone monitoring instrument. IEEE Trans Geosci Remote Sensing 2006, 44: 1093–1101.View ArticleGoogle Scholar
- Levelt PF, Hilsenrath E, Leppelmeier GW, van den Oord GHJ, Bhartia PK, Tamminen J, de Haan JF, Veefkind JP: Science Objectives of the Ozone Monitoring Instrument. IEEE Trans Geosci Remote Sensing 2006, 44: 1199–1208.View ArticleGoogle Scholar
- Longchamp C, Bonadonna C, Bachmann O, Skopelitis A: Characterization of tephra deposits with limited exposure: the example of the two largest explosive eruptions at Nisyros volcano (Greece). Bull Volcanol 2011, 73: 1337–1352. doi:10.1007/s00445–011–0469–9 doi:10.1007/s00445-011-0469-9 10.1007/s00445-011-0469-9View ArticleGoogle Scholar
- Lopez T, Carn S, Werner C, Kelly P, Fee D, Kelley P, Doukas M, Pfeffer M, Webley P, Cahill C, Schneider D: Evaluation of Redoubt Volcano’s sulfur dioxide emissions by the Ozone Monitoring Instrument. J Volcanol Geotherm Res 2012, 259: 290–307. doi:10.1016/j.jvolgeores.2012.03.002 doi:10.1016/j.jvolgeores.2012.03.002View ArticleGoogle Scholar
- Lube G, Cronin SJ, Thouret J-C, Surono : Kinematic characteristics of pyroclastic density currents and controls on their avulsion from natural and engineered channels. Geol Soc Am Bull 2011, 123: 1127–1140. 10.1130/B30244.1View ArticleGoogle Scholar
- Macedonio G, Costa A, Longo A: A computer model for volcanic ash fallout and assessment of subsequent hazard. Comp Geosci 2005, 31: 837–845. 10.1016/j.cageo.2005.01.013View ArticleGoogle Scholar
- Mangeney-Castelnau A, Vilotte J, Bristeau MO, Perthame B, Bouchut F, Simeoni C, Yernini S: Numerical modeling of avalanches based on Saint-Venant equations using a kinetic Scheme. J Geophys Res 2002, 108: 2537. doi:10.1029/2002JB002024 doi:10.1029/2002JB002024Google Scholar
- Marzocchi W, Sandri L, Selva J: BET_VH: a probabilistic tool for long-term volcanic hazard assessment. Bull Volcanol 2010, 73: 705–716. doi:10.1007/s00445–010–0357–8 doi:10.1007/s00445-010-0357-8View ArticleGoogle Scholar
- Mastin LG: Plumeria 2.3.1. 2011. . Accessed 16 Apr 2013 https://vhub.org/resources/1194 . Accessed 16 Apr 2013Google Scholar
- McLennan M, Kennell R: HUBzero: a platform for dissemination and collaboration in computational science and engineering. Comp Sci Eng 2010, 12: 48–52.View ArticleGoogle Scholar
- Palma JL: Titan2D Viewer. 2012. . Accessed 20 May 2013 https://vhub.org/resources/titanview . Accessed 20 May 2013Google Scholar
- Palma JL: Energy cone. 2013a. . Accessed 20 May 2013 https://vhub.org/resources/econe . Accessed 20 May 2013Google Scholar
- Palma JL: DEM converter for Titan2D. 2013b. . Accessed 20 May 2013 https://vhub.org/resources/dem4titan . Accessed 20 May 2013Google Scholar
- Patra AK, Bauer AC, Nichita CC, Pitman EB, Sheridan MF, Bursik MI, Rupp B, Webber A, Stinton AJ, Namikawa LM, Renschler CS: Parallel adaptive numerical simulation of dry avalanches over natural terrain. J Volcanol Geotherm Res 2005, 139: 1–21. 10.1016/j.jvolgeores.2004.06.014View ArticleGoogle Scholar
- Pitman EB, Nichita CC, Sheridan M, Bursik MI: Computing granular avalanches and landslides. Phys Fluids 2003, 15: 3638–3646. 10.1063/1.1614253View ArticleGoogle Scholar
- Rose WI: AshFall: A graduate course in volcanology with substantial meteorological content. 2011. . Accessed 16 Apr 2013 https://vhub.org/resources/411 . Accessed 16 Apr 2013Google Scholar
- Rosenberg MJ: E-Learning: Strategies for delivering knowledge in the digital era. New York: McGraw Hill; 2001.Google Scholar
- Savage SB, Hutter K: The motion of a finite mass of granular material down a rough incline. J Fluid Mech 1989, 199: 177–215. 10.1017/S0022112089000340View ArticleGoogle Scholar
- Sheridan MF, Macías JL: Estimation of risk probability for gravity-driven pyroclastic flows at Volcan Colima, Mexico. J Volcanol Geotherm Res 1995, 66: 251–256. 10.1016/0377-0273(94)00058-OView ArticleGoogle Scholar
- Shih W-C, Tseng S-S, Yang C-T: Wiki-based rapid prototyping for teaching-material design in e-learning grids. Comput Educat 2008, 51: 1037–1057. 10.1016/j.compedu.2007.10.007View ArticleGoogle Scholar
- Siebert L, Simkin T, Kimberly P: Volcanoes of the World. 3rd edition. Berkeley: Smithsonian Institution, University of California Press; 2010.Google Scholar
- Stefanescu ER, Bursik MI, Patra AK: Effect of digital elevation models on Mohr-Coulomb geophysical flow model output. Nat Haz 2012, 62: 635–656. 10.1007/s11069-012-0103-yView ArticleGoogle Scholar
- Stinton AJ, Sheridan MF, Patra AK, Dalbey KR, Namikawa LM: Integrating variable bed friction into the TITAN2D mass-flow model: applications to the Little Tahoma Peak Avalanches. Acta Vulcanol 2004, 16: 153–163.Google Scholar
- Surono JP, Pallister J, Boichu M, Buongiorno MF, Budisantoso A, Costa F, Andreastuti S, Prata F, Schneider D, Clarisse L, Humaida H, Sumarti S, Bignami C, Griswold J, Carn SA, Oppenheimer C, Lavigne F: The 2010 explosive eruption of Java’s Merapi volcano – a ‘100-year’ event. J Volcanol Geotherm Res 2012, 241–242: 121–135.View ArticleGoogle Scholar
- Thomas HE, Watson IM, Carn SA, Prata AJ, Realmuto VJ: A comparison of AIRS, MODIS and OMI sulphur dioxide retrievals in volcanic clouds. Geomat Nat Haz Risk 2011, 2: 217–232. 10.1080/19475705.2011.564212View ArticleGoogle Scholar
- Tonini R, Sandri L, Selva J, Marzocchi W: Bayesian event tree for volcanic hazard. 2012. . Accessed 16 Apr 2013 https://vhub.org/resources/betvh . Accessed 16 Apr 2013Google Scholar
- Volentik ACM, Connor CB, Connor LJ, Bonadonna C: Aspects of volcanic hazards assessment for the Bataan nuclear power plant, Luzon Peninsula, Philippines. In Volcanic and Tectonic Hazard Assessment for Nuclear Facilities. Edited by: Connor CB, Chapman N, Connor L. New York: Cambridge University Press; 2009.Google Scholar
- Volentik ACM, Bonadonna C, Connor CB, Connor LJ, Rosi M: Modeling tephra dispersal in the absence of wind: Insights from the climatic phase of the 2450 BP Plinian eruption of Pululagua volcano (Ecuador). J Volcanol Geotherm Res 2010, 193: 117–136. 10.1016/j.jvolgeores.2010.03.011View ArticleGoogle Scholar
- Yu B, Dalbey K, Webb A, Bursik M: Numerical issues in computing inundation areas over natural terrains using Savage-Hutter theory. Nat Haz 2009, 50: 249–267. 10.1007/s11069-008-9336-1View ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.