A data management and visualization framework for community vulnerability to hazards

Science Center Objects

USGS research in the Western Geographic Science Center has produced several geospatial datasets estimating the time required to evacuate on foot from a Cascadia subduction zone earthquake-generated tsunami in the U.S. Pacific Northwest. These data, created as a result of research performed under the Risk and Vulnerability to Natural Hazards project, are useful for emergency managers and communi...

USGS research in the Western Geographic Science Center has produced several geospatial datasets estimating the time required to evacuate on foot from a Cascadia subduction zone earthquake-generated tsunami in the U.S. Pacific Northwest. These data, created as a result of research performed under the Risk and Vulnerability to Natural Hazards project, are useful for emergency managers and community planners but are not in the best format to serve their needs. This project explored options for formatting and publishing the data for consumption by external partner agencies and the general public.



The project team chose ScienceBase as the publishing platform, both for its ability to convert spatial data into web services and for its designation as an official platform for the public release of USGS data.  Because the travel time map datasets are large vector files, the team planned to experiment with different extents, projections, and data resolutions to determine usability for internal use and for use by external partners.  For the general public who reside in or near areas exposed to tsunamis and who are not likely to use web services, the project team experimented with Cesium, an innovative new JavaScript 3D mapping package, to create a web mapping application for exploration and visualization of the data.

Accomplishments

The accomplishments for this project are described in the following sections.



External Partner Needs Assessment



The original travel time maps were created as continuous value rasters, and for project purposes were binned into 1-minute increments and converted to vector format.  The project team uploaded these vector shapefiles to ScienceBase, created web feature services, and then contacted colleagues at the California Earthquake Clearinghouse to solicit feedback on the usefulness of the data.  This external partner evaluated the product and made the following observations.

  • The use of web services in ArcMap requires the Interoperability Extension, which many partners won’t have.
  • The 1-minute resolution time map is useful when zoomed in, but is slow to load, so multiple services with times dissolved to 5- and 10-minute increments in a scale-dependent map would be ideal.
  • The web feature service is nice for querying the map but is a slow format, and some users will prefer a simple image or tile service to use as an overlay in the field.
  • If an image or map service is used that cannot be queried, create an accompanying semi-transparent layer with time labels.
  • Serving the travel time maps by county is a good choice because many partners work at this level.
  • Create an inventory of the web services with links and make this available to external partners so they know how to get to the data.

Datasets and Metadata in ScienceBase

The project team created metadata and uploaded eight travel time map vector shapefiles to ScienceBase (Wood and Jones, 2017). The maps are at the county level for Del Norte and Humboldt Counties in northern California, and the project team left them in the original 1-minute increments of travel time to safety. The team created web feature services in ScienceBase, loaded the services into the ArcMap desktop application, and tested the ability to query the maps for travel times at various locations. The team initially projected the maps to Web Mercator so they would work well with a commercial web base map, but ended up using World Geodetic System (WGS) 84 to meet the requirement of their Cesium mapping application.  The use of the ArcMap Interoperability Extension to access the services and the initial load time is a deterrent to use, but once in place, the services work well.  Though the team appreciated the suggestions of the external partner, they realized that the creation of maps and metadata for so many different representations of the information was beyond the scope of this project.

 

Mapping Application Testbed

The project team decided that a 3D viewing platform would be the most appropriate platform for this project because the extra dimension of height is a critical aspect of tsunami risk and evacuation.  Viewing in 3D allows the user to identify the role elevation plays in tsunami risk, as well as evacuation potential. The team wanted to impart this information to the public. Few browser-based 3D mapping libraries exist, but an emerging JavaScript library, Cesium, presented an opportunity to satisfy the requirements and also explore emerging technologies in visualization.  The capabilities of this software were tested, allowing the team to identify best practices, which are outlined below, for using this software to present hazard exposure information.

  • Input data must be projected to WGS 84.
  • Datasets with a high number of features (over 10,000) or large file sizes can be displayed, though the browser’s memory usage increases heavily.
  • The preferred input data format for all but the largest datasets is topoJSON, providing seamless integration into the Cesium viewer and the ability to access data attributes and alter styling.
  • By default, polygons are loaded into the 3D scene and projected flat against the spheroid of the Earth, but by changing a setting, the polygons can be “draped” over real-world terrain. This “draping” improves appearance but reduces performance, interfering with the ability to pan and interact smoothly with a map containing a large dataset. 

Overall, Cesium had many promising features and may prove to be a viable alternative to the traditional web map (figure 2).



Note:  This description is from the Community for Data Integration 2016 Annual Report.