Abstract

As experiments and simulations in Earth System Science grow larger and more complex, dataset volumes are growing explosively. Web-based visualization and analysis of these datasets is becoming a challenge due to large amount of data and the limit of network bandwidth. The Live Access Server (LAS) is a highly configurable Web server designed to provide flexible access to visualization and analysis products generated from geo-referenced scientific datasets. In this presentation, we introduce a new capability of LAS for interactively visualizing large datasets by utilizing the view-based refresh queries of Google Earth and the automatic decimation capabilities of Ferret. When viewing a high resolution dataset on a global scale it is wasteful of bandwidth to handle the full resolution data. With the striding capability, Ferret selects every nth point along an axis, where n is the striding value. The striding values are dynamically computed based upon the size of the area of interest. The automatic striding approach minimizes the volume of data that need be touched to visualize a large geographic area. Higher resolutions are utilized for smaller areas to reveal the fine structures. The LAS provides this behavior using Google Earth as the user interface. As users zoom or pan on Google Earth, Google Earth interacts with LAS through a Network Link, which contains the URL of a LAS server. When the view inside Google Earth stops for a set number of seconds, it makes a request to a LAS server, sending the LAS server the latitude and longitude boundaries (bounding box) of the area currently in view. The LAS server uses that information to compute stride values, instruct Ferret to generate a visualization for that particular geographic area, and send the resulting image back to Google Earth.

Presentations

Poster
(6 MB)