Map Projections | Features | News | College of Earth, Ocean, and Atmospheric Sciences | Oregon State University
Today’s web maps (such as those by OpenStreetMap, Google, Bing, or Yahoo!) have two major shortcomings. The Mercator projection is an unfortunate choice, as it generates a distorted world view at small scales. On the Mercator world map, Greenland appears huge in comparison to Brazil. In reality Brazil is about four times larger than Greenland. Current web maps are restricted to predefined zoom levels, that is, the scale of the map cannot be chosen freely, and smooth animations between map scales are not possible. This is unfortunate, as it has been shown that a smooth animation helps map readers orient more efficiently when adjusting the scale of a map. The goal of this research is to develop web maps that take advantage of the latest web techniques, such as HTML5 Canvas and WebGL, which allow for very fast geometric manipulation and rendering of vector data on the web client. This allows us to adjust the projection of the web map to the current scale and the area displayed. An adaptive composite map projection scheme has been developed and a first proof-of-concept prototype has been implemented using HTML5 Canvas and WebGL. The map projection seamlessly adapts to map scale and the area shown on the map. Fabian Gschwend, ETH Zurich, developed a prototype for smoothly interpolating two sets of vector data with changing scale. He uses two cartographically generalized river networks provided by Natural Earth and generates an intermediate data set on the fly. To try the prototype, zoom in to about 80% of the zoom slider, then adjust the Generalization slider to blend between the two data sets. Loading this page take a moment, as all pre-processing for finding matching vertices between the two data sets is done by the browser. The adaptive composite map projection will be improved and extended. A smoother transitions towards the web Mercator projection will be developed. Vector interpolation will be extended to other data types besides rivers, and the method will be improved. New methods for scale-adaptive cartographic symbolization for vector features and text labels will be explored. Source.