Geospatial data processing and visualization has matured into an important research area. Although a number of tools and devices exist for acquiring raw data, visual analysis and exploration of raw datasets (census data or LANDSAT imagery) acquired in this manner remains a challenge. One crucial tradeoff in such visualizations is choice between the quantity of data presented, and important details that are associated with individual (or clustered) data instances. This dissertation presents solutions to three topical problems in the field of geospatial visualization. The first solution is an extension to the CGA Shape Grammar that allows a domain scientist to quickly author glyph-based visualizations on a geospatial domain. This work extends the grammar framework by adding new commands that process geospatial databases and provide geo-referencing functionality for the glyphs. The second solution presents a method for visualizing Route Maps that preserves the overview, and provides detail information at points of interest along the route. The optimal placement of the detail lenses is calculated with the help of a backtracking based optimization algorithm. Usability studies that validate the usefulness of this approach are presented. The third solution provides a novel method for designing time-constrained tours for tourist activities. The described approach combines a time-dependent scheduling method for creating the trip itinerary with a novel layout algorithm to display sightseeing routes at given attractions. Route details, including turn points and landmarks around the route, are shown in detail while preserving the overall path shape and local context relationships. Several examples of trip brochures created with the system are also provided.