We propose two novel map navigation techniques, called Anchored Zoom (AZ) and Anchored Zoom and Tilt (AZT). In these techniques, the zooming and tilting of a virtual camera are automatically coupled with users' panning displacements so that the anchor point determined by users always remains in a viewport. This allows users to manipulate a viewport without mode-switching among pan, zoom, and tilt while maintaining a sense of distance and direction from the anchor point. We conducted an experiment to evaluate AZ and AZT and compare them with Pan & Zoom (PZ)  and Speed-dependent Automatic Zooming (SDAZ)  in off-screen target acquisition tasks and spatial recognition tests. Results showed that our proposed techniques were more effective than those of competitors in reducing time to reach off-screen objects while maintaining users' sense of distance and direction as well as PZ.