Naturally enough, the Atlas proper is divided into four major parts, one for each satellite. Global maps at 1-kilometer resolution introduce each satellite. These are in cylindrical projection, in which latitude increases at a constant rate from pole to pole, and are reduced in scale to fit the page. These are followed by a set of five orthographic global maps from various perspectives (including leading and trailing hemispheres), simulating the views a passing astronaut might have. Following planetary mapping convention, each satellite is then divided into 15 quasirectangular maps of roughly equal size called quadrangles (Figure 2.1). These show the full 1-kilometer resolution detail of the global maps. These maps are named and numbered according to convention. For example, quadrangle “Je9” refers to “J”upiter satellite “E”uropa quadrangle “9.”

No two-dimensional map can fully represent a three-dimensional surface without distorting either feature sizes or shapes. The quadrangles come in three map projections. Polar maps are in polar stereographic projection. Equatorial maps are in mercator, while mid-latitude maps are in lambert conformal conic projection. All three projections are conformal in that they preserve shapes fairly well. None of these projections preserves areas, but each map size is sufficiently small that this distortion is minimized.

Following each quadrangle map, all Galileo and Voyager targeted high-resolution mosaics that occur within that quadrangle are presented. With a few exceptions, these are shown at their original resolution, which can vary between ∼6 meters to roughly 500 meters (plus a few at up to 850 meters).