- The selection process for map projections is a mystery to many mapmakers and GIS users. Map projections ought to be selected based on the map’s geographic extent and the required distortion properties, with the goal of minimizing the distortion of the mapped area. Despite some available selection guidelines, the selection of map projections is not yet automated. Automated selection would help mapmakers and GIS users to better select a projection for their map. The overall goal of this dissertation is to take a step towards this automation and explore user preferences with an objective to provide additional criteria for selecting world map projections. An additional goal is to optimize automatic map projection selection for web maps. The results presented in this work are mathematical models (new map projections for world maps, polynomial equations for selecting standard parallels) and new selection criteria for world maps. They improve our knowledge about map projection selection for world maps and web maps. As a result of the research presented in this doctoral dissertation, we know more about the map projection preferences of map-readers and have improved techniques for adapting map projections for scalable web maps and GIS software. Altogether, four concrete research questions were addressed.
The first research question explores user preferences for world map projections. Many small-scale map projections exist and they have different shapes and distortion characteristics. World map projections are mainly chosen based on their distortion properties and the personal preferences of cartographers. Very little is known about the map projection preferences of map-readers; only two studies have addressed this question so far. This dissertation presents a user study among map-readers and trained cartographers that tests their preferences for world map projections. The paired comparison test of nine commonly used map projections reveals that the map-readers in our study prefer the Robinson and Plate Carrée projections, followed by the Winkel Tripel, Eckert IV, and Mollweide projections. The Mercator and Wagner VII projections come in sixth and seventh place, and the least preferred are two interrupted projections, the interrupted Mollweide and the interrupted Goode Homolosine. Separate binominal tests indicate that map-readers involved in the study seem to like projections with straight rather than curved parallels, and meridians with elliptical rather than sinusoidal shapes. The results indicate that map-readers prefer projections that represent poles as lines to projections that show poles as protruding edges, but there is no clear preference for pole lines in general. The trained cartographers involved in this study have similar preferences, but they prefer pole lines to represent the poles, and they select the Plate Carrée and Mercator projections less frequently than the other participants.
The second research question introduces the polynomial equations for the Natural Earth II projection and tests user preferences for its graticule characteristics. The Natural Earth II projection is a new compromise pseudocylindrical projection for world maps. It has a unique shape compared to most other pseudocylindrical projections. At high latitudes, the meridians bend steeply toward short pole lines resulting in a map with highly rounded corners that resembles an elongated globe. Its distortion properties are similar to most other established world map projections. The projection equation consists of simple polynomials. A user study evaluated whether map-readers prefer Natural Earth II to similar compromise projections. The 355 participating general map- readers rated the Natural Earth II projection lower than the Robinson and Natural Earth projections, but higher than the Wagner VI, Kavrayskiy VII, and Wagner II projections.
The third question examines how Wagner's transformation method can be used for improving map projections for scalable web maps, and its integration into the adaptive composite map projections schema. The adaptive composite map projections schema, invented by Bernhard Jenny, changes the projection to the geographic area shown on a map. It is meant as a replacement for the commonly used web Mercator projection, which grossly distorts areas when representing the entire world. The original equal-area version of the adaptive composite map projections schema uses the Lambert azimuthal projection for regional maps, and three alternative projections for world maps. In this dissertation, it is explored how the adaptive composite map projections schema can include a variety of other equal-area projections when the transformation between the Lambert azimuthal and the world projections uses Wagner's method. In order to select the most suitable pseudocylindrical projection, the distortion characteristics of a pseudocylindrical projection family were analyzed, and a user study among experts in the area of map projections was carried out. Based on the results of the distortion analysis and the user study, a new pseudocylindrical projection is recommended for extending the adaptive composite map projections schema. The new projection is equal-area throughout the transformation to the Lambert azimuthal projection, has better distortion characteristics than small-scale projections currently included in the original adaptive composite map projections schema, and aligns with map-readers' preferences for world map projections.
The last research question explores how the selection of the standard parallels of conic projections can be automated. Conic map projections are appropriate for mapping regions at medium and large scales with east-west extents at intermediate latitudes. Conic projections are appropriate for these cases because they show the mapped area with less distortion than other projections. In order to minimize the distortion of the mapped area, the two standard parallels of conic projections need to be selected carefully. Rules of thumb exist for placing the standard parallels based on the width-to- height ratio of the map. These rules of thumb are simple to apply, but do not result in maps with minimum distortion. There also exist more sophisticated methods that determine standard parallels such that distortion in the mapped area is minimized. These methods are computationally expensive and cannot be used for real-time web mapping and GIS applications where the projection is adjusted automatically to the displayed area. This article presents a polynomial model that quickly provides the standard parallels for the three most common conic map projections: the Albers equal- area, the Lambert conformal, and the equidistant conic projection. The model defines the standard parallels with polynomial expressions based on the spatial extent of the mapped area. The spatial extent is defined by the length of the mapped central meridian segment, the central latitude of the displayed area, and the width-to-height ratio of the map. The polynomial model was derived from 3825 maps--each with a different spatial extent and computationally determined standard parallels that minimize the mean scale distortion index. The resulting model is computationally simple and can be used for the automatic selection of the standard parallels of conic map projections in GIS software and web mapping applications.