Issue link: https://read.dmtmag.com/i/46658
As the Geo-Web mindset gains acceptance and data storage becomes ubiquitous, more maps will take on image characteristics. 10-degree lat/long gridding steps representing approximately 692-mile movements along the equator. So much for a conceptual review of lat/long, keeping in mind that there are a lot of geographic and mathematical considerations in implementing the coordinate system. Thankfully, they were hammered out years ago, resulting in a set of universal standards that form the foundation for contemporary GPS use. Points, Lines, Polygons The next conceptual step involves extending the paper-map paradigm to grid-based data layers. Tradi- tional mapping holds that there are three fundamen- tal map features: discrete points, lines and polygons. With the advent of the digital map, a fourth feature type emerges: continuous surfaces. The lat/long grid forms a surface for geographic referencing that's analogous to a digital image with a "dot" (pixel) for every location in the viewed area. In the case of a grid-map layer, a map value that identifies the characteristic/condition at a location replaces the pixel value denoting color. Like the image on a computer screen, a point feature is represented by a single dot/cell containing that feature; a line feature by a series of connected cells; and a polygon feature by all of the cells defining it—interior and boundary. A surface is rep- resented by the entire gridded continuum and, like an elevation surface, depicts continuous changes in a map variable. Whereas points, lines and polygons have sharp, abrupt boundaries, surfaces form gradients of change. Practical Problems The final conceptual leap is shown in Figure 2. Like Google Earth's registration of satellite images to lat/long referencing, mapped data of all types can be stored, retrieved and analyzed based on the stored coordinates for the records, thereby forming a "Universal DBMS Key" that can link seemingly disparate database files. The process is similar to a date/time stamp, except the "where information" provides a spatial context for joining datasets. Demographic records can be linked to resource records that, in turn, can be linked to business records, etc. In practice, extensive data are stored in "Virtual Tiles" for efficient storage, access and processing. A user identifies a boundary extent, then an "Analysis Frame" resolution (cell size) and pertinent data layers that are automatically extracted into a "Map Stack" for grid-based map analysis and modeling. Solution map(s) resulting from analysis can be exported in vector, raster or DBMS form, and the map stack is retained for subsequent processing or deleted and reconstructed as needed. Figure 2. The lat/long grids are used to construct a map stack of geo-registered map layers that are preconditioned for map analysis and modeling. So what's holding back this seemingly utopian mapping world? The short answer is "some practical and legacy considerations." On the practical front, the geographic stretching and pinching of the grid cells with increasing latitude confounds map analysis at global scales and lacks the precision necessary for detailed cadastral/survey- ing applications. On the legacy front, the approach relies more on DBMS and image-processing mindsets than on traditional mapping paradigms and geographic principles underlying most flagship GIS packages. It's like a Rorschach inkblot. Since the Middle Ages, we have thought of lat/long as intersecting lines, whereas the new perspective is flipped to a continuum of grid cells. Current thinking is more like a worldwide egg crate, with the grid spaces as locations for placing map values that indicate the characteristics/conditions of a multitude of map variables anywhere in the world. As the Geo-Web mindset gains acceptance and data storage becomes ubiquitous, more maps will take on image characteristics—a raster map where lat/long grid cells replace pixels, and map values amenable to analysis replace color codes. Although the vector map model will continue, it's the raster model (lat/long grids, in particular) that takes us well beyond traditional mapping. Author's Notes: For more information on grid-based mapped data considerations, see the online book, Beyond Modeling III, Topic 18, Understanding Grid-based Data, at www.innovativegis.com/Basis/MapAnalysis. OCT O BER 2O11 / WWW . GEOPLA CE . C OM 11