{"id":597,"date":"2014-04-21T15:23:54","date_gmt":"2014-04-21T20:23:54","guid":{"rendered":"http:\/\/homepages.uc.edu\/~yaozo\/wordpress\/?p=597"},"modified":"2014-04-21T15:23:54","modified_gmt":"2014-04-21T20:23:54","slug":"beyond-mapping-iii-topic-18-understanding-grid-based-data","status":"publish","type":"post","link":"https:\/\/zhuoyao.net\/index.php\/2014\/04\/21\/beyond-mapping-iii-topic-18-understanding-grid-based-data\/","title":{"rendered":"Beyond Mapping III Topic 18: Understanding Grid-based Data"},"content":{"rendered":"

Grids and Lattices Build Visualizations<\/a>\u00a0<\/em>\u2014\u00a0describes Lattice and Grid forms of map surface display<\/em><\/p>\n

Maps Are Numbers First, Pictures Later<\/a><\/em>\u00a0<\/em>\u2014\u00a0discusses the numeric and geographic characteristics of map values<\/em><\/p>\n

Contour Lines versus Color Gradients for Displaying Spatial Information<\/a>\u00a0<\/em>\u2014\u00a0discusses the similarities and differences between discrete contour line and continuous gradient procedures for visualizing map surfaces<\/em><\/p>\n

Setting a Place at the Table for Grid-based Data<\/a>\u00a0\u2014\u00a0describes the differences between individual file and table storage approaches<\/em><\/p>\n

VtoR\u00a0and Back!<\/a>\u00a0\u2014\u00a0describes various techniques for converting between vector and raster data types<\/em><\/p>\n

Normalizing Maps for Data Analysis<\/a>\u00a0<\/em>\u2014\u00a0describes map normalization and data exchange with other software packages<\/em><\/p>\n

Comparing Apples and Oranges<\/a>\u00a0<\/em>\u2014\u00a0describes a Standard Normal Variable (SNV) procedure for normalizing maps for comparison<\/em><\/p>\n

Correlating Maps and a Numerical Mindset<\/a>\u00a0<\/em>\u2014\u00a0describes a\u00a0Spatially\u00a0Localized Correlation procedure for mapping the mutual relationship between two map variables\u00a0<\/em><\/p>\n

Multiple Methods Help Organize Raster Data<\/a>\u00a0<\/em>\u2014\u00a0discusses different approaches to storing raster data<\/em><\/p>\n

Use Mapping \u201cArt\u201d to Visualize Values<\/a>\u00a0<\/em>\u2014\u00a0describes procedures for generating contour maps<\/em><\/p>\n

What\u2019s\u00a0Missing\u00a0in Mapping?<\/a>\u00a0\u2014\u00a0discusses the need for identifying data dispersion as well as average in Thematic Mapping<\/em><\/p>\n

 <\/p>\n

Note<\/em>:<\/em>\u00a0The processing and figures discussed in this topic were derived using\u00a0MapCalcTM<\/sup>\u00a0software.\u00a0\u00a0See\u00a0www.innovativegis.com<\/a>\u00a0to download a free MapCalc Learner version with tutorial materials for classroom and self-learning map analysis concepts and procedures.<\/em><\/p>\n

\u00a0<\/em><\/p>\n

<Click here<\/a>> right-click to download a printer-friendly version of this topic (.pdf).<\/em>
\n<\/em>
\n<\/em><\/strong>(Back to the Table of Contents<\/a>)<\/em><\/strong>
\n______________________________
\n<\/em><\/p>\n

<\/em><\/strong><\/p>\n

Grids and Lattices Build Visualizations<\/em><\/strong>\u00a0\u00a0<\/strong><\/p>\n

(GeoWorld, July 2002, pg. 26-27)<\/strong><\/em><\/p>\n

(return\u00a0to top of Topic<\/a>)<\/strong><\/em><\/p>\n

 <\/p>\n

For thousands of years, points, lines and polygons have been used to depict map features.\u00a0\u00a0With the stroke of a pen a cartographer could outline a continent, delineate a highway or identify a specific building\u2019s location.\u00a0\u00a0With the advent of the computer, manual drafting of these data has been replaced by the cold steel of the plotter.<\/p>\n

 <\/p>\n

In digital form these spatial data have been linked to attribute tables that describe characteristics and conditions of the map features.\u00a0\u00a0Desktop mapping exploits this linkage to provide tremendously useful database management procedures, such as address matching, geo-query and routing.\u00a0\u00a0Vector-based<\/em>\u00a0data forms the foundation of these techniques and directly builds on our historical perspective of maps and map analysis.<\/p>\n

 <\/p>\n

Grid-based<\/em>\u00a0data, on the other hand, is a relatively new way to describe geographic space and its relationships.\u00a0\u00a0Weather maps, identifying temperature and barometric pressure gradients, were an early application of this new data form.\u00a0\u00a0In the 1950s computers began analyzing weather station data to automatically draft maps of areas of specific temperature and pressure conditions.\u00a0\u00a0At the heart of this procedure is a new map feature that extends traditional points, lines and polygons (discrete objects) to continuous surfaces.<\/p>\n

 <\/p>\n

Figure 1.<\/em>\u00a0\u00a0Grid-based data can be displayed in 2D\/3D lattice or grid forms.<\/em><\/p>\n

 <\/p>\n

The rolling hills and valleys in our everyday world is a good example of a geographic surface.\u00a0\u00a0The elevation values constantly change as you move from one place to another forming a continuous spatial gradient.\u00a0\u00a0The left-side of figure 1 shows the grid data structure and a sub-set of values used to depict the terrain surface shown on the right-side.<\/p>\n

 <\/p>\n

Grid data are stored as an organized set of values in a matrix that is geo-registered over the terrain.\u00a0\u00a0Each grid cell identifies a specific location and contains a map value representing its average elevation.\u00a0\u00a0For example, the grid cell in the lower-right corner of the map is 1800 feet above sea level.\u00a0\u00a0The relative heights of surrounding elevation values characterize the undulating terrain of the area.<\/p>\n

 <\/p>\n

Two basic approaches can be used to display this information\u2014lattice and grid.\u00a0\u00a0The\u00a0lattice<\/em><\/strong>\u00a0display form uses lines to convey surface configuration.\u00a0\u00a0The contour lines in the 2D version identify the breakpoints for equal intervals of increasing elevation.\u00a0\u00a0In the 3D version the intersections of the lines are \u201cpushed-up\u201d to the relative height of the elevation value stored for each location.\u00a0\u00a0The\u00a0grid<\/em><\/strong>\u00a0display form uses cells to convey surface configuration.\u00a0\u00a0The 2D version simply fills each cell with the contour interval color, while the 3D version pushes up each cell to its relative height.<\/p>\n

The right-side of figure 2 shows a close-up of the data matrix of the project area.\u00a0\u00a0The elevation values are tied to specific X,Y\u00a0coordinates (shown as yellow dots).\u00a0\u00a0Grid display techniques assume the elevation values are centered within each grid space defining the data matrix (solid back lines).\u00a0\u00a0A 2D grid display checks the elevation at each cell then assigns the color of the appropriate contour interval.<\/p>\n

 <\/p>\n

Figure 2.<\/em>\u00a0\u00a0Contour lines are delineated by connecting interpolated points of constant elevation along the lattice frame.<\/em><\/p>\n

 <\/p>\n

Lattice display techniques, on the other hand, assume the values are positioned at the intersection of the lines defining the reference frame (dotted red lines).\u00a0\u00a0Note that the \u201cextent\u201d (outside edge of the entire area) of the two reference frames is off by a half-cell*.\u00a0\u00a0Contour lines are delineated by calculating where each line crosses the reference frame (red X\u2019s) then these points are connected by straight lines and smoothed.\u00a0\u00a0In the left-inset of the figure note that the intersection for the 1900 contour line is about half-way between the 1843 and 1943 values and nearly on top of the 1894 value.<\/p>\n

 <\/p>\n

Figure 3 shows how 3D plots are generated.\u00a0\u00a0Placing the viewpoint at different look-angles and distances creates different perspectives of the reference frame.\u00a0\u00a0For a 3D grid display entire cells are pushed to the relative height of their map values.\u00a0\u00a0The grid cells retain their projected shape forming blocky extruded columns.<\/p>\n

 <\/p>\n

Figure 3.<\/em>\u00a0\u00a03D display \u201cpushes-up\u201d the grid or lattice reference frame to the relative height of the stored map values.<\/em><\/p>\n

 <\/p>\n

3D lattice display pushes up each intersection node to its relative height.\u00a0\u00a0In doing so the four lines connected to it are stretched proportionally.\u00a0\u00a0The result is a smooth wireframe that expands and contracts with the rolling hills and valleys.\u00a0\u00a0Generally speaking, lattice displays create more pleasing maps and knock-your-socks-off graphics when you spin and twist the plots.\u00a0\u00a0However, grid displays provide a more honest picture of the underlying mapped data\u2014a chunky matrix of stored values.<\/p>\n

____________________<\/p>\n

Author’s Note<\/em><\/strong>:\u00a0\u00a0Be forewarned that the alignment difference between grid and lattice reference frames is a frequent source of registration error when one \u201cblindly\u201d imports a set of grid layers from a variety of sources.<\/em><\/p>\n

Contour Lines versus Color Gradients for Displaying Spatial Information<\/em><\/strong><\/p>\n

(GeoWorld, November 2012)<\/strong><\/em><\/p>\n

(return\u00a0to top of Topic<\/a>)<\/strong><\/em><\/p>\n

 <\/p>\n

In mapping, there are historically three fundamental map features\u2014points<\/em>,\u00a0lines<\/em>\u00a0and\u00a0areas<\/em>\u00a0(also referred to as polygons or regions)\u2014that partition space into discrete spatial objects (discontinuous).\u00a0 When assembled into a composite visualization, the set of individual objects are analogous to arranging the polygonal pieces in a jigsaw puzzle and then overlaying the point and line features to form a traditional map display.\u00a0\u00a0In paper and electronic form, each separate and distinct spatial object is stored, processed and displayed individually.<\/p>\n

 <\/p>\n

By definition line features are constructed by connecting two or more data points.\u00a0\u00a0Contour lines<\/em>\u00a0are a form of displaying information that have been used for many years and are a particularly interesting case when displaying information in the computer age.\u00a0\u00a0They represent a special type of spatial object in that they connect data points of equal value.\u00a0 A contour map is typically composed of a set of contour lines with intervening spaces generally referred to as\u00a0contour intervals<\/em>.<\/p>\n

 <\/p>\n

In general usage, the word \u201cline\u201d in the term \u201ccontour line\u201d has been occasionally dropped, resulting in the ill-defined terms of \u201ccontour\u201d for an individual contour line and \u201ccontours\u201d for a set of contour lines or contour intervals depending on whether your focus is on the lines or the intervening areas.<\/p>\n

 <\/p>\n

The underlying concept is that contour lines are lines that are separate and distinct spatial objects that connect data points of equal value and are displayed as lines to convey this information to the viewer.\u00a0Historically these values represented changes in elevation to generate a contour map of a landscape\u2019s undulating terrain, formally termed\u00a0topographical relief<\/em>.<\/p>\n

 <\/p>\n

Today, contour lines showing equal data values at various data levels are used to represent a broad array of spatial data from environmental factors (e.g., climatic temperature\/pressure, air pollution concentrations) to social factors (e.g., population density, crime rates) to business intelligence (e.g., customer concentration, sales activity).\u00a0\u00a0Within these contexts, the traditional mapping focus has been extended into the more general field of data visualization.<\/p>\n

 <\/p>\n

With the advent of the computer age and the associated increased processing power of the computer, a fourth type of fundamental map feature has emerged\u2014surfaces<\/em>\u2014an uninterrupted gradient (continuous). \u00a0This perspective views space as a continuously changing variable analogous to a magnetic field without interruption of discrete spatial object boundaries.\u00a0 In electronic form, a continuous surface is commonly stored as two-dimensional matrix (x,y) with a data value (z) at each grid location.\u00a0 These data are often displayed as a 3-dimentional surface with the\u00a0x,y-coordinates orienting the visualization plane in space and the z-value determining the relative height above the plane.<\/p>\n

 <\/p>\n

In general, a 2-dimensional rendering of surface data can be constructed two ways\u2014using a continuous\u00a0color gradient<\/em>\u00a0or by generating a set of\u00a0contour lines\u00a0<\/em>connecting points of equal data value (see figure 1).<\/p>\n

 <\/p>\n

As shown in the figure, the color gradient uses a continuum of colors to display the varying data across the surface.\u00a0\u00a0\u00a0The contour line display, on the other hand, utilizes lines to approximate the surface information.<\/p>\n

 <\/p>\n

Contour intervals can be filled with a color or can be displayed as a solid background (left-side of figure 1).\u00a0\u00a0\u00a0For displays of contour lines, the lines convey the information to the view as a set of lines of constant values.\u00a0\u00a0The contour intervals do not have constant values as they depict a range values.<\/p>\n

 <\/p>\n

 <\/p>\n

Figure 1.<\/em>\u00a0Surface displays using Contour Lines and Color Gradient techniques.<\/em><\/p>\n

 <\/p>\n

Figure 2 shows some of the visual differences in the two types of display approaches.\u00a0 The top portion of the figure depicts the positioning (Data Points<\/em>) and storage of the spatial information (Data Values<\/em>) that characterize surface data.\u00a0\u00a0A reference grid is established and the data points are positioned at specific\u00a0x,y-locations within the reference grid.\u00a0 This arrangement is stored as a matrix of data values in a computer with a single value corresponding to each grid cell.\u00a0\u00a0Both contour lines and continuous color gradient displays use this basic data structure to store surface data.<\/p>\n

 <\/p>\n

Differences between the two approaches arise, however, in how the data is processed to create a display and, ultimately, how the information is displayed and visualized.\u00a0 Typically, color gradient displays are generated by assigning a color from a continuous spectrum to each grid cell based on the magnitude of the data value (lower left side of figure 2).\u00a0 For example, the 1843 value is assigned light green along the continuous red to green color spectrum.\u00a0\u00a0The color-fill procedure is repeated for all of the other grid spaces resulting in a continuous color gradient over an entire mapped area thereby characterizing the relative magnitudes of the surface values.<\/p>\n

 <\/p>\n

Contour line displays, on the other hand, require specialized software to calculate lines of equal data value and, thus, a visualization of a surface (right side of figure 2).\u00a0 Since the stored data values in the matrix do not form adjoining data that are of a constant data value, they cannot be directly connected as contour lines.<\/p>\n

 <\/p>\n

 <\/p>\n

Figure 2.<\/em>\u00a0Procedures for creating surface displays using Color Gradient and Contour Line techniques.<\/em><\/p>\n

\u00a0<\/em><\/p>\n

For example, the derivation of a \u201c1900 value\u201d contour line involves interpolating the implied data point locations for the line from the centers of the actual gridded data as shown in the figure.\u00a0\u00a0The two data values of 1843 to 1943 in the extreme lower right portion of the reference grid bracket the 1900 value.\u00a0\u00a0Since the 1900 value is slightly closer to 1943, an\u00a0x,y-point is positioned slightly closer to the 1943 grid cell.\u00a0\u00a0Similarly, the data bracket for the 1990 and 1894 grid cells position a point very close to the 1894 cell.\u00a0\u00a0The procedure is repeated for all data values that bracket the desired 1900 contour line value, resulting in a series of derived\u00a0x,y-points that are then connected and displayed as a constant line that represents \u201c1900\u201d.<\/p>\n

 <\/p>\n

Figure 3 illustrates fundamental differences between the color gradient and contour line data visualization techniques.\u00a0\u00a0The left side of the figure identifies the data values defining the surface that contain the 1900 contour line discussed above.\u00a0\u00a0Note that these values are not constant and vary considerably from 1894 to 1994, so they cannot be directly connected to form a contour line at constant value.<\/p>\n

 <\/p>\n

While the \u201clumpy bumpy\u201d nature of a continuous data surface accurately portrays subtle differences in a surface, it has to be generalized to derive a contour line connecting points of constant data value.\u00a0\u00a0The right side of the figure portrays the contour line floating at a fixed height of 1900 (note: the pink \u201c1900 line\u201d is somewhat enhanced to see it better<\/em>).<\/p>\n

 <\/p>\n

 <\/p>\n

Figure 3.<\/em>\u00a0The\u00a0x,y-coordinates position surface information that can be displayed as continuously varying grid cells of relative z-value heights (color gradient display) or as derived lines of constant value (contour line display).<\/em><\/p>\n

\u00a0<\/em><\/p>\n

So what is the take-home from all this discussion?\u00a0\u00a0First, surface data (continuous) and contour lines (discontinuous) are not the same in either theory or practice but are fundamentally different spatial expressions.\u00a0\u00a0A continuous color gradient is effective when the underlying surface is a mathematical model or where the interval between physical measurements is very small.\u00a0\u00a0In these cases, the continuous color gradient is limited only by the resolution of the display device.<\/p>\n

 <\/p>\n

Contour lines, on the other hand, are effective to display coarse data.\u00a0\u00a0A contour line display creates a series of lines at constant values that somewhat \u201capproximate\u201d the gradient and provide the viewer with points of equal value that serve as references for the changing surface.<\/p>\n

 <\/p>\n

For example, if a mountain region of Colorado has elevation measurements taken every 100 feet, it would be impossible to display the elevation of each individual square inch.\u00a0\u00a0In this scenario, a series of contour lines describes the elevations in the region, with the understanding that the contour lines are neither precise nor capture small bumps or depressions that occur between the 100 foot measurements.\u00a0\u00a0If more detailed measurements are taken, contour lines can be drawn closer together, thus depicting greater detail of the underlying surface.<\/p>\n

 <\/p>\n

However, drawing contour lines close together is ultimately limited both visually and computationally.\u00a0\u00a0When contour lines are too close together, the lines themselves obscure the map, creating a useless blur.\u00a0In addition, the processing time required to compute and display a large number of contour lines renders the method impractical.<\/p>\n

 <\/p>\n

This illustrates the beauty of the continuous color gradient and contour line techniques\u2014they are not same, hence they have unique advantages and limitations that provide different data visualization footholds for different spatial applications.<\/p>\n

 <\/p>\n

 <\/p>\n

Setting a Place at the Table for Grid-based Data<\/em><\/strong><\/p>\n

(GeoWorld, June 2013)<\/strong><\/em><\/p>\n

(return\u00a0to top of Topic<\/a>)<\/strong><\/em><\/p>\n

 <\/p>\n

First the bad news\u2014spatial data structure, formats and storage schemes tend to be a deep dive into a quagmire of minutia from which few return.\u00a0\u00a0The good news is that is the realm of vector data and this month\u2019s journey is into the Land of Grids where things are quite regular and straight forward (for the most part).<\/p>\n

 <\/p>\n

At its conceptual foundation, grid-based map data is simply a matrix of values with a specified number of rows and columns (configuration<\/em>) that is referenced to a specific location on the earth\u2019s surface (geo-registered<\/em>).\u00a0\u00a0It\u2019s as if you tore out a spreadsheet and stretched and tugged until\u00a0its rectangular boxes formed squares (grid cell<\/em>) to cover a portion of the landscape.\u00a0\u00a0The value assigned to each grid summarizes the characteristic or condition of a spatial variable (map layer<\/em>) at that location.\u00a0\u00a0Additional aligned worksheets piled on top\u00a0(map stack<\/em>) describe other important geographically-based variables\u2014just a stack of numbers towering over a continuous set of regularly-spaced grid cells (analysis frame<\/em>) depicting what occurs at every location throughout an area (see author\u2019s note 1).<\/p>\n

 <\/p>\n

There are no differing types of map features, irregular shapes, missing puzzle pieces, complex topology about neighboring pieces, overlapping borders, serially linked files or other confounding concepts to wade through.\u00a0\u00a0It is just a set of organized grid cells containing information about every location.\u00a0\u00a0The only major difference in storage approaches is whether the matrices are stored as individual \u201cfiles\u201d or as a \u201ctable.\u201d<\/p>\n

 <\/p>\n

Figure 1 depicts the data organization for\u00a0Esri\u2019s\u00a0Grid ASCII export format as a single independent \u201cfile<\/em><\/strong>.\u201d\u00a0\u00a0The first six records in the file identify the geo-registration and configuration information (bottom left-side of the figure).\u00a0\u00a0Note that the geo-registration is identified in decimal degrees of latitude and longitude for the lower-left corner of the matrix.\u00a0\u00a0The configuration is identified as the number of columns and rows with a specified cell size that defines the matrix.\u00a0\u00a0In addition a \u201cno data\u201d value is set to indicate locations that will be excluded from processing.<\/p>\n

 <\/p>\n

 <\/p>\n

Figure 1.<\/em>\u00a0Storing grid maps as independent files is a simple and flexible approach that has been used for years.<\/em><\/p>\n

 <\/p>\n

The remainder of the grid map file contains blocks of values defining individual rows of the analysis frame that are read left to right, top to the bottom, as you would read a book.\u00a0\u00a0In the example, each row is stored as a block of 25 values that extends for 25 blocks for a total of 625 values.\u00a0\u00a0In a typical computer, the values are stored as double-precision (64 bit) floating point numbers which provides a range of approximately\u00a010\u2212308<\/sup>\u00a0to 10308<\/sup>\u00a0(really small to really big numbers) capable of handling just about any map-ematical\u00a0equation.<\/p>\n

 <\/p>\n

Each grid map is stored as a separate file and in the export format no legend or display information is provided.\u00a0\u00a0However in native ArcGIS grid format, this ancillary information is carried and individual grid maps can be clipped, adjusted and resized to form a consistent map stack for analysis stored as a group of individual flat files.<\/p>\n

 <\/p>\n

Figure 2 illustrates the alternative \u201ctable<\/em><\/strong>\u201d format for storing grid-based mapped data.\u00a0\u00a0In the table, each grid layer is stored as a separate column (field) with their common grid cell locations identified as an informational line (records).\u00a0\u00a0Each line in the table identifies all of the spatial information for a given grid location in a project area.\u00a0\u00a0In the example, strings of 25 values in a column define rows in the matrix read left to right.\u00a0\u00a0However in this example, the matrix origin and the geo-registration origin align and the matrix rows are read from the bottom row up to the top.\u00a0\u00a0The assignment of the matrix origin is primarily a reflection of a software developer\u2019s background\u2014lower-left for science types and upper-left for programmer types.\u00a0\u00a0Configuration information, such as 99.94 feet for the cell size, is stored as header lines while the geo-registration of a grid cell\u2019s centroid can be explicitly stored as shown or implicitly inferred by a map value\u2019s position (line#) in the string of numbers.\u00a0\u00a0When map-ematical\u00a0analysis creates a new map its string of map values is added to the table as a new column instead of being stored in an individual file.<\/p>\n

 <\/p>\n

\u00a0<\/em><\/p>\n

Figure 2.<\/em>\u00a0\u00a0Storing groups of commonly geo-referenced and configured grid maps in a table provides consistency and uniformity that facilitates map analysis and modeling.<\/em><\/p>\n

 <\/p>\n

The relative advantages\/disadvantages of the file versus table approach to grid map storage have been part of the GIS community debate for years.\u00a0\u00a0Most GIS systems use independent files because of their legacy, simplicity and flexibility.\u00a0\u00a0The geo-referencing and configuration of the stored grid maps can take any form and \u201con-the-fly\u201d processing can be applied to account for the consistency needed in map analysis.<\/p>\n

 <\/p>\n

However, with the increased influence of remotely sensed data with inter-laced spectral bands and television with interlaced signal processing, interest in the table approach has been on the rise.\u00a0\u00a0In addition, relational database capabilities have become much more powerful making \u201ctable maintenance\u201d faster, less problematic and more amenable to compression techniques.<\/p>\n

 <\/p>\n

Figure 3 shows the organizational structure for an early table-based system (MapCalc).\u00a0\u00a0The top portion of the table identifies the common geo-registration and configuration information shared by all of the map layers contained in the map set.\u00a0\u00a0In the example, there are 19 map layers listed whose values form an embedded table (bottom-left) as described in figure 2.<\/p>\n

 <\/p>\n

\u00a0<\/em><\/p>\n

Figure 3.<\/em>\u00a0Combining all three basic elements for storing grid maps (geo-registration\/configuration, map values and legend\/display) sets the stage for fully optimized relation database tables for the storage of grid-based data.\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0\u00a0<\/em><\/p>\n

 <\/p>\n

The right-side of the figure depicts the storage of additional legend and display information for two of the map layers.\u00a0\u00a0The three display considerations of\u00a0Data Type<\/em>\u00a0(continuous or discrete),\u00a0Display Type<\/em>(lattice or grid) and\u00a0Display Mode<\/em>\u00a0(2D or 3D) are identified and specific graphics settings, such as on\/off grid mesh, plot cube appearance, viewing direction, etc., are contained in the dense sets of parameter specifications.<\/p>\n

 <\/p>\n

Similarly,\u00a0Legend Labeling<\/em>\u00a0and associated parameters are stored, such as assigned color for each display category\/interval.\u00a0\u00a0When a map layer is retrieved for viewing its display and legend parameters are read and the appropriate graphical rendering of the map values in the matrix is produced.\u00a0\u00a0If a user changes a display, such as rotating and tilting a 3D plot, the corresponding parameters are updated.<\/p>\n

 <\/p>\n

When a new map is created through map analysis its map name, map type and derivation command is appended to the top portion of the table; its map values are appended as a new field to the table in the middle portion; and a temporary set of display and labeling parameters are generated depending on the nature of the processing and the data type generated.\u00a0\u00a0While all this seems overwhelming to a human, the \u201cmechanical\u201d reading and writing to a structured, self-documenting table is a piece-of-cake for a computer.<\/p>\n

 <\/p>\n

The MapCalc table format for grid-based data was created in the early 1990s.\u00a0\u00a0While it still serves small project areas very well (e.g., a few thousand acre farm or research site), modern relational database techniques would greatly improve the approach.<\/p>\n

 <\/p>\n

For the technically astute, the full dB design would consist of three related tables:\u00a0Map Definition Table<\/strong>\u00a0containing the\u00a0MapID\u00a0(\u201cprimary key\u201d for joining tables) and geo-registration\/configuration information, a\u00a0Grid Layer Definitions Table<\/strong>\u00a0containing the\u00a0MapID,\u00a0LayerID\u00a0and legend\/display information about the data layer, and a third\u00a0Cell Definition Table<\/strong>\u00a0containing\u00a0MapID\/LayerID\u00a0plus all of the map values in the matrix organized by the primary key of\u00a0MapID<\/em>\u00a0\u00e0\u00a0RowNumber<\/em>\u00a0\u00e0\u00a0ColumnNumber<\/em>\u00a0(see author\u2019s note 2).<\/p>\n

 <\/p>\n

Organizing the three basic elements for storing grid maps (geo-registration\/configuration, map values and legend\/display) into a set of linked tables and structuring the relationships between tables is a big part of the probable future of grid-based map analysis and modeling.\u00a0\u00a0Also, it sets the stage for a follow-on argument supporting the contention that latitude\/longitude referencing of grid-based data is the \u201cuniversal spatial database key\u201d (spatial stamp) that promises to join disparate data sets in much the same way as the date\/time stamp\u2014sort of the Rosetta Stone of geotechnology.<\/p>\n

_____________________________<\/p>\n

Author\u2019s Notes<\/em><\/strong>: 1) For a discussion of basic grid-based data concepts and terms, see the online book Beyond Mapping III, Topic 18, \u201cUnderstanding Grid-based Data\u201d posted at\u00a0www.innovativegis.com\/basis\/MapAnalysis\/<\/a>.\u00a0\u00a02) A generalized schema for a full grid-based database design is posted as an\u00a0appended discussion<\/a>.<\/em><\/p>\n

 
\nVtoR<\/em><\/strong>\u00a0and Back!\u00a0<\/em><\/strong>(Pronounced \u201cV-tore\u201d)<\/em><\/strong><\/p>\n

(GeoWorld, December 2012)<\/strong><\/em><\/p>\n

(return\u00a0to top of Topic<\/a>)<\/strong><\/em><\/p>\n

 <\/p>\n

The previous section described considerations and procedures for deriving contour lines from map surfaces.\u00a0\u00a0The discussion emphasized the similarities and differences between continuous\/gradient\u00a0map surfaces<\/em>\u00a0(raster) and discontinuous\/discrete\u00a0spatial objects<\/em>\u00a0identifying points, lines and polygons (vector).<\/p>\n

 <\/p>\n

Keep in mind that while raster treats geographic space as a continuum, it can store the three basic types of discrete map features\u2014a point as a single grid cell, a line as a series of connecting cells and a polygon as a set of contiguous cells.\u00a0\u00a0Similarly, vector can store generalizations of continuous map surfaces as a series of contour lines or contour interval polygons.<\/p>\n

 <\/p>\n

Paramount in raster data storage is the concept of a data layer in which all of the categories have to be mutually exclusive in space.\u00a0\u00a0That means that a given grid cell in a Water map layer, for example, cannot be simultaneously classified as a \u201cspring\u201d (e.g., category 1) and a \u201cwetland\u201d (e.g., category 2) at the same unless an additional category is specified for the joint condition of \u201cspring\u00a0and\u00a0wetland\u201d (e.g., category 12).<\/p>\n

 <\/p>\n

Another important consideration is that each grid cell is assumed to be the same condition\/characteristic throughout its entirety.\u00a0\u00a0For example, a 30m grid cell assigned as a spring does not infer a huge bubbling body of water in the shape of a square\u2014 rather it denotes a cell location that contains a spring somewhere within its interior.\u00a0\u00a0Similarly, a series of stream cells does not imply a 30m wide flow of water that moves in a saw-tooth fashion over the landscape\u2014 rather it identifies grid cells that contain a stream somewhere within their interiors.<\/p>\n

 <\/p>\n

While raster data tends to generalize\/lower the\u00a0spatial precision<\/em>\u00a0of map object placement and boundary, vector data tends to generalize\/lower the\u00a0thematic accuracy<\/em>\u00a0of classification.\u00a0\u00a0For example, the subtle differences in a map surface continuum of elevation have to be aggregated into broad contour interval ranges to store the data as a discrete polygon.\u00a0\u00a0Or, as in the case of contour lines, store a precise constant value but impart no information for the space between the lines.<\/p>\n

 <\/p>\n

Hence the rallying cry of \u201cVtoR\u00a0and back!\u201d by grid-based GIS modelers echoes from the walls of cubicles everywhere for converting between vector-based spatial objects and raster-based grids.\u00a0\u00a0This enables them to access the wealth of vector-based data, then utilize the thematic accuracy and analytical power of continuous grid-based data and upon completion, push the model results back to vector.<\/p>\n

 <\/p>\n

Figure 1 depicts the processing steps for a frequently used method of converting vector polygons to contiguous groupings of grid cells.\u00a0\u00a0It uses a point file of grid centers and \u201cpoint in polygon\u201d processing to assign a value representing a polygon\u2019s characteristic\/condition to every corresponding grid cell.\u00a0\u00a0In essence it is a statistical technique akin to \u201cdot grid\u201d sampling that has been used in aerial photo interpretation and manual map analysis for nearly 100 years.\u00a0\u00a0While fast and straight forward for converting polygon data to cells, it is unsuitable for point and line features.\u00a0\u00a0In addition, if large cell-sizes are used, small polygons can be missed.<\/p>\n

 <\/p>\n

 <\/p>\n

Figure 1.<\/em>\u00a0Basic procedure for centroid-based vector to raster conversion (VtoR).<\/em><\/p>\n

 <\/p>\n

The top portion of figure 2 depicts an alternative approach that directly \u201cburns\u201d the vector features into a grid analogous to a branding iron or\u00a0woodburning\u00a0tool.\u00a0\u00a0In the case of a point feature the cell containing its\u00a0x,y\u00a0coordinates is assigned a value representing the feature.\u00a0\u00a0The tricky part comes into play if there is more than one point feature falling into a grid cell.\u00a0\u00a0Then the user must specify whether to simply note \u201cone or more\u201d for a binary map or utilize a statistical procedure to summarize the point values (e.g, #count, sum, average, standard deviation, maximum, minimum, etc.).<\/p>\n

 <\/p>\n

For line features there are two primary strategies for\u00a0VtoR\u00a0conversion\u2014 fat and thin.\u00a0\u00a0Fat identifies every grid cell that is cut by a line segment, even if it is just a nick at the corner.\u00a0\u00a0Thin, on the other hand, identifies the smallest possible set of adjoining cells to characterize a line feature.\u00a0\u00a0While the \u201cthin\u201d option produces pleasing visualization of a line, it discards valuable information.\u00a0\u00a0For example, if the line feature was of a stream, then water law rights\/responsibilities are applicable to all of the cells with a \u201cstream running through it\u201d regardless of whether it is through the center or just in a corner.<\/p>\n

 <\/p>\n

 <\/p>\n

Figure 2.<\/em>\u00a0Basic procedures for direct calculation-based vector to raster conversion (VtoR) and the reverse (RtoV).<\/em><\/p>\n

\u00a0<\/em><\/p>\n

The upper-right inset in figure 2 illustrates the conversion of adjacent polygons.\u00a0\u00a0The \u201cfat\u201d edges containing the polygon boundary are identified and geometry is used to determine which polygon is mostly contained within each edge cell.\u00a0\u00a0All of the interior cells of a polygon are identified to finish the polygon conversion.\u00a0\u00a0On rare occasions the mixed cells containing boundary line are given a composite number identifying the two characteristics\/conditions.\u00a0\u00a0For example, if a soil map had two soil types of 1 and 2 occurring in the same cell, the value 12 might be stored for the boundary condition with 1 and 2 assigned to the respective interior cells.\u00a0\u00a0If the polygons are not abutting but scattered across the landscape, the boundary cells (either fat or thin) are assigned the same value as the interior cells with the exterior cells assigned a background value.<\/p>\n

 <\/p>\n

The lower portion of figure 2 depicts approaches for converting raster to vector.\u00a0\u00a0For single cell locations, the\u00a0x,y\u00a0coordinates of a cell\u2019s centroid is used to position a point feature and the cell value is used to populate the attribute table.\u00a0\u00a0On rare occasion where the cell value indicates the number of points contained in a cell, a random number generator is used to derive coordinates within the cell\u2019s geographic extent for the set of points.<\/p>\n

 <\/p>\n

For gridded line features, the\u00a0x,y\u00a0coordinates of the of the\u00a0centroids\u00a0(thin) are frequently used to define the line segments.\u00a0\u00a0Often the set of points are condensed as appropriate and a smoothing equation applied to eliminate the saw-tooth\u00a0jaggies.\u00a0\u00a0A radically different approach converts the sides of the cells into a thin polygon capturing the area of possible inclusion of the grid-based line\u2014 sort of a narrow corridor for the line.<\/p>\n

 <\/p>\n

For gridded polygons, the sides at the edges of abutting cells are used to define the feature\u2019s boundary line and its points are condensed and smoothed with topology of the adjoining polygons added.\u00a0\u00a0For isolated gridded polygons, the outside edges are commonly used for identifying the polygon\u2019s boundary line.<\/p>\n

 <\/p>\n

 <\/p>\n

Figure 3.<\/em>\u00a0Comparison of the basic approaches for identifying contour lines and contour interval boundaries.<\/em><\/p>\n

\u00a0<\/em><\/p>\n

Figure 3 deals with converting continuous map surfaces to discrete vector representations.\u00a0\u00a0A frequently used technique that generates true contour lines of constant value was described in detail in the previous section.\u00a0\u00a0The procedure involves identifying cell values that bracket a desired contour level, then interpolating the\u00a0x,y\u00a0coordinates for points between all of the cell value pairs and connecting the new points for a line of constant value.\u00a0\u00a0The black lines in figure 3 identify the set interpolated 200 foot contour lines for a data set with values from 500 to 2500.<\/p>\n

 <\/p>\n

Another commonly used technique involves slicing the data range into a desired number of contour intervals.\u00a0\u00a0For example, 2500-500 \/ 10 = 200 identifies the data step used in generating the data ranges of the contour intervals (color bands) shown in the figure.\u00a0\u00a0The first contour range from 500 to 700 \u201ccolor-fills\u201d with red all of the grid cells having values that fall within this range; orange for values 700 to 900; tan for values 900 to 1100; and so forth.\u00a0\u00a0The 3D surface shows the contour interval classification draped over the actual data values stored in the grid.<\/p>\n

 <\/p>\n

The added red lines in the enlarged inset identify the edges of the grid cell contour interval groupings.\u00a0\u00a0As you can see in the enlarged 3D plot there are numerous differing data values as the red border goes up and down with the data values along the\u00a0sawtooth\u00a0edge.\u00a0\u00a0As previously noted this boundary can be smoothed (dotted red) and used for the borders of the contour interval polygons generated in the\u00a0RtoVconversion.<\/p>\n

 <\/p>\n

The bottom line (pun intended) is that in many mapped data visualizations the boundary (border) outlining a contour interval is not the same as a contour line (line of constant value).\u00a0\u00a0That\u2019s the beauty of grid vs. vector data structures\u2014 they are not the same, and therefore, provide for subtly and sometimes radically different perspectives of the patterns and relationships in spatial information\u2014\u201cV-tore and back!\u201d is the rallying cry.<\/p>\n

 
\nMaps Are Numbers First, Pictures Later<\/strong><\/em><\/p>\n

(GeoWorld, August 2002, pg. 20-21)<\/strong><\/em><\/p>\n

(return\u00a0to top of Topic<\/a>)<\/strong><\/em><\/p>\n

 <\/p>\n

The unconventional view that \u201cmaps are numbers first, pictures later<\/em>\u201d forms the backbone for taking maps beyond mapping.\u00a0\u00a0Historically maps involved \u201cprecise placement of physical features for navigation.\u201d\u00a0\u00a0More recently, however, map analysis has become an important ingredient in how we perceive spatial relationships and form decisions.<\/p>\n

 <\/p>\n

Understanding that a digital map is first and foremost an organized set of numbers is fundamental to analyzing mapped data.\u00a0\u00a0But exactly what are the characteristics defining a digital map?\u00a0\u00a0What do the numbers mean?\u00a0\u00a0Are there different types of numbers?\u00a0\u00a0Does their organization affect what you can do with them?\u00a0\u00a0If you have seen one digital map have you seen them all?<\/p>\n

 <\/p>\n

In an introductory\u00a0GIS\u00a0course, concepts of \u201cvector\u201d and \u201craster\u201d seem to dominate discussion of what a digital map is, and isn\u2019t.\u00a0\u00a0Within this context, the location of map features are translated into computer form as organized sets of X,Y\u00a0coordinates (vector) or grid cells (raster).\u00a0\u00a0Considerable attention is given data structure considerations and their relative advantages in storage efficiency and system performance.<\/p>\n

 <\/p>\n

However this geo-centric view rarely explains the full nature of digital maps.\u00a0\u00a0For example consider the numbers themselves that comprise the X,Y\u00a0coordinates\u2014how does number type and size effect precision? A general feel for the precision ingrained in a \u201csingle precision floating point\u201d representation of Latitude\/Longitude in decimal degrees is*\u2026<\/p>\n

\u00a0<\/em><\/p>\n

1.31477E+08 ft = equatorial circumference of the earth<\/p>\n

1.31477E+08 ft \/ 360 degrees = 365214 ft\/degree length of one degree Longitude<\/p>\n

Single precision number carries six decimal places, so\u2014<\/p>\n

365214 ft\/degree * 0.000001= .365214 ft *12 =\u00a04.38257 inch\u00a0<\/strong>precision<\/p>\n

 <\/p>\n

Think if \u201cdouble-precision\u201d numbers (eleven decimal places) were used for storage\u2014you likely could distinguish a dust particle on the left from one on the right.<\/p>\n

 <\/p>\n

In analyzing mapped data, however, the characteristics of the attribute values are even more critical.\u00a0\u00a0While textual descriptions can be stored with map features they can only be used in geo-query.\u00a0\u00a0For example if you attempted to add Longbrake Lane to\u00a0Shortthrottle\u00a0Way all you would get is an error, as text-based descriptors preclude any of the mathematical\/statistical operations.<\/p>\n

 <\/p>\n

 <\/p>\n

Figure 1.<\/em>\u00a0Map values are characterized from two broad perspectives\u2014numeric and geographic\u2014then further refined by specific data types.<\/em><\/p>\n

 <\/p>\n

So what are the numerical characteristics of mapped data?\u00a0\u00a0Figure 1 lists the data types by two important categories\u2014numeric and geographic.\u00a0\u00a0You should have encountered the basic numeric data types in several classes since junior high school.\u00a0\u00a0Recall that\u00a0nominal<\/em><\/strong>\u00a0numbers do not imply ordering.\u00a0\u00a0A 3 isn\u2019t bigger, tastier or smellier than a 1, it\u2019s just not a 1.\u00a0\u00a0In the figure these data are schematically represented as scattered and independent pieces of wood.<\/p>\n

 <\/p>\n

Ordinal<\/em><\/strong>\u00a0numbers, on the other hand, do imply a definite ordering and can be conceptualized as a ladder, however with varying spaces between rungs.\u00a0\u00a0The numbers form a progression, such as smallest to largest, but there isn\u2019t a consistent step.\u00a0\u00a0For example you might rank different five different soil types by their relative crop productivity (1= worst to 5= best) but it doesn\u2019t mean that soil 5 is exactly five times more productive than soil 1.<\/p>\n

\u00a0<\/em><\/p>\n

When a constant step is applied,\u00a0interval<\/em><\/strong>\u00a0numbers result.\u00a0\u00a0For example, a 60o<\/sup>\u00a0Fahrenheit spring day is consistently\/incrementally warmer than a 30\u00a0oF<\/sup>\u00a0winter day.\u00a0\u00a0In this case one \u201cdegree\u201d forms a consistent reference step analogous to typical ladder with uniform spacing between rungs.<\/p>\n

 <\/p>\n

A\u00a0ratio<\/em><\/strong>\u00a0number introduces yet another condition\u2014an absolute reference\u2014that is analogous to a consistent footing or starting point for the ladder, analogous to zero degrees \u201cKelvin\u201d defined as when all molecular movement ceases.\u00a0\u00a0A final type of numeric data is termed \u201cbinary<\/em><\/strong>.\u201d\u00a0\u00a0In this instance the value range is constrained to just two states, such as forested\/non-forested or suitable\/not-suitable.<\/p>\n

 <\/p>\n

So what does all of this have to do with analyzing digital maps?\u00a0\u00a0The type of number dictates the variety of analytical procedures that can be applied.\u00a0\u00a0Nominal data, for example, do not support direct mathematical or statistical analysis.\u00a0\u00a0Ordinal data support only a limited set of statistical procedures, such as maximum and minimum.\u00a0\u00a0Interval and ratio data, on the other hand, support a full set mathematics and statistics.\u00a0\u00a0Binary maps support special mathematical operators, such as .AND. and .OR.<\/p>\n

 <\/p>\n

Even more interesting (this interesting, right?) are the geographic characteristics of the numbers.\u00a0\u00a0From this perspective there are two types of numbers.\u00a0\u00a0\u201cChoropleth<\/em><\/strong>\u201d numbers form sharp and unpredictable boundaries in space such as the values on a road or cover type map.\u00a0\u00a0\u201cIsopleth<\/em><\/strong>\u201d numbers, on the other hand, form continuous and often predictable gradients in geographic space, such as the values on an elevation or temperature surface.<\/p>\n

 <\/p>\n

Figure 2 puts it all together.\u00a0\u00a0Discrete<\/em><\/strong>\u00a0maps identify mapped data with independent numbers (nominal) forming sharp abrupt boundaries (choropleth), such as a covertype map.\u00a0\u00a0Continuous<\/em><\/strong>\u00a0maps contain a range of values (ratio) that form spatial gradients (isopleth), such as an elevation surface.\u00a0\u00a0This clean dichotomy is muddled by cross-over data such as speed limits (ratio) assigned to the features on a road map (choropleth).<\/p>\n

 <\/p>\n

Discrete maps are best handled in 2D form\u2014the 3D plot in the top-right inset is ridiculous and misleading because it implies numeric\/geographic relationships among the stored values.\u00a0\u00a0What isn\u2019t as obvious is that a 2D form of continuous data (lower-right inset) is equally as absurd.<\/p>\n

 <\/p>\n

Figure 2.<\/em>\u00a0\u00a0Discrete and Continuous map types combine the numeric and geographic characteristics of mapped data.<\/em><\/p>\n

\u00a0<\/em><\/p>\n

While a contour map might be as familiar and comfortable as a pair of old blue jeans, the generalized intervals treat the data as discrete (ordinal, choropleth).\u00a0\u00a0The artificially imposed sharp boundaries become the focus for visual analysis.\u00a0\u00a0Map-ematical<\/em>\u00a0analysis of the actual data, on the other hand, incorporates all of the detail contained in the numeric\/geographic patterns of the numbers …where the rubber meets the spatial analysis road.<\/p>\n

Normalizing Maps for Data Analysis<\/strong><\/em>\u00a0<\/strong><\/em>\u00a0\u00a0<\/strong><\/p>\n

(GeoWorld, September 2002, pg. 22-23)<\/strong><\/em><\/p>\n

(return\u00a0to top of Topic<\/a>)<\/strong><\/em><\/p>\n

 <\/p>\n

The last couple of sections have dealt with the numerical nature of digital maps.\u00a0\u00a0Two fundamental considerations remain\u2014data normalization and exchange.\u00a0\u00a0Normalization<\/em><\/strong>\u00a0involves standardizing a data set, usually for comparison among different types of data.\u00a0\u00a0In a sense, normalization techniques allow you to \u201ccompare apples and oranges\u201d using a standard \u201cmixed fruit scale of numbers.\u201d<\/p>\n

The most basic normalization procedure uses a \u201cgoal\u201d\u00a0to adjust map values.\u00a0\u00a0For example, a farmer might set a goal of 250 bushels per acre to be used in normalizing a yield map for corn.\u00a0\u00a0The equation,Norm_GOAL<\/strong>\u00a0= (mapValue\u00a0\/ 250) * 100<\/strong>,\u00a0derives the percentage of the goal achieved by each location in a field.\u00a0\u00a0In evaluating the equation, the computer substitutes a map value for a field location, completes the calculation, stores the result, and then repeats the process for all of the other map locations.<\/p>\n

 <\/p>\n

Figure 1 shows the results of goal normalization.\u00a0\u00a0Note the differences in the descriptive statistics between the original (top) and normalized data (bottom)\u2014a data range of 2.33 to 295 with an average of 158 bushels per acre for the original data versus .934 to 118 with an average of 63.3 percent for the normalized data.<\/p>\n

 <\/p>\n

However, the histogram and map patterns are identical (slight differences in the maps are an artifact of rounding the discrete display intervals).\u00a0\u00a0While the descriptive statistics are different, the relationships (patterns) in the normalized histogram and map are the same as the original data.<\/p>\n

 <\/p>\n

 <\/p>\n

Figure 1.<\/em>\u00a0\u00a0Comparison of original and goal normalized data.<\/em><\/p>\n

\u00a0<\/em><\/p>\n

That\u2019s an important point\u2014 both the numeric and spatial relationships in the data are preserved during normalization.\u00a0\u00a0In effect, normalization simply \u201crescales\u201d the values like changing from one set of units to another (e.g., switching from feet to meters doesn\u2019t change your height).\u00a0\u00a0The significance of the goal normalization is that the new scale allows comparison among different fields and even crop types based on their individual goals\u2014 the \u201cmixed fruit\u201d expression of apples and oranges.\u00a0\u00a0Same holds for normalizing environmental, business, health or any other kind of mapped data.<\/p>\n

 <\/p>\n

An alternative \u201c0-100\u201d\u00a0normalization forces a consistent range of values by spatially evaluating the equation\u00a0Norm_0-100 = (((mapValue\u00a0\u2013 min) * 100) \/ (max \u2013 min)) + 0<\/strong>.\u00a0\u00a0The result is a rescaling of the data to a range of 0 to 100 while retaining the same relative numeric and spatial patterns of the original data.\u00a0\u00a0While goal normalization benchmarks a standard value, the 0-100 procedure rescales the original data range to a fixed, standard range (see Author\u2019s note).<\/p>\n

 <\/p>\n

A third normalization procedure, standard normal variable\u00a0(SNV), uses yet another approach.\u00a0\u00a0It rescales the data based on its central tendency by applying the equation\u00a0Norm_SNV\u00a0= ((mapValue<\/strong>\u00a0– mean) \/ stdev) * 100<\/strong>.\u00a0\u00a0The result is a rescaling of the data to the percent variation from the average.\u00a0\u00a0Mapped data expressed in this form enables you to easily identify \u201cstatistically unusual\u201d areas\u2014 +100% locates areas that are one standard deviation above the typical value; -100% locates areas that are one standard deviation below.<\/p>\n

 <\/p>\n

Map normalization is often a forgotten step in the rush to make a map, but is critical to a host of subsequent analyses from visual map comparison to advanced data analysis.\u00a0\u00a0The ability to easily export the data in a universal format is just as critical.\u00a0\u00a0Instead of a \u201cdo-it-all\u201d\u00a0GIS\u00a0system,\u00a0data exchange<\/em><\/strong>\u00a0exports the mapped data in a format that is easily consumed and utilized by other software packages.<\/p>\n

 <\/p>\n

 <\/p>\n

Figure 2.<\/em>\u00a0The map values at each grid location form a single record in the exported table.<\/em><\/p>\n

 <\/p>\n

Figure 2 shows the process for grid-based data.\u00a0\u00a0Recall that a consistent analysis frame is used to organize the data into map layers.\u00a0\u00a0The map values at each cell location for selected layers are reformatted into a single record and stored in a standard export table that, in turn, can be imported into other data analysis software.<\/p>\n

 <\/p>\n

Figure 3 shows the agricultural data imported into the JMP statistical package (by SAS).\u00a0\u00a0Area (1) shows the histograms and descriptive statistics for the P, K and N map layers shown in figure 2.\u00a0\u00a0Area (2) is a \u201cspinning 3D plot\u201d of the data that you can rotate to graphically visualize relationships among the map layers.\u00a0\u00a0Area (3) shows the results of applying a multiple linear regression model to predict crop yield from the soil nutrient maps.\u00a0\u00a0These are but a few of the tools beyond mapping that are available through data exchange between\u00a0GIS\u00a0and traditional spreadsheet, database and statistical packages\u2014a perspective that integrates maps with other technologies.<\/p>\n

 <\/p>\n

 <\/p>\n

Figure 3.<\/em>\u00a0\u00a0Mapped data can be imported into standard statistical packages for further analysis.<\/em><\/p>\n

 <\/p>\n

Modern statistical packages like JMP \u201caren\u2019t your father\u2019s\u201d stat experience and are fully interactive with point-n-click graphical interfaces and wizards to guide appropriate analyses.\u00a0\u00a0The analytical tools, tables and displays provide a whole new view of traditional mapped data.\u00a0\u00a0While a map picture might be worth a thousand words, a gigabyte or so of digital map data is a whole revelation and foothold for site-specific decisions.
\n______________________<\/p>\n

Author\u2019s Note<\/strong>:\u00a0the generalized rescaling equation is\u2026<\/em><\/p>\n

Normalize a data set to a fixed range of\u00a0Rmin<\/em>\u00a0to\u00a0Rmax<\/em>= (((X-Dmin<\/em>) * (Rmax<\/em>\u00a0\u2013\u00a0Rmin<\/em>)) \/ (Dmax<\/em>\u00a0\u2013\u00a0Dmin<\/em>)) +\u00a0Rmin<\/em><\/p>\n

\u2026where\u00a0Rmin\u00a0and\u00a0Rmax\u00a0is the minimum and maximum values for the rescaled range,\u00a0Dmin\u00a0and\u00a0Dmax\u00a0is the minimum and maximum values for the input data and X is any value in the data set to be rescaled.<\/em><\/p>\n

 <\/p>\n

 <\/p>\n

Comparing Apples and Oranges<\/em><\/strong><\/p>\n

(GeoWorld, April, 2011)<\/strong><\/em><\/p>\n

(return\u00a0to top of Topic<\/a>)<\/strong><\/em><\/p>\n

 <\/p>\n

How many times have heard someone say \u201cyou can’t compare apples and oranges,<\/em>\u201d they are totally different things.\u00a0\u00a0But in GIS we see it all the time when a presenter projects two maps on a screen and uses a laser pointer to circle the \u201cobvious\u201d similarities and differences in the map displays.\u00a0\u00a0But what if there was a quantitative technique that would objectively compare each map location and report metrics describing the degree of similarity?\u00a0\u00a0\u2026for each map location?\u00a0\u00a0\u00a0\u2026for the entire map area?<\/p>\n

 <\/p>\n

Since maps have been \u201cnumbers first, pictures later\u201d for a couple of decades, you would think \u201cocular subjectivity\u201d would have been replaced by \u201cnumerical objectivity\u201d in map comparison a long time ago.<\/p>\n

 <\/p>\n

A few years back a couple of Beyond Mapping columns described grid-based map analysis techniques for comparing discrete and continuous maps (Statistically Compare Discrete Maps<\/em>, GeoWorld, July 2006 and\u00a0Statistically Compare Continuous Map Surfaces<\/em>, GeoWorld, September 2006).\u00a0\u00a0An even earlier column described procedures for normalizing mapped data (Normalizing Maps for Data Analysis<\/em>, GeoWorld,\u00a0September\u00a02002).\u00a0\u00a0Given these conceptual footholds I bet we can put the old \u201capples and oranges\u201d quandary to rest.<\/p>\n

 <\/p>\n

Consider the maps of Elevation and Slope shown in figure 1.\u00a0\u00a0I bet you eyes are quickly assessing the color patterns and \u201cseeing\u201d what you believe are strong spatial relationships\u2014dark greens in the NW and warmer tones in the middle NE.\u00a0\u00a0But how \u201cprecise and consistent\u201d can you be in describing the similarity?\u00a0\u00a0\u2026in delineating the similar areas?\u00a0\u00a0\u2026what would you do if you needed to assess a thousand of these patches?<\/p>\n

 <\/p>\n

 <\/p>\n

Figure 1.<\/em>\u00a0Elevation and Slope like apples and oranges cannot be directly compared.\u00a0<\/em><\/p>\n

 <\/p>\n

Obviously Elevation (measured in feet) and Slope (measured in percent)\u00a0are\u00a0not the same thing but they are sort of related.\u00a0\u00a0It wouldn\u2019t make sense to directly compare the map values; they are apples and oranges after all, so you can\u2019t compare them \u2026right?<\/p>\n

 <\/p>\n

That\u2019s where a \u201cmixed fruit\u201d scale comes in.\u00a0\u00a0As depicted in the top portion of figure 2, Elevation on the left and Slope on the right have unique raw data distributions that cannot be directly compared.<\/p>\n

 <\/p>\n

The middle portion of the figure illustrates using the\u00a0Standard Normal Variable (SNV)<\/em>\u00a0equation to \u201cnormalize\u201d the two maps to a common scale.\u00a0\u00a0This involves retrieving the map value at a grid location subtracting the Mean from it, then dividing by the Standard Deviation and multiplying by 100.\u00a0\u00a0The result is a rescaling of the data to the percent variation from each map\u2019s average value.<\/p>\n

 <\/p>\n

 <\/p>\n

Figure 2.<\/em>\u00a0Normalizing maps by the Standard Normal Variable (SNV) provides a foothold for comparing seemingly incomparable things.\u00a0\u00a0<\/em><\/p>\n

 <\/p>\n

The rescaled data are no longer apples and oranges but a mixed fruit salad that utilizes the standard normal curve as a common reference, where +100% locates areas that are one standard deviation above the typical value and -100% locates areas that are one standard deviation below.\u00a0\u00a0Because only scalar numbers are involved in the equation, neither the spatial nor the numeric relationships in the mapped data are altered\u2014like simply converting temperature readings from degrees Fahrenheit to Celsius.<\/p>\n

 <\/p>\n

 <\/p>\n

The middle\/lower portion of figure 2 describes the comparison of the two SNV normalized maps.\u00a0\u00a0The normalized values at a grid location on the two maps are retrieved then subtracted and the absolute value taken to \u201cmeasure\u201d how far apart the values are.\u00a0\u00a0For example, if Map1 had a value of -100 (one Stdev below the mean) and Map 2 had a value of +200 (two Stdev above the mean) for the same grid location, the absolute difference would be 300\u2014indicating very different information occurring at that location.<\/p>\n

 <\/p>\n

Figure 3 shows the SNV comparison for the Elevation and Slope maps.\u00a0\u00a0The red areas indicate locations where the map values are at dramatically different positions on the standard normal curve; blue tones indicate fairly similar positioning; and grey where the points are at the same position.\u00a0\u00a0The median of the absolute difference is 52 indicating that half of the map area has differences of about half a standard deviation or less.<\/p>\n

 <\/p>\n

 <\/p>\n

Figure 3.<\/em>\u00a0The absolute difference between SNV normalized maps generates a consistent scale of similarity that can be extended to different map variables and geographic locations.<\/em><\/p>\n

 <\/p>\n

In practice, SNV Comparison maps can be generated for the same variables at different locations or different variables at the same location.\u00a0\u00a0Since the standard normal curve is a \u201cstandard,\u201d the color ramp can fixed and the spatial pattern and overall similarities\/differences among apples, oranges, peaches, pears and pomegranates can be compared.\u00a0\u00a0All that is required is grid-based quantitative mapped data (no qualitative vector maps allowed).<\/p>\n

_____________________________<\/p>\n

Author\u2019s Note<\/strong>:\u00a0\u00a0For more information on map Normalization and Comparison see the online book Beyond Mapping III, posted at\u00a0www.innovativegis.com<\/a>, Topic 18,\u00a0Understanding Grid-based\u00a0Data\u00a0\u00a0and<\/strong>\u00a0Topic 16,\u00a0Characterizing Patterns and Relationships<\/strong>.<\/em><\/p>\n

 <\/p>\n

 <\/p>\n

Correlating Maps and a Numerical Mindset<\/em><\/strong><\/p>\n

(GeoWorld, May 2011)<\/strong><\/em><\/p>\n

(<\/strong><\/em>return<\/em><\/strong>\u00a0to top of Topic<\/em><\/strong><\/a>)<\/strong><\/em><\/p>\n

 <\/p>\n

The previous section discussed a technique for comparing maps, even if they were \u201capples and oranges.\u201d\u00a0\u00a0The approach normalized the two sets of mapped data using the\u00a0Standard Normal Variable<\/em>equation to translate the radically different maps into a common \u201cmixed-fruit\u201d scale for comparison.<\/p>\n

 <\/p>\n

Continuing with this statistical comparison theme (maps as numbers\u2014bah, humbug), one can consider a measure of\u00a0linear correlation<\/em>\u00a0between two continuous quantitative map surfaces.\u00a0\u00a0A general dictionary definition of the term correlation is \u201cmutual relation of two or more things\u201d that is expanded to its statistical meaning as \u201cthe extent of correspondence between the ordering of two variables; the degree to which two or more measurements on the\u00a0same group of elements show\u00a0a tendency to vary together.\u201d<\/p>\n

 <\/p>\n

So what does that have to with mapping?\u00a0\u00a0\u2026maps are just colorful images that tell us what is where, right?\u00a0\u00a0No, today\u2019s maps actually are organized sets of number first, pictures later.\u00a0\u00a0And numbers (lots of numbers) are right down statistic\u2019s alley.\u00a0\u00a0So while we are severely challenged to \u201cvisually assess\u201d the correlation among maps, spatial statistics, like a tireless puppy, eagerly awaits the opportunity.<\/p>\n

 <\/p>\n

Recall from basic statistics, that the\u00a0Correlation Coefficient (<\/em>r)<\/em>\u00a0assesses the linear relationship between two variables, such that its value falls between -1<\u00a0r<\/em>\u00a0<\u00a0+1.\u00a0\u00a0Its sign indicates the direction of the relationship and its magnitude indicates the strength.\u00a0\u00a0If two variables have a strong positive correlation,\u00a0r<\/em>\u00a0is close to +1 meaning that as values for x increase, values for y increase proportionally.\u00a0\u00a0If a strong negative correlation exits,\u00a0r<\/em>\u00a0is close to -1 and as x increases, the values for y decrease.<\/p>\n

 <\/p>\n

A perfect correlation of +1 or -1 only occurs when all of the data points lie on a straight line.\u00a0\u00a0If there is no linear correlation or a weak correlation,\u00a0r<\/em>\u00a0is close to 0 meaning that there is a random or non-linear relationship between the variables.\u00a0\u00a0A correlation that is greater than 0.8 is generally described as\u00a0strong<\/em>, whereas a correlation of less than 0.5 is described as\u00a0weak<\/em>.<\/p>\n

 <\/p>\n

The\u00a0Coefficient of Determination (<\/em>r<\/em>\u00a0<\/em>2<\/sup>)\u00a0<\/em>is a related statistic that summarizes the ratio of the explained variation to the total variation.\u00a0\u00a0It represents the percent of the data that is the closest to the line of best fit and varies from 0\u00a0<\u00a0\u00a0r<\/em>\u00a0<\/em>2<\/sup>\u00a0<\u00a01.\u00a0 It is most often used as a measure of how certain one can be in making predictions from the linear relationship (regression equation) between variables.<\/p>\n

 <\/p>\n

With that quickie stat review, now consider the left side of figure 1 that calculates the correlation between Elevation and Slope maps discussed in the last section.\u00a0\u00a0The gridded maps provide an ideal format for identifying pairs of values for the analysis.\u00a0\u00a0In this case, the 625\u00a0Xelev<\/sub>\u00a0and\u00a0Yslope<\/sub>\u00a0values form one large table that is evaluated using the correlation equation shown.<\/p>\n

 <\/p>\n

The spatially aggregated result is\u00a0r<\/em>\u00a0= +0.432, suggesting a somewhat weak overall positive linear correlation between the two map surfaces.\u00a0\u00a0This translates to\u00a0r<\/em>\u00a0<\/em>2<\/sup>\u00a0= 0.187, which means that only 19% of the total variation in\u00a0y<\/em>\u00a0can be explained by the linear relationship between\u00a0Xelev<\/sub>\u00a0and\u00a0Yslope<\/sub>.\u00a0 The other 81% of the total variation in\u00a0y<\/em>\u00a0remains unexplained which suggests that the overall linear relationship is poor and does not support useful regression predictions.<\/p>\n

 <\/p>\n

The right side of figure 1 uses a spatially disaggregated approach that assesses spatially localized correlation.\u00a0\u00a0The technique uses a roving window that identifies the 81 value pairs of\u00a0Xelev<\/sub>\u00a0and\u00a0Yslope<\/sub>\u00a0within a 5-cell reach, then evaluates the equation and assigns the computed\u00a0r<\/em>\u00a0value to the center position of the window.\u00a0\u00a0The process is repeated for each of the 625 grid locations.<\/p>\n

 <\/p>\n

 <\/p>\n

Figure 1.<\/em>\u00a0Correlation between two maps can be evaluated for an overall metric (left side) or for a continuous set of spatially localized metrics (right side).\u00a0<\/em><\/p>\n

 <\/p>\n

For example, the spatially localized result at column 17, row 10 is\u00a0r<\/em>\u00a0= +0.562 suggesting a fairly strong positive linear correlation between the two maps in this portion of the project area.\u00a0\u00a0This translates to\u00a0r<\/em>\u00a0<\/em>2<\/sup>= 0.316, which means that nearly a third of the total variation in\u00a0y<\/em>\u00a0can be explained by the linear relationship between\u00a0Xelev<\/sub>\u00a0and\u00a0Yslope<\/sub>.<\/p>\n

 <\/p>\n

Figure 2 depicts the geographic distributions of the spatially aggregated correlation (top) and the spatially localized correlation (bottom).\u00a0\u00a0The overall correlation statistic assumes that the\u00a0r<\/em>\u00a0= +0.432 is uniformly distributed thereby forming a flat plane.<\/p>\n

Spatially localized correlation, on the other hand, forms a continuous quantitative map surface.\u00a0\u00a0The correlation surrounding column 17, row 10 is\u00a0r<\/em>\u00a0= +0.562 but the northwest portion has significantly higher positive correlations (red with a maximum of +0.971) and the central portion has strong negative correlations (green with a minimum of -0.568).\u00a0\u00a0The overall correlation primarily occurs in the southeastern portion (brown); not everywhere.<\/p>\n

 <\/p>\n

The bottom-line of spatial statistics is that it provides spatial specificity for many traditional statistics, as well as insight into spatial relationships and patterns that are lost in spatially aggregated of non-spatial statistics.\u00a0\u00a0In this case, it suggests that the red and green areas have strong footholds for regression analysis but the mapped data needs to be segmented and separate regression equations developed.\u00a0\u00a0Ideally, the segmentation can be based on existing geographic conditions identified through additional grid-based map analysis.<\/p>\n

 <\/p>\n

 <\/p>\n

Figure 2.<\/em>\u00a0Spatially aggregated correlation provides no spatial information (top), while spatially localized correlation \u201cmaps\u201d the direction and strength of the mutual relationship between two map variables (bottom)\u00a0<\/em><\/p>\n

 <\/p>\n

It is this \u201cnumerical mindset of maps\u201d that is catapulting GIS beyond conventional mapping and traditional statistics beyond long-established spatially aggregated metrics\u2014the joint analysis of geographic and numeric distributions inherent in digital maps provide the springboard.<\/p>\n

_____________________________<\/p>\n

Author\u2019s Note<\/strong>:\u00a0\u00a0For more information, see the online book Beyond Mapping III, posted at\u00a0<\/em>www.innovativegis.com<\/em><\/a>, Topic 16,\u00a0Characterizing Patterns and Relationships<\/strong>\u00a0and Topic 28,\u00a0Spatial Data Mining in Geo-Business<\/strong>.<\/em><\/p>\n

 <\/p>\n

 <\/p>\n

Multiple Methods Help Organize Raster Data<\/em><\/strong><\/p>\n

(GeoWorld, April 2003, pg. 22-23)<\/strong><\/em><\/p>\n

(return\u00a0to top of Topic<\/a>)<\/strong><\/em><\/p>\n

 <\/p>\n

Map features in a vector-based mapping system identify discrete, irregular spatial objects with sharp abrupt boundaries.\u00a0\u00a0Other data types\u2014raster images, pseudo grids and raster grids\u2014treat space in entirely different manner forming a spatially continuous data structure.<\/p>\n

 <\/p>\n

For example, a\u00a0raster image<\/em><\/strong>\u00a0is composed of thousands of \u201cpixels\u201d (picture elements) that are analogous to the dots on a computer screen.\u00a0\u00a0In a geo-registered B&W aerial photo, the dots are assigned a grayscale color from black (no reflected light) to white (lots of reflected light).\u00a0\u00a0The eye interprets the patterns of gray as forming the forests, fields, buildings and roads of the actual landscape.\u00a0\u00a0While raster maps contain tremendous amounts of information that are easily \u201cseen,\u201d the data values simply reference color codes that afford some quantitative analysis but are far too limited for the full suite of map analysis operations.<\/p>\n

 <\/p>\n

 <\/p>\n

Figure 1.<\/em>\u00a0\u00a0A vector-based system can store continuous geographic space as a pseudo-grid.<\/em><\/p>\n

 <\/p>\n

Pseudo grids and raster grids are similar to raster images as they treat geographic space as a continuum.\u00a0\u00a0However, the organization and nature of the data are radically different.<\/p>\n

 <\/p>\n

A\u00a0pseudo grid<\/em><\/strong>\u00a0is formed by a series of uniform, square polygons covering an analysis area (figure 1).\u00a0\u00a0In practice, each grid element is treated as a separate polygon\u2014it\u2019s just that every polygon is the same shape\/size and they all\u00a0abut\u00a0each other\u2014with spatial and attribute tables defining the set of little polygons.\u00a0\u00a0For example, in the upper-right portion of the figure a set of discrete point measurements are stored as twelve individual \u201cpolygonal cells.\u201d\u00a0\u00a0The interpolated surface from the point data (lower-right) is stored as 625 contiguous cells.<\/p>\n

 <\/p>\n

While pseudo grids store full numeric data in their attribute tables and are subject to the same vector analysis operations, the explicit organization of the data is both inefficient and too limited for advanced spatial analysis as each polygonal cell is treated as an independent spatial object.\u00a0\u00a0A\u00a0raster grid<\/em><\/strong>, on the other hand, organizes the data as a listing of map values like you read a book\u2014left to right (columns), top to bottom (rows).\u00a0\u00a0This implicit configuration identifies a grid cell\u2019s location by simply referencing its position in the list of all map values.<\/p>\n

 <\/p>\n

In practice, the list of map values is read into a matrix with the appropriate number of columns and rows of an analysis frame superimposed over an area of interest.\u00a0\u00a0Geo-registration of the analysis frame requires an X,Y\u00a0coordinate for one of the grid corners and the length of a side of a cell.\u00a0\u00a0To establish the geographic extent of the frame the computer simply starts at the reference location and calculates the total X, Y length by multiplying the number of columns\/rows times the cell size.<\/p>\n

 <\/p>\n

Figure 2.<\/em>\u00a0\u00a0A grid-based system stores a long list of map values that are implicitly linked to an analysis frame superimposed over an area.<\/em><\/p>\n

 <\/p>\n

Figure 2 shows a 100 column by 100 row analysis frame geo-registered over a subdued vector backdrop.\u00a0\u00a0The list of map values is read into the 100×100 matrix with their column\/row positions corresponding to their geographic locations.\u00a0\u00a0For example, the maximum map value of 92 (customers within a quarter of a mile) is positioned at column 67, row 71 in the matrix\u2014 the 7,167th<\/sup>\u00a0value in the list ((71 * 100) + 67 = 7167).\u00a0\u00a0The 3D plot of the surface shows the spatial distribution of the stored values by \u201cpushing\u201d up each of the 10,000 cells to its relative height.<\/p>\n

 <\/p>\n

In a grid-based dataset, the matrices containing the map values automatically align as each value list corresponds to the same analysis frame (#columns, # rows, cell size and geo-reference point).\u00a0\u00a0As depicted on the left side of figure 3, this organization enables the computer to identify any or all of the data for a particular location by simply accessing the values for a given column\/row position (spatial coincidence used in point-by-point overlay operations).<\/p>\n

 <\/p>\n

Similarly, the immediate or extended neighborhood around a point can be readily accessed by selecting the values at neighboring column\/row positions (zonal groupings used in region-wide overlay operations).\u00a0\u00a0The relative proximity of one location to any other location is calculated by considering the respective column\/row positions of two or more locations (proximal relationships used in distance and connectivity operations).<\/p>\n

 <\/p>\n

There are two fundamental approaches in storing grid-based data\u2014individual \u201cflat\u201d files and \u201cmultiple-grid\u201d tables (right side of figure 3).\u00a0\u00a0Flat files store map values as one long list, most often starting with the upper-left cell, then sequenced left to right along rows ordered from top to bottom.\u00a0\u00a0Multi-grid tables have a similar ordering of values but contain the data for many maps as separate field in a single table.<\/p>\n

 <\/p>\n

Figure 3.<\/em>\u00a0\u00a0A map stack of individual grid layers can be stored as separate files or in a multi-grid table.<\/em><\/p>\n

 <\/p>\n

Generally speaking the flat file organization is best for applications that create and delete a lot of maps during processing as table maintenance can affect performance.\u00a0\u00a0However, a multi-gird table structure has inherent efficiencies useful in relatively non-dynamic applications.\u00a0\u00a0In either case, the implicit ordering of the grid cells over continuous geographic space provides the topological structure required for advanced map analysis.
\n_________________<\/p>\n

Author’s Note<\/em><\/strong>:\u00a0\u00a0\u00a0Let me apologize in advance to the \u201cgeode-ists\u201d readership\u2014yep it\u2019s a lot more complex than these simple equations but the order of magnitude ought to be about right \u2026thanks to Ken Burgess, VP R&D, Red Hen Systems for getting me this far.
\n<\/em><\/p>\n

Use Mapping \u201cArt\u201d to Visualize Values<\/em><\/strong>\u00a0<\/strong><\/em>\u00a0\u00a0<\/strong>
\n<\/strong>(GeoWorld, June 2003, pg. 20-21)<\/strong><\/em><\/p>\n

(return\u00a0to top of Topic<\/a>)<\/strong><\/em><\/p>\n

 <\/p>\n

The digital map has revolutionized how we collect, store and perceive mapped data.\u00a0\u00a0Our paper map legacy has well-established cartographic standards for viewing these data.\u00a0\u00a0However, in many respects the display of mapped data is a very different beast.<\/p>\n

 <\/p>\n

In a\u00a0GIS, map display is controlled by a set of user-defined tools\u2014not the cartographer\/publisher team that produced hardcopy maps just a couple of decades ago.\u00a0\u00a0The upside is a tremendous amount of flexibility in customizing map display; the downside is a tremendous amount of flexibility in customizing map display.<\/p>\n

 <\/p>\n

The display tools are both a boon and a bane as they require minimal skills to use but considerable thought and experience to use correctly.\u00a0\u00a0The interplay among map projection, scale, resolution, shading and symbols can dramatically change a map\u2019s appearance and thereby the information it graphically conveys to the viewer.<\/p>\n

 <\/p>\n

While this is true for the points, lines and areas comprising traditional maps, the potential for cartographic effects are even more pronounced for contour maps of surface data.\u00a0\u00a0For example, consider the mapped data of phosphorous levels in a farmer\u2019s field shown in figure 1.\u00a0\u00a0The inset on the left is a histogram of the 3288 grid values over the field ranging from 4.2 to 53.2 parts per million (ppm).\u00a0\u00a0The table describes the individual data ranges used to generalize the data into seven contour intervals.<\/p>\n

 <\/p>\n

Figure 1.<\/em>\u00a0\u00a0An Equal Ranges contour map of surface data.<\/em><\/p>\n

 <\/p>\n

In this case, the contour intervals were calculated by dividing the data range into seven\u00a0Equal Ranges<\/em><\/strong>.\u00a0\u00a0The procedure involves: 1] calculating the interval step as\u00a0(max \u2013 min) \/ #intervals= (53.2 \u2013 4.2) \/ 7 = 7.0 step<\/em>, 2] assigning the first contour interval\u2019s breakpoint as\u00a0min + step = 4.2 + 7.0 = 11.2<\/em>, 3] assigning the second contour interval\u2019s breakpoint as\u00a0previous breakpoint + step = 11.2 + 7.0 = 18.2<\/em>, 4] repeating the breakpoint calculations for the remaining contour intervals (25.2, 32.2, 39.2, 46.2,\u00a053.2<\/em>).<\/p>\n

 <\/p>\n

The equally spaced red bars in the plot show the contour interval breakpoints superimposed on the histogram.\u00a0\u00a0Since the data distribution is skewed toward lower values, significantly more map locations are displayed in red tones\u2014 41 + 44 = 85% of the map area assigned to contour intervals one and two.\u00a0\u00a0The 2D and 3D displays on the right side of figure 1 shows the results of \u201cequal ranges contouring\u201d of the mapped data.<\/p>\n

 <\/p>\n

\\<\/p>\n

Figure 2.<\/em>\u00a0\u00a0Equal Count and +\/- 1 Standard Deviation contour maps.<\/em><\/p>\n

 <\/p>\n

Figure 2 shows the results of applying other strategies for contouring the same data.\u00a0\u00a0The top inset uses\u00a0Equal Count<\/em><\/strong>\u00a0calculations to divide the data range into intervals that represent equal amounts of the total map area.\u00a0\u00a0\u00a0This procedure first calculates the interval step as\u00a0total #cells \/ #intervals= 3288 \/ 7 = 470 cells<\/em>\u00a0then starts at the minimum map value and assigns progressively larger map values until 470 cells have been assigned.\u00a0\u00a0The calculations are repeated to successively capture groups of approximately 470 cells of increasing values, or about 14.3 percent of the total map area.<\/p>\n

 <\/p>\n

Notice the unequal spacing of the breakpoints (red bars) in the histogram plot for the equal count contours.\u00a0\u00a0Sometimes a contour interval only needs a small data step to capture enough cells (e.g., peaks in the histogram); whereas others require significantly larger steps (flatter portions of the histogram).\u00a0\u00a0The result is a more complex contour map with fairly equal amounts of colored polygons.<\/p>\n

 <\/p>\n

The bottom inset in figure 2 depicts yet another procedure for assigning contour breaks.\u00a0\u00a0This approach divides the data into groups based on the calculated mean and\u00a0Standard Deviation<\/em><\/strong>.\u00a0\u00a0The standard deviation is added to the mean to identify the breakpoint for the upper contour interval (contour seven = 13.4 + 5.21= 18.61 to max<\/em>) and subtracted to set the lower interval (contour one = 13.4 – 5.21= 8.19 to min)<\/em>.<\/p>\n

 <\/p>\n

In statistical terms the low and high contours are termed the \u201ctails\u201d of the distribution and locate data values that are outside the bulk of the data\u2014 sort of unusually lower and higher values than you normally might expect.\u00a0\u00a0In the 2D and 3D map displays on the right side of the figure these locations are shown as blue and pink areas.<\/p>\n

 <\/p>\n

The other five contour intervals are assigned by forming equal ranges within the lower and upper contours (18.61 – 8.19 = 10.42 \/ 5 = 2.1 interval step<\/em>) and assigned colors red through green with a yellow inflection point.\u00a0\u00a0The result is a map display that highlights areas of unusually low and high values and shows the bulk of the data as gradient of increasing values.<\/p>\n

 <\/p>\n

Figure 3.<\/em>\u00a0Comparison of different 2D contour displays.<\/em><\/p>\n

 <\/p>\n

The bottom line is that the same surface data generated dramatically different 2D contour maps (figure 3).\u00a0\u00a0All three displays contain seven intervals but the methods of assigning the breakpoints to the contours employ radically different approaches.\u00a0\u00a0So which one is right?\u00a0\u00a0Actually all three are right, they just reflect different perspectives of the same data distribution \u2026a bit of the art in the \u201cart and science\u201d of\u00a0GIS.<\/p>\n

What\u2019s\u00a0<\/em><\/strong>Missing<\/em><\/strong>\u00a0in Mapping?<\/em>\u00a0<\/em><\/strong>\u00a0\u00a0<\/strong>
\n<\/strong>(GeoWorld, April 2009)<\/strong><\/em><\/p>\n

(return\u00a0to top of Topic<\/a>)<\/strong><\/em><\/p>\n

 <\/p>\n

We have known the purpose of maps for thousands of years\u2014precise placement of physical features for navigation<\/em>.\u00a0\u00a0Without them historical heroes might have sailed off the edge of the earth, lost their way along the Silk Route or missed the turn to Waterloo.\u00a0\u00a0Or more recently, you might have lost life and limb hiking the Devil\u2019s Backbone or dug up the telephone trunk line in your neighborhood.<\/p>\n

 <\/p>\n

Maps have always told us where we are, and as best possible, what is there.\u00a0\u00a0For the most part, the historical focus of mapping has been successfully automated.\u00a0\u00a0It is the \u201cWhat\u201d component of mapping that has expanded exponentially through derived and modeled maps that characterize geographic space in entirely new ways.\u00a0\u00a0Digital maps form the building blocks and map-ematical\u00a0tools provide the cement in constructing more accurate maps, as well as wholly new spatial expressions.<\/p>\n

 <\/p>\n

For example, consider the left-side of figure 1 that shows both discrete (Contour) and continuous (Surface) renderings of the Elevation gradient for a project area.\u00a0\u00a0Not so long ago the only practical way of mapping a continuous surface was to force the unremitting undulations into a set of polygons defined by a progression of contour interval bands.\u00a0\u00a0The descriptor for each of the polygons is an interval range, such as 500-700 feet for the lowest contour band in the figure.\u00a0\u00a0If you had to assign a single attribute value to the interval, it likely would be the middle of the range (600).<\/p>\n

 <\/p>\n

But does that really make sense?\u00a0\u00a0Wouldn\u2019t some sort of a statistical summary of the actual elevations occurring within a contour polygon be a more appropriate representation?\u00a0\u00a0The average of all of the values within the contour interval would seem to better characterize the \u201ctypical elevation.\u201d\u00a0\u00a0\u00a0For the 500-700 foot interval in the example, the average is only 531.4 feet which is considerably less than the assumed 600 foot midpoint of the range.<\/p>\n

 <\/p>\n

Our paper map legacy has conditioned us to the traditional contour map\u2019s interpretation of fixed interval steps but that really muddles the \u201cWhat\u201d information.\u00a0\u00a0The right side of figure 1 tells a different story.\u00a0\u00a0In this case the polygons represent seven Districts that are oriented every-which-way and have minimal to no relationship to the elevation surface.\u00a0\u00a0It\u2019s sort of like a\u00a0surrealist\u00a0Salvador\u00a0Dali painting with the Districts melted onto the Elevation surface indentifying the coincident elevation values.\u00a0\u00a0Note that with the exception of District #1, there are numerous different elevations occurring within each district\u2019s boundary.<\/p>\n

 <\/p>\n

Figure 1.<\/em>\u00a0Visual assessment of the spatial coincidence between a continuous Elevation surface and a discrete map of Districts.<\/em><\/p>\n

 <\/p>\n

One summary attribute would be simply noting the\u00a0Minimum\/Maximum<\/em>\u00a0values in a manner analogous to contour intervals.\u00a0\u00a0Another more appropriate metric would be to assign the\u00a0Median<\/em>\u00a0of the values identifying the middle value for a metric that divides the\u00a0total frequency into two halves.\u00a0\u00a0However the most commonly used statistic for characterizing the \u201ctypical condition\u201d is a simple\u00a0Average<\/em>\u00a0of all the elevation numbers occurring within each district.\u00a0\u00a0The \u201cThematic Mapping\u201d procedure of assigning a single value\/color to characterize individual map features (lower left-side of figure 2) is fundamental to many GIS applications, letting decision-makers \u201csee\u201d the spatial pattern of the data.<\/p>\n

 <\/p>\n

The discrete pattern, however, is a generalization of the actual data that reduces the continuous surface to a series of stepped mesas (right-side of figure 2).\u00a0\u00a0In some instances, such as District #1 where all of the values are 500, the summary to a typical value is right on.\u00a0\u00a0On the other hand, the summaries for other districts contain sets of radically differing values suggesting that the \u201ctypical value\u201d might not be very typical.\u00a0\u00a0For example, the data in District #2 ranges from 500 to 2499 (valley floor to the top of the mountain) and the average of 1539 is hardly anywhere, and certainly not a value supporting good decision-making.<\/p>\n

 <\/p>\n

So what\u2019s the alternative?\u00a0\u00a0What\u2019s better at depicting the \u201cWhat component\u201d in thematic mapping?\u00a0\u00a0Simply stated, an honest map is better.\u00a0\u00a0Basic statistics uses the\u00a0Standard Deviation (StDev)<\/em>\u00a0to characterize the amount dispersion in a data set and the\u00a0Coefficient of Variation (Coffvar= [StDev\/Average] *100)<\/em>\u00a0as a relative index.\u00a0\u00a0Generally speaking, an increasing\u00a0Coffvar\u00a0index indicates increasing data dispersion and a less \u201ctypical\u201d Average\u2014 0 to 10, not much data dispersion; 10-20, a fair amount; 20-30, a whole lot; and >30, probably too much dispersion to be useful (apologies to statisticians among us for the simplified treatise and the generalized but practical rule of thumb).\u00a0\u00a0In the example, the thematic mapping results are good for Districts #1, #3 and #6, but marginal for Districts #5, #7 and #4 and dysfunctional for District #2, as its average is hardly anywhere.<\/p>\n

 <\/p>\n

So what\u2019s the bottom line?\u00a0\u00a0What\u2019s missing in traditional thematic mapping?\u00a0\u00a0I submit that a reasonable and effective measure of a map\u2019s accuracy has been missing (see Author\u2019s Notes).\u00a0\u00a0In the paper map world one can simply include the\u00a0Coffvar\u00a0index within the label as shown in left-side of figure 2.\u00a0\u00a0In the digital map world a host of additional mechanisms can be used to report the dispersion, such as mouse-over pop-ups of short summary tables like the ones on the right-side of figure 2.<\/p>\n

 <\/p>\n

 <\/p>\n

Figure 2.<\/em>\u00a0Characterizing the average Elevation for each District and reporting how typical the typical Elevation value is.<\/em><\/p>\n

 <\/p>\n

Another possibility could be to use the brightness gun to track the\u00a0Coffvar\u2014with the display color indicating the average value and the relative brightness becoming more washed out toward white for higherCoffvar\u00a0indices.\u00a0\u00a0The effect could be toggled on\/off or animated to cycle so the user sees the assumed \u201ctypical\u201d condition, then the\u00a0Coffvar\u00a0rendering of how typical the typical really is.\u00a0\u00a0For areas with aCoffvar\u00a0greater than 30, the rendering would go to white.\u00a0\u00a0Now that\u2019s an honest map that shows the best guess of typical value then a visual assessment of how typical the typical is\u2014sort of a warning that use of shaky information may be hazardous to\u00a0you\u00a0professional health.<\/p>\n

 <\/p>\n

As Geotechnology moves beyond our historical focus on\u00a0\u201cprecise placement of physical features for navigation\u201d\u00a0<\/em>the ability to discern the good portions of a map from less useful areas is critical.\u00a0\u00a0While few readers are interested in characterizing average elevation for districts, the increasing wealth of mapped data surfaces\u00a0is apparent\u2014 from a realtor wanting to view average home prices for communities, to a natural resource manager wanting to see relative habitat suitability for various management units, to a retailer wanting to determine the average probability of product sales by zip codes, to policemen wanting to appraise typical levels of crime in neighborhoods, or to public health officials wanting to assess air pollution levels for jurisdictions within a county.\u00a0\u00a0It is important that they all \u201csee\u201d the relative accuracy of the \u201cWhat component\u201d of the results in addition to the assumed average condition.<\/p>\n

_____________________________<\/p>\n

 <\/p>\n

Author\u2019s Notes<\/strong>:\u00a0see\u00a0http:\/\/www.innovativegis.com\/basis\/MapAnalysis\/Topic18\/Topic18.htm<\/a>\u00a0for an online discussion of the related concepts, structures and considerations of grid-based mapped data.\u00a0\u00a0Discussion of the differences between map Precision and Accuracy is at\u00a0http:\/\/www.innovativegis.com\/basis\/MapAnalysis\/MA_Intro\/MA_Intro.htm<\/a>, \u201cDetermining Exactly\u00a0Where\u00a0is What.\u201d<\/em><\/p>\n

 <\/p>\n

(return\u00a0to top of Topic<\/a>)<\/strong><\/em><\/p>\n

 <\/p>\n","protected":false},"excerpt":{"rendered":"

Grids and Lattices Build Visualizations\u00a0\u2014\u00a0describes Lattice and Grid forms of map surface display Maps Are Numbers First, Pictures Later\u00a0\u2014\u00a0discusses the numeric and geographic characteristics of… <\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[2],"tags":[],"class_list":["post-597","post","type-post","status-publish","format-standard","hentry","category-arcgis"],"_links":{"self":[{"href":"https:\/\/zhuoyao.net\/index.php\/wp-json\/wp\/v2\/posts\/597","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/zhuoyao.net\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/zhuoyao.net\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/zhuoyao.net\/index.php\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/zhuoyao.net\/index.php\/wp-json\/wp\/v2\/comments?post=597"}],"version-history":[{"count":0,"href":"https:\/\/zhuoyao.net\/index.php\/wp-json\/wp\/v2\/posts\/597\/revisions"}],"wp:attachment":[{"href":"https:\/\/zhuoyao.net\/index.php\/wp-json\/wp\/v2\/media?parent=597"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/zhuoyao.net\/index.php\/wp-json\/wp\/v2\/categories?post=597"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/zhuoyao.net\/index.php\/wp-json\/wp\/v2\/tags?post=597"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}