Cartographic generalization

Cartographic generalization is the process of controlling the amount of detailed information portrayed in a map. This also includes using the appropriate scale, purpose, and medium of the map. This form of Generalization commonly consists of reducing the visual detail of data by reducing the map scale when the map purpose suggests the need for a simpler design. It is impossible to represent every detail of the world on a map; therefore, every map has been generalized to some extent. Generalization of maps has become necessary due to automatic production of maps on the web, and the increased amount of detailed GIS data available.

Generalization is closely related to the concept of visual hierarchy because visual detail helps to emphasize the most important map elements while less detailed features tend to attract less attention. After generalizing map data, the importance of what is remaining on the map must outweigh the insignificance of items that were generalized.

Generalization is important in GIS applications other than cartography because increased data detail requires more storage space, data entry time, and processing time. This would encourage GIS users to use less detailed data, but issues are also created in generalizing (especially aggregation) data, such as the modifiable areal unit problem. When thousands or even millions of records or pixels are being used, generalized data can make a big difference in geoprocessing run time if fewer digits or characters are in the attributes of the data.

Contextual Generalization & Selection Process
Map generalization can take many forms and is designed to reduce the complexities of the real world by strategically reducing ancillary and unnecessary details. One way that geospatial data can be reduced is through the selection process. The cartographer can select and retain certain elements that he/she deems the most necessary or appropriate. In this method, the most important elements stand out while lesser elements are left out entirely. For example, a directional map between two points may have lesser and un-traveled roadways omitted as not to confuse the map-reader. The selection of the most direct and uncomplicated route between the two points is the most important data, and the cartographer may choose to emphasize this.

Good generalization requires understanding and characterizing the geography of the mapped area, and involves finding patterns in the data, and abstracting them. In particular:


 * Features – Generalization of features cannot be done exclusively as it involves relationships with neighbors.
 * Layers – Generalization of layers cannot be done exclusively as layers are dependent with one another across many classes.
 * Variation – Generalization cannot be done to achieve uniformity since features of one class have different surroundings.
 * Partition – Generalization cannot be done across the whole map at once. It requires localization or partitioning.
 * Topological properties – Generalization cannot be done based solely on geometry. It also involves topology and the attributes related to topology.

=Generalization Operations= Cartographers have developed many techniques or operations for removing, subduing, or enhancing visual detail in maps. Many of these have been performed intuitively by cartographers over the entire history of the craft, but they have been studied and developed further since the advent of computerized cartography. A number of conceptual frameworks have been proposed for listing and classifying the various operators.

Content Operations
These are techniques that alter the set of thematic topics or layers, or the set of features within each topic, portrayed on the map. The operators are arranged below according to the ScaleMaster Typology outline.

Add
The add operator inserts new features to the map display that are only appropriate for representation at larger scales. Such layers may be useless, and even deceiving, at small scales, but can be included in the representation once the scale has been changed to a sufficient level. It has been known by several other names in the past, including "select," "preselection," and "class selection." It is the opposite of the omission operator of Raisz (1962) and Keate (1989). This technique can be used in both cartography and remote sensing.

Eliminate
The eliminate operator can also be referred to as 'omit.'It can be implemented to remove features when they become unnecessary or illegible at a certain scale. It evolved from the omission operator, instituted by Raisz in 1962 and has also been known as “refine” at several points over the years including as recently as 2007 with Regnauld and McMaster. It is helpful to use this operator when objects do not portray a clear message or fulfill the purpose for which they were created. There are four conditions that generally indicate when the eliminate operator should be used. If: There is also a special case that has been identified by researchers in which a subset of features is eliminated from a larger whole to help reinforce the order of the visual hierarchy.
 * the data has a resolution that is to coarse for the viewing scale that results in a mismatch with other layers.
 * the data has too detailed of a resolution, thus providing unnecessary information.
 * there are too many layers at a given scale which creates illegibility.
 * only the most significant features in a grouping are required to convey a specific message.

Reclassify
During the reclassification process, the organization of features is changed based on attributes in order to improve legibility within the map. This can be accomplished in one of three ways: Reclassifying more specifically deals with taking input cell values and replacing them with new output cell values. This will in many instances do as was mentioned above; simplify or change the interpretation of raster data by changing a single value to a new value. It is important to note that the word reclassify is used to emphasize that the data can be reclassified multiple time and scales. Before the use of the term reclassify the term most often used to describe this function was classify. The term reclassify better describes this operation because of the need to reclassify data multiple times.
 * A revision to the total number of classes represented
 * A revision to the composition of existing classes
 * A combination of both of the above

Reorder
Reorder is the adjustment to the stacking position of features on a map. Frequently features will obscure one another, making the map difficult to read and understand. An example of reorder is a lake that is being split up by different line features. The lines above the lake create the appearance of multiple polygon features rather than one lake. Using the reorder operator would fix this conflict by stacking the lake feature on top of the line features in the GIS. This will eliminate confusion about the lake, improve the legibility of the feature and increase the aesthetic appeal. In cartography, reordering is required when features are in conflict with one another, or when the use of transparency or displace operators do not give a satisfactory solution to the overlap.

Geometry Operations
These operators alter the shape of the features portrayed on the map to be more or less detailed.

Aggregate
Aggregation is a method of map generalization which combines features of similar characteristics into a single feature of increased dimensionality that covers the spatial extent of the original features (i.e. points-to-line, points-to-polygon, or lines-to-polygon). . Data aggregation is used for summarizing, partitioning, and simplifying data, and is also useful when preserving confidentiality. Two challenges exist when aggregating data; one is knowing how homogeneous or densely-spaced data must be in order to be combined in one feature, and the other is deciding the boundary of the aggregate area.

While using this tool does help remove noise repetition, caution should be taken when obscuring the properties of the various components, often inhibiting data analysis. The aggregate operator may also be referred to as combination, regionalization, or area conversion.

Collapse
The collapse operation, also known as point conversion, involves the conversion of geometry, such the simplification of a polygon to a line, or a line to a point. This allows for an otherwise complex feature, such as an urban area, to be reduced to a point that is resymbolized with a geometric form like a circle or a square that has a different (often smaller) size. In cartography this operator helps to de-clutter a map, especially when there is an increase in scale, making details harder to see. This only applies when the object only serves as a geographical reference. Otherwise, the detail being removed from the object would change the purpose of the map.

In relation to other operators, collapse performs the opposite function as aggregate and differs from the adjust shape operator due to its change in symbol size.

Merge (Amalgamation)
Amalgamation and merging take multiple objects and combine them into a single feature to generalize the map. Amalgamation deals with joining of polygons, while merge deals with joining of lines. Amalgamation and merging are necessary with a reduction in scale or scope, because the features become harder to distinguish as they are clustered closer together on a map. There are two types of amalgamation: noncontinuous and continuous. In noncontinuous amalgamation, features that are too small to be seen are grouped together to be represented. An example of noncontinuous amalgamation is bunching the buildings of a school together and representing them with a single building. Continuous amalgamation occurs when features of similar attributes are grouped together. Merging takes multiple lines and joins them, often finding the average placement and using that for the new line.

The merge operator has also been termed dissolution and fusion.

In the past, a distinction was made between the specific definitions of “merge” and “amalgamation,” but more recently, the term “merge” is generally used to describe both terms.

Displace
To displace a feature means to alter its absolute location to preserve its unique identity. Displacing a feature may be an effective cartographic choice when faced with a few issues. One may want to move a feature slightly out of the way of another to allow both features to be clearly distinguished by the map viewer. For example, in the first map on the right of Prince George's County, Maryland, the three main arterial routes to Washington, D.C. - I-95, U.S. Route 1, and the Baltimore-Washington Parkway - are difficult to distinguish in their true locations. However, when the features are slightly moved, as in the second map below it, the three routes can be more easily distinguished. While this technique compromises the accuracy of absolute location, it clarifies relative location. Therefore, this generalization technique should be used with caution.

The displace operator makes it possible to adjust the location of features to avoid overlapping at certain scales. While two objects may have negative space separating them at a large (fine) scale, this space could be reduced to nothing at a smaller scale, making the symbols difficult to discern.

The displace operator can keep the symbols/features separate from each other by increasing the distance between them when zooming out. This can help manage the problems that arise from displaying different features that are close to each other, even at various scales.

The displace operator may also be referred to as conflict resolution.

Exaggerate
Exaggeration is a process whereby an object is relatively enlarged so that its featured characteristics are not lost when drawn at a different scale. This enlargement leads to a distortion in the representation of the object when compared to its true shape or size. An example of this is to make a small slender spit of land larger in order to see its characteristics more clearly on a small scale map. This operation is often used to show small physical features more clearly when a map is reduced in size or viewed at a different scale. Often when making a three-dimensional rendering of a mountain or other topographical feature, the Z dimension (Z dimension = altitude or height) will be exaggerated to show more clearly the relationship of the terrain with the nearby area. The exaggerate operator is most often used to exaggerate only the portion that needs to be viewed more clearly. In the example of the spit of land, only the spit would be exaggerated in its geometry, leaving the rest of the land in its original geometry. In the example of exaggerating the mountain peaks, the Z dimension is exaggerated, but the X and Y dimensions are left the same.

Simplify
Generalization is a process that not only removes and selects data but also simplifies it. Simplification (or point reduction) is a technique where shapes of retained features are altered to enhance visibility and reduce complexity. Smaller scale maps typically have features that are more simple than larger scale maps because they simply exhibit more area.

An example of simplification is to scale and remove points along an area. Doing this to a mountain would reduce the detail in and around the mountain, but would ideally not detract from the user interpreting the feature as a mountain. Often in a map, roads or trails are simplified, as seen in the image on the right. Small curves are removed from the road as they can make the map appear more complex and are not required for the reader to understand the map. Lines and polygons can both be simplified, but points cannot.

Combine
Multiple features may be merged or combined into a single representative example when their separation is irrelevant to the map focus. A mountain chain may be isolated into several smaller ridges and peaks with intermittent forest in the natural environment, but shown as a contiguous chain on the map, as determined by the scale. The map reader has to remember that because of scale limitations, combined elements are not concise depictions of natural or manmade features.

Smooth
Smoothing is the removal of small variations in the geometry of a feature to improve its appearance. According to ESRI, the Smooth tool "smooths sharp angles in lines to improve aesthetic or cartographic quality." With this operation it adds intermediate points between the original points, creating a more aesthetically pleasing smooth line that reduces the sharpness of angles between line segments. The purpose of smoothing is to create linework that is less complicated and less visually jarring. Smoothing is another way of simplifying the map features, but it involves several other characteristics of generalization that lead to feature displacement and locational shifting.

Symbology Operations
These techniques use symbology that alters the appearance of detail and do not actually alter the data being portrayed.

Adjust Color
The adjust color operator alters the hue, value, or saturation of a feature so that it remains legible across multiple scales. A change in scale may adjust the color distribution on the map enough to produce situations of simultaneous contrast and color illegibility not present in larger-scale versions. Therefore, the adjust color operator may be implemented for two reasons: (1) to increase the position of a feature in the visual hierarchy by increasing its contrast or distinctiveness or (2) to increase the position of surrounding features in the visual hierarchy by decreasing the resymbolized feature’s contrast or distinctiveness.

The adjust color operator is a symbology operator that can be done largely by hand without the use of specific tools. In ArcMap, colors can be changed as preferred within the layer's specific properties. Another preferred method is to export a map that is then used in the Adobe suite such as Photoshop or Illustrator. Photoshop, in particular, has tools that allow the user to change all instances of a single color into the specified color. Adobe's color selections may be preferrable to other programs.

Adjust Pattern
The adjust pattern operator reduces the complexity of a symbol by changing its pattern or texture. This adjustment to the fill pattern or the line pattern will improve legibility. Texture is defined by Caivano as having three dimensions: This operator is different than both the exaggerate operator and the typify operator because the pattern is completely independent of the underlying feature geometry below it.
 * directionality of the texture units,
 * size of the texture units,
 * and density of the texture units.

Enhance
Enhancement is also a method that can be employed by the cartographer to illuminate specific elements that aid in map reading. While many of the aforementioned generalizing methods focus on the reduction and omission of detail, the enhancement method concentrates on the addition of detail. Enhancement can be used to describe the true character of the feature being represented and is often used by the cartographer to highlight specific details about his or her individual knowledge that would otherwise be left out. An example includes enhancing the detail about specific river rapids so that the map reader may know the facets of traversing the most difficult sections beforehand. Enhancement can be a valuable tool in aiding the map reader to elements that carry significant weight to the map’s intent and purpose.

Rotate
The rotate feature changes the orientation of one feature in relation to other features. Rotation is the 360-degree shift in the orientation of an object. The rotate operator differs from other operators such as the displace operator in that it will change orientation where the displace operator only changes location. It also differs from the exaggerate operator because it moves the entire object and not just parts of it. One of the many uses for the rotate operator is to align buildings with roads after they have been collapsed.

Adjust Size
The adjust size operator changes the size of a symbol to keep the symbol legible at different scale sizes. This is most commonly used for point symbols but can also be used with line weights for both lines and polygons. This operator should not be confused with the exaggerate operator. Adjust size does not change the underlying geometry for any of the features while the exaggerate operator does. The adjust size operator is also called exaggeration, magnification, or enlargement.

Adjust Shape
The adjust shape operator replaces a symbol that has a complex, irregular shape with one that is more compact for legibility. During a change in scale, it is often necessary to swap detailed, unambiguous symbols for simplified geometric primitives whose interpretations are reliant upon a legend. While point symbols are the most common example of shape change, it may also be extended to the symbols along lines and polygons; the symbology used to represent fronts on weather maps are an example of a shape variation for lines. The adjust shape operator differs from the simplify, smooth, and collapse operators in that the underlying geometry is not altered.

Typify
The typify operator can also be referred to as distribution refinement. When typifying data, a set of graphics (points, lines or polygons) is replaced with a smaller set of the same graphics. Unlike aggregation, there is no change in dimensionality. Many typification tools only create new pattern swatches, rather than altering or manipulating the spatial data. The image below gives an example of a typification. A set of graphics is replaced with a smaller set of graphics.



Refine
Refinement is another form of re-symbolization similar to collapse, though it is an operation that involves reducing a multiple set of features such as roads, buildings and other urban structures to a simplified representation, rather than a conversion to a geometric shape.

Adjust Transparency
The adjust transparency operator performs an adjustment of the symbol opacity to improve the legibility of the feature or underlying features. It modifies the degree to which one feature conceals another. When the transparency is turned on, multiple features can be displayed at the same time. With an increase in transparency, the multiple feature layers are more visible to the map reader. With a decrease in transparency, the uppermost features conceal the bottommost features.

=Automated Generalization= With the development of geographical databases (like ArcGIS), automated mapmaking became possible and thus the need for automated generalization become necessary. Geographical databases allowed cartographers to digitally produce maps, a majority of modern-day cartographers have adopted the use of these geographical databases. With the widespread use and adoption of these databases, the technology to automate generalization became possible. Though automated generalization has always been in competition with manual cartographers, the roots of automated generalization are found in manual old school paper and pencil cartography. In order to understand the process of map generalization, extensive studies were conducted on the manual cartographic generalization practices. These studies resulted early in different generalization operators. By now there is no clear classification of operators available and it is doubtful if a comprehensive classification will evolve in the future. Several automated generalization models from researchers exist and are referenced frequently in a connection with cartographic generalization. (See Generalization for models.)

=References=


 * Buttenfield, B. P., & McMaster, R. B. (Eds.). (1991). Map Generalization: making rules for knowledge representation. New York: John Wiley and Sons.
 * Campbell, J. (2001). Map Use and Analysis (4th ed.). New York: McGraw Hill.
 * Harrie, L. (2003). Weight-setting and quality assessment in simultaneous graphic generalization. Cartographic Journal, 40(3), 221-233.
 * Krygier, J., & Wood, D. (2005). Making Maps: A Visual Guide To Map Design for GIS . New York: The Guilford Press.
 * Lonergan, M., & Jones, C. B. (2001). An iterative displacement method for conflict resolution in map generalization. Algorithmica, 30, 287-301.
 * Li, Z. (2006). Algorithmic Foundations of Multi-Scale Spatial Representation. Boca Raton: CRC Press.
 * Mackanaess, W.A., Ruas, A., & Sarjakoski, L.T. (eds)(2007). Generalisation of Geographic Information: Cartographic Modelling and Applications. Oxford: Elsevier.
 * McMaster, R.B. & Shea, K.S. (1992) Generalization in Digital Cartography. Washington, DC: Association of American Geographers.
 * Qi, H., & Zhaloi, L. (2004). Progress in studies on automated generalization of spatial point cluster. IEEE Letters on Remote Sensing, 2994, 2841-2844.
 * Töpfer, F. T., & Pillewizer, K. (1966). The principles of selection. The Cartographic Journal, 3, 10-16.
 * Petcu, D., Zaharie, D., Neagul, M., Panica, S., Frincu, M., Gorgan, D., Stefanut, Teodor, Bacu, V. Remote Sensed Image Processing on Grids for Training in Earth Observation. InTechOpen, an Open Access Publisher in the fields of Science, Technology and Medicine. Retrieved 10 October 2011.
 * Brewer, C.A., Roth, R.E., Stryker, M. The ScaleMaster Typology: Literature Foundation ScaleMaster.