Scribd Logo
Upload

Welcome to Scribd!

  • Book 3

    Uploaded by

    Jayasri Arunachalam
    8 views40 pages

    Original Title

    Book-3

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    Report this Document

    Copyright:

    © All Rights Reserved
    Download as PDF or read online from Scribd
    Flag for inappropriate content
    Download as pdf
    8 views40 pages

    Book 3

    Uploaded by

    Jayasri Arunachalam

    Copyright:

    © All Rights Reserved
    Download as PDF or read online from Scribd
    Flag for inappropriate content
    Download as pdf
    of 40
    CHAPTER . 6 VECTOR-BASED GIS DATA PROCESSING ~ g1 INTRODUCTION jaa processing is the manipulation a eis of 5 eographic database for spatial probe ides a wide spectrum of processes they al data conversion, database creation’ ste tm Guery and management, overlay and ae afl data Qo the development of sophisticated tamale omputer programming, Since the vector ions conceptually logically, and physically dig, line raster data model, the objectives jy it Mets of vector-based geographic data process. ieetialy different from those of raster-based i hat we learned in Chapter 5, This chapter ves ne methods of geographic data processing cate found ina pea vector Gi, i ‘aye incl 62 CHARACTERISTICS oF yecTOR-BASED GIS para PROCESSING veartased GIS data processing makes use of spatial Kerepesented as discrete entities on well-defined ‘ee sin digital databases. Therefore, it is useful to re- ane properties of vector data and the vector layer (aoexplain the major characteristics of using these jjaand layers in geographic data processing. 62.1 VECTOR DATA AND THE VECTOR LAYER \icordata in a digital geographic database depict indi- ‘ial spatial features as discrete entities using three basic tial elements: points, lines, and polygons. In the ‘aha, these entities are identified as feature classes, ‘hperaining to a particular theme, such as drainage, ae topography, transportation, land parcels, and nirative boundaries. Vector data, therefore, are Pe GS appcations that focus primarily on the char- Pog linvidual or individual classes of spatial fea- Tha geattCular geographic area of interest. inci ttl graphic database, feature clases are rignceini2ed as layers (Figure 6.1). Since each ey, aes to have only one type of graphical el- Piha classes are also organized according to es thay as type by default. This means that fea- ‘eth dite 2 @ Particular theme but are repre- “egg on ttt types of graphical elements must be t Febiah according to their respective types suey NEMS: For example, although spatial “With the germs: Fivers, and lakes are all asso- ‘ame theme, that is, drainage, they must “morro bate Posing 185 be identi ih ilferen aye’ S2¥Ste Teature classes and stored on Bon Tenures me BeeUS they ae point ine and poy ectively, In acs eS eres ly. In this particular example, Springs, rive) * Fivers, and lakes €5, and draingrid l@kes are independent feature dass, 1 isa collection of feature clases Convent explained nel the topologically structured layer as rat for veer taPlet 3 has been the standard date for * GIS. In recent years, several other vector logical dat notoniy a mat of ArcView GIS and is commonly used erally ag ord ne 0mm Software products, but more gen- (ESR, Toph 20 interchange format among GIS users created by 1; Spatially indexed and tiled files, which are pack ad map library functions of various GIS software ae ioe as Arcinfo LIBRARIAN (ESRI, 1996c); ase aa files, which are created by special spatial data-~ Tosatnanagement systems, such as ArcStorm (ESRI, a a and object-relational databases, which are spatial jane SYstems used by GIS software packages based on. object-oriented technology, such as Oracle Spatial, ArcSDE, ate 8 (ESRI, 2000; Zeiler, 1999). : ecause the formats of these vector different from the topologically Saeed acs ‘used by conventional vector GIS, spatial data processing using these data types tends to be done differently. It is not within the scope of this chapter to cover individual data-processing methods designed to work with these ‘new vector data formats. However, although the actual methods for data processing using data in shapefile, map librarian, and object-relational database formats may be different, the concepts behind the construction of these methods are similar. In this regard, the vector-based func- tions to be discussed in this chapter will be generally use- ful to GIS users engaged in vector-based data processing, no matter what specific vector format their data are in. 6.2.2. CHARACTERISTICS OF VECTOR-BASED GIS DATA PROCESSING Vector GIS data processing is different from raster pro- cessing in several ways. One of the most important dis- tinctions between these two approaches of GIS data processing is that whereas raster processing is based en- tirely on the layers, vector processing uses both data layers and individual features. As vector data are oF- ganized by layers, vector GIS data processing can be designed to work on layers, either individually or of Jectively in very much the same way as raster 12}°° are processed. However, since spatial ise ais dependently identified and represented 09.5 i" layer it is also possible to apply data PCT vers tions atthe individual fearure evel Tyee nox and features in vector-based data pe re h, Vector ase GIS Data Proeessing 186 BE Ohapere o nt 0 brags L a ; _ in | —i.. Undershoot | Intersection (no nade) Missing abet | ‘Duplicate!multipte labels FIGURE 6.11 | Examples of digitizing errors, Whether i an error often depends on hac seperate how it looks. For example. a cul ee eae arc that must not be regarded as a digitizing. 196 EB Chapter 6 Vector Based GIS Data Processing Setting the Editing Environment This process sets up the working environment for a data-editing session. Although the procedures used by an individual GIS may mands. These errors may ay matically by setting user so? ec ‘defineg gor erances in the topology bug datas 4 ing oye vary, the objectives of setting the editing environment fr example, uses the cone rar are essentially the same. In ArcInfo, an important con- gl ing tolerances in conjunction, wig? cept in setting the editing environment is tolerance, and BUILD commands (which inyoy which is a user-defined measurement to facilitate the ogy building process) to Tesolve nest editing process. Common tolerances include edit toler- resulting from digitizing ( Figure g ia ie ance, which enables the operator to select a graphical el- fying a fuzzy tolerance, nodes that ans ement more quickly by picking without having to to one another will be Tesolved ints actually click on it (ie, the graphical element will be se- tersection, By specifying a dang, small dangling nodes will be removed inating the need todo this mange lected when the screen cursor is clicked at a small dis- tance away from it) (Figure 6.12a); weed tolerance, which Temoves digitized arcs that are shorter than a specified length (Figure 6.12b); and grain tolerance, which filters Out digitized points in curves, circles, and other Smoothed arcs that are too close together (Figure 6.12¢) Topology Building This is probably the most im- Portant process in graphical data editing. As a postdig- izing process, however, topology building actually serves two interrelated purposes. Topology building is normally When the errors in a layer havi corrected, either manually or aut. and dangling tolerances, topolo errors have been introduced or ing process, they will show up of topology building. These e and topology rebuilt once agai cepted only when it is shown 'Y 2 repetitive € been iden fomaticaliy yo BY MUs be rp overlooked in again in the neg TOTS Must be rey in. The layer wits, to be totally ee, * Building the Topological Structure and Relationships Sor the Graphical Elements on a Layer. The actual Process of topology building is dependent on the type of the graphical elements on a layer (Fig- ure 6.13). In general, this includes the creation of Point, line, and polygon topology by assigning an internal identifier to each graphical element iden- tified and the creation of attribute tables, Error Identification and Autontated Corrections. It digitizing errors exist in the digital datafile, they tizing and topological errors. Data Editing and Error Corrections Graphics editing includes procedures that enable the open, select, delete, copy, and add graphical elements « as to change their properties, on a vector lye.» ure 6.15). Data editing is an integral part of ogy building process as noted earlier. Itisalo when errors are detected as the result of quai ance measures in data conversion (e.g., by verify will be highlighted by the topology building com. data in a check plot). FIGURE 6.12 Edit, weed, and grain colerancsia staphical data editing (@) Edit tolerance Cursors able to select Point x when itis clicked within the edit tolerance (©) Weed tolerance Digitized arcs smaller than weed tolerance are filtered () Grain tolerance b-+ Grain tolerance ee. oe | hind yon too | me. _ | close to one another | os \ are removed + rs I 3 8 | eee 6. Vector Data Input Functions er m | a © a a| a 8 eee Leen + Dangling nodes + Dangling ares ise + Arc attribute table (AAT) + Other tables «Deiat reve Danae «Daan ae | — +o “ID for nodesiares attribute ible neta ‘Dermmee ea) SRatetpng | TOM | tAcopog — ncuRe 6.13 Talserormed during topology building. The types of tasks performed in topology building depend on the types of features on a fear layer. Topology building on a polygon layer, for example, is far more complex than topology building on a point layer. Data editing is an error-prone process, particularly macros as well as customized Graphic User Interface enomplex map layers are involved (e, (GUI) described earlier (ESRI, 1995b). lee ee tTY Maps). It is good practice to back up Joining Adjacent Layers The procedure of data ver from time to time during a data-editing conversion described here applies to a single map with °-This allows dat _ . Meade a that have been deleted by mis-_rputtiple layers, For applications that requige more than ie fte¥ered by reloading the backup data, thus Prey Pl a 8 the more h one map, a further step of work is necessary to ensure "tg the data fren me eonSUMIng process of red that all the layers can be joined together to form a con- ning prow the map. Data editing is also a time- Syaye 8 PX0cess becal tinuous geographic database. This usually involves three faa, a 'stheretore use it must be performed layer sequential processes that include: important to make use of ever} m ‘Beerhe pl hd software tool inthe system to au. + Bdge Matching. ‘The purpose of this step is to en- "hoiogy a These include the use of the vari- sure that linear features across layer boundaries 'ng and editing tolerances, editing match with one another (Figure 6.16a, b and c). > ne 198 chapter 6 veaor-Based GIS Data Proesing B-+— - | Ot © Node 1 Danating length if [+ {4 Fay wlrane ee & (a) Digitized graphics before applying CLEAN command. © tie © Node (new nodes are ereated at all line intersections) FIGURE 6.14 Using fuzzy and dangling tolerances with command to automatically correc digi Arcinfo. Dangling arcs smaller than the dang a tolerance are deleted and all n (©) Vector data after using the CLEAN command odes within thf, tolerance ae forced to become a single nn FIGURE 6.15 Examples of graphical editing to correct Alighig errors. Graphical editing includes a vate of such as selecting, adding, deleting, moving graphg elements, and editing of label text. omin | P Yaad = wee [Ot Dy | é tend ine to polygon boundary La] | | | Movenode oc yon | \ ogre | \ 99 63 Veator Data input Functions WA 1 (h) Due to digitizing errors i ine features {do not always match ( Across adjacent layers, : Eg © fiom a inal layers remain separate files The process of map-join Physically merge the two layers to create a single layer, Q\ Np 2 (01 TYE final ste isto dssove the neat ne ‘ividing the two original iiacent layers. The topology of tne Fesulting layer is rebuilt te form a continuous layer, WORE 6.16 ‘eiaga continuous layer from indi sige continuous layer. The sa ‘dually digitized layers, This diagram illustrat sembic ata process 5 how two adjacent layers are merged to ime techniques can be applied to merge ‘multiple adjacent layers in ve ctor-based *MpJoining, The Purpose of this step is to phys- tables. The objectives of attribute data conversion are to ‘allyjoin adjacent layers to form one single layer create the database tables Containing descriptive data ‘Faure 616d). maaaclated with the graphics ana link them logically by * Disoh Redundant Polygon Boundaries, The aim Means of the common feature identifiers in the feature his process isto ‘merge Polygons on adjacent attribute tables (Figure 6.17), ea removing their boundaries. The topol- In GIS, descriptive attribute data are Managed either nina its layers rebuilt, formasin- internally as Ick the database managemers compo- NOUS layer (Figure 6.16e) Rent of the GIs itself or. “xternally by a commercial data. se management system (DBMS). In Arcinto, for cxample, internal database tables are INFO files. Un- these files are not generat. cally but must be created by Srerclational orobject. the rear it the INFO environment. attribute duc stored Na are stored in database in &xternal databases are processed and managed by a 200 Bearers voatem tat Amite data converwen Defining tae serwtre 4 abe Togetogs ‘ Soha ' Pobveon ‘ setae Lamece ‘ al + aie + tae ‘ between graph © doginal data The graphical digital data conver PEORENS Were et ap © $e * aphical elements on the vestor layer The attribute tga ny efting proveat generate a + comtaine the deveriptive properties of the graphical elements These types of © fer ate Liked Senaifiers of the graphical elements in the fies DEMS 5 Are pparted by the GIS. which in and (3) | king the file to its ansoctated layer sadote xP ern data files. Although BMS is used. These inctide. include Oracte, Sybase DB2. ary he actual peon among different systems, this xequu to both internal and external dures may ve ce of wk le databave lies ute of a data tle defined by the FROM WHERE Be number of times, if any, this line intersects with 2 gon boundaries (Figure 6.23). If the number « sections is odd, then the point is within ¢ otherwise it is outside. There are many GIS applications use of the point-in-polygon test. Notable exam dude the determination of customer locati different sales territories in market analysis. allocation analysis of hospitals, fire stations, 322 social or community facilities in planning the matching between surveyed sites of natura entificimportance (e.g., nests of protected bit! ¥ and forest types in resource management : $52 BUFFERING By definition, a buffer is a zone with a specie surrounding a spatial feature. The process o! buffer, or buffering, is an essential fu based geographic data processing, Since t ical elements of vector GIS are points, lines. and F°" the bes fIGURE 6.23 poinvin-polygon search by the plumb-line algorithm, {nthismethod, a point is within a polygon if its plumb line hasan odd number of intersections with the boundaries of the polygon, such as point a and b, which cross the polygon ‘oundares 5 and 3 times, respectively: otherwise the point {soutside the polygon, such as c d, and e, which cross the oiygon boundaries 0, 2, and 6 times, respectively. buffers can be generated around each of these elements. Ineach case, the buffer generated takes the shape from the feature, For a point buffer, itis a circle with a spec- ified radius (as the buffer distance) drawn surrounding the point (Figure 6.24a). For a line buffer, it isa band with a specified distance created on both sides of,the line conforming to the line’s curve (Figure 6.24b). For a polygon buffer, it is a belt of a specified buffer dis- tance from the edge of the polygon surrounding the polygon and conforming to its shape (Figure 6.24c). When the buffer zone is constructed inward from the polygon boundary, it is sometimes called a setback (Figure 6.24d). Buffer zones are polygons, no matter whether they are generated from a point, line, or polygon feature. In St0graphic data processing, the buffer distance is a user- defined parameter that is used by the system to generate the buffer zones. For highly convoluted line features, such as a meandering river or complex polygon bound- aries, buffer zones generated tend to overlap. However, boundaries of overlapping buffer zones are automatical- y dissolved to produce a single coherent buffer zone for i entire spatial feature (Figure 6.25). Buffering is a very important function in vector- based geographic data processing. Like its raster-based shumtetpart, vector-based buffering is used primarily ‘r neighborhood analysis that aims to evaluate the aracteristics of an area surrounding a specific loca- Yor linear spatial feature. The objective of bulfering } delineate the area of interest for a neighborhood \ ili, "*-Pasea Toplogical Functions 207 @ Butter distance (2) Point bute Ps =e Bulfer distance () Line butfer (© Polygon butter QO Butter distance (@ Polygon setback FIGURE 6.24 Point, line, and polygon butters. analysis, which is then used in conjunction with other geographic data-processing functions to search spatial features that fall inside or outside this area of interest and analyze their spatial patterns and interrelationships. ‘Common applications of buffering indude, for example, identification of properties within a certain distance of a zoning change application in urban development, de- lineation of areas around natural features where human activities are restricted in natural resource management, and determination of areas affected by location, move- ment, or spread of hazardous materi mental management. : In most GIS applications, it is sufficient to assume 2 uniform buffer distance. However, it is also possible to produce variable-distance buffer zones. Some applica- tions require different buffer distances for different parts of the spatial features of interest. For example, Xiang (1993) conducted research on the use of variable- distance buffers for rivers for water quality management Data Freeney 208 BE chapter o vector Based GIS a) Original spatial feature (b) Butler zones ate generated by each ine segment and node of the spatial feature, (€) Boundaries of individual buffer zones ar dissolved to form ‘coherent buffer zone for the entire spatial feature, FIGURE 6.25 The process of buffering a line feature. Buller distances are determined with reference to phys- ical, ecological, and socioeconomic conditions of the re- gion. He developed a formula to compute riparian buffer distance using a detention time model of buffer effectiveness developed by Phillips (1989), which incorporates soil, hydrological features, land cover, and topography. Xiang (1993) extracted soil, hydrologic, and topographic map data within an 805-meter distance from the river (using the CLIP command in ArcInfo). From the topographic data, slope angles were computed using a Triangulated Ir- regular Network (TIN) model. All these map layers were overlaid to form a new layer. Buffer distances were com- puted according to the riparian butfer distance formula he developed. These buffer distances were then applied to produce variable-distance buffer zones in different parts of the river (Figure 6.26). 6.6 LAYER-BASED TOPOLOGICAL FUNCTIONS Layer-based topological analysis functions, as the name implies, are those that work on the entire layer rather than on individual spatial features. Map layers that are FIGURE 6.26 a ‘An example of variable-distance butter zones along a river. (Source: Xiang, 1993) used as input to this particular class of geogray processing functions must be topologically » The output layers from these functions are top. ly structured on completion of the function. ts topological functions, as a rule, are computat tensive in nature. 6.6.1 RECLASSIFICATION Reclassification is a database simplification proce:~ aims at reducing the number ies data on a layer. This process is made up of tos work that Involve a nontopological process uss tribute data and a topological process using git data. The first step makes use of the nontop0=: methods of attribute data query to select attrib of a particular value (or range of values) by metic operator (=) or a relational operator (El = then assign a new value to them according 10” classification scheme (Figure 6.27). After reclassification, some polygons will same attribute values as their neighbors. The boundaries between polygons with identical values become redundant and have to be *' order to maintain database integrity. This isdo™* second step in a topological process that make the adjacency relationship to determine ne polygons with like attribute values and 10 Fe" redundant boundaries in between. As the rope the original layer has been changed after 1° tion, it must be rebuilt 10 generate new attribu! and to establish the topological relationships". graphical elements in the new layer (Figure °° Sony om Sant jeuy cay esi cal sinilty map sedis Satefajcent eslae siedaneterisies 6.27 “sof reclassification. 662 “AYER-BASED. OVERLAY ANALYSIS “ologicg) Sion ot yea is probably the most important ina wrbased geographic data processing. seta ses gto" equivalent of cartograph- m5 tens S that we learned in Chapter 5 upgRetecompleyy hOweVET topological overlay Sa -topoig td CoMputationally demand- Sig thave ge o8'Cal overlay” implies that the ate 42M ale : the oven) acY been topologically structured, By of the combined er ed rom a he topo © ov °verlay operation will have to ‘be » Se 64 Layer Based T be updated. The resulting attribute tab layer consists of the old and nev attributes [OFT TT logical or mathematical operation on the oid anes CE ure 6.28) The method of topological overlay can be ME ily understood by explaining it from three perspective namely: (1) the types of overlay. (2) the topolog!t 8" overlay operators, and (3) the topological ov process. erlay Types of Topological Overlay In topological ca © . lay, features of two layers are intersected to create BEN output features, and attributes of the two layers are combined to describe the new output features. The new output features vary according to the nature of the input layers. Because in vector GIS, layers can be ae up of points, lines, or polygons, fir e types of topolog- ical overlay are possible: (a) overlaying a point layer on polygon layer, (b) overlaying a line layer on a poly gon layer, and (c) overlaying a polygon layer on a poly gon layer (Figure 6.29) Of these three types of topological overlay. the polygon-on-polygon overlay is the most complex. It is computationally intensive because of the need to com- pute the coordinates of the intersecting points and then to compare them with the list of coordinates of the two intersecting lines (Figure 6.30). This type of overlay is particularly useful for modeling applica tions. By overlaying soil types with land-use parcels, for example, it is possible to associate soil types with land use, and hence determine which types of land use have occupied fertile soil and where they are located. Topological Overlay Operators For polygon-on- polygon overlays, different operators are used to com- bine two or more input layers into a single output layer. (opological overlay operators that are commonly imple- mented in a typical vector-based GIS such as Arcinfo in- clude UNION, INTERSECT, IDENTITY, CLIP, ERASECOV, and SPLIT Figure 6.29) (ESRI, 1994a). The definitions of these operators are as follows: (/ UNION overlays polygons and keeps all areas in both layers. * INTERSECT overlays polygons but keeps only those portions of the first input layer features falling within the second input layer polygon * IDENTITY overlays polygons and keeps all input layer features. + CLIP cuts out a piece of the first input layer using the second input layer as a cookie cutter, -ASECOV erases a part of the first input layer using the second input layer. * SPLIT divides the first input layer into a number of smaller layers based on the second input layer. 210 BE chapter 6 Vector-Based GIS Data Processing structured Topology of new map is built ‘automatically Topological Overlay Process Topological overlay is a four-step process that includes (1) defining the ob- jectives and data analysis criteria, (2) applying topo- logical overlay operators, (3) postoverlay processing and interpreting the results, and (4) the removal of sliv- 2r polygons. Defining the Objectives and Data Analysis Criteria or example, we may seek to create a land suitability ap for agricultural development. Three layers are -ailable: soil, topography (slope angles), and land ver (Figure 6.31a). The land assessment criteria are ulated as follows: * Areas with loam soil, two degrees of slope, and covered by grass are the most suitable for agricul- tural development (using a suitability code = 2). Areas with gravel, three degrees of slope, and covered by bush are unsuitable for agricultural development (using a suitability code = 0). Any other areas are suitable (using a suitabili- ty code = 1). ‘Maps must be topologically Performed on ‘one or more layers ‘New attribute table contains combined old attributes FIGURE 6.28 The topological overlay process Applying Topological Overlay Operators De; ing on the objective of topological overlay, the o tors can be applied as explained. In this parti example, only the UNION operator is needed. Ii plied two times as follows: * to overlay soil ($) and topography (T) to pro a new layer S-T, creating a new attribute combining soil and topography (slope angles) ure 6.31b) * to overlay S-T layer with the land cover (L)! to produce a new layer $-T-L and associated attribute tables (Figure 6.31¢) Postoverlay Processing and Interpreting the Re This step of work requires the use of database ma ment tools to change the structure of the attribute! by adding an item called “suitability code” to the$ feature table (Figure 6.31). It is possible to add the to this item manually with SQL queries and joi" on the criteria noted earlier. Another approach 51 attribute data manipulation commands to sele“ 2 & ) UNION aurbute file of the S-T-L layer the type of soil, ppog- i, and land cover that meet the stipulated criteria, hd then add the correct suitability code. In Arcinfo, for Brample, this can be achieved using the RESELECT and CULATE commands in the TABLE module. The final Erfoute table will incorporate characteristics of all three Dut layers and an item of suitability code. A look-up Ba ©20 be used to control the display of a land suit- wity map. ILis also possible to extract only those poly- Pes that are most suitable for agricultural development liability code 2). Therefore, the end product of ereical overlay isa land suitability map from which ‘ost suitable locations may be selected. Using the 3h)" Of statistical computation (see Section 6.4.2), 4 be showing the statistics of different categories of land- Suitability may also be generated. \{emoval of Sliver Polygons Sliver polygons, also Ht spurious polygons, are very small polygons that are fa tated in topological overlay as a result of the mis- te 8s of identical features on different layers (Fig- He 6.32), This can be due to a variety of factors, such Fandom errors in line digitizing and imperfection in INTERSECT 4 Lwo8Kjod uo wo8soq — FIGURE 6.29 ‘Types of topological overlay operations. registering the layers. There are different ways to elim- inate these sliver polygons. These include: + Using Fuzzy Tolerance. Fuzzy tolerance is the min- imum distance between coordinates on the out- put layer. Setting a suitable fuzzy tolerance will cause the sliver polygon to disappear when co- ordinate points that are closer than the fuzzy tol- erance are merged. To determine the correct fuzzy tolerance, make a hard copy of the over- lay and measure the amount of sliver with a scale” This will give an initial estimate of the fuzzy tolerance to use in “cleaning” the overlay. Ifsliv- ers persist, increase or decrease the fuzzy toler- ance until the slivers are eliminated. Using the Builtin Functions of GIS. Many vector- based GIS software packages have a built-in function to eliminate the sliver polygons result- ing from topological overlay. In ArcInfo, for ex- ample, the command ELIMINATE, which merges selected polygons by dropping the longest shared border between them, can be used to remove sliv- ers based on one or more criteria, These criteria can be the area (sliver the shape (sliver polygons eo bi etna row), and the number of ge" nye, cpg two bounding arcs). SS stv ae + Using a Template in Digitizin, * two methods just describes! con, polygons as a postoverlay pra Hich ke | provides a preventive meaguacs® tir | sliver polygon will occurin rae ett takes advantage of the meine a ing template as explained in sect, polygons (e.g. lakes) that fs layers in the overlay process are same digitizing template, they fectly and the occurrence of sini! therefore avoided (Figure 6 33),"" iy + Eien ov 6.6.3 NETWORK ANAtYsi5 In the context of GIS, a NetWOFK i ae ol gy, that are interconnected. Common exampig gh include highways, railways, ty streets river, ya tion routes (e.g. transit, school buses, gabayeas ca and mail delivery), and utility distribution gs eas, electricity, telephone, water supply, and serge; tively, these networks form the infrastrucure im. ‘Gery. They provide the means for the movemen, and goods, the delivery of services, the Now irs. F16¥ To find coordinate Coonsinat lis for: )——— Compare each pair ff coordinates with ‘one another in the te lists end (Coodinate list forbs (ay) (ey Compute coordinates ‘at intersecting point ——- @ FIGURE 6.30 ‘Computational intensity in polygon overlay. —4 fecal ‘Topography >; 7 and cover — ] | a Land assessment rier ena) ven sable @) cube ©) A | er eh 2 | | sutaiiy : goo seep 2. (4) TPO FIGURE 6:31 ological overlay. (b) Topological overlay: step 1 (c) Topological overlay: s6P fa) An example UH / No sliver because [take boundary and | shoreline come from | \ the same template / Land suitability map 4 a Very aia ll ts | Map showing most stable location only 7 | ruta =—— Crate a-tempia) [sea FIGURE 6.33 ‘The method of using a digitizing temp sliver polygons in topological overlay 214 WE Chapter 6 Vecor-Based GIS Data Procesing and energy, as well as the communication of information (Haggett and Chorley, 1969; Kansky, 1963). Despite the obvious geographic importance of net- works, however, GIS in general is not optimized for network data proceing i more oriented toward re- source and environmentaf applications that emphasize the use of polygons. Dealing with networks requires sophisticated processes and data structure that may not be available in the core functionality of most GIS soft- ware packages. Network data-processing procedures are usually provided in the form of software extensions that are obtainable from the GIS vendors or third-party software developers. In Arcinfo, for example, the native ‘commands and the georelational data model are not capable of handling network data processing other than simple graphical display and attribute data query. Ad- vanced network processing can be done by using a spe- cial software extension called ARC NETWORK (ESRI 1996a). ArcView GIS, similarly, provides an extension known a Network lyst that allows the user to solve network-t€lated spatial problems such 4s finding the best routes, defining travel costs, and modeling traffic flowyOrmsby and Alvi, 1999). ‘he concepts and techniques of network analysis in vector-based geographic data processing can be ex- ‘Street segment. | FIGURE 6.34 Characteristics of a network layer. plained from three perspectives: (2) network analysis algorithms Whe ney, of network analysis. ” and (3) at The Network Layer A network ig topologically structured line segment, 8 tionally, turns (Figure 6.34). Aline oo as an edge, is the basic spatial unig peau # identified by a segment identifier ang | "n° lated attributes such as street name eps 384; limit in a street network: and secign ea By and carrying capacity ina power trangg o> A junction, or an intersection as it issomeetaay we a point or node in the network where {8S meet, It is identified by a junction idea line segments leading to and from it in the tum is a numerical value specifying the yo") quired to travel through a junction. Ina ae for example, a value of 5 seconds can be asi 7 tum if there is a stop sign at the particularig®? and a value of 15 seconds is assigned if theres je fic light. These values represent the average °° i it will take for a car to tum from one streetgg street in normal road and traffic conditions can be set to different values for differen cous B SHEE Sweet game | LED Speen segment iD | Steet @) |e) [1196 College MS “Junction Junction | Time for tura , rithms ‘Network analysis al- ‘i pais SBT ively sophisticated mathe- civel 8 ‘ oes (see Chapter 10, Section ‘ 20S axial problem using networks, such 88 Pl Pe th between wo points, requires Bes sputation to solve. Itis not i 00S com to discuss the different Aes 0 ms that have been used in GIs nll 8 51 40 demonstrate the complexity of is aininghow the shortest path be- alt Pe getermined in GIS. ining it is relatively easy (0 visually see pbetween (v0 points, Aand B, on st) In ClS-based network analysis, fe simple because the system has £0 ise are all possible paths between the Mp oe order to arrive at an answer, Several Fm een developed to solve thls problem satis ly, The best known of these algorithms, aie py EW. ifkstra, is used by the Network An- seen of ArcView GIS (Ormsby and Alvi, 1999). The algorithm begins with the by entering itas the first item in a “rea (Figure 6.36a). It then consi cent to this point and 66 Layer Based Topological Functions M215 FIGURE 6.35 Determining the shortest path between A and B. starting point. A. ched nodes” table {ders all the junctions adja~ places them in a “scanned nodes” —— el 7 FIGURE 6.36 ni T pete Changing values in the “reached Cumalatve | Previous node ‘Node | CumUIAHYE | Pyviousnode | nodes” and “scanned nodes” tables at fifercot stages during the execution of Dijkstra’s shortest-path algorithm. (a) Contents of tables at beginning i ) Seamed nodes able I Cumulative | Previous node Node | Cumulative | Previous node al : +] ao a ? 720 x ipl) mo A Q | ms A EL Ltr 2 A a )R from A. [Contes of ables after reaching P and (e) Contents of tables after reaching M and N from R Previous node Node | Cumulative oa eT | ~ & Q 105, ee M 184 Liew 100 R Previous node a>> (ey Contents oF ables after reaching B from X and ¥ Preview ode ] Cumulative] previous node a + B 1250, | x B | sss ik | a * | | ——* | 216 BY Chapters: Veor Based IS Data Processing table, From this scanned nodes table, the system identi- fies the node with the shortest distance from A and moves it to the reached nodes table (Figure 6.36b). The process is repeated using the latest reached nodes (i-¢., Ry N. and so on) until the destination point B is reached by all possible paths (Figure 6.36c). On completion of this the reached nodes table will contain the cu- Imutativedistanes from point Ato all the nodes in the network, and the scanned nodes table will con- tain cumulative distances of all possible paths from point A to point B (Figure 6.36d). The shortest path between the starting and destination points is obtained by selecting the smallest cumulative distance in the scanned nodes table. In order to find the sequence of junctions that are included in the selected path, it is nec- essary to trace the node identifiers in the “previous node” column of the reached nodes table. This description of the shortest-path algorithm prob- ably represents the simplest scenario to which it is ap- plied. The process will be much more complicated if the shortest travel time or least travel cost between the two points is required. Since travel time and travel cost are not necessarily proportional to distance directly due to, for example, different speed limits and different values of turns at different junctions, itis necessary to perform the analysis using the time and cost associated with each individual street segment and junction, rather than by simply multiplying the total distance by an average travel time or average cost. The algorithm used in this example determines the shortest path between two specific junctions only. If ¢ shortest path between any two points is located on. ¢ segments, rather than right at the junctions, it is necessary to create the junction by breaking the line segments concerned first and rebuilding the network topology before the algorithm is applied. It should also be noted that despite its complexity, the creation of the reached nodes and scanned nodes tables often repre- sents only the first step in network analysis. Practical- ly all the network-based spatial analysis tasks listed next require further processing using more sophisticated mathematical or statistical models. Applications of Network Analysis There are many spatial problems that require the use of network analy- sis for their solution. These include, for example, + Pathfinding. ‘This analysis finds the shortest path (in terms of physical distance or least cost) that can be followed to visita series of locations in a network * Allocation, This analysis assigns one or more por. tions of a network to be served by a facility or business location. * Tracing. This analysis finds all portions of the network that are connected with the movement of a particular feature, for example, a truck. + Spatial interaction. This analysis a, accessibility of a location and the n't, occur between different locations a m nique known a5 gravity modeting tn used in economics, geography, «4's urban planning (see Chapter 10, gf © Distance Matrix Calculation. ‘Thig ont ates a distance matrix between gia, locations in the network, ey + Location-Allocation Modeling. This sy, most complex network analysis, i 4% mining simultaneously the locatign and planned facilities, as well asthe !"" demand to these facilities, a tee Wy It 6.7 VECTOR-BAszE, OUTPUT FUNCTION, Output functions of vector-based GIS are data, procedures that generate maps, charts, andy, reports, as well as transformation of digital ds, one form or scale to another. The purpose of they tions is to provide the necessary means to pres results of the data analysis processes discussed ex: end users in an effective manner. Vector-basedo, functions share many common concepts ani niques with raster-based methods explained in: ter 5, in particular the use of color for displaying on the computer screen and producing hard-cop using printers and plotters (Section 5.4). Hon: vector-based output functions are also differen = their raster-based counterparts in several ways. following sections, we will provide an overview characteristics of the output functions that ate = in a typical vector-based GIS software package. detailed explanation of GIS output functions i broader perspective of information communicati: be given in Chapter 7. ' 6.7.1 METHODS OF PRESENTING VECTO BASED SPATIAL INFORMATION Since GIS has its roots in cartography and m0! is invariably characterized by the ability 10 8 high-quality maps in both digital and hard? Depending on the nature of the information results of vector-based data-processing tasks sented in various forms. The most comme? "| methods include the following (Figure 6.37 / a * Graduated Symbols. Graduated symb® points drawn to show magnitude of 62 at locations where they are measufel that show volumes or ranks for lineat™ such as roads, utility lines, and rivers \ 6.7 Vector Based Output Huncore + Locations ‘+ Amociation between | « Difficult to read if RY + Lines ‘size of symbols and too many features ~ | PAT [a | Seem | cet , ” Pets | De) fogs Lots 7 | - += Difficult 10 4 ee associate colors i ayn «Ares swith valves they mt [e i =o Bey on | ty, i | "ay . vin | | | ST tam] | -Gmmon | sarees | q] Fc [ peewee | "mac acane ts jcc hy | | ie | pe ! | * Continuous | + Easy to show '* Difficutt to read ree] ees | phenomene | “raofcange |" abet iniual i | a ioe tof | | = + Diao 9 ae | sie oon ~ | L | Soe, rf ia pe xnnethods of presenting results of vector-based geographic data analysis. an * Giluated Colors, This method is most suitable for ‘toving discrete areas by which data are sum- ‘ute, using no more than seven to ten colors, Belduker colors to mean more or greater, and Ube coors i ie] toons to mean fewer or smaller. Tage Method is commonly used to dis # Ni summarized by area oF discrete loca Ss in terms “Coasts Of both quantity and categories. et] ue gf COMOUT lines are best for show- PY pauanet228e in data values across an area aly continu 8. rainfal te ‘ous phenomena, such as ‘mperature, and barometric vai "rspective Views SE 0 sho This method Stscrete ContiNUOUS surtaces oF mag. "itaeg.“"* teas by which spatial data are 6.7.2. MAP COMPOSITION Although the methods of presenting vector-based spa- tial information are relatively simple in concept, the ac- tual generation of the output products still requires a considerable amount of effort in planning and design. GIS software packages today generally provide a very rich set of tools that allow the user to compose a map to meet different data-processing needs (Figure 6.38) However, the quality of the end product is dependent not merely on the individual user's ability to master these map composition tools in GIS, but also the user's technical training in cartographic design. Therefore, users who do not have a cartographic background should take extra care when generating output prod to ensure that these products are generated according to standard cartographic practices in the use of eee eer re Pulldown Color palete FIGURE 6.38 ‘Map composition graphical interface and tools of ArcView GI color, map layout, and symbology, as well as map scale and size. From the data-processing perspective, map compo- sition is a repetitive design process. Even experienced cartographers have to test different alternatives before arriving at the final result. In the digital cartographic environment of GIS, its relatively easy for the user to try different ways of presenting the results of spatial in- formation. As a general guideline, the user should pay particular attention to the following issues in the map composition process: + The Order of Layer Display or Plotting. The normal sequence is to display or plot layers pertaining to the topography first, followed by layers showing man-made features, and finally annotation. This means that highways and buildings, for exam. ple, are often displayed or plotted on top of con- tour and drainage, and place names are displayed or plotted on top of all layers of graphical data. Use of Fonts. Make sure that font sizes are pro- portional to the cartographic prominence of the features that they describe and that the align- Map composite Fill palette 1S. To view this photo in color, see the color insert ments of labels are appropriate to: of their associated features. Use of Symbols. The symbols supplied ware packages are sometimes not enous resent all types of features to be ds particularly in thematic or special-purpos cations such as geological mapping. Ins the user should know where to obtain essary symbols and how to incorporate the GIS for deployment in map compos & 6.7.3 VECTOR-RASTER FORMAT CONES” The increasing integration of GIS and ote"! computer-based applications, particule ™ Publishing and word processing, has led 0", ing importance of vector-raster format comer function in vector-based geographic da¥8 Pe Since vector GIS data most often use POP mats that are not accepted by desktoP pul word-processing software packages, it" js __ ut products into a _pased Outpt yect Lath such as GIF and JPEG, so aster {07 ed into reports and presen- 190 41 vee gc! gpa ore is also required 4 conversion i 4 apt Soret Pe cror-based geographic data i rtd us FV end users by means of issemnan atial data on the Internet 6 yp 10 NOW Pe cplayed in raster format, smited 2, and protocols for using vec- are tana f vector-based output prod- onversiO’ “res that they can be viewed x o7T9 ine Web browser, thus avoiding jireclY ectY P user {0 install a product-specific ae gin or helper program, necessary 8 by which vector-based output eset ot to raster format. These in. an De oe g the built-in conversion func- sn ole, using for ofa GIS software package itself, using re a facility of a commercial graphics pro- panes y means of a screen capture package. <16pee individual spatial features are lost in igen "jon process the resulting output prod- ro display only. Their use as input for fur- “i eapoic data analysis is rather limited. i 68 APPLICATION PROGRAMMING us Gsioday use the “tool-box” approach to soft- vutdsgn. The major advantage of this approach is ‘elebity in application development. Since GIS vine packages are developed as general-purpose ‘kis for managing graphical and attribute data, ‘Rabe easly adapted to any application such as “uit of natural resources, municipal admin- tnd transportation planning. The trade-off, cesta aight out-of-the box GIS are diffi- eb aggte et 25Y for users to master the hun- Vlg tO Procedures that are found in a “te yengordet © make GIS easier to use, all ag cpab ve their products with pro- ications, h3t allows users to build their aah Gsenigi ns SUL, application program- “tay, PA of vector-based GIS data pro- ‘a eaion pro eben peeing isa substantial topic that “ey sion 1S own right. In the’ following age te of gag ted t0 an introductory Sete sing ee eee from the per- "gee Porat gest explain the char We with then SN software-specific ‘*n examine the emerging is limited 9 68 Application Programming 21 Concepts pertaining to software components and describe how they can be applied to geographic data processing With special reference to vector-based procedures. 6.8.1 BUILDING DATA-PROCESSING FUNCTIONS USING SCRIPTING LANGUAGES A scripting or macro language is an interpreted com- puter programming language. An interpreted language translates each command line, entered from either the keyboard or a program file, to machine language and executes it before moving to the next line. Scripting languages have simpler structure than conventional computer programming languages. This makes them relatively easy to learn and use. A basic function of ap- Plication programming is to group frequently used com- mand sequences into individual special-purpose tasks so that they can be used repeatedly. At a more advanced level, application programs may also be used to create Prompts for data entry, validate input data, return error messages, build user-defined commands and menus, and develop a point-and-click interface for specific ap- plications (Figure 6.39). ‘There are many advantages of using custom appli- cation programs in vector-based geographic data pro- cessing. These include, for example, enforcing the standardization of procedures by using a common interface, and thus making possible the development and deployment of Custom applications on top of the host GIS soft- ware package improving user productivity by automating fre- quently used and repetitive procedures, as well as by reducing typing and mouse clicking by stor- ing command sequences in a file minimizing the need to memorize commands and their functions, thus enabling less experi. enced users to master the skill of using the sys- tem more quickly * reducing data input error by providing the user with possible options enhancing the native capabilities of the GIS soft. ware by developing application-specific com- mands, menus, and graphical interface windows enabling the integration of geographic data pro- cessing with computer-based applications Scripting languages, therefore, serve a wide var of purposes in vector-based geographic data proces: ing. Conventionally, scripting was the only w which GIS applications were developed. Adv. software development methodology have m able powerful programming tools by which cations can be developed more cost-effec: Tesult, GIS application development is a 220 WB chapter str Maced 18 Oma Prvcesons Customize data processing 10018 added tothe yop wer interface to enhanc the antag Interface window of “Register and Transform” process 7 created using Avenue, the scripting language of ArcView GIS FIGURE 6.39 Example of using Avenue scripting language in ArcView GIS to enhance repetitive tasks. Scripting languages allow users to develop new d by a specific group of users. The functionality of the complex processes based on sophisticated mathe the standard data-processin, lata-processing tools that are used 's¢ tools ranges from relative ‘matical solutions, 18 100ls and 0 atone in a particular ely simple data editing procedures tox away from the use of scripting languages for particular i (GS soltware packages, 1 the open system develop- conventional seftarece development stratege snent approach using industry standards as described aim mainly at generating reusable source coie. ainiiheneaeecin component strategy focuses instead on «te usable preprogrammed software modules that2"> nerally known as “components,” “custom ct 68.2 GIS APPLICATION DEVELOPMENT USING beg sttomation objects.” Components mus RE COMPONENT TECHNOLOGY bedded within a container program in order fet the data-processing functions for which the! tended (Figure 6.40), Components are «te low-level computer languages such as C ult but the container programs can be writen in? Cly of languages including Visual Basic, Viu3. Delphi, PowerBuilder, and some others. or software is therefore characterized b ia code) reusability,” as opposed to “source 0 ce ‘ity” in conventional software application de (Anderson et al., 1996). the advent of component software is one of the portant developments in soltware engineering in re. en years, The concept of component software can be aplained Irom two perspectives. First, component soft, ware is the product of a particular software develop- tent mettiodology. The component software develo. nent methodology evolves from the attempts to search for nicthods that can expedite application development anid minimize development cost through the develop, most im- } map control loaded Mapes Base L00ID0X foihe Visual 5 6.40 ol sing Map added to a Visual Basic form objects components in Visual Basic to develop an application that uses map data. A form is the MapObjects ical i of Microsoft Windows : visual Basic programming to design and create the graphical interface eal oes proves the necesarysofwate components that alow the uses Toad a map toa Tors ofa Visual 2 ee rovides the components to build the tools (eg, 200m, pan, and edity necessary to manipulate the data gant also provide sen the Visual Basic program is run. sod component software is a software develop- ~uditecture. Components are developed in com- wih industry-adopted programming standards tne both the interface among the components “chad their behavior when embedded in appli- ‘yams. The most widely accepted component “uesare Microsoft's Object Linking and Embed- «tm Obet Model (OLE/COM, which has also “any called OLE controls and OCX; it has now ‘sieX controls) and Object Management ‘quest Broker Architecture to standard software archi- components can run under tying geaBeS 38 noted above. This is “itheimpiem eee Behind the concept of ration of interoperable ge- “ing, POcessing ign ae "8(See Chapter 1) (OCIS, 19965 ns “ 2 Common rant Object Re ct The adherence mat Feason why i SPU Langtry 00 “the @p PRtam, Mend ser mn Which is oriented to- id the application developer, component software is intended primarily for the ap- plication developer. There are two basic approaches to using component software in GIS application develop- ment (Hartman, 1997): component suites and component parts. Component suites are collections of components that are assembled into an integrated comprehensive system. Component technology has now been used by GIS software vendors to develop large-scale systems and to reengineer existing ones. ArcInfo Version 8, for example, has been designed and developed on the com- Ponent software development methodology. This al- lows geographic data-processing functions of Arcinfo to be embedded within non-GIS in what is called the Open Development Environment (ODE) (ESRI 1998). In Windows NT, the ODE is implemented by making ArcInfo functions (e.g., data editing and display) ac- cessible via OCX using industry-standard software de- velopment languages (i.e., Visual Basic, Visual C++, Delphi, and PowerBuilder). In UNIX, ODE is implement- €d as shared libraries for development environments such

    You might also like