城市空间信息技术 第十一章 矢量数据分析 胡嘉骢 不动产学院 博士 副教授 城市规划系主任 手机 : （ ） QQ:

2 CHAPTER 12 VECTOR DATA ANALYSIS 12.1 Buffering 12.2 Overlay 12.3 Distance Measurement 12.4 Pattern Analysis 12.5 Map Manipulation

3 CHAPTER 12 VECTOR DATA ANALYSIS Scope of GIS analysis varies among disciplines and users Two approaches to software packaging –Set of basic tools used by most users –Extensions designed for specific applications

4 Vector Data Model Uses points and their x- and y-coordinates to construct spatial features of points, lines, and polygons Accuracy of analysis depends on the accuracy of feature shape and location Topology also a factor in data analysis

Buffering Creates two areas: one within and one beyond a specified distance of a selected feature Area within called the buffer zone May buffer around points, lines, or polygons

6 Figure 12.1 Buffering around points, lines, and areas.

Variations in Buffering Buffer distance or size need not be constant A feature may have more than one buffer zone Buffer does not need to be on both sides of a line Boundaries dissolved or not dissolved Buffer distance measurement unit

8 Figure 12.2 Buffering with different buffer distances.

9 Figure 12.3 Buffering with four rings.

10 Figure 12.4 Buffer zones not dissolved (top) or dissolved (bottom).

Applications of Buffering City ordinance: no liquor store or porno shop within 1000 feet of school or church Two-mile buffer along streams to minimize sedimentation from logging Forest restrict oil & gas drilling within 500 feet of roads Stream buffers to protect streams from agriculture or construction

Overlay Combines geometries and attributes of two or more feature layers Geometry of output represents geometric intersection of input layers Attributes also combined Spatial registration required

13 Figure 12.5 Overlay combines the geometry and attribute data from two layers into a single layer. The dashed lines are not included in the output.

Feature Type and Overlay Point-in-polygon overlay Line-in-polygon overlay Polygon-in-polygon overlay

15 Figure 12.6 Point-in-polygon overlay. The input is a point layer (the dashed lines are for illustration only and are not part of the point layer). The output is also a point layer but has attribute data from the polygon layer.

16 Figure 12.7 Line-in-polygon overlay. The input is a line layer (the dashed lines are for illustration only and are not part of the line layer). The output is also a line layer. But the output differs from the input in two aspects: the line is broken into two segments, and the line segments have attribute data from the polygon layer.

17 Figure 12.8 Polygon-on-polygon overlay. In the illustration, the two layers for overlay have the same area extent. The output combines the geometry and attribute data from the two layers into a single polygon layer.

Overlay Methods Based on Boolean connectors AND (intersect), OR (union), and XOR (symmetrical difference)

19 Figure 12.9 The Union method keeps all areas of the two input layers in the output.

20 Figure The Intersect method preserves only the area common to the two input layers in the output. (The dashed lines are for illustration only; they are not part of the output.)

21 Figure The Symmetric Difference method preserves only the area common to only one of the input layers in the output. (The dashed lines are for illustration only; they are not part of the output.)

22 Figure The Identity method produces an output that has the same extent as the input layer. But the output includes the geometry and attribute data from the identity layer.

Overlay of Shapefiles Shapefiles allow polygons to have multiple components which may overlap Essentially an editing tool. Not overlay tools because they do not perform geometric intersections nor combine attribute data from different layers

Slivers Common error from overlaying polygon layers Cluster tolerance Minimum mapping unit

25 Figure The top boundary has a series of slivers. These slivers are formed between the coastlines from the input layers in overlay.

26 Figure A cluster tolerance can remove many slivers along the top boundary (A) but can also snap lines that are not slivers (B).

Error Propagation in Overlay Generation of errors due to inaccuracies of input layers Positional errors Identification errors

Applications of Overlay Useful for query and modeling purposes Areal interpolation involves transferring known data from one set of polygons to another

29 Figure An example of areal interpolation. Thick lines represent census tracts and thin lines school districts. Census tract A has a known population of 4000 and B has The overlay result shows that the areal proportion of census tract A in school district 1 is 1/8 and the areal proportion of census tract B, 1/2. Therefore, the population in school district 1 can be estimated to be 1500, or [(4000 x 1/8) + (2000 x 1/2)].

Distance Measurement Measuring straight line (Euclidean) distances between features Can be used directly for data analysis Can also be used as inputs to data analysis

Pattern Analysis Quantitative analysis of spatial distribution of features Can reveal whether a pattern is random, dispersed, or clustered

Nearest Neighbor Analysis Uses distance between each point and its closest neighbor to determine whether the pattern is random, regular, or clustered Nearest neighbor statistic is the ratio of observed average distance to expected average for a hypothetical random distribution

33 Figure A point pattern showing deer locations.

Moran’s I for Measuring Spatial Autocorrelation Considers both the point locations and variation of an attribute Measures relationship among values of a variable according to spatial arrangement of values Is variation of a variable at a location related to variation at another location?

35 Figure A point pattern showing deer locations and the number of sightings at each location.

36 Figure Percent Latino population by block group in Ada County, Idaho. Boise is located in the upper center of the map with small sized block groups.

37 Local Indicators of Spatial Association Figure Z scores for the Local Indicators of Spatial Association (LISA) by block group in Ada County, Idaho.

G-Statistic for Measuring High/Low Clustering Moran’s I can only detect presence of clustering of similar values Cannot tell whether cluster is made of high or low values G-statistic can separate clusters of high values from those of low values

39 Figure Z scores for the local G-statistics by block group in Ada County, Idaho.

Applications of Pattern Analysis Hot spot analysis useful for crime scene analysis Spatial autocorrelation –SIDS study –Analyzing temporal changes –Validating standard statistical tests

Map Manipulation Basic GIS tools –Dissolve –Clip –Append –Select –Eliminate –Update –Erase –Split

42 Figure Dissolve removes boundaries of polygons that have the same attribute value in (a) and creates a simplified layer (b).

43 Figure Clip creates an output that contains only those features of the input layer that fall within the area extent of the clip layer. (The dashed lines are for illustration only; they are not part of the clip layer.)

44 Figure Append pieces together two adjacent layers into a single layer but does not remove the shared boundary between the layers.

45 Figure Select creates a new layer (b) with selected features from the input layer (a).

46 Figure Eliminate removes some small slivers along the top boundary (A).

47 Figure Update replaces the input layer with the update layer and its features. (The dashed lines are for illustration only; they are not part of the update layer.)

48 Figure Erase removes features from the input layer that fall within the area extent of the erase layer. (The dashed lines are for illustration only; they are not part of the erase layer.)

49 Figure Split uses the geometry of the split layer to divide the input layer into four separate layers.

谢 谢！