GEOG 473 Study Guide - Final Guide: Contour Line, Surface Weather Analysis, Polynomial Regression
29 May 2018
School
Department
Course
Professor
GEOG 473 Final Review
I. Line Data Analysis
A. Linear objects in GIS
• Linear geographic objects: Geographic objects that can be represented by a line or
set of lines
• Line features have length but not area at a given scale. The length of them is
dependent on the scale of measurement
• Length of a line feature:
B. Fractal Dimension of Linear objects
• D is a value between 1 and 2 (A higher D value indicates a more complicated linear
object.) When D=2, the line turns into an area.
C. Linear direction/orientation
Linear direction:
• The value is between -90 and 90 degrees, but the resulting angle can be between 0
and 360 degrees.
• The angle is its exact value when the numerator and denominator are positive, but
you have to add the result by 360 if only the numerator is negative, or by 180 if
either the denominator only or both are negative.
Linear Orientation:
• Orientation can be between 0 and 180 degrees.
• When the value is negative, add it by 180. Otherwise, the value is the same when
positive.
II. Network Analysis
A. What is a network?
• A network is an interconnected set of points and lines that represent possible routes
from one location to another
• Consists of junctions and edges
• Graph Theory is the math behind network analysis
• Routing: the process of selecting the path from origin to destination along a network
B. Types of network
• Directional: Each edge has an associated flow direction
• Undirectional: each edge has no associated direction of flow
1. Methods of routing
• Brute force: You’ll have to list all the acyclic paths between the locations along
the edges, compute the distance of each path, and pick the smallest one
• Hierarchy: The network’s edges have a hierarchy attribute, which reclassifies the
edges to different ranks (Ex. Types of Roads). Find the closest entry/exit point to
next level, then find the entry/exit point combination that results in shortest path
III. Area Data Analysis: Geometric Measures
A. Measures of Individual objects
find more resources at oneclass.com
find more resources at oneclass.com
• Perimeter (P): the sum of the Euclidean lengths of the line segments that define the
polygon boundary
• Centroid (the center of gravity): the point at which a polygon would balance if the
polygon were cut out
• Mean center:
a. Calculating Area
1. Vertices are coded clockwise and all y coordinates must be non-negative. (
)
2. Drop vertical lines to the x axis from the 1st and 2nd vertices and a trapezoid is
created
3. Find the area of the trapezoid:
4. Find the trapezoid area defined by the 2nd and 3rd vertices
5. Continue around the polygon, finding the area of each trapezoid.
6. The polygon area is the sum of the areas of all trapezoids
• Calculating the area of polygons is not easy for computers. It can definitely fail for
polygons with intersecting arcs, negative values, and/or have very large coordinate
numbers
• Centroid (center of gravity): The point at which a polygon would balance if the
polygon were cut out in plywood of uniform thickness.
• The central point can be found using the skeleton method using what remains when
a polygon contracts each of its straight edges moving inward at a constant rate.
(More like peeling away layers of a polygon until it is reduced to a point)
B. Shape
• The shape of an area object may reflect generating processes
• Compactness: how different the shape is than a circle that has the same perimeter as
a polygon; If the value is closer to 1 then the polygon is circular, but if it’s closer to
0, then it is more of a line
1. Cohesion index
•
; The closer the value is to 1, the more the map becomes a single polygon
2. Shannon’s Diversity Index (How different a set of values are)
•
• SHDI = 0; No diversity. A value of 1 signals complete evenness of the frequencies
of the types
IV. Spatial Autocorrelation (Area data analysis)
A. Local Indicator of spatial autocorrelation LISA (2 requirements are met)
1. the LISA for each location gives an indication of the extent of spatial clustering of
similar or dissimilar values.
find more resources at oneclass.com
find more resources at oneclass.com
Document Summary
When d=2, the line turns into an area: linear direction/orientation, line features have length but not area at a given scale. The length of them is: length of a line feature: (cid:1838)= (cid:1856)(cid:3036) (cid:3036)=(cid:2869) (cid:1830)= log(cid:4666)(cid:1840)(cid:2869)(cid:4667) log(cid:4666)(cid:1840)(cid:2870)(cid:4667) log(cid:4666)(cid:1838)(cid:2869)(cid:4667) log(cid:4666)(cid:1838)(cid:2870)(cid:4667);(cid:1840)=(cid:1866)(cid:1873)(cid:1865)(cid:1854)(cid:1857)(cid:1870) (cid:1867)(cid:1858) (cid:1871)(cid:1857)(cid:1859)(cid:1865)(cid:1857)(cid:1866)(cid:1872)(cid:1871),(cid:1838)=(cid:1845)(cid:1855)(cid:1853)(cid:1864)(cid:1857) (cid:1867)(cid:1858) (cid:1877)(cid:1853)(cid:1870)(cid:1856)(cid:1871)(cid:1872)(cid:1855)(cid:1863) Find the closest entry/exit point to next level, then find the entry/exit point combination that results in shortest path. A value of 1 signals complete evenness of the frequencies: shannon"s diversity index (how different a set of values are, cohesion index (cid:3038)(cid:3036)=(cid:2869) of the types. It can measure how the variable values spatial joins. Surface analysis: representing a surface, raster model: rectangular array of identical square cells , each of which has an associated value, vector model: it is more compact than a raster model and includes contours and. In a gis, contours are usually converted to a raster, a dem or a tin before used to construct a surface.
