GEOGRAPHY 275: SEMINAR IN GIS

SPATIAL ANALYSIS

Definitions

methods applied to spatial data that:
add value
support decisions
reveal patterns and anomalies that are not immediately obvious


turning raw data into information

ways in which the sender tries to inform the receiver by:

adding greater informative content and value
revealing things the receiver might not otherwise see
methods whose results depend on the locations of objects
are not invariant under relocation

rotation, translation, scaling, inversion, ...

A collaboration between the computer and the human mind
are maps "mere"?
The Snow map
Dr Snow
the pump
The Openshaw map

Types of spatial analysis

the Tomlin scheme
local
focal
global
zonal
prospects for vector data
A six-way scheme
queries and reasoning

measurements

transformations

descriptive summaries

optimization

hypothesis testing

Queries and reasoning
real-time answers to simple questions

based on alternative views

catalog view
map view

table view

other views:
scatterplot
histogram
linked views
ESDA
Measurements
area, centroid

distance, length

shape

slope, aspect

Transformations
buffering

point in polygon

polygon overlay

field case
object case

vector case
raster case

slivers

Acres 1 2 3 4 5 1+2+5 1+2+3+5 1+2+3+4+5
0-1 0 0 0 1 2 2640 27566 77346
1-5 0 165 182 131 31 2195 7521 7330
5-10 5 498 515 408 10 1421 2108 2201
10-25 1 784 775 688 38 1590 2106 2129
25-50 4 353 373 382 61 801 853 827
50-100 9 238 249 232 64 462 462 413
100-200 12 155 152 158 72 248 208 197
200-500 21 71 83 89 92 133 105 99
500-1000 9 32 31 33 56 39 34 34
1000-5000 19 25 27 21 50 27 24 22
 >5000 8 6 7 6 11 2 1 1
Totals 88 2327 2394 2149 487 9558 39188 90599
spatial interpolation
IDW
bilinear interpolation
kriging
density estimation
Descriptive summaries
centers
centroid
bivariate median
median

properties

minimizing
dispersion

histograms and pie charts

spatial dependenc

fragmentation and fractal dimension

Optimization
point location

route location

optimum paths

Hypothesis testing
samples drawn randomly and independently from a larger population
what is the population?

randomly and independently?

can I conceive of a larger population that I want to make inferences about?

are my data acceptable as a random and independent sample of that population?

discard data until independence holds

abandon inferential tests