GIS Spatial Analysis and Vector Operations

Set of techniques depending on location and attributes

Selecting features from a target using spatial relationships of input

Clip: Cutting out one layer using another like a cookie-cutter without combining attributes

Dissolve: Aggregating features based on a single attribute

ADDED FIELDS:
Point Distance – find distance from points to points (police station, crime)
Near – distance, fid, angle, coords of nearest point/line (benchmarks, river lines)
Buffer – create zone of interest around an entity
Thiessen Polygons – catchment, closer to its midpoint than any other midpoint

Creating a new spatial layer taking the lowest dimension of inputs. Georeferenced

Within, intersect

Union (OR) - all attributes combined, all polygons
Clip (NOT) – cookie cutter, output has attributes of input only
Intersect (AND) – spatial extent of input, combined attributes

Query, Reclass, Distance/Proximity, Neighborhood

Attribute Query/Cell-Based – query raster cells based on value, “find cells with elev=1m”
Location-Based – using coords, “what is land cover at x,y?”

Using rules to create a new raster image

Boolean – two-coded (0,1)
Weighted – assigned weights (priority) based on purpose

Euclidean distance – shortest, hypotenuse
Manhattan – distance traveled by right angles
Proximity – concentric equidistant zone around a point A, distance to B is the
corresponding radius of circle intersecting B
Buffer – distance contours
Thiessen Polygon – zone of influence

new spatial layer from old, georeferenced, same AOI
AND – multiply
OR – addition
XOR – add then NOT

local, focal, zonal, global

per cell; single layer (arithmetic, exponential, log), multiple (min, max, ave)

Focal Max – highest value to center
Focal Majority – most often
Focal Variety – unique
Filtering – new value for center is weighted average
Resampling – associate cell values in one grid to another
Surface Analysis – DEM, slope, terrain, intervisibility

– aggregates values in a zone (areas having common attribute values)

Requirements: zone layer (shapes, location), value (inputs), output
Composition: zone (2 or more cells with same value can be connected or disconnected) composed of region (connected cells in a zone)

– output raster values may be functions of all cells of input raster dataset
Euclidean Distance (hypotenuse), weighted distance (accumulated travel cost)

State problem – goal, issues
Break it down – identify data, tools, approach
Explore datasets tools – method, data manipulation and workflow
Perform analysis – decide and use
Validate – accuracy assessment, ground truthing
Present – draw conclusion, decision support

– combine datasets to produce maps that incorporate diverse information sources
Polygon Overlay (Areal Weighting) – estimate target zone using weights of overlapping
Vector vs Raster – Spatial Join vs Map Algebra

Does not account for discontinuities and fuzziness

Indexed Overlay – adds criteria values, ordinal (0 to M) degrees of favorability
Weighted Linear Combination – specific weights to each criteria, relative importance
Suitability Analysis – site selection, best location is where

Method for multicriteria decision making of a group, , know weights in an overlay analysis

L1: goal L2: criteria L3: alternatives
Step 1: determine relative weights of decision criteria
Step 2: determine relative rankings (priority) of alternatives

scale of 7+-2

Mathematical method using power iterations to determine AHP weights until convergence

Measure of inconsistency caused by inadequate model structure

Measure of judgment consistency compared to a sample of random judgment/ CR must be 0.10 or less to be considered consistent and trustworthy

= into a consensus weight vector

Geometric Mean of Judgment – pairwise using mean, then eigenvector for inconsistent judgment
Geometric Mean of Individual Priorities – mean of weight vectors directly
Arithmetic Mean of Individual Priorities – simple mean of weight vectors

Verbal to numeric scale = untested assumptions
Mathematical inconsistency, pairwise comparisons can be excessively large