gis midterm 2

geocoding

the process of using an address locator to enter address text or a table of addresses to find a corresponding address location in a geographic database

reverse geocoding

turning coordinates back into an address

examples of geocoding

1. crime pattern

2. animal-vehicle collisions

3. population at risk

4. pollution monitoring

what are the three address formats used for geocoding purposes?

1. place name

2. street address

3. zip code

what are the address elements?

a) house number

b) street name

c) street type

d) postal code

what is the purpose of address elements?

they help in the geocoding search, pinpointing an address to a particular location

what question can you answer through geocoding?

“what can geocodes do for my business?”

what are the four steps in the geocoding workflow?

1. build/obtain reference

2. select address locator style

3. build address locator

4. perform address matching

geocoding components

1. input data (the description you want to locate)

2. a locator (using matching rule to search)

3. reference data (contains known features and the attributes that describe them)

step 1 of geocoding: build/obtain reference

reference data should cover the geographic area for which you want to match addresses; text addresses and locators that you eventually geocode are matched against the attribute values of the reference features (ex. Street Centerline data)

what role does the Locator tool play in reference data?

it identifies address elements in the input data and uses a set of rules to match input data to reference data and places features on a map

step 2 of geocoding: select address locator style

a locator role is a template for building an address locator; select a locator role that is compatible with the types of input addresses you want to match (street, place name, city, zone, or other); locator role must also be compatible with the fields representing address elements in your reference data (single or multiple role locator)

locator

a tool that ties the location descriptions in the input data to geographic coordinates to place features on a map

step 3 of geocoding: build an address locator

an address locator is built as a dataset; there are two basic steps:

1. selecting an address locator role

2. loading reference data from feature classes

address locator options (storage)

can be stored locally, published as a service using ArcGIS Enterprise, or shared as a locator package on ArcGIS Online

step 4 of geocoding: perform address matching

the result is one or many point locations on the map with the found location (matched = M, unmatched = U, tied = T)

matching score

calculated for each address match; assigned a score from 0-100 - the higher the score, the greater the agreement between the address details, but not a perfect match

what are some causes of matching errors?

input: are the street name (spelling), direction, town, and ZIP correct?

reference data: does the locator include the street and address range? (geographic coverage, currency, street presence, and number range)

setup: do the locator role and field mappings fit the data?

geocoding sources

1. online geocoding

2. US Census TIGER Files

3. US Census Zip Code Tabulation Areas (ZCTAs)

4. Parcel Address Points (local jurisdictions)

5. Street Center Lines (local jurisdictions)

online geocoding

• pros: free and easy to access

• cons: loss of privacy, requires Internet, may not be accurate

US Census TIGER Files

• pros: free and easy to download, uniform across jurisdiction lines, street address formatting works well

• cons: incomplete data, placement of address points are approximate

US ZCTAs

• approximation for census purposes

• don’t reflect actual ZIP Code areas and aren’t kept up to date

Parcel Address Points (local jurisdictions)

• pros: accurate placement of residential location (parcel positional data is often very good, e.g. ± 5 meters or less)

• cons: may need to contact agencies to get up to date data, may not be available/expensive, data ends at jurisdictional boundaries

Street Center Lines (local jurisdictions)

• pros: potential to be more up to date, often decent accuracy to meet city infrastructure needs (typically ± 10 meters or less)

• cons: may need to contact agencies to get up to date data, data ends at jurisdictional boundaries

geocoding for large areas

1. Geographic Commons web service (uses TIGER maps, across US)

2. ESRI’s StreetMap North America (across US and Canada, uses 2005 TomTom maps - more accurate than TIGER)

3. ESRI’s StreetMap Premium (uses HERE and TomTom commercial street datasets, updates often)

geoprocessing

applying a suite of tools with specified inputs and settings to produce a result; allows you to model and automate processes using ModelBuilder or Python

benefits of geoprocessing

• automate GIS work

• ran batch processing

• create custom tools

• customize sets of geoprocessing tools

• take advantage of geoprocessing history

overlay

the process of taking two or more different thematic maps of the same area and placing them on top of one another to form a new map; combines not only the spatial features of the dataset, but also the attribute information

point-in-polygon overlay

creates an output point layer that includes all the points occurring within a spatial extent of the overlay layer; requires a point input layer and a polygon overlay layer

polygon-on-point-overlay

creates a polygon layer from those input polygons that overlay features in a point layer; the opposite of point-in-polygon operation; requires a polygon input layer and point overlay layer

line-on-line overlay

output from this operation is a point(s) located at the intersection(s) of two linear datasets; requires line features from the input and overlay layer

line-in-polygon overlay

a technique in which each line that has any part of its extent within the overlay polygon layer will be included in an output line layer (although these lines will be truncated at the boundary of the overlay); similar to point-in-polygon overlay

polygon-on-line overlay

polygon features that overlay lines are selected and subsequently preserved in an output layer; opposite of the line-in-polygon operation

polygon-in-polygon overlay

polygon input and overlay layers are combined to create an output polygon layer with the extent of the overlay

union

preserves all features, attribute information, and spatial extents from an input layer and employs the OR operator

intersection

contains common features and attributes from both the input and overlay layers and employs the AND operator

identity

creates an output layer with the spatial extent of the input layer but includes attribute information from the overlay

symmetrical difference

contains those areas common to only one of the feature datasets and employs the XOR operator

map scale

representation at a chosen map ratio

measurement accuracy

closeness to the real location or value

coordinate precision

numeric detail in stored coordinates

buffer distance zones

creates zones within a stated distance; planar: distance on flat coordinate system, geodesic: distance accounting for Earth curvature

dissolve by a shared field

group features using a selected field

append to an existing target

adds records to an existing target dataset; matches geometry types of map fields deliberately

merge into a new output

combines compatible datasets into a new output

recoding

assigns new categories through an explicit lookup or range rule; retain original values of traceability

interpolation

estimates an unknown value between known observations, using an assumed relationship

sample interpolation

measured points support estimates at unsampled locations (ex. temperature, rainfall, and elevation); a smooth map DOESN’T guarantee accuracy

first law of geography connection

nearby places often have related values; “everything is related to everything else, but near things are more related than distant things”

spatial autocorrelation

describes how a variable relates to itself across locations; shade, barriers, and direction can change local similarity

geostatistics

a class of statistics used to analyze and predict the values associated with spatial or spatiotemporal phenomena

geostatistical analyst

provides a set of tools for modeling spatial phenomena

geostatistical interpolation workflow

1. explore: check values, locations, and sampling patterns

2. predict: choose a method and its settings

3. validate: compare hidden observations with predictions

how to pick a chart?

• histogram: how are values distributed?

• line chart: how do values change over time?

• QQ plot: how does the distribution compare with a reference?

deterministic methods

• IDW

• natural neighbor

• spline and radial basis functions

• local and global polynomial

geostatistical methods

• kriging and cokriging

• moving window kriging

specialized tools

• topo to raster and barrier methods

• simulations generates possible realizations

IDW

• gives closer observations more influence (higher power)

• spreads influence across more distant observations (lower power)

• uses validation, not visual smoothness alone (select power)

ex. a ridge may separate temperature patterns; a river isn’t automatically an air-temperature barrier

Kriging

• uses a fitted spatial dependence model and the arrangement of observations to estimate weights

• clustered observations can contain overlapping information

• supports prediction standard errors under its assumptions

ex. ten thermometers on one block may add less information than ten spread across an neighborhood

spatial structure: modeling before prediction

1. summarize pairwise differences by separation distance

2. fit a suitable semivariogram or covariance model

3. use that model and the sample geometry in kriging

semivariance

half the average squared distance for pairs at a similar distance

nugget

measurement error and/or unresolved small-scale variation

site selection

if you want to locate where something SHOULD be

raster reclassification - single layer analysis

often used to simplify or change the interpretation of raster data by changing a single value to a new value, or grouping the ranges of values into single values (ex. assigning a value of 1 to cells that have values of 1-50, 2 to cells that range from 51-100, etc.)

important to weighted suitability analysis because raster data will have differing values for the phenomena they represent

raster reclassification - multiple layer analysis

relational raster overlay - pixel or grid cell values in each raster datasets are combined using relational operators to produce a new value in the output dataset (cells meeting the criteria, or true cells = 1; cells not meeting the criteria, or false cells = 0)

local operations

operations performed on a single, target cell on single OR multiple rasters

neighborhood operations

operations performed on a central, target cell AND surrounding cells

zonal operations

• operations performed on groups of cells of similar value OR like features, called zones

• create a new output raster dataset that contains input cell values for each zone in the zonal raster

global operations

operations performed over the entire extent of a dataset

terrain/elevation analysis

vector or raster dataset that contains an attribute value for every locale throughout its extent

slope map

a map depicting rasterized slope values throughout its extent

aspect map

a map depicting rasterized aspect values throughout its extent

hillshade map

a map showing relative relief based on elevation of the desired area, the illumination source of which can be rotated and tilted to any desired angle for viewing

digital elevation model (DEM)

a digital model or 3D representation of a terrain’s surface - commonly for a planet (including Earth), moon, or asteroid

uses of DEM

1. slope: steepness (ex. compare gradients along candidate trails)

2. aspect: downhill-facing direction (ex. identify south-facing fields); azimuth angle

3. drainage and contributing area: flow direction/accumulation/watershed (ex. map the area draining to a chosen culvert)

4. visibility from a viewpoint: geodesic viewshed uses terrain and observer height (ex. map terrain visible from a lookout tower)

raster data processing

a set of transformation steps performed on raster data to extract meaningful information using a wide range of algorithms:

• visual appearance of the raster data can be altered

• transform the coordinate system

• create a new dataset

ModelBuilder

a visual programming environment for geoprocessing workflows; add datasets and tools, then connect them in the order the analysis requires

variable

data or a value, such as distance

tool

an operation on inputs

connector

the relationship linking an input or result to a tool

group

a visual container for related elements

environment setting hierarchy

override priority the most specific, supported, explicit setting wins

1. model process: a setting for one tool inside the model

2. model: a setting saved with the workflow

3. tool run: a setting when launching the model as a tool

4. project/application: defaults passed to the workflow

ex. a process-level output coordinate system overrides the model default

ArcGIS Online

• cloud-based mapping, analysis, and presentation

• create interactive maps and web applications

• publish maps and layers from desktop

• share maps, data, and workflows with the public or specific individuals and organizations