Transcript Keynote - Spatial Information Systems (Basis)

GIS Technology in Transition
Moving Maps to Mapped Data, Spatial Analysis and Beyond
2003 Northwest GIS User Group Meeting
September 16, 2003 – Skamania Lodge, Stevenson, Washington
Presented by
Joseph K. Berry
GIS is more different than it is similar to
traditional mapping and data analysis
Berry & Associates // Spatial Information Systems
2000 South College Ave, Suite 300, Fort Collins, CO 80525
Phone: (970) 215-0825 Email: [email protected]
…visit our website at www.innovativegis.com/basis
Historical Setting and GIS Evolution
Traditional Mapping
manually drafted map
Computer Mapping
automates the cartographic process (70s)
Spatial Database Management
links computer mapping techniques with
traditional database capabilities (80s)
GIS Modeling
representation of relationships within
and among mapped data (90s)
(Berry)
Where is What
…and Wow
WHERE – Digital Map
Mapped data can be queried by interacting
with the map (where) or database (what)
1) Select forest type Aspen, SP1= Aw
2) Select tall Aspen stands, Height > 20m
WHAT -- database
Query Builder
Map Display
Connectivity and
Map Delivery
SDT, www.spatialdatatech.com
Indelix, www.idelix.com
(Berry)
Where is What and Wow to…
Why and So What
WHERE IS WHAT
Vector-based processing provides Mapping
and Geo-Query capabilities that repackage
existing spatial data as reports and displays
WHY AND SO WHAT
Discrete Objects
Descriptive Mapping
Grid-based
processing
provides Map Analysis
capabilities that derive
new information on
relationships within and
among mapped data
Continuous Surfaces
Prescriptive Mapping
(Berry)
Simple Erosion Model
…GIS Modeling involves logical sequencing of map
analysis operations
…a Command
Macro Language
consists of a
graphical
interface for
entering, editing,
executing,
documenting,
storing and
retrieving a GIS
Model
Script
Logic
(Berry)
Variable-Width Buffer (Sediment loading)
Simple Buffer
Effectively far away, though
right near a stream
…how can that be?
…what about different soils?
…what about roughness?
…or time of year?
(Berry)
Micro Terrain Analysis
“Map-ematics”
Calculation of slope considers
the arrangement and magnitude
of elevation differences
Characterizing Slope (and Aspect)
A digital terrain surface is formed by assigning
an elevation value to each cell in an analysis
grid. The “slant” of the terrain at any location
can be calculated– inclination of a plane fitted
to the elevation values of the immediate vicinity.
Characterizing Surface Flow
The relative amount of water passing through
each grid cell is determined by simulating a
drop of water landing in each cell and
proceeding downhill by the steepest path. The
number of paths crossing each location
identifies the total uphill confluence.
(See Map Analysis, “Topic 11” for more information)
(Berry)
Map Analysis
Surface Modeling maps the spatial distribution and pattern of point data…
 Map Generalization— characterizes spatial trends (e.g., titled plane)
 Spatial Interpolation— deriving spatial distributions (e.g., IDW, Krig)
 Other— roving window/facets (e.g., density surface; tessellation)
Data Mining investigates the “numerical” relationships in mapped data…
 Descriptive— aggregate statistics (e.g., average/stdev, similarity, clustering)
 Predictive— relationships among maps (e.g., regression)
 Prescriptive— appropriate actions (e.g., optimization)
Spatial Analysis investigates the “contextual” relationships in mapped data…
 Reclassify— reassigning map values (position; value; size, shape; contiguity)
 Overlay— map overlay (point-by-point; region-wide; map-wide)
 Distance— proximity and connectivity (movement; optimal paths; visibility)
 Neighbors— ”roving windows” (slope/aspect; diversity; anomaly)
(Berry)
Map Analysis
Surface Modeling maps the spatial distribution and pattern of point data…
 Map Generalization— characterizes spatial trends (e.g., titled plane)
 Spatial Interpolation— deriving spatial distributions (e.g., IDW, Krig)
 Other— roving window/facets (e.g., density surface; tessellation)
Data Mining investigates the “numerical” relationships in mapped data…
 Descriptive— aggregate statistics (e.g., average/stdev, similarity, clustering)
 Predictive— relationships among maps (e.g., regression)
 Prescriptive— appropriate actions (e.g., optimization)
Spatial Analysis investigates the “contextual” relationships in mapped data…
 Reclassify— reassigning map values (position; value; size, shape; contiguity)
 Overlay— map overlay (point-by-point; region-wide; map-wide)
 Distance— proximity and connectivity (movement; optimal paths; visibility)
 Neighbors— ”roving windows” (slope/aspect; diversity; anomaly)
(Berry)
Spatial Interpolation (Geographic Distribution)
“Surface Modeling” is similar to slapping a big chunk of modeler’s clay
over the “data spikes,” then taking a knife and cutting away the excess to
leave a continuous surface that encapsulates the peaks and valleys implied
by the spatial pattern of the field samples
…nearby things are more alike than distant things
(Berry)
Mapping the Variance
Non-Spatial statistics seeks
the “typical” condition and
applies uniformly
throughout geographic
space-- AVERAGE
Spatial Statistics seeks to
map the variance
Spatial Interpolation is
similar to throwing a
blanket over the “data
spikes” to conforming to
the geographic pattern
of the data.
(Berry)
Map Analysis
Surface Modeling maps the spatial distribution and pattern of point data…
 Map Generalization— characterizes spatial trends (e.g., titled plane)
 Spatial Interpolation— deriving spatial distributions (e.g., IDW, Krig)
 Other— roving window/facets (e.g., density surface; tessellation)
Data Mining investigates the “numerical” relationships in mapped data…
 Descriptive— aggregate statistics (e.g., average/stdev, similarity, clustering)
 Predictive— relationships among maps (e.g., regression)
 Prescriptive— appropriate actions (e.g., optimization)
Spatial Analysis investigates the “contextual” relationships in mapped data…
 Reclassify— reassigning map values (position; value; size, shape; contiguity)
 Overlay— map overlay (point-by-point; region-wide; map-wide)
 Distance— proximity and connectivity (movement; optimal paths; visibility)
 Neighbors— ”roving windows” (slope/aspect; diversity; anomaly)
(Berry)
Visualizing Spatial Relationships
Interpolated Spatial Distribution
Phosphorous (P)
What spatial
relationships do
you see?
…do relatively high levels
of P often occur with high
levels of K and N?
…how often?
…where?
(Berry)
Clustering Maps
…groups of “floating balls” in data space identify locations in the field
with similar data patterns– data zones
(Berry)
The Precision Ag Process (Fertility example)
As a combine moves through a field 1) it uses GPS to check its location then 2) checks
the yield at that location to 3) create a continuous map of the yield variation every few
feet. This map 4) is combined with soil, terrain and other
Steps 1) – 3)
maps to derive a 5) “Prescription Map” that is used to
6) adjust fertilization levels every few feet in the field.
Step 5)
On-the-Fly
Yield Map
Step 4)
Prescription Map
Map Analysis
Cyber-Farmer, Circa 1992
Farm dB
Zone 3
Zone 2
Zone 1
Variable Rate Application
Step 6)
(Berry)
Spatial Data Mining
…making sense out of a map stack
Mapped data that
exhibits high spatial
dependency create
strong prediction
functions. As in
traditional statistical
analysis, spatial
relationships can be
used to predict
outcomes
…the difference is
that spatial statistics
predicts where
responses will be
high or low
(Berry)
Precision Ag to Precision Conservation
From a Field perspective to Watershed, Landscape and Ecosystem perspective
Precision Conservation
Precision Ag
Wind Erosion
Chemicals
SURFACE MODELING
SPATIAL DATA MINING
Soil
Erosion
Runoff
Leaching
Leaching
Leaching
SPATIAL ANALYSIS
3-dimensional
2-dimensional
Interconnected Perspective
Isolated Perspective
(Berry)
Map Analysis
Surface Modeling maps the spatial distribution and pattern of point data…
 Map Generalization— characterizes spatial trends (e.g., titled plane)
 Spatial Interpolation— deriving spatial distributions (e.g., IDW, Krig)
 Other— roving window/facets (e.g., density surface; tessellation)
Data Mining investigates the “numerical” relationships in mapped data…
 Descriptive— aggregate statistics (e.g., average/stdev, similarity, clustering)
 Predictive— relationships among maps (e.g., regression)
 Prescriptive— appropriate actions (e.g., optimization)
Spatial Analysis investigates the “contextual” relationships in mapped data…
 Reclassify— reassigning map values (position; value; size, shape; contiguity)
 Overlay— map overlay (point-by-point; region-wide; map-wide)
 Distance— proximity and connectivity (movement; optimal paths; visibility)
 Neighbors— ”roving windows” (slope/aspect; diversity; anomaly)
(Berry)
Spill Migration Modeling
Elevation
Surface
Overland Flow Model
1) Pipeline
X
2) Spill Point #1
X
5) Flowing
Water
3) HCA
4) Report
HCA Impact
The Pipeline is positioned on the Elevation surface
2) Flow from Spill Points along the pipeline are simulated
1)
High Consequence Areas (HCA) are identified
4) A Report is written identifying flow paths that cross HCA areas
3)
5)
Overland flow is halted when Flowing Water is encountered (Channel Flow Model)
(Berry)
Types of Surface Flows
Common sense suggests that “water flows downhill”
however the corollary is “…but not always the same way.”
(Berry)
Characterizing Overland Flow and Quantity
Intervening terrain and conditions form Flow Impedance and Quantity
maps that are used to estimate flow time and retention
(Berry)
Simulating Different Product Types
Physical properties combine with terrain/conditions to model
the flow of different product types
Flow Velocity is a function of—
Specific Gravity (p), Viscosity (n) and Depth (h) of product
Slope Angle (spatial variable computed for each grid cell)
(Berry)
Characterizing Impacted Areas
The minimum time for flows from
all spills… characterizes the impact
for the High Consequence Areas
Flows from spill 1, 2 and 3
Drinking water HCA
Impacted portion of the Drinking water HCA
(Berry)
Modeling Stream Channel Flow
3)
Impacted
HCA Times
HCA
4)
Report of Impacted HCA’s
Base
Point
Out= 9.86 hr
0 hr
HCA
In= 11.46 hr
HCA
HCA
X = 12.10 + .36 = 12.46 hr
away from Base Point
2)
X
13.1 hr
Channel Flow
Model
HCA
.12
.27
X
.14
1)
Channel Flow Time
10.1 hr
10.8 hr
.25
.72
1)
9.6 hr
11.2 hr
8.4 hr
11.2 hr
.12
13.1 hr
7.3 hr
13.6 hr
Overland Flow
Entry Time
Overland Flow
(2.5 hours)
.78
HCA
2.5 + (12.46 -11.46) =
3.5 hours total
Channel Flow times along stream network segments are added
2)
3)
4)
Overland Flow time and quantity at entry is noted
Impacted High Consequence Areas (HCA) are identified
Report is written identifying flow paths that cross HCA areas
(Berry)
Modeling Customer Flow
…customer flow along a road
network is similar to water
flowing in a stream channel
…a Travel-time Map identifies
the time to travel from
anywhere to a store
(Berry)
Competition Analysis
… travel-time
surfaces for two
different stores
… can be compared for relative travel-time advantage
(Berry)
Transmission Line Siting Model
Existing Powerline
Goal – identify the best route for an electric
transmission line that considers various criteria
for minimizing adverse impacts.
Proposed
Substation
Houses
Criteria – the transmission line route should…
 Avoid areas of high
Roads
housing density
Sensitive Areas
 Avoid areas that are far from
roads
Elevation
 Avoid areas within
or near sensitive areas
 Avoid areas of high visual
exposure to houses
Houses
(Berry)
Routing and Optimal Paths
PROPOSED
SUBSTATION
AVOID AREAS OF HIGH
VISUAL EXPOSURE
TO HOUSES
(END)
EXISTING
POWERLINE
(START)
ELEVATION
MOST
PREFERRED
ROUTE
ACCUMULATED
COST
SURFACE
HOUSES
VISUAL
EXPOSURE
TO HOUSES
DISCRETE
COST
MAP
Visual Exposure levels (0-40 times
seen) are translated into values indicating
relative cost (1=low to 9=high) for siting a
transmission line at every location in the
project area.
Step 1.
Accumulated
Cost from the
existing powerline to
all other locations is
generated based on
the Discrete Cost
map.
Step 2.
The steepest
downhill path from
the Substation over
the Accumulated
Cost surface
identifies the “least
cost path”—
Step 3.
Most Preferred Route
avoiding areas of high
visual exposure
(Berry)
Considering Multiple Criteria
Step 3
Discrete Cost
AVOID AREAS OF HIGH HOUSING DENSITY
HOUSING
HOUSING
DENSITY
AVOID
AREAS
OF HIGH
HOUSING
DENSITY
AVOID AREAS THAT ARE FAR FROM ROADS
ROADS
PROXIMITY
TO ROADS
AVOID
AREAS
THAT ARE
FAR FROM
ROADS
AVOID AREAS IN OR NEAR SENSITIVE AREAS
SENSITIVE
AREAS
PROXIMITY
TO
SENSITIVE
AREAS
HOUSING
Base
Maps
START
STARTING
LOCATION
AVG_COST
VISUAL
EXPOSURE
TO HOUSES
AVOID
AREAS
OF HIGH
VISUAL
EXPOSURE
Derived
Maps
Cost/Avoidance
Maps
ENDING
LOCATION
ACUMM_COST
ACCUMULATION
SURFACE
BEST_ROUTE
MOST
PREFERRED
ROUTE
Step 3
Steepest Path
AVERAGE
COST
AVOID
AREAS
IN OR NEAR
SENSITIVE
AREAS
AVOID AREAS OF HIGH VISUAL EXPOSURE
ELEVATION
END
Step 2
Accumulated Cost
Criteria – the transmission line route should
avoid…




Areas of high housing density
Areas that are far from roads
Areas within or near sensitive areas
Areas of high visual exposure to houses
(Berry)
Considering Multiple Criteria
Step 3
Steepest Path
AVOID AREAS OF HIGH HOUSING DENSITY
END
BEST_ROUTE
START
AVOID AREAS THAT ARE FAR FROM ROADS
ACUMM_COST
AVG_COST
Step 3
Steepest Path
AVOID AREAS IN OR NEAR SENSITIVE AREAS
Step 2
Accumulated Cost
AVOID AREAS OF HIGH VISUAL EXPOSURE
Base
Maps
Derived
Maps
Cost/Avoidance
Maps
Criteria – the transmission line route should
avoid…




Areas of high housing density
Areas that are far from roads
Areas within or near sensitive areas
Areas of high visual exposure to houses
(Berry)
Step 1
Discrete Preference Map
… identifies the relative preference of locating a
transmission line at any location throughout a project
area considering multiple criteria
Least
Most
Preferred
…average of the
four individual
preference maps
(Berry)
Step 2
Accumulated Preference Map
… identifies the preference to construct the
preferred transmission line from a starting
location to everywhere in a project area
Splash Algorithm – like tossing a stick into a pond with waves
emanating out and accumulating costs as the wave front moves
(Berry)
Step 3
Most Preferred Route
… the steepest downhill path over the
accumulated preference surface identifies the
most preferred route — minimizes areas to avoid
Preferred
Route
(Berry)
Siting Model Flowchart (Model Logic)
Model logic is captured in a flowchart where the boxes represent
maps and lines identify processing steps leading to a spatial solution
Rankings
Weights
Avoid areas of…
High Housing
Density
Far from Roads
In or Near
Sensitive Areas
High Visual
Exposure
…but what is high
housing density and
how important is it?
…etc?
(Berry)
Calibrating Map Layers (Relative Preferences)
Model calibration refers to establishing a consistent scale from 1
(most preferred) to 9 (least preferred) for rating each map layer
1 for 0 to 5 houses
…group consensus is that
low housing density is
most preferred
The Delphi Process
is used to achieve
consensus among
group participants.
It is a structured
method involving
iterative use of
anonymous
questionnaires and
controlled feedback
with statistical
aggregation of
group response.
(Berry)
Weighting Map Layers (Relative Importance)
Model weighting establishes the relative importance among map
layers (model criteria) on a multiplicative scale
…group consensus is that housing density is very important (10.38 times more important than sensitive areas)
HD * 10.38
R * 3.23
SA * 1.00
VE * 10.64
The Analytical Hierarchy Process (AHP) establishes relative importance among by
mathematically summarizing paired comparisons of map layers’ importance.
(Berry)
Generating Alternate Routes (changing weights)
The model is run using three
different sets of weights for the
map layers—
…to generate three alternative
routes (draped over Elevation)
(Berry)
Transitioning Beyond Mapping
Where is What and Wow
mapping, geo-query, delivery
and display…
Surface Modeling
maps the spatial distribution
and pattern of point data…
Data Mining
investigates the “numerical”
relationships in mapped data…
Spatial Analysis
investigates the “contextual”
relationships in mapped data…
(Berry)
GIS Technology in Transition
GIS technology is transitioning from
Where is What and Wow
…to Why and
So What
…we’ve covered a lot,
any questions?
…for more importation online, see
(Berry)