1

GEOG 300 - Introduction to GIS

March 12, 2003

- Analysis functions continued- Overlay continued; boundary operations; examples of analysis within projects

Map Overlay in GIS (review)

Definition: - the process of taking two different thematic maps of the same area and overlaying them one on top of the other to form a new map layer.

“The ability to integrate data from two sources using map over is perhaps the key GIS analysis function”.

The Concept of Map Overlay

• Map overlay addresses the relationship of the intersection and overlap between spatial features.

• Map overlay combines the spatial and attribute data of two input themes.

Three input feature types, overlay cover is always polygon:

1)Point-in-polygon, points are output2)line-in-polygon, lines are output3)polygon-in-polygon polygons are output

Overlay Methods:

UNION, INTERSECT, IDENTITY

Union• = “or” Boolean operator.• creates new coverage by overlaying two polygon coverages. • the output coverage contains;

-the combined polygons-attributes of both coverages

• all coverages must be polygons

union

What is the combination of the stands and fire area ?

Intersect

• = ‘and’ Boolean operator. • creates a new coverage by overlaying two sets of features. • the output coverage contains only those portions of features that are in the area occupied by both the input and intersect coverages.

intersect

Where do the stands and the fire intersect?

‘and’ decreases selected set or output, while ‘or’ - see Union - increases

Identity• creates a new coverage by overlaying two sets of features. • the output coverage contains

-all of the input features -keeps only those portions of identity coverage features that overlap the input coverage.

Where did fire occur in the stands database ?

identity

2

Different Ways of Looking at Polygon Overlay - illustrations

1. Bolean Overlay (exclusionary variables -and,or)- illustration of different ways to pose question that will result in same outcome

2. Mathematically based overlay - where some input factors may be weighted more than others

UNION is ….

OR

INTERSECT is ...

AND

Example that follows: how one might use UNION or INTERSECT to identify land parcels that are suitable for development …. Two ways of asking ...1) Where are areas where development cannot occur (UNION)?

2) Where are areas where development can occur (INTERSECT)?

EndangeredSpecies

Prime Agricultural

Land

FederalLand

Bolean (Logical) Overlay Process - Potential for Urban Development

Union

No PotentialIFEndangered SpporPrime Ag LandorFed Land

Good PotentialIFNot Endangered SppandNot Prime Ag LandandNot Fed Land

Intersect

No Potential(union)

GoodPotential(intersect)

(VS)Conditions

Areas where development not possible …

1) endangered sp. Habitat;

2) prime agricultural land;

3) federal land eg national park

Two approaches: Another example … also includes looking at weighted overlay

‘Problem statement’ --

Identify areas where it is possible for housing development to occur.

Criteria:

1) areas of lower slope are better than areas with steeper slope

2) must be outside of riparian zone by at least 1 mile

3) areas of greater distance from incinerator are better than closer areas

Creation of input coverages to model reflects criteria:

Create slope coverage, RECLASSIFY values to create slope classes ...

For purposes of model -

Higher class values reflect more desirable land:

Class Slope Value

1 <15% 3

2 16 - 30% 1

3 >=30% 0

TinPOLYGON

with

Percent_slope

tinarc

310

SLOPE - criteria 1- areas of lower slope better than areas of higher slope

Creation of input coverages to model reflects criteria:

RIPARIAN AREAS - criteria 2 - housing must be outside of riparian zones by 1 mile

TRIM twtrreselect riparian buffer

Reselect and reclassify values

For purposes of model -Higher class values reflect more desirable land:

Class Riparian Value

1 No 1

2 Yes 0

10 0

1

3

Creation of input coverages to model reflects criteria:

Distance to incinerator - criteria 3 - areas greater distance are favoured over areas closer

incineratorbuffer

Incinerator

buffer

310

For purposes of model -Higher class values reflect more desirable land:

Class Distance Value

1 <1 miles 0

2 1.1 - 2 1

3 > 2 3

Reselect and reclassify values

UNION all resultant coverages (keeps all input features)

Slopes

+Riparian

union

+

Union1

Incinerator

union

Interpreting the resultsMATHEMATICAL OPERATION

AD

DITI

ON

MUL

TIPLI

CA

TION

OR

310

10 0

1

310

4

5

51

7 4

53

3

2

0

3

30

9 0

3

1

0

0

Some references of interest (1):

Agee, J.K., S.C.F. Stitt, M. Nyquist, and R. Root. 1989. “A geographic analysis of historical grizzly bear sighting in the north cascades.” Photogrammetric Engineering & Remote Sensing. Vol. 55 (11), pp. 1637 - 42.

Bian, Ling and Eric West. 1997. “GIS Modeling of Elk Calving Habitatin a Prairie Environment with Statistics.” Photogrammetric Engineering & Remote Sensing. Vol. 63 (2), pp. 161 - 7.

Breininger, D.R., M.J. Provancha, and R.B. Smith. 1991. “Mapping Florida scrub jay habitat for purposes of land-use management.” Photogrammetric Engineering & Remote Sensing. Vol. 57 (11), pp. 1467 - 74.

Clark, J.D., J.E. Dunn, and K.G. Smith. 1993. “A multivariate model of female black b0ear habitat use for a geographic information system.” Journal of Wildlife Management. Vol. 57 (3), pp. 519 - 57.

Some references of interest (2):

Jensen, John R., Sunil Narumalani, Oliver Weatherbee, and Keith S. MorrisJr. 1992. “Predictive Modeling of Cattail and Waterlily Distribution in a South Carolina Reservoir Using GIS.” Photogrammetric Engineering & Remote Sensing. Vol. 58 (11), pp. 1561- 68.

Mladenoff, D.J., T.A. Sickley, R.G. Haight, and A.P. Wydeyen. 1995. “A regional landscape analysis and prediction of favorable gray wolf habitat in the Northern Great Lakes region.” Conservation Biology. Vol. 9(2), pp. 279 - 294.

Pereira, J.M.C. and R.M. Itami. 1991. “GIS-based habitat modeling using logistic multiple regression: A study of the Mt. Graham red squirrel.”Photogrammetric Engineering & Remote Sensing. Vol. 57 (11), pp. 1475-1486.

Problems in Map Overlay

• Remember … Themes to be overlaid must be spatially registered before the overlay process (quality control in pre-process).

• Digitizing errors can cause slivers.• Propagation of errors: errors will accumulate during the

map overlay process (ie something is coded incorrectly in attribute table)

• Processing creates many polygons … Consider other commands, CLIP, ERASE, SPLIT, MERGE, DISSOLVE, etc., as essential part of post-processing …. ‘Boundary operations’

4

First, Slivers (review)• Slivers may occur as a result of imperfect digitization of

spatial features (review);• can also be created during Overlay -- unions, etc.• In some cases, slivers can be eliminated by selecting only

those features that have a large size; minimum mapping unit.• Use ELIMINATE command at the arc level

Eliminate

• Merges selected polygons with neighboring polygons that have the largest shared border between them, or that have the largest area.

• Uses same kind of syntax as the Arc:reselect command, with a prompt for a logical expression. Output is new coverage

• Most often used to remove ‘sliver’ polygons, but can be used to combine polygons, by ‘absorbing’one type by another ‘type’ ….

eg. ARC: ELIMINATE CARVEGE VEGE2: resel veg_type = ‘Mesophyll Vine Forest’

The 5 vine forest polygons (dark blue wide stripes) have been merged with those neighbouring polygons with the greatest shared boundary - perhaps this vegetation type, for purposes of analysis, was better combined with neighbouring vegetation types.

Boundary Operations - Coverage Manipulation

• Dissolve merges adjacent polygons, lines or regions which have the same value for a specified item.

- In the case of a polygon coverage, it removes arcs between polygons containing equal values for the dissolve item. In the case of arcs, nodes between arcs are removed.

• DISSOLVE is used to create a simplified coverage from one which is more complex. While the input coverage may contain information concerning many feature attributes, the output coverage contains information only about the dissolve item.

Map Coverage Manipulation• Dissolve

– Removes boundaries between polygons that have the same value of a selected attribute.

• extracts a subregion of a coverage that intersects with a secondcoverage, called a clip coverage

• the features of the input coverage, but not the clip coverage, are retained

• therefore the feature attribute table for the output coverage contains the same items as the input coverage attribute table.

• topology is rebuilt for the output coverage

• input features can be polygons, lines or points.

• Clip cover must be a polygon coverage.

• the output coverage has the same feature type as the input coverage.

Boundary Operations - Coverage ManipulationCLIP

5

Map Coverage Manipulation• Clip

– Includes only those features of the input theme that are within the spatial extent of the clip theme.

clip theme(Upper Salmon River Watershed)

input theme(Idaho counties)

result(watershed divided into counties)

Clip - 2

Unit of interest is one watershed

Desired output for data management, analysis and display

steps

Create clip coverage (must be polygon) in arcedit by digitzing polygon around watershed (follow height of land ?)

CLIP - arc command which requires using ‘clip’ coverage (boundary of interest) and coverage where features stored … to output new feature coverage containing only subset of interest

ERASE is very similar to Clip, except that the input coverage features that overlap with the erase coverage polygons are erased rather than preserved.

• Again, the input coverage features can be polygons, lines, or points; but erase coverage features must be polygons.

• Output coverage features are of the same class as the input coverage features.

Map Coverage Manipulation• Erase coverage- removes part of the inside of a coverage

The boundary of an existing coverage is used to erase features from another coverage and create a new coverage.

• Erase coverage -2- this may be used as part of buffer analysis steps

streams

Sewer line

Criteria requires that housing is outside of streams, but within a set distance of sewer lines (black background chosen to show difference).

You may create buffer polygons around streams and sewer lines as first step ...

Continued ….

Use erasecov to delineate areas that are outside of streams, but within specified distance of sewer lines ...

Output contains only area meeting both stipulations from above

6

Map Coverage Manipulation• Merge

– Creates a new theme by piecing together two or more themes;

– however, the boundaries between the themes remain intact.

Separate themesSingle theme

Idaho counties

Snake River Plain Aquifer