Transcript 9MapAttributes

Mapping “what?”
Instead of “where?”
Two types of geographic data:
Location

Attributes
Horizontal location
 Vertical location

Vegetation types

Soil types

Land cover

Number of people
Attributes:
characteristics of a
place
Back to a previous topic …

Reference maps

Thematic maps

Emphasize location


Place names are
attributes, but Location
data is key

Emphasize patterns
Need location data, but
attribute data is key
Leading candidate, by
county
Mapping attributes
Sources of attribute data
 Focus on census data

Examples of attributes:
How would
you collect
data for maps
of these
phenomena?

Vegetation types

Soil types

Land cover (forested, grass, asphalt, etc)

Land use (wilderness, recreational,
residential, commercial, etc.)

Number of people

People’s typical ages, incomes, etc!

Temperature

Precipitation
Sources of attribute data:
Satellite-based scanners

Remote sensing:

Land cover

Vegetation types

Soil types

Precipitation

Change over time, for
any of these topics
Aerial photography
More sources of attribute data:

Ground surveys



Vegetation, land cover
Road type, land use
Count-based
surveys


Population (human)
Population (animal)
Count-based Surveys:
some definitions

Census


All members of a population
Sampling

Inferences about whole population based
on some
Example: Decennial U.S. Census

Why count?

It’s in the Constitution

Mandate of Government Agencies

US Census Bureau: “to be the preeminent
collector and provider of timely, relevant, and
quality data about the people and economy of
the United States.”
2000 Reapportionment
US Census … Issue 1

Challenges to getting a complete count:

Time-consuming, laborious, costly

Dishonest or uncompleted responses

Homeless, transient populations, illegal
immigrants
Challenges of a complete count

Non-response follow-up
enumerations (NRFU) =
expensive

Visit every house without a
response

Non-respondents usually
urban, poor, minority

Solution?
Stratified sampling
Traditional sampling
approach: “Random sample”
sample site

More efficient and
accurate

Widely used

Example: survey of
tree species in a forest
Stratified sampling
“Stratified sample”

biodiversity
low
medium
high
sample site
The technique:

Divide a region into
homogeneous regions

Assign sample sites to
each stratum in
proportion to what each
area is thought to
contribute
Complete Count v. Sampling

Supreme Court ruling on use of sampling for
census NRFU’s:
WASHINGTON (CNN) – The U.S. Supreme Court
on Monday ruled out the use of statistical sampling
to adjust the 2000 census to make up for an
expected undercount.
The 5-4 ruling was a defeat for the Clinton
administration, which had hoped statistical sampling
would add population -- and subsequently House
members -- to areas that traditionally vote
Democratic.
-AP, Jan 25, 1999
Issue 2: Reporting census data

4
Aggregation: combining
counts into spatial units

Rather than recording
precise location of
individuals

Less costly

Preserves confidentiality
3
10 8
6
12
2
US census aggregation units

States


Counties
 Census Tracts
– Census Block
Groups
 Census Blocks
Challenge: changes

Before 1990, census blocks
and tracts only in some
areas

As population increases,
units are divided
Challenge: Lost Detail

Can aggregation lead to misrepresentation?
8
8
8
8
8
8
8
8
Accurate representation
difficult to achieve
Usually convenient
regions are used
Conclusions on Count-based
Surveys:
Not “totally accurate”
 Reporting and map representation
challenges
 Next Lecture…
