Geographic Information Systems

1
Geographic Information Systems Methodology for
Assessing Socioeconomic And Environmental Impacts
Studies. By Osei Yeboah and Upton
Hatch Auburn University.
2

• Current Research Activities
• 1. Collection and Conversion of Socioeconomic
• Data into Spatial Format.
• Data Type Examples Age, Income, Population,
• Income, Employment, Business, Housing, and
• Race.
• Scale County, Zip Code, and Watershed Levels.
• Time Period 1970-2000, 2001-2025.

3

• Environmental Justice Issues
• Effects of Interstate Highways and the
• Distribution Of Superfund Sites.
• Toxic Waste Sites in Alabama (Zip
• Code Level).
• Spatial Distribution of Pulp and Paper
• Mills in the Southeast.

4

• 3. Valuation of Technology
• Cost-Effectiveness Analysis of the Use
• of Remote Sensing Technology in
• Environmental Impact Studies Relative
• to the Conventional Methods.

5
(No Transcript)
6
Results of Multivariate Logit Regression
1980 Estimated Asymptotic Variable
Coefficient Std. Errors PBLACK -0.0014 0.0126 P
HISP -0.5808 0.3672 PDENS -0.0009 0.0012 PUEM -1
.9126 2.2158 HVAL -0.00005 0. 00002 HINC
0.0001 0.00005 PMANU -2.6359 2.2084
7
Estimated Asymptotic Variable
Coefficient Std. Errors PURB 0.0289 0.0104
ROAD 1.0262 0.3974 Observation 403 Pe
rcent of Right Predictions 90.81 Log Likelihood
Test Statistics 38.69 Maddala
R2 0.090 pless than 0.05 pless
than 0.01
8
Results of Multivariate Logit Regression
1990 Estimated Asymptotic Variable
Coefficient Std. Errors PBLACK -0.0137 0.0121
PHISP -0.4136 0.2508 PDENS -0.0011 0.0011 PUE
M 0.9443 2.2791 HVAL -0.00004 0.00002
HINC 0.00007 0.00004
9
Estimated Asymptotic Variable
Coefficient Std. Errors PMANU -3.5653 2.2273
PURB 0.0327 0.0101 ROAD 0.9857 0.3961
Observation 403 Percent of right
Predictions 90.57 Log Likelihood Test
Statistics 38.31 Maddala R2 0.092
10
(No Transcript)
11

• Scaling County Level Data Down
• to Zip Code Level
• 1990 Zip Codes Shape Files are Used to
• Approximate the Proportions of data in each
• Zip Code that form the County Data.
• County Data Rescaled to Cell Level and the
• Number of Cells in Zip Code Estimated.
• Each Zip Code is then Converted to a Shape
• File.
• Software Two Themes in Arcview.

12
(No Transcript)
13
Results of the Final Probit Model Estimated A
symptotic Variable Coefficient Std.
Errors PBLACK -0.0045 0.0057 PHISP -0.1267 0.1
105 RIVER 0.3950 0.1797 INCAP -0.1540 0.06
46 PURB 0.0155 0.0040
PMANU -0.4695 0.9658 PUEM 5.1975 10.7110 WAG
E -0.8775 1.1993
14
Observation 454 Percentage of Right
Predictions 86.57 Log Likelihood Test
Statistics 29.98 Maddala R2 0.12
pless than 0.05 pless than 0.01