Table 1. The geographic data sets with descriptions and sources for the logistic regression model used to predict the probability of development

Data Description

Geographic Data Set

Source

Random integer grid

RANDOM

Created in GRID TM and masked to include only areas within the parish

Points remained forested or changed to urban (0 or 1 respectively)

CHANGE

Change detection analysis performed in GRID using Stennis Space Center image (1981) and the USGS Southern Science Center image (1993)

Area of contiguous pine or mixed forest cover (ha)

AREA

Polygon coverages created from Stennis Space Center (for 1981) and USGS Southern Science Center (for 1993) satellite images using ArcInfo

Population density by census block

POPDEN

Block locations were obtained TIGER TM 92 line files and changed to a coverage using Arc/Info TM. block areas determined from Arc/Info. STF1A CD from the US Census Bureau provided total population in each block in 1990. St. Tammany Parish Office of Information Services Provided 2003 population estimates by Census Tracts. Population density was calculated using Arc/Info

Euclidean distance from sample point to nearest state of federal highway (km)

ROADS

Federal and state highways identified through Wessex First Street TM. GRID was used to determine Euclidean distances

Euclidean distance to the major access points to New Orleans (km)

ACCESS

Digitized points from the screen guided by the parish roads information obtained from Wessex First Street. GRID was used to determine Euclidean distances

Euclidean distance to interchanges on limited access highways (km)

INTERCHANGE

Interchanges were identified through Wessex First Street. GRID was used to determine Euclidean distances