Formalization of Factors

Modeling urban pressure on the ALR involves the simplification of highly complex economic and cultural factors that affect land use. In order to formalize these factors in a simplified manner, three entities were selected based on the available data.

Population Density

The measure of density, or the ratio of individuals per area is a factor that strongly influences urban growth. Studies have shown that due to advances in transportation and communication technologies, individuals have more choices when it comes to choosing a residential location, and tend to favour less urban settings within 30 miles of urban centres (La Gro 1994 ).

The spatial incidence of high concentrations of people represent the interests and economic forces of urbanism

Road Networks

As stated previously, The presence of road networks have a significant impact on Agricultural land. The higher costs and safety issues involved in using roads shared by urban dwellers is problematic.

Patch/Parcel size

Isolated and fragmented patches are more prone to non-agricultural land use conversion. Author Marin notes that smaller agricultural land parcels near urban areas tend to devalued and much lower in price, than small parcels at a greater distance (Marin 2007). A lower value for farmland yeilds a greater uncertainty for future farm-use.

 

Methodology

The spatial analysis for this project was carried out using IDRISI's built-in Multi-Criteria Analysis Decision Wizard utilizing non-Boolean "fuzzy" sets to approximate the negative impacts of urban activity on ALR areas.

In order to run the decision support module, the available data required prosessing to create the requisite constraints and factors.

Constraint: ALR Area

Because the area of interest is the Agricultural Land Reserve itself, a Boolean image was created using the assign module to isolate the anaysis within its boundaries:

Factor 1: ALR area patch size

Based on the assumption that smaller patches are more suseptable to conversion to urban land use a raster image of area sizes was generated, and then rescaled as a fuzzy set, with larger patches ranked low (0) and smaller patches ranked high (255).

Factor 2: Population Density Distance

A population density image was created by overlaying a raster image of DA areas (km2) in the 2006 Census data with an image displaying population per DA. The resulting data were reclassed into four groups, low to high population density. The classification pertaining to the highest densities was converted into a Boolean image, from which an image showing the linear distance from these areas was generated. Finally, the linear distance was rescaled inside the MCE module (monotonically decreasing, linear) to represent the fuzzy density distance factor (near 255, and far 0).

Factor 3: Road Network

The 2006 road network image was rasterized from the original vector data in IDRISI, and using the distance operation, an image displaying linear distance from the road network was created. The distance data were rescaled (linear, monotonically decreasing) in the MCE wizard to represent the fuzzy road distance factor (near 255, and far 0).

Factor Weights:

Following the creation of fuzzy sets the three factors were weighted by measuring their relative importance within a pairwise comparison table. The weighting scheme ranks population density as the the most influential factor, followed by road netwrk distance and patch size. An image was generated as output from the MCE module, and subsequently reclassed and aggregated into 4 risk areas.

 

Analysis