Application of geographic information systems in site selection and location analysis.
Geographic Information Systems (GIS) are powerful computer-based tools for integrating and analyzing spatial data from multiple sources. GIS permits geographically referenced information to be stored, edited, manipulated, and analyzed to generate interpretive maps and related statistics relevant for decision making. Applications in real estate spatial analysis are gradually taking off with the introduction of personal computer--based GIS softwares. Some applied studies using GIS in land use planning include Domenico,(1) Breedlove,(2) Daumiller,(3) and Brandt and Elliott.(4) The studies by Weber(5) and Hart and Robbins(6) present how GIS could be applied in real estate market analysis and in the real estate valuation process, respectively.
The focus of this article is the application of GIS in the area of real estate market analysis. Specifically, the application of GIS in site selection and location analysis for residential subdivision development as well as in highest and best use analysis are demonstrated. Site analysis determines the ability of a site's specific physical and geographic characteristics to satisfy the operational and functional objectives of a particular land use. The focus of location analysis is generally the evaluation of existing linkages. Linkage relationships develop because of the need to move people and goods between locations. For example, people constantly travel between residential locations and other facilities for retail, employment, recreation, education, and entertainment purposes. Linkages, among other factors, are thus important contributors to the value of any location.
In addition, an important step in the real estate valuation process is highest and best use analysis. In the determination of the highest and best use of a site, four criteria are considered. They include criteria are considered. They include physical supportability, legal permissibility, financial feasibility, and maximal productivity.(7) The numerous site and location variables and their complex nature coupled with the need for joint evaluation make the use of a GIS tool helpful in the analysis of the first two criteria. In the next section, the physical and legal constraints applicable in site selection and in highest and best use analysis are discussed.
PHYSICAL AND LEGAL CONSTRAINTS ON DEVELOPMENT
The main components of the physical product of real estate are site-specific physical elements (e.g., natural site attributes, site infrastructure, structural improvements); and locational characteristics (e.g., linkages). Linkages examine the spatial relationships that exist between different land uses. The strong interrelationships among location and site factors necessitate their joint evaluation. Location and site analyses serve three main purposes: 1) determination of development potential and costs for a site; 2) identification of competitive differentials of the subject property; and 3) evaluation of site selection.(8)
Zoning imposes constraints on real estate development. Thus the zoning for a parcel of land is a primary consideration. In many communities it is difficult to effect zoning change, particularly in antigrowth communities with burdensome land use restrictions. These constraints add to the high cost of real estate development. It is therefore necessary to consider the current zoning of available parcels.
ANALYSIS OBJECTIVE AND CRITERIA
The analysis objective of the study is to determine the optimum location for the development of a rural suburban residential community. The area of the analysis will be limited to Durham County, North Carolina. This community should 1) be located in an area with low to moderate traffic activity; 2) not be subject to prior zoning restrictions; 3) have a gently sloping 2- to 3-degree terrain preference; 4) contain a small stream that would add appreciably to the aesthetic quality of the community; 5) not be plagued by excessive railroad noise; and 6) be accessible to major employment, retail, and recreation areas. The proximity of the proposed development to primary and secondary schools, shopping malls, recreational facilities, and employment centers all affect the decision-making process when a residential community is planned.
Data acquisition and preparation
After identifying the criteria necessary to satisfy the analysis objective, it was necessary to identify data sources that would aid in the interpretation of environmental and economic restrictions enumerated by this analysis. An accurate boundary map with at least four detailed reference points was entered into the GIS database. The reference points are required to accurately superimpose additional layers of physical information. If each layer is geographically referenced to the same points, maximum location accuracy is enhanced.
The county boundary was the first data source input into the GIS system. This boundary was input through a digitizing tablet. That is, a base map containing the perimeter that demarcates the county was input by tracing over the network with a device that electronically stores the X and Y grid coordinates in computer memory and subsequently in disk storage, thus permitting quick and accurate redrawing of the boundary at will. A soil map (see Figure 1) describing the soil location, type, name, slope, and texture was then entered into the GIS system. In the next step, a land use zoning restriction layer was input into the system, as shown in Figure 2. Polygons that mark the boundaries of the different areas legislated for specific land use activities were assigned shading symbols for future reference.
Road and railroad infrastructure were then considered. The road network infrastructure shown in Figure 3 was the next data source input into the GIS system. Further, attributes such as the road name or number, the number of lanes, the traffic flow rates, and any other information considered pertinent were attached to various road segments via an interactive database system. This provided a means to query this layer of information at a later time. For example, a map showing only two-lane roads using a designated symbolics could easily be drawn. Because railroad noise level was a concern, it was necessary to input a layer of railroad locations in another file, shown in Figure 4. To complete the data input of physical environmental attributes for the area of analysis, lakes and streams were digitized into separate files, with stream order assigned for future reference (see Figure 5).
In an attempt to discern the extent to which certain linkages might influence the selection of one location over another that has also satisfied all the physical requirements, selected point features could have been input into the GIS. These include employment centers, recreational facilities, primary and secondary schools, and shopping malls. Figures 6 and 7 show the major employment and retail centers, respectively. In this analysis only distance from a particular linkage is illustrated. Weighted values, however, might become a part of a gravity algorithm to more accurately portray the impact of such an establishment. Measured distance is typically used as a proxy to determine the cost of movement between sites that perform different functions.
Once the various physical features and linkages were input and referenced to the same geographical scale, each file with any accompanying database was double-checked for content and accuracy. After these checks were performed, spatial interrogation of the layers of information was possible.
The data layers supplied to the GIS were refined at this stage of the analysis by manipulating features within and between the layers. Techniques such as extracting, overlaying, buffering, and merging were used to further enhance the information commonly obscured by its spatial nature and overwhelming volume.
Initially, the roads file was queried to extract only the two-lane highways from the full data set. The assumption was that this road type would limit traffic flow rate and thus represent a road type conducive to residential development. This assumption was made primarily for illustrative purposes. Ideally, traffic flow rates would be incorporated to more accurately portray the real-world scenario that is controlled by temporal qualifications. State or local departments of transportation are typically sources of this type of data.
This segment of the road network was then buffered. Buffering is a technique that permits a designated area (e.g., a polygon) along a road (e.g., an arc) to be added spatially to a map image. The area specified extended to some 1,200 meters on either side of the road. To further insulate the proposed community from excessive road noise, 250 meters on both sides of the road network were eliminated through a second buffering procedure. The new data file created by this procedure now includes the roads and the adjacent defined area.
The stated criterion that the newly defined area contain no sections previously zoned for a land use other than residential was satisfied through an extracting procedure. Here, the land use file containing all of the different legislated zones for the county and their attributes was superimposed onto the buffered roads file. The resulting output file was composed of only the polygons zoned for residential development (see Figure 8).
Optimum soil slope for residential development is a gentle (i.e., 10% to 15%) slope. In addition, the land texture should permit easy drainage of water. It thus became necessary to merge soil information layers with the previously refined layer. Merging the soil layer and extracting only those soils with the proper qualifications yielded a series of polygons representing those areas with the desired soil conditions. Figure 9 shows the resulting polygons. Similarly, the streams file was merged with this file to produce a layer of polygons within 1,000 meters of a first-or second-order stream, as shown in Figure 10. Proximity to these smaller streams strengthened the attractiveness of the parcel of land.
Finally, to assure minimal annoyance from railroad traffic noise, any remaining property within 1,000 meters of existing railroad tracks was deleted by extracting these areas. As Figure 11 illustrates, the polygons that persist in the northern and central regions of the county are those that satisfy all of the criteria under the physical parameters initially set for the analysis.
Attention was then directed to the linkages that further influence the selection of the best location for a residential community. Distance from the resulting areas is the critical variable. It is a proxy for the travel or frictional costs associated with moving between two linked establishments. In general, individuals try to minimize these costs. Tables 1 and 2 show the distances from the northern and central land parcels, respectively. The distances to the major employment areas of the county are shown in Table 3. By merging the areas that meet the physical and legal requirements with the files that contain the locations of the county's major employment and shopping centers, a map illustrating the spatial separation was constructed (see Figure 12). Other linkages such as secondary and primary schools, cultural centers, and recreational establishments could be similarly incorporated into the spatial analysis.
TABLE 1 Northern Proposed Community Distance to Retail Centers (in feet)
Retail Center ID Regional Community Neighborhood Number Center Center Center 1 -- 22,305 -- 2 -- -- 36,145 3 -- -- 14,804 4 21,743 -- -- 5 -- -- 13,142 6 -- -- 24,715 7 -- -- 36,408 8 -- 22,558 -- 9 -- -- 24,352 10 -- 33,418 -- 11 -- 19,827 -- 12 -- 20,040 -- 13 -- -- 24,187 14 -- -- 18,903 15 -- 15,804 -- 16 21,159 -- -- 17 -- 28,496 -- 18 -- -- 24,426 19 -- -- 32,529 20 -- -- 26,743 21 -- 28,621 -- 22 -- 27,250 -- 23 -- 12,988 -- 24 -- -- 28,839 25 -- 24,614 -- 26 27,723 -- -- 27 -- -- 33,475 28 -- -- 26,458 29 -- 21,445 -- 30 -- 22,372 -- 31 -- -- 21,708 32 -- 28,577 -- 33 -- 14,291 -- 34 -- -- 28,308 35 -- 30,793 --
TABLE 2 Central Proposed Community Distance to Retail Centers (in feet)
Retail Center ID Regional Community Neighborhood Number Center Center Center 1 -- 3,059 -- 2 -- -- 18,164 3 -- -- 6,775 4 2,583 -- -- 5 -- -- 7,158 6 -- -- 5,537 7 -- -- 17,597 8 -- 4,702 -- 9 -- -- 5,262 10 -- 14,463 -- 11 -- 1,154 -- 12 -- 3,689 -- 13 -- -- 11,409 14 -- -- 665 15 -- 3,757 -- 16 1,558 -- -- 17 -- 10,317 -- 18 -- -- 12,749 19 -- -- 14,358 20 -- -- 9,654 21 -- 9,804 -- 22 -- 8,362 -- 23 -- 6,505 -- 24 -- -- 10,333 25 -- 5,598 -- 26 8,877 -- -- 27 -- -- 15,032 28 -- -- 7,255 29 -- 3,993 -- 30 -- 3,974 -- 31 -- -- 2,672 32 -- 7,611 -- 33 -- 9,752 -- 34 -- -- 11,851
TABLE 3 Distance to Major Employment Centers (in feet)
Land Use Proposed Proposed Polygon Northern Central Number Community Community 1 11,941 6,244 2 9,504 8,969 3 8,682 10,156 4 8,907 10,216 5 4,361 14,143 6 2,164 16,570 7 4,022 18,890 8 4,673 17,937 9 4,190 21,280 10 6,996 21,280 11 4,456 21,904 12 1,742 20,025 13 3,588 21,734 14 4,377 21,812 15 9,389 23,133 16 12,261 29,620 17 6,222 24,714 18 8,172 25,916 19 8,481 26,509 20 10,089 26,493 21 10,634 28,990 22 11,770 29,741 23 12,629 30,003 24 13,372 29,167 25 15,505 33,626 26 14,399 32,709 27 13,589 31,495 28 14,928 32,508 29 15,633 32,279 30 16,920 33,921
This article demonstrates how GIS could be applied in site selection and location analysis for residential developments. In light of the numerous variables and their complex interrelationships, GIS is invaluable. Land parcels are identified here that could be closely studied for possible residential development. The identified areas meet the physical and legal requirements, and linked sites for employment and retailing were incorporated into the analysis.
(1.)Cindy Domenico and Steve Dunbar, "Property Value Mapping and Assessment Database Analysis Using GIS," Proceedings of the Tenth Annual ESRI User Conference (1990).
(2.)Michael J. Breedlove, "Utilization of a Parcel-Based Geographic Information System for Small Area Demographic Estimates," Proceedings of the Tenth Annual ESRI User Conference (1990).
(3.)Gerry Daumiller, "Land Use Allocation for Estimating Study Area Populations," Proceedings of the Tenth Annual ESRI User Conference (1990).
(4.)R.C. Brandt and H.A. Elliot, "Sludge Application Site Screening Using a Geographic Information System," Working Paper 89--2001 (Pennsylvania State University, Department of Agricultural Engineering, 1989).
(5.)Bruce R. Weber, "Application of Geographic Information Systems to Real Estate Market Analysis and Appraisal," The Appraisal Journal (January 1990): 127--132.
(6.)Patricia Hart and Michael L. Robbins, "Revolution in Spatial Analysis for Real Estate Decision Making," Working Paper (Washington, D.C.: The American University, 1990).
(7.)American Inst. of Real Estate Appraisers, The Appraisal of Real Estate, 9th ed. (Chicago: American Inst. of Real Estate Appraisers, 1987), 244.
(8.)Neil Carn, Joseph Rabianski, Ronald Racster, and Maury Seldin, Real Estate Market Analysis: Techniques & Applications (Englewood Cliffs, N.J.: Prentice Hall, 1988), 97--100.
|Printer friendly Cite/link Email Feedback|
|Author:||Barnett, Albert P.; Okoruwa, A. Ason|
|Date:||Apr 1, 1993|
|Previous Article:||Case study: an alternative technique to the land residual method.|
|Next Article:||Forecasting likely job growth or decline with the net deficit technique.|