GIS planning and application-GIS service area analysis based on the Harf Model

1,GISService Area Analysis

Location factors are a crucial factor in business analysis. Therefore, routine service area analysis is very important for store site selection. The service area refers to the main area of customer distribution. The sales volume or service turnover of the store exceeds that of its competitors. For existing stores, the service area analysis can be used to evaluate the market potential and business performance. For new stores, the service area can be used to discover Business Opportunities behind competitors, so as to determine the optimal site selection. In addition, the service area analysis helps enterprises determine the key areas of advertising coverage, reveal the weak areas with fewer customers, and propose enterprise expansion plans.

Common Methods for dividing service areas include analogy, neighborhood, and gravity. Analogy is a non-geographic method. Regression analysis is commonly used. The neighborhood method and gravity method are both geographic methods. It can be achieved by using GIS technology. The Harf model is a model based on the gravity method.

2,Harf Model

The Harf model is an economic scholar of the University of California. Professor Harf proposed in 1963 to predict the scale of the business districts in the urban area. It believes that the psychological identity of consumers engaged in Shopping Behaviors on the store is the root cause of the impact on the size of the store's business area. The size of the store's business area is related to whether the consumer chooses the store for shopping. Generally, consumers are more willing to shop in attractive stores. These attractive malls usually have a large area, strong product selectivity, high brand awareness, and more attractive promotions. On the contrary, if the distance to the store is far away and the transportation system is not smooth enough, consumers will hesitate. Therefore, the core argument of the Harf model is that the size of the business area of the store is proportional to the attractiveness of the shopping place to the consumers, and the time distance between the consumers and the places they feel. The greater the attractiveness of various factors in a store's shopping place, the larger the business area of the store. The longer the consumer's time from departure to the business area, the smaller the business area of the store. The formula for the Harf model is:

Pij indicates the probability of choosing store J for consumers. s indicates the store size, D indicates the distance, and Beta> 0 indicates the friction coefficient.

3. Application case -- how to generate the store Service Area

• Tool download and repair

The ArcGIS software does not have a built-in hav model tool, but it can be downloaded from the ESRI official website (including tools and sample data):Http://arcscripts.esri.com/details.asp? Dbid = 15999.This tool is compiled by a Python script and cannot be directly run in software version 10 after being downloaded because the Python code is version 9.3, after version 10, the arcpy site package replaces the previous arcgisscripting module. Therefore, you need to modify the code several times:

(1) Add import arcpy;

(2) Replace GP. extent with arcpy. Env. extent;

(3) Replace extent. xmin with extent. xmin.

• Application Cases

 

Consumers select the probability of the mall to generate a service area and predict the new mall. As shown in:

Figure 1 mall location and census data

 

Add the downloaded tool to arctoolbox, double-click it, and set it according to the following parameters:

Figure 2 tool parameter settings

Description of main parameters:

Parameter Name	Description
Storelocations	Enter the location of the mall. There must be at least two elements.
Storename Field	The unique name field that identifies a mall.
Storeattractiveness Field	Attractive fields of a mall, such as turnover, mall area, and commodity quantity
Studyarea	Study area.
Distancefriction Coefficient	Friction Coefficient, indicating the degree to which gravity degrades with distance. The default value is 2.
Generatemarket Areas	If the default value is none, a random vertex is generated in study area to indicate the location information of the consumer. If the following two parameters are set, you can select origin.
Originlocations	Consumer location information or census data (such as street data ).
Salespotential Field	Predicted consumption potential field, which is multiplied by the probability that the consumer selects a mall to obtain the predicted consumption potential of the mall.
Potentialstore locations	The location of the new mall to be predicted. In ArcMap, you can add new points by interacting with the map.

 

Output result:This tool generates the service area of each mall and the probability that consumers choose each Mall.

Figure 3 service area of each mall 

Figure 4 probability of consumers selecting mall 1