An aptitude map is the answer to the question: Where is the best location for a new resource? The result you looking is a map representing all the possible sites (ranked from the most to the least likely) that would be suitable for the installation a new resource. This map is a classified aptitude map since it depicts a relative in the map, taking into consideration the criteria you enter in in your study.
We will follow the ArcGIS Spatial Analyst tutorial
in parallel with
http://help.arcgis.com/en/arcgisdesktop/10.0/help/index.html#//00nt00000002000000 ): using the conventional tools based on the Boolean logic, and using the tools based on fuzzy criteria (fuzzy numbers). Therefore we can appreciate that what these tools, actually, bring along, and how this approach differs from the classic process.
The required data appears on the ArcGIS Desktop CD. After having executed the ArcGIS installation in the additional installation Components dialog box, activate the option ArcGIS Installation data tutorial. In the Setup wizard ArcGIS data tutorials, activate the option installation of Spatial Analyst data (the path access default installation is C: \ arcgis \ ArcTutor \ SpatialAnalyst). To follow this example you will need a Spatial Analyst license.
These data, mostly raster, allow fulfilling the exercise following the classic approach (Boolean). The tools we introduce work on vector layers. Therefore it is necessary to vectorise the raster layers to follow the approach based on fuzzy criteria ( fuzzy numbers ). Using this approach, you do not need a Spatial Analyst license but you need to have three logical tools fuzzy installed.
To avoid the vectorisation task, we put at your disposal the necessary data for this exercise here .
ESRI tutorial scenario
The city of Stowe, Vermont, USA, has known a substantial population increase. The demographic data suggest that this increase is related to families with children who move to the region, enjoying the many leisure centres located in the surroundings. It has been decided that a new school must be created to relieve existing schools and, as urbanist, you have been asked to search for a potential site.
Once the problem has been stated, we split the problem to identify the necessary steps to solve it. These steps correspond to the objectives to be achieved.
When we define the objectives, we have to ponder how we will evaluate them. How is it going to be evaluated the best area for the new establishment ? In this example of site search it is better to locate it near recreational installations, because most families that have moved to the agglomeration have young children who wish to have recreational activities. Besides it is important to be away from schools already present in order to spread them in the whole city. The establishment must be built on appropriate land, relatively flat. Other goals to be included in such a research could be: a fairly large area adequate for the school and its grounds, or an area with the highest density of children of a given age, but this model is simplified for the example.
To achieve these goals, we must identify the following information:
- Where are the locations whose terrain is relatively flat ?
- Is land use in these locations adequate?
- Are these locations relatively close to the recreational installations?
- Are they far enough from the existing schools?
Where are the locations>whose terrain is relatively flat ?
To find the areas whose land is relatively flat, you must create a map that shows the slope of the land. At this point, the process model deals with the calculation of the slope of the ground. Based on the Altitude layer, the tool Slope of the Spatial Analyst toolbox, allows to calculate the slope for each raster cell. We have vectorised this result, with the tool Conversion- Tools> From Raster -> Raster to polygon. You will find this layer in the exercise data with the name SlopePolys.
Is Land use in these locations adequate?
You must determine the characteristics of a usage type of suitable ground on which to build school. This process is subjective and a function of your problem. At this point, the arable land is considered as the less expensive building lots and, therefore, privileged. These are followed by the infertile lands, then shrub areas, forests, and previously built areas. No process model is involved here, only the identification of the data set input ground and determining the type of use the most suitable soil. The land use data have been vectorised with the same command ( Conversion- Tools> From Raster -> Raster To polygon). You find the result in the layer OccupationSolPolys.
Are these locations close enough to recreational installations?
You reckon it is better to build the school near the recreational activities installational . Therefore you must create a map that displays the distances between these facilities to locate, eventually, the school in the surrounding areas . Here, the process model involves calculating distances for the recreational facilities. To calculate distances in a straight line (Euclidean distance) we use the tool Spatial Analyst -> Distance -> Euclidean distance, then we vector the result to obtain the layer DistanceRecre.
Are they far enough away from the existing schools ?
It is better to locate the school away from institutions already present, to avoid encroaching on their capture areas. Therefore you must create a map showing the distance to existing schools. At this point, the process model calculates the distance to existing schools. To calculate distances in straight right ( Euclidean distance ) we use the tool Spatial Analyst -> Distance ->Euclidean distance , then we vector the result to obtain the layer DistanceSchools .
To start the production of the aptitude card according to the classical procedure tutorial of Spatial Analyst tutorial you should have the following configuration:
The starting configuration for the Spatial Analyst tutorial
to start the production of the aptitude map according to the procedure based on fuzzy logic, you should have the following configuration :
Starting configuration for the Logic fuzzy tutorial
Thank you for this