The ESRI Guide to GIS Analysis, vol. 1 (second edition, 2020) by Andy Mitchell
Chapter 4
Density maps show where the highest concentration of features is and are particularly useful when looking for patterns in areas that vary in size, such as census tracts and counties. Before creating a map, the kind of data being used and whether the density of features or feature values will be mapped should be considered. Density can either be mapped graphically (calculating density values for each area or dot mapping) or by density surface. If the data have already been summarized by area or the objective of the map is to compare administrative or natural areas with defined borders, density should be mapped by defined area. If the objective of the map is to see the concentration of point or line features, a density surface should be created. When mapping density for defined areas, calculating a density value for each area involves dividing the total number of features/total value of the features by the area of each polygon while creating a dot density map involves specifying how many features each dot represents and how big the dots are. Density surfaces, however, are created in GIS using raster layers. The specified cell size, search radius, calculation method, and units affect how the GIS calculates the density surface and, ultimately, what the patterns will look like. Density surfaces can either be displayed using graduated colors (usually displayed using shades of a single color with common classification schemes including natural breaks, quantile, equal interval, or standard deviation) or contours (connecting points of equal density). While the patterns on a density map are partially dependent on how the density surface was created, it is important to remember that there may not actually be any features where the highest density is. To see a better picture of what is going in a place, the locations of features from which the density surface was calculated should also be mapped with the density surface or on a separate map.
Chapter 5
Mapping what is inside an area allows for the monitoring of what is occurring inside it or the comparison of several areas based on what is inside each. Depending on the number of areas, what type of features are inside the areas (discrete or continuous), and the information needed from the analysis (list, count, or summary), an area boundary can be drawn on top of the features, the features inside can be selected using a area boundary, or the area boundary and features can be combined to create summary data in order to find what’s inside. A map that shows the boundary of the area and the features is good for seeing whether one or a few features are inside or outside a single area. This method requires a dataset containing the boundary of the area or areas and a dataset containing the features. Creating a map by selecting the features inside an area is good for getting a list or summary of features inside a single area or group of areas being treated as one. A dataset containing the areas and a dataset with the features are needed for this method which involves specifying the area and the layer containing the features so that the GIS may select a subset of the features inside the area. To overlay the areas and features, the GIS either combines the area and the features to create a new layer with the attributes of both or compares the two layers to calculate summary statistics for each. Overlaying the areas and features is good for finding the features that are in each of several areas or finding out how much of something is in one or more areas. Data containing the areas and a dataset with the features are needed for this method. When selecting features inside an area, GIS can be used to create a report of the selected features (count, frequency, sum, average, median, standard deviation). There are also key differences between overlaying areas with discrete features, continuous categories or classes, and continuous values.
Chapter 6
What is occurring within a set distance or traveling range of a feature is understood through finding what is nearby. There are also three methods to find what is nearby: measuring straight-line distance, measuring distance or cost over a network, or measuring cost over a surface. Selecting a method entails determining the information needed from the analysis (list, count, or summary) and defining and measuring “near” which can be based on a set distance or on travel to or from a feature. Using straight-line distance means specifying the source feature and the distance before the GIS finds the area or surrounding features within the distance. This method is good for creating a boundary or selecting features at a set distance around the source. A layer containing the source feature and a layer containing the surrounding features are required to find what is nearby using straight-line distance. Using the area covered by segments of the network within the distance or cost to find the surrounding features near each source is known as measuring distance or cost over a network. If the objective is to find what is within a travel distance or cost of a location using the locations of the source features, a network layer, and a layer containing the surrounding features, this is a suitable method. Measuring cost over a surface begins by specifying the location of the source features and a travel cost. Then, the GIS creates a new layer showing the travel cost from each source feature. This approach is good for calculating overland travel cost and it requires a layer containing the source features and a raster layer representing the cost surface. When calculating cost over a geographic surface, it is important to acknowledge that cost refers not only to monetary value but also to factors like time, effort, or resource expenditure required to traverse a landscape.