White Week 3

Chapter 4).

Mapping density is crucial for identifying patterns relative to where features are concentrated. Concentrated areas of crime for example will need action by law enforcement. Mapping density is useful for mapping areas (census tracts or counties) of different sizes and showing patterns rather than details about individual features. In order to map density, we can either shade defined areas based on a density value or create a density surface. Generally features of GIS are mapped using a density surface. The other thing we can do is map already summarized data by defined areas like administrative boundaries. We can either map density features or feature values that we talked about last week. A feature value goes beyond just the feature location like the number of employees at each business. Density by defined summarized area does not show actual feature locations but rather represents a specific number of features. Density = # of features / area of polygon. We may need to use a conversion factor to keep units consistent. On the other hand, the GIS creates a density surface as a raster layer and can be locations or linear features like roads. Map density surface if you have individual features or sample points. Density by defined area seems simpler but density surface mapping looks better but is more complex to do. ArcGIS imposes density by defined area on the maps by shading. For a dot density map, we map each area or dot based on a total count or amount. Two different size census tracts with the same population would be the same color on a shaded map but a dot density map shows that the smaller tract has a higher density with the same number of dots in a smaller space. Dot maps are not calculated density values but the actual total numbers or values for each area. The four parameters of cell size, search radius, calculation method, and units are important considerations for calculating density values. I’m not too comfortable with calculating cell size and density converse and so I’ll definitely have to refer back to the book if we have to do this. A tip to remember is to use a value for units that reflect what features we are mapping Square meters for plants or trees and square miles for businesses. Another cool thing is that we can create a density surface using center points from data summarized by defined areas. When displaying a density surface, we employ graduated colors by creating custom class ranges or allow the GIS to do this through a standard classification scheme. Overall, the patterns of the map depend on the creation of the density surface and its parameters. 

Chapter 5).

Finding what’s inside allows us to evaluate if something occurs within an area or identify comparable information for different/multiple areas. Comparison here is important because it can be crucial in some cases to know what surrounding areas have within them or not. In order to find what’s inside we can either draw or utilize area boundaries. The number of areas and the types of features in those areas are fundamental. A grouping of zipcodes would be several areas combined. Identify each area with a name like the name of the watershed for example. We can use the GIS to get a list, count, or summary of the features within an area. We can include features that fall completely inside (for amounts), partially inside (for lists and counts), or the portion of each feature inside the area. The overlaying the areas and features approach seems good by showing what features are inside and summary details but takes longer to process. If you’re overlaying an area on data that’s summarized by area, we should make sure the summarized areas fall completely inside. This is good to do with multiple areas or single areas that need summaries of continuous data, or discrete features including only the portion inside the area. To distinguish areas when actually making the map, label them and or draw in a different shade. When selecting features inside an area and using the results it is good to know what a frequency is. A frequency is the number of features with a given value or range of values. A bar chart can be shown for numbers and a pier chart for proportions of a whole or percentages. The summary of a numeric attribute of a feature can be a sum, average, median, or standard deviation. A sidenote is that we can show what is inside the area only but it is good to show the features outside of the area as well for contextual information. I like the look of showing features inside the area with a darker color and features outside with a lighter shade of that same color. When overlaying areas with continuous categories or classes the GIS will generally select the modeling whether it be vector or raster methods based on the data we have.  Pay attention to slivers which are very small areas that are there or emerge after overlay. Remove them at first or have the GIS remove after mapping. Sometimes Raster overlay is the default or the GIS converts it to raster because it is simpler. The GIS also creates a table to analyze the results of the overlay for the raster method. Geez vector overlay seems much more difficult.  

Chapter 6).

Finding what’s nearby is super good for considering events in an area, finding the area served by something, or the features affected by something (homes impacted by flooding). What occurs within set distance or traveling ranges is critical for many uses of GIS. In order to find and evaluate what’s nearby, we can measure straight line distance, measure distance or cost over a network, or measure cost over a surface. In cases where no movement between the source and surrounding features, measure using straight-line distance. If there is movement, travel can be measured over a geometric network like a street or over land. Finding what’s nearby can also be done by measuring costs which include time, money and or effort. This relevance of the curvature of the Earth comes into play here again in that calculating distance under the conditions of a flat Earth uses the planar model while doing this under the conditions of a spherical Earth uses the geodesic model. This distortion only occurs again when the area is large but for small areas of interest it doesn’t apply and planar modeling can be done. It is important to consider whether we will need a list, count or summary, and how many distance or cost ranges are needed. If we want to know how many streets are within 1, 2, and 3 miles of a fire station, we can use inclusive rings or distinct bands.