Chapter 4, “Mapping Density,” got me thinking about how we present information, especially when you’re dealing with varying areas. It’s not enough to just count things up, sometimes we need a bit more context, like how spread out something is. I think that’s why the book dives into density mapping, the distribution of features or values per unit area. There’s a difference between saying “there are 100 houses in this neighborhood” and “there are 100 houses per square mile in this neighborhood.” What I found particularly interesting was how you can map density in a few different ways. The book highlights two methods: mapping density for defined areas, like census tracts, which is where you calculate density within existing boundaries. Then, there’s creating a density surface, which involves creating a continuous surface that shows how density changes across an entire area, even without defined boundaries. It’s like taking something that’s usually summarized by area, and making it a landscape you can view. It seems to me that these two methods would be used for different purposes and I’m wondering which gets used for what application and why? Mapping density is about more than just visualizing data. It’s about taking into account the context of the data and how it’s distributed. It’s a really valuable tool to see the patterns that might be hidden at first glance. I wonder if density mapping could be applied to research in neuroscience, since it deals with location based data sometimes.
Chapter 5 seems to build off the idea that location matters, but in a different way. Now instead of looking at where things are, we’re looking at what things are contained by. The book’s main question here is, Why map what’s inside? and, in my understanding, it’s because it allows us to explore relationships between features. The book outlines three approaches for this: drawing areas and features (where we manually create areas to select features), selecting features inside an area (where we use existing boundaries), and overlaying areas and features, which sounds like the most complex one. The overlay method, as I understand it, combines two layers of features to see how they interact spatially. I’m starting to think this is where GIS really shines because it creates new relationships between features that wouldn’t exist in the real world otherwise. I’m curious how you all go about choosing which of these three methods to use? I imagine that using drawn areas is more appropriate when you need to be more precise with your selection, and that using existing boundaries is better for broader analysis. How do you decide which features to overlay, if that makes sense?
Chapter 6 seems to add a layer of complexity to our spatial analysis by focusing on closeness. I think the main idea is that we can learn a lot from the distance and relationships between features. The book asks, Why map what’s nearby?, and the answer I think is that it allows us to explore how features interact, or how they might influence one another. Three ways of exploring this include using straight line distance, measuring distance or cost over a network, and calculating cost over a geographic surface. It seems like the first one is the most basic, just measuring distance as the crow flies, so to speak, while the second one takes into account that movement is often confined to networks, like roads. The last one, where you calculate cost over a geographic surface, is a bit more abstract, where you take into account the “cost” of travel, which is interesting. I’m realizing that “cost” doesn’t always mean money, and that different types of cost can be included in research. It seems to me that GIS is very useful for understanding and calculating all these different types of distance. I’m also thinking about the different applications for these analyses. It seems that you could use straight line distance to do quick analyses, or when the network isn’t important. You could use network analyses to find optimal routes, and surface analysis to calculate the cost of travelling across different topographies. I am wondering if there are times when you would use a network analysis to find straight line distance, or is that redundant?