But I don't want to assume that you do or don't know what GIS is. Perhaps there are less people than I think who has heard of this software technology. It's called a few different things, but generally I think most would call it Geographic Information Systems (GIS). To a certain extent it's digital mapping. But to just call it mapping software is to really miss it's power as an analytic tool. For the thing that makes GIS software more than just Street Maps 2010 is that there is a database that is associated with the shape information.
Essentially it works like this, there is a file containing shapes (either points, lines or polygons), and these shapes have a coordinate system based on a cartographic projection. This gives each shape a spatial definition on the Earth. This spatial definition allows one to overlap (or layer) multiple types of shape files over top of each other. Finally there is a spreadsheet "behind" the shapes that holds data about these shapes.
For example one could have a shape file of cuyahoga county municipalities. There are fifty nine cities, villages and townships that comprise this file. The file then can have a spreadsheet that has demographic information like population totals, income info, size of city and so on. We then could color coordinate the shapes to show what cities have, say, high crime rates compared to those with low. Or we could layer this map with one of streets or parcels and start to compare one layer to another. I could take a map of zoning districts and then take my parcel layer and "ask" the software to tell me where all of the vacant properties are that are zoned for heavy industrial uses. Or I could define boundaries for a retail center's trade area and use census data to determine how many people live within five miles of the shopping center; and then estimate the median income for that never before defined area. GIS technology is used to create the directions we get from an internet mapping sites.
The name GIS gives the impression that it's a highly technical sort of software. Like it's really an IT sort of thing. And I suppose it can be in certain applications. But for my position within my government department, it's an analytical tool. While there is a level of knowledge and experience to using the software, the real "skill" comes in by applying cartographic skills and analytical ability to take raw data, show it in a spatial context, and then interpret and communicate what it tells us about a place. This last bit is why GIS is so important to the world of urban planning. For some that is a fair definition of what we do with comprehensive plans.
--------------------------------------------------------------------------------------------------
But now I must digress, for the real reason I posted this blog was not to talk about my thoughts on GIS and to define it for all. (Although I just spent a lot of time talking about that.) The point is that I had an interesting map I made this evening and wanted to share it.
It's what I would call a down and dirty map, but the results are interesting for those interested in the history and growth of the county.
There is a difference between a good map and a bad map. More are bad rather than good. The thing is that all maps must lie; most commonly as a lie of omission. Because if we include every piece of information about a place on the map it would be unreadable. And after all when one makes a map, it is meant to be read. So the need is to selectively choose what information to show and stress to the reader. This can be quite subjective with certain topics.
I've uploaded a map, I would say that this is a C- map. I'm making a lot of assumptions when I made this map. Most importantly I'm assuming that the intended recipient will be able to make out what it is that I'm conveying with little to no guides. There are hints, but only a scant few. And it really isn't pretty. But here it is:
It's a map of Cuyahgoa County and it shows all of the residential property within.
First there are no labels, no guides to any real landmarks, no north arrow (but it would be absurd to have a map with this small of a scale to NOT orient up as north), no scale. There aren't any road names or highways or rail to give any clues.
There are however city boundaries, but that only helps if one is familiar enough with the county to know where what is since those too are unlabeled.
The colors indicate what range of years that the house was built. The thematic key indicates that there are six colors and a dark grey that indicates residential property without a building (vacant). The remaining light grey is not residential property at all.
The ranges chosen were made at the discretion of the maker based on his opinion and his knowledge of the data source. Breaks were chosen at 1900, 1929, 1945, 1960 and 1990.
Thematic maps like this can show interesting things. I tried to break the ranges in a way that corresponded with different eras/generations. The problem with the data set is that while I would have preferred to have a break around 1880 the accuracy for the pre-1900 parcels leave a lot to be desired. There are several thousand houses listed as built in 1900 and I think that a great many could have really been built up to ten years before or after.
The other interesting break I think is the 1929 to 1945 range. Of course this is during the depression and second world war. We are told that there was little to no construction during this period, which on the whole is a relatively true statement. However we see by the results that there was more going on in this period than we are lead to believe. Most certainly the bulk of these homes would be built in the little period after the 1937 "recession within a depression" and before the war. In fact my neighborhood in South Euclid fits into that period.
The second effort shows a little different story:
Here I added a seventh range to split out those before 1880 (pink) from those between 1880 and 1900 (red). The 1880 to 1900 represents the first half of the ascendancy of the city of Cleveland. The second half being the era after Mayor Johnson leading into the depression.
I also changed the newest range from the dark purple to teal to help bring out where the most recent developments have been. Finally I tweaked the post war range to include the 60's with the 50's. This evened out the length that the ranges represent.
But here is another point on thematic maps. This is in no way a scientific exercise. I chose the ranges, I didn't use any standard deviations or other scientific methods to create the breaks. I chose the breaks. And my bias is a factor. I wanted to show how the bulk of our housing in the county was created in the first two decades of the 20th century and immediately following the second world war. I chose breaks that made sense to me. And this sort of decision making process must always be questioned when one reads a map that is trying to tell you something. Ask "who's telling me this and why/what is their motive?"
I'm posting this for someone who is interested in where the oldest housing stock is within the county. And I think this should meet his curiosity. Hopefully he'll let me know.
But I like the second version better.
There are a few real surprises for me here.
ReplyDeleteI didn't realize how many pre-1880 (henceforth "older) houses there were in the Broadway area (mostly on the south side of Broadway). Bedford should have been obvious, but it wasn't. I hadn't noticed so many older places in the Clark-Fulton area, though I probably should have. The group of older houses around Five Points, in Collinwood, is surprising, and really worth further investigation.
The biggest surprise that I see are is the presence of so many older houses on larger pieces of land outside Cleveland and the inner-ring suburbs. Most are on older roads, so I suspect that that data is reasonably accurate. Still, I am shocked to see that ther are so many still sitting on parcels that must clearly be several acres. I would expect this in the areas where such buildings might be part of "gentleman's farms", but not in many of the rest of these cities.
When I'm searching the Multiple Listing Service, one way that I try to find older homes that might slip through the cracks is to look at houses listed as being built between 1900 and 1920 that are smaller than 1200 square feet, which excludes most of the houses actually built during that period. I suspect a search like this would be unbearable if not for the fact that the included photographs allow one to see which houses are interesting rather quickly.
This is very interesting - thank you!