Just another My blog Sites site
In case you’re interested – a fairly grouchy article on why not to use emojis in maps can be found here.
Here is a link to open data collections concerning Mecklenburg County. It might be useful for the assignment where you overlay data on a QGIS map.
Davidson College has a long history of book collecting, but the Little Library is a relatively new addition to the campus landscape. Over the years, college libraries have been housed in the Old Chambers building, Phi and Eu halls, in the Carnegie guest house (which was once a Carnegie library) and in Sloan. The Little Library of today was built in 1974, but not before considerable debate over where it should be located.
One of the artifacts of this debate is this map of possible library sites. Each was accompanied by a list of pros and cons. Location one – south of the President’s house and the guest house, was “near dorms and classrooms” but might “necessitate the removal [of] good trees.” Location two – just north of the current location – would “crowd Duke & East dorms.” Location three, despite “good access from [the] road” was both “too far from dorms and classrooms” and “would cause removal of Jackson Court.” Other locations were too far from classrooms, and one – directly across Concord road from campus “present[ed] road crossing problem for students.” The committee went on to note that “solution of tunnel or bridge could be expensive.”
Given how much the campus has expanded since then, the thought of a bridge over or tunnel under Concord is a bit laughable, but it’s interesting to imagine what life on campus might have been like – and how we might have used space differently – had the library been placed somewhere else.
I had very similar issues with Maptiler when compared to my classmates (particularly Joe). It is very simple- there are only a few pages to click through making it far less intimidating than QGIS. In fact, I thought the entire process would be super simple based on how simple it was getting to the page where I was forced to do the overlapping. Before I could get to that page, I realized that I would have to rely on a googlemap image of davidson college. This image did not fit well with the map I chose, although I had high hopes. My map is too wide making everything a bit off of scale. When I finally found the best overview shot, and was ready to lay my map over, I realized I’d have to rely more on my map directory to understand how much the campus has changed since 1969. A lot of the buildings were moved, some changed names and new ones were built. This narrowed my reference points down to Belk, Chambers, Duke, Baker Drive and Concord road. When I finally, was ready to have my map rendered I was immediately disappointed at how different the two maps were. I played around with adding new points and tried my best to make the existing choices very precise but, my best map still isn’t a perfect overlay.
I think that the Maptiler can improve by having a few tools to help manipulate the overlay map especially if it is meant to fit a googlemap image. Because googlemap images can be as close or as far as the user wants, being able to rotate, cut and reshape the image that is being overlayed will significantly improve the precision. When I tried adding nearly 7 points, I became frustrated at how distorted my map became, making it impossible to understand. Finding the middle road would be the best, and that falls on the user to decide how many points to use. 3 is too little and 7 is too many.
https://54ab69828fac3cefb7996d0a0a971b8d58d0c923.googledrive.com/host/0Bzi_ze1VzfJbd2NmSkxJb1lidU0/
Map Tiler’s Interface is very user friendly at the beginning. The addition of points to match up with the given map is very intuitive. However, the map result can encounter noticeable problems if the map picture is not entirely to scale. For example, many of the buildings on Davidson campus to not match up with their counterparts on the two maps in relation to the streets. In addition, if a map is fairly un-detailed in a certain area, or simply lacks easily identifiable points for detail, the map will generally be very skewed, even if there are many points in a clustered area. Roads are the best point of reference, I’ve found, as they often have very clear intersections. They are also much less subject to distortion than buildings, which can have very many differences in scale between the two maps.
The other issue with the program is the fact that the output system can be a little confusing. It seems one has to devote an entire separate folder to the rendering files or the program will become confused. I would have rather the system create a folder in the destination file rather than just simply putting each file in the folder (or on the desktop) individually.
However, once one gets accustomed to the quirks, Map Tiler is very straightforward and simple to use. It gives a very good product for very little effort and can be easily taught to just about any level of user with minimal guidance.
Attempt 1: https://33d8b2c24eb605307ea0635c48868382a9c76f54.googledrive.com/host/0B3qWZoVoqkj0OTlOaGJJb0ZSNjg/
**for attempt 1 the map doesn’t show the first time but if you reload the page it usually shows.
Attempt 2:
https://fde27f0d66c1f67b7b15ffe8436a7c6284d9733d.googledrive.com/host/0B3qWZoVoqkj0bjRJMkxzNHNoOWM/
Geocoding this map using MapTiler demonstrated a couple of challenges of map geocoding. In class we discussed the overwhelming amount of map data available for urban places, and the underwhelming amount available for rural ones. My geocoding experience presents tangible evidence of this phenomena. In my first attempt of geocoding this historical map of Davidson College and its surrounding areas the historical map fit nicely in the areas of Davidson with roads and buildings (bottom right), but was warped tremendously in the surrounding natural areas (at the top). Although Davidson College is not quite an urban center, the differences between the inhabited and uninhabited areas are striking. Without roads or buildings it is very hard to geocode a map. It took me a second try to guess where areas in the historical map corresponded to google maps for the northern part of Davidson’s property, and still my second attempt is not quite perfect. While this small experiment shows that the differences in information for rural versus urban landscapes is quite stark, it also shows that historical mapping processes can inform current map makers of the past human-environment relationships.
Using geocoding as a tool, it would be interesting to measure the “warpness” of historical maps as Davidson college expands into nature. By mapping a series of historical maps on a current google map and viewing the areas being warped, one could distinguish the areas in which the College expands into the surrounding natural area. Do they map the natural areas or leave them sparse? How quickly does the college expand its infrastructure? Is there a way to remove all of the warp in these historical maps? These results could show not only the expansion of the college, but also the fluctuations of valued spaces.
What at first seemed to be a mistake, might just demonstrate one of the benefits of geocoding techniques. By showing warp, MapTiler allows for map makers to develop certain theories about historical maps – in this case about Davidson College’s relationship with the surrounding land.
While projects in the digital humanities range from thick mapping data onto Maps, such as Hyper Cities, to reinterpreting social interactions through mobile devices, such as location-based social networks described by Adriana de Souza Silva and Jordan Frith in Mobile Stories, all employ a unique digital medium that is tailored to best convey the message. One of the many niches that the digital humanities has carved for itself is representing historical documents through a digital medium specifically by placing historical information on maps. For example, in this link, a physical map of Davidson College’s campus from 1978 has been layered on top of Google Maps using MapTiler.
By placing this piece of Davidson history in the context of an immediately understandable satellite image, the viewer is fluidly transported from 2014 to 1978 and vice versa. In addition to helping the user understand the context of the map, MapTiler includes a slider that adjusts the image’s transparency and affords the viewer the opportunity to compare Davidson’s campus from the past to the present and see how new and old buildings are spatially aligned. As can be seen in this map, which represents an even older version of Davidson’s campus, the campus of Davidson College has changed dramatically: Chambers has burned and been rebuilt, dorms and sports fields have been added, and the library has been relocated a number of times.
MapTiler allows the user to attach geographic coordinates to a map and spatially situate images on top of satellite images. From a humanities standpoint, MapTiler provides a useful way to place a historical image in its current context; from a user’s standpoint, MapTiler makes this humanities project a simple undertaking. With this user-friendly tool, adding raster images (like the older map of Davidson’s campus) over Google Maps is as easy as dropping matching points onto each image, uploading the file to a host server, and using the tool to spot the difference between different images and time periods.
Link to the geocoded map of Davidson from 1939/1940
I took an image of a map of Davidson College from 1939-1940 and used MapTiler to mark points on this map that corresponded with a current cartographic view of Davidson College similar to what someone would see in Google Maps. While geocoding, I only found it to be a challenge when a building on one side didn’t correspond to a building on the old map of Davidson College. I wasn’t sure if this would skew my results, but it turned out the way that I hoped it would. Overall, the program is very straightforward for the purposes that I was using it.
My geocoded result shows the accuracy of the traditional map as well as change over time (click on the link above and use the slider in the top right to see the differences between the two layers). The map tells the story of a growing college. Like people, the college has changed in certain ways and has remained the same in other ways as time has gone on. An example of accuracy is shown by Main Street (once called U.S Highway No. 21), which lines up almost perfectly with the Main Street in the bottom layer; Concord Road does as well. Differences include additions of dorms, academics buildings, and relocation of certain buildings such as the fraternities (currently in Patterson Court but once in Jackson Court-bottom right of old map). The most noticeable consistency is Chambers (large building in the center) as well as the paths on campus, which we can only see through the top layer since the trees in the bottom layer block our view.
The story that the map is telling isn’t limited to just showing position change of buildings, but we can speculate as to why certain buildings moved. For example, the library used to be far from Chambers (top left building on top layer map), but now it is located right behind Chambers. This position change shows clustering of academic buildings, which the college currently does. The increase of population accounts for additional dormitories, expanding the campus, and possibly moving the fraternities to Patterson Court so that the faculty could have Jackson Court. These two observations lead me to a realization from this ‘story in the form of a map’: necessity brings change, and change is made possible by money (we might even be able to say that potential needs arise when the money is there).
With this information, I could use a map like this in the future to estimate trends of how the campus might evolve. Obviously, Chambers will most likely remain where it is; it is the Sun while other buildings are planets that can enter an orbit from Chamber’s ‘gravitational pull’. Furthermore, people determine what is necessary to change, which means that the layout of Davidson College is very much a social construct. People are predictable, so the evolution of Davidson College should be as well.
Here is my map:
http://heganoo.com/node/9335
It’s not possible to export these, but it’s a cool resource for georectifying old maps: http://www.bl.uk/maps/georeferencingmap.html