Spatial Analysis of Buried Treasure Distribution in America
PDF link is here
10 thoughts on “Spatial Analysis of Buried Treasure Distribution in America”
Comments are closed.
Geospatial Analysis: Multidisciplinary Applications
A virtual conference showcasing final student e-projects in the course Introduction to GIS and Remote Sensing, Vanderbilt University
PDF link is here
Comments are closed.
I also find your conclusions regarding the positive relationship between highway location and buried treasure interesting- its supposed to be hard to find, which usually means hard to get to! I am curious what the timeline looks like for the discoveries of these treasures, and if highways were present at their finding and at their burial. I would be intrigued to see some amateur or even professional treasure hunters put this information to use as well. Thank you for sharing!
Yeah, the highway proximity correlation was interesting to me as well. I thought the exact opposite would be the case. As for the timeline, most of these treasures were buried in the 1700s or 1800s and found fairly recently (in the past 50 years). Highways would not have existed when they were buried but would have been there when they were discovered. I hope someone is able to put this information to use and maybe find some treasure! Thank you for your comment!
Gabby, interesting initial exploration here–I wonder where you got the information about the buried treasure (more specifically–as in what periodicals, etc) and how representative your sample might be. Also, how correlated are highways? How did you find this correlation? Likewise for the others. The inverse correlation on income is also interesting, and I wonder why you think that might be?
I got the treasure location data from new articles of found treasure and locations where treasure hunters with metal detectors found coins. I also searched by state, making sure that I got at least three points per contiguous state. There are definitely more that I missed, I wanted to avoid overwhelming myself with research so I limited it to 150 (which might have skewed the data because of the popularity of websites on the first couple pages on Google). The highway correlation (well, more probability from measuring the distance from points to highways) was an 83.3% chance of the treasure point being located on or very close to a highway. I found the correlation by using the Measure tool in ArcMap to find the distance between the point and the closest part of the nearest highway to it, the other correlations were obtained by overlaying the population density and median income by county with the point layer and recording the data associated with the county where the point was located.
I think the inverse correlation with income may be due to the fact that a lot of the treasure locations were in former ghost towns or in areas where there was a natural disaster, so people may not have wanted to move there afterward unless they had to for some reason. The people who did move there are more likely to be lower-income (which trends with the population density of those areas as well).
Fun project! It might fit nicely with a suitability analysis, since it seems like you have a few variables and criteria in mind as to what makes a good hiding spot. Since presumably there are some valuable treasures hidden in places we don’t know about, it would be interesting to create a suitability model and compare the results to the data points you collected via preliminary online research. If you can create a model accurate enough, it might give some good ideas on where to look next 🙂 Did you stumble across any treasure hunters who have used GIS to aide in their searches?
That is a really good idea, I hadn’t thought of that! Thank you for letting me know. And I did not, though I feel like maybe this project could help them.
Interesting project, Gabby. This reminds of helping in a national treasure hunt back in 2005, which led us to a small state park in Arkansas. It was an organized hunt, commercialized puzzle, to go along with a children’s book called A Treasures Trove, but it was similar. Of the 12 treasures, most were located a bit off the beaten path. (I found one of the 12 treasures, a jeweled caterpillar brooch). This also reminds me of how pirates’ treasure would be located on remote islands. Ideally, they would be looking for a less populated island, perhaps with lower living standards, but not so far away from main shipping routes that they could not get back to their treasure. Do you think the same type of spatial analysis would hold true for “pirate treasure” in the Caribbean?
I’ve heard of that one and that’s really cool that you were able to find one of the treasures! It makes sense that pirates would want to take into account population and shipping routes (those are similar to the metrics I was using!) I believe that with a couple tweaks to account for the lack of some modern conveniences, this type of spatial analysis could definitely work for pirate treasure.
This is a very interesting topic. I like the idea of covering different layers with spatial information and apply the visual comparison. I think it would be better to incorporate some of the statistical analysis such as conducting regression of population density as well as highway density on treasure using county data points to support your opinion. Overall, it is very innovative GIS project! Thanks!
You’re probably right about the statistical analysis, it could definitely help me get accurate results with respect to population and distance from highways. Thank you for lettng me know!