�ؾ���

This summer I worked with Professor of Geography William Meyer on a project titled, “State, Local, and City Parks and the Worthless Lands Thesis.” The “worthless lands” thesis, proposed by the historian Alfred Runte, holds that national parks in the United States are situated on lands that are useless for conventional economic practices such as agriculture, grazing, and mining. Our project assessed this thesis at the state, local, and city levels by comparing the location of protected areas to the location of areas with little economic potential, and by examining sources regarding the rationale for the designation of parkland.

We began with a focus on elevation, slope, and distance from the main urban center in the City of Syracuse, Onondaga County, and New York State. Using ArcGIS, a geographic information system software, we mapped protected areas and the three “worthless” variables in each of the three study areas. Our initial methods for analysis involved the division of the study areas into groups based on the “worthless” variables, such that we could test the degree to which low and high value groups coincided with protected areas. We had some trouble with data distribution, but found that the thesis generally held true: protected areas seemed to correlate with high elevation, high slope, and high distance from the urban center.

We then ran further tests focusing on elevation, slope, and soil survey data (flooding frequency for Syracuse; farmland desirability class for Onondaga and New York): adopting a new method for analysis, we set aside the top decile (top 10%) of each study area based on elevation, and did the same for slope and for the soil data. For each study area, we tested the location of the top elevation decile, top slope decile, and top flooding frequency decile (or bottom farmland class decile) against the location of protected areas to assess the degree to which the protected areas coincided with each given “worthless” variable. The percentage of protected areas that coincided with the top elevation decile, for example, could be tested against the 10% figure which represented the proportion of the overall study area that the top elevation decile made up.

Our findings, like earlier, supported Runte’s thesis. After that, we used the same method to test the location of impervious surfaces in Syracuse against the location of the “worthless” variables. Impervious surfaces, the most fully built up areas, presumably indicate high land value, so the expectation would be for impervious land to correlate negatively with “worthless” land. Such a pattern held true, validating our selection of “worthless” variables.

We also ran the analysis in Syracuse with parks built after 1950 to see whether transportation changes affected these patterns, and the results were fairly similar to our initial findings. Our focus then shifted to the nine northeasternmost United States: Maine, New Hampshire, Vermont, New York, Massachusetts, Connecticut, Rhode Island, Pennsylvania, and New Jersey. We ran the analysis as before, testing elevation, slope, and farmland class, and the findings favored Runte’s thesis, although there were some exceptions. Next, we ran the analysis on Los Angeles, testing elevation, slope, and flooding frequency against the protected areas in the City of Los Angeles, Los Angeles County, and the Los Angeles metropolitan area (defined for our purposes as Los Angeles County and Orange County).

Our findings here again supported Runte’s thesis, and our continued work will assess patterns in the six most populous United States cities (New York, Los Angeles, Chicago, Houston, Phoenix, and Philadelphia), and in each of New York’s five boroughs.

I’m thankful to Professor Meyer and Myongsun Kong for their guidance and support this summer.