The map is temporarily down for maintenance while I transition from ESRI to open source mapping libraries. Please contact me for an offline copy of the data set.
This analysis used data from OpenStreetMap, the US Census, and the Maine GeoLibrary to determine the amount of time it takes to walk from the center of each Census block to the nearest conserved property.
75% of Mainers lack access to conservation land within a 10-minute walk.
Communities identified as “disadvantaged” by the Climate Equity and Justice Screening Tool (CEJST) are 24% more likely to lack access to conserved land within a 10-minute walk.
Utilizing the CEJST equity metrics revealed inequities in access that were not visible based on purely demographic data.
Diagram showing relationships between different data frames used in this analysis.
This analysis is a step towards quantifying disparities in access to the outdoors, but it is by no means a comprehensive analysis.
This analysis uses the Maine Conserved Lands layer as the database of conservation lands. However, doing so limits the analysis to lands that have been formally (and accurately) registered in the state database. Future avenues of study could include utilizing other metrics of "green space" that are more inclusive. For instance, many cemetaries serve similar values as community parks, but were not included in this analysis.
CEJST is a relatively new initiative developed by the Biden Administration (but drawing on a decade of agency work under previous administrations). It assesses eight different categories of "indicators of burden" and labels a tract as "disadvantaged" if it hits a predefined threshold in one or more categories. While this method is already being used by federal agencies, it has yet to be studied extensively.
This analysis connects Census blocks and conserved lands to the OpenStreetMap network by assigning each block/land to the network node that is closest to the block/tract centroid. This means we are assuming that 1) the center of the block or land is a good approximation for the rough "location" and 2) the block or land is sufficiently small that it can be assigned to a single node. This may not always be accurate, particularly in the case of larger blocks or parcels.
This analysis did not attempt to filter lands with regards to whether they are explicitly open to the public. Thus, even though a land may be within a walk distance, that does not guarantee that the land has trails or otherwise can be accessed by the general public.
Furthermore, many private lands in Maine are open to the public. These lands are not captured by this analysis.