Before the afternoon thunderstorms started rolling in, we (the trees team) spent the afternoon with Alex Beasley of Trees Atlanta. After arriving at their headquarters near Grant Park, Alex drove us around a few surrounding neighborhoods to find potential tree planting locations. Our goal was to experience and understand his scouting process, which will help us create the best tools. It can only be considered a success if Trees Atlanta and the City of Atlanta find it useful; it ultimately doesn’t matter if only we think it is.
Prior to today, we knew that the process was mainly the following steps:
(1) Drive around looking for possible planting locations
(2) Mark possible planting locations on a paper map printed from Google Maps. Possible planting locations are empty public spaces, such as grass between the sidewalk and road, or private property such as apartment complexes.
As Alex talked us through his process, we learned that location of power lines and local temperatures are both important factors to consider – trees need room to grow without interfering with existing development, and higher local temperatures likely mean less existing shade. We are looking into using power line data and urban heat maps to narrow down viable planting locations.
Over the past week, we have continued working on visualizations using d3 (a Javascript library) and analyses using ArcGIS. I have primarily been working with parcel data from the City of Atlanta and Urban Tree Canopy (UTC) data generated by Tony (one of our mentors). In a nutshell, parcel data breaks land down by property (e.g. single residential home, cemetery). I’ve broken it down by UTC percentage.
I’ve also started looking at attributes including property type and financial information to help Trees Atlanta find new planting locations and the city determine which parcels of land they might want to purchase and preserve. Another interesting dataset, especially to help Alex, is impervious surfaces such as buildings or parking lots or roads. It’s easy to tell by looking at an overlay of impervious surfaces and UTC percentage where potential planting locations might be. Low impervious cover combined with low tree cover is a good indicator that there are many planting locations within the parcel.
These images show the same location 4 different ways (Google Maps, Alex’s drawing, UTC impervious, satellite). This will be interesting to explore, as we start to develop and iterate on a tool that Trees Atlanta and the City of Atlanta will incorporate into their process. Finding out which representations of the data are most helpful for various tasks is possibly the most important consideration – otherwise it’s just numbers that no one knows what to do with.