5.2 Imaging
5.3 Statistics
Statistically, the Audubon volunteer data is challenging to work with due to its small scale. So far, the purpose of the Audubon volunteer data is to serve as a point of reference for light and radar analysis depending on where spikes in bird death occur. When we see a spike in found birds and bird death in the Audubon volunteer data, this should call for a more geographically thorough analysis of light and radar data on that date (or the night before). Moving forward, we hope to be able to correlate spikes in volunteer data and evaluate the possibility of finding patterns across individual volunteer walkers. Generally speaking, however, this correlative analysis of death spikes with radar and light data will be investigated from July 8th through July 15th.
Furthermore, we plan to run quasi-experiments comparing overall bright nights to dark nights, as well as looking at specific high-brightness buildings (such as the Empire State Building) - comparing when these buildings are on / off. We also expect to compare results between building topographies - in particular, determining whether or not building height is a factor in explaining bird density in a particular area. While it may be possible to determine that on days with high bird death counts, the previous night's bird densities were comparatively high as well, and even that the area was also comparatively bright that night, attributing causality to the increased light levels will require experimental setups such as these in areas where high bird death counts were found.
5.4 Visualization
The framework of the time-dependent visualization of the brightness and bird mass data has been established. The code for interpolating values of interest across our scene of interest has been written and primarily awaits data from the imaging and radar portions of the project. Formatting compatibility between the other outputs and these inputs will be discussed and established by the end of the first week of July. For the remaining weeks of the project, through mid July, the visualization of these data will be refined to clearly communicate the relationship between these data.