Sam Fuchs (scf73)
I've designed a simple, static visualization that uses the San Francisco trees dataset to illustrate the growth of trees over time. In the process of designing, building and iterating on this visualization, I've added and subtracted elements to generate this final product.
Many of the trees in the San Francisco Trees dataset have the date of their planting recorded. For these trees, we can calculate almost exactly their age. This allows us to examine the relationship between tree age and size. In a sense, this represents the growth of trees over time. While the data is of course noisy, we can clearly see a pattern when we examine the mean diameter for trees planted in each year. This visualization allows us to watch trees grow over time.
The largest concern with this design concept is that the planting date is not recorded for all of the trees in the dataset. Because there is no way to assume a planting date for these trees, I've chosen to exclude them from the visualization, which means that this visualization only represents 9,300 trees across San Francisco. This is, however, a significant number of trees, and makes for a clear and visually appealing plot.
The major plot shows all of the trees with planting dates from the dataset, organized like so: the x-axis uses the planting date of the tree to compute its age in years from planting, and the y-axis represents the tree's diameter at breast height. Although no unit is provided for the y-axis, I'm relatively confident that the trees are measured in centimeters. Because the density of these points makes the trend within the plot difficult to interpret unaided, I've added a trendline which follows the mean diameter of trees planted in each year. This trendline helps us to see how the tree diameter increases with time.
The two diagrams in the right-hand column depict two subsets of the data. In the plot on the top right, a map is drawn showing the locations of the largest trees around San Francisco. Many of these are immediately recognizable in the plot on the left as outliers. By mapping the locations of these trees, we gain a better understanding of those outliers and how they are distributed throughout the city. We can imagine a San Francisco resident who looks at the plot on the left and immediately wonders at the fact that there are trees in the city that are nearly a meter in diameter. This plot would show them where those trees are, perhaps so they see them when they are nearby.
The scatterplot on the lower right-hand side is similar to the plot on the left, except that it examines the opposite subset of the data: the trees which are less than 25cm in diameter. Because there are so many more of these trees, the pattern of growth is more pronounced, particularly as major outliers are removed. Additionally, I use a different line interpolation scheme, which smoothes out peaks and valleys more aggressively, and, as a result, the correlation between increasing age and increasing diameter is more obvious. This plot is designed to more clearly depict how the dataset illustrates tree growth.