Falloff in D.C. construction 2006-2007
1 August 2008
The D.C. construction market provides a good example of a complex geotemporal (location + time) system. As in any city, construction increases and decreases over both time and space. The particular ebbs and flows of construction in the D.C. area, though, are particular to that metropolis. The dataset for this visualization comes from DCStat, a pioneer in open city data, and includes all completed construction projects reported by the District Department of Transportation from 2004 to 2007.
A quick view of this dataset in the SpatialKey Map Comparison template shows the construction hotspots across the city and the time chart (a combined timeline/histogram) reveals the linear and seasonal trends over time. The Map Comparison template uses a visualization technique called small multiples (in this case, only two multiples are used). Rather than showing one view, and requiring users to switch back-and-forth between different areas or time periods, you can view and manipulate two maps at the same time. This template is especially useful for visualizing data where specific geographic or temporal comparisons are meaningful. In the D.C. construction dataset, comparing year-to-year reveals a marked decrease from 2006 to 2007 (see screenshot above). Moreover, the trend is not a simple decrease, as some areas decreased more than others, and some neighborhoods actually experienced an increase in construction activity. The biggest decrease took place in the Mount Pleasant neighborhood, while the Georgetown and Palisades neighborhoods actually experienced a growth in construction.
Such trends may or may not be visible in our other visualization templates, but the Map Comparison Template is particularly well suited at revealing them. Another powerful aspect of this template is the live linking of the two views. By mousing over an area of D.C., the same area is highlighted on the other map (if in view), with both maps revealing the specific number of construction sites for that area. Live linking is also demonstrated on the time charts — mousing over a histogram bar in one chart reveals the tooltip on both charts, displaying different values if the maps are showing different locations or scales.
No Comments Yet