Through my work at the Center for Integrated Data Analytics, I’ve created many interactive web-based data visualizations. I highlight a few of the publicly-available works below.
Descriptions of the apps are my own – they are not representative of the US Geological Survey, or any related stakeholders including, but not limited to, the EPA, the National Water Census, NHDPlus, or the Great Lakes Restoration Initiative. All the mentioned applications are the product of extensive teamwork. While none of the work I reference is entirely my own, I have elected to highlight applications where I played an instrumental role in data visualization. This portfolio is a snapshot of the apps as of 3/7/2015. The screenshots I provide and the behavior I describe may fall out of sync with the referenced apps as developers implement changes.
The California Drought, Visualized
This application narrates the California drought from 2011 to present. As users scroll through time, interwoven prose provides context for interpreting charts and maps representing available water resources.
My contribution was the implementation of the scroll-sensitive reservoir bar chart.
Note: reservoirs occasionally overflow their published capacity
The bar chart compares reservoirs’:
- maximum capacities represented by bar width
- percent capacities at a point in time represented by bar height
- elevations represented by the ordering of the bars in ascending elevation from left to right. This highlights the seasonal contribution of high-altitude snow as a source of water for lower elevations.
Most of this California Drought work was accomplished as part of a two-day hackathon.
The National Water Census Data Portal
This recently released application provides workflows for exploring various water-related national data sets. The most complex workflow is the Water Budget Workflow. The Water Budget Workflow provides users with a map they can use to drill down to their area of interest.
Once the area of interest is selected, users can visualize and download the data sets necessary to compute water budgets. The first data sets visualized are watershed-level precipitation and evapotranspiration.
Next, users can select water usage data. The water usage data is available at a county level. This is problematic because a watershed can span multiple counties. Someday it would be nice to get all the water budget datasets onto the county level, or all of them onto the watershed level, but presently there is no way to automatically pick the most representative county for the water usage of a watershed. The amount of overlap between the county and the watershed can serve as a good guideline, but it is not definitive.
To help users overcome this mismatch, the app presents a map and a table that show the spatial extent and approximate quantity of overlap between the watershed and the counties. The user can choose to use the app’s overlap heuristics, or they can choose a county according to their own criteria.
After representative pairings have been chosen, area-weighted water usage data is plotted in inches per day. This choice of unit for water usage is consistent with precipitation, thus eliminating the need to perform manual unit conversion when constructing water budgets.
Even though the app enables large permutations of possible user selections, the app can create a link at any point in the workflow so that others can reproduce and interact with another user’s analysis. This feature is reminiscent of Google Maps’ link-sharing functionality.
This application provides a geospatial web interface for exploring a hydrologic flow model in the context of the Great Lakes.
Users can select a location of interest via an interactive map. Once a site is selected, users plot model forecasts and hindcasts for a reach or a catchment at that site.
Users can zoom to a time period of interest.
Statistics are dynamically computed on the selected site’s flows and yields. The results can be added to the plots. Here we see decile lines: