waterviz

WaterViz.com aims to provide an intuitive view of the current and historic state of water in the conterminous United States. To do so, it presents all of the relevant information in a mapped form.

A walk-through video is here and a video showing how WaterViz.com can be used to facilitate analysis is here.

Hydrographic data drawn from the National Hydrography Dataset Plus allows the map to display data from all rivers and active U.S. gauge stations for 2007 onwards. Gauges and rivers are colored and sized based on how their discharge rate ranks against a thirty year history. Hovering over rivers and stations will display their current discharge rate and stage height, along with links for more information. Current discharge rates are drawn from the National Water Information System.

To facilitate understanding of conditions, the National Land Cover Database, 2011 edition has been used to generate a base layer showing a 20-classification land-use through 13 distinct zoom levels.

To facilitate understanding of the relationships between land-use and water, the National Hydrography Dataset Plus was burned into raster format and proximity-to-water rasters were developed. These were used in conjunction with the National Land Cover Database edition 2001 and 2011 editions to quantify how land-use has changed with regards to proximity to water over the past decade.

This information is displayed on a per-county basis (using the Census TIGER/Line files) where counties are coloured from white to deep red depending on how much near-shore development they have incurred.

Put another way, the foregoing provides an analysis of riparian buffer zones. The lack of such zones has much to do with poor water quality and dead zones downstream and policy action to expand buffer zone size and quantity has become a hot topic recently.

Hurricane data has been overlaid in order to facilitate understanding of the causes for high water volumes.

Future improvements to the project include optimizing server-side database queries and caching. While the project is set up to update data in real-time, updating has been disabled for the next couple of weeks as a proof-of-concept to ensure speedy performance. Additionally improvements will be made to the user interface to facilitate deeper exploration of the data.

In terms of project stats, the server holds 50 gigabytes of explorable data. To perform county-level buffer strip analysis over 11 terabytes of intermediate products were generated and a few hundred hours of supercomputing time were expended. Being a student can have its benefits!

Technical details are available on the project's GitHub page.

Project Information

Resources