Viers, JH, NR Santos. 2014. Hydrolapse Videography: a coupled hydroinformatic stack for improved visual assessment of river dynamics.11th Hydroinformatics Conference. International Water Association. New York, NY.

A central problem for managers of river ecosystems is understanding hydrodynamics in an intelligible fashion for effective decision-making. Too often decisions are based on some combination of incomplete hydrologic time series, poorly parameterized models, photographic snapshots of the place in question, and human imagination. Because river ecosystems are dynamic, with ever changing physical dimensions, inferences are made to make up for unknowns, which remain knowledge gaps. This is particularly true in the pursuit of ecogeomorphology, which seeks to understand the interplay between physical processes and ecological responses in river systems over time. Increasing deployment of affordable sensors, however, is beginning to provide information to fill such gaps, but their utility remains constrained by human interpretation. We developed “hydrolapse videography” – a digital, time-encoded coupling of hydrologic information records with corresponding timelapse imagery – as a means to provide river managers a rich, but intelligible, data stream that is both quantitative and qualitative in nature. Hydrolapse videography couples still images from timelapse cameras with hydroinformatical graphs of key monitoring variables sourced from either local instruments or remote data providers (e.g., USGS gages) at a fixed time interval (typically hourly). Video composites are created by a combination of Python scripting and matplotlib for graphing, Perl for data pairing, and FFmpeg for video composition (typically 10 fps). Completed videos, typically capturing an entire water year, allow for diagnostic viewing of major hydrologic events, and simultaneous assessment of corresponding environmental variables, such as discharge and stream temperature. Further, the hydrolapse approach allows managers to comprehend river behavior, such as changing habitat dimensions in lateral extent and stage, as function of flow regime. Future versions will include fully automated data retrieval from standard data providers (e.g., WaterML) and processing workflow in a Python application stack.