2009 BES Annual Meeting Presentation and Poster Abstracts
Integrating Real-Time Sensor Networks, Data Assimilation, and Predictive Modeling to Quantify the Urban Water Cycle
Co-Authors: Andrew Miller, Michael McGuire, Philip Larson, Christiane Runyan, Ed Doheny, Jonathan Dillow, Jeffrey Campbell, Aditi Bhaskar, Garth Lindner, Alimatou Seck
Abstract: Our goal is to build a real-time observing system of sensors, wireless telemetry, and advanced visualization and modeling tools for simultaneous display of the temporal and spatial patterns of all components of the hydrologic cycle at sites throughout the Baltimore metropolitan region. Within the past two years we have (1) worked with USGS to convert existing BES stream gages to real-time and have added 9 real-time gages to the Gwynns Falls observing system; (2) replaced BES single tipping-bucket raingages with real-time pairs at each site; (3) deployed an eddy covariance station on the roof of the UMBC Physics building streaming data for real-time ET calculation; (4) installed 8 shallow riparian wells with automatic water level recorders in Dead Run; (5) “adopted” abandoned homeowner bedrock wells and outfitted them with recording pressure transducers; and (6) identified and ingested regional legacy data into our spatio-temporal databases. We have completed testing of soil moisture, temperature and tensiometer probes; we are testing nitrate and other water quality sensors to add to the network. Telemetry includes GOES satellite, cell phone, and radio transmission. Our visualization lab includes a bank of LCD screens providing a 4.5 x 14.5-ft display composed of 16.6 million pixels. A group of high-speed workstations is available for intensive computing. We are building local-scale hydraulic models of Red Run and Dead Run, coupled three-dimensional distributed hydrologic models of Dead Run, Gunpowder-Patapsco, and Chesapeake Bay watersheds, and revising an older SWMM model of the Gwynns Falls using more modern data sets.